Save cached_read_* in superblock

Instead of it, just do not verify checksums of currently mutated objects.
When clean data modification during flush runs in parallel to a read request,
that request may read a mix of old and new data. It may even read a mix of
multiple flushed versions if it lasts too long... And attempts to verify it
using temporary copies of metadata make the algorithm too complex and creepy.

.gitea/workflows/test.Dockerfile

@@ -10,6 +10,9 @@ RUN set -e -x; \
     ln -s /root/fio-build/fio-*/ ./fio; \
     ln -s /root/qemu-build/qemu-*/ ./qemu; \
     ls /usr/include/linux/raw.h || cp ./debian/raw.h /usr/include/linux/raw.h; \
+    cd mon; \
+    npm install; \
+    cd ..; \
     mkdir build; \
     cd build; \
     cmake .. -DWITH_ASAN=yes -DWITH_QEMU=yes; \

.gitea/workflows/test.yml

@@ -71,7 +71,7 @@ jobs:
     steps:
     - name: Run test
       id: test
-      timeout-minutes: 3
+      timeout-minutes: 10
       run: /root/vitastor/tests/test_add_osd.sh
     - name: Print logs
       if: always() && steps.test.outcome == 'failure'
@@ -190,24 +190,6 @@ jobs:
           echo ""
         done
 
-  test_failure_domain:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: /root/vitastor/tests/test_failure_domain.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
   test_interrupted_rebalance:
     runs-on: ubuntu-latest
     needs: build
@@ -280,7 +262,7 @@ jobs:
           echo ""
         done
 
-  test_minsize_1:
+  test_failure_domain:
     runs-on: ubuntu-latest
     needs: build
     container: ${{env.TEST_IMAGE}}:${{github.sha}}
@@ -288,115 +270,7 @@ jobs:
     - name: Run test
       id: test
       timeout-minutes: 3
-      run: /root/vitastor/tests/test_minsize_1.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
-  test_move_reappear:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: /root/vitastor/tests/test_move_reappear.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
-  test_rebalance_verify:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: /root/vitastor/tests/test_rebalance_verify.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
-  test_rebalance_verify_imm:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: IMMEDIATE_COMMIT=1 /root/vitastor/tests/test_rebalance_verify.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
-  test_rebalance_verify_ec:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: SCHEME=ec /root/vitastor/tests/test_rebalance_verify.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
-  test_rebalance_verify_ec_imm:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: SCHEME=ec IMMEDIATE_COMMIT=1 /root/vitastor/tests/test_rebalance_verify.sh
-    - name: Print logs
-      if: always() && steps.test.outcome == 'failure'
-      run: |
-        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
-          echo "-------- $i --------"
-          cat $i
-          echo ""
-        done
-
-  test_rm:
-    runs-on: ubuntu-latest
-    needs: build
-    container: ${{env.TEST_IMAGE}}:${{github.sha}}
-    steps:
-    - name: Run test
-      id: test
-      timeout-minutes: 3
-      run: /root/vitastor/tests/test_rm.sh
+      run: /root/vitastor/tests/test_failure_domain.sh
     - name: Print logs
       if: always() && steps.test.outcome == 'failure'
       run: |
@@ -442,6 +316,132 @@ jobs:
           echo ""
         done
 
+  test_minsize_1:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: /root/vitastor/tests/test_minsize_1.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_move_reappear:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: /root/vitastor/tests/test_move_reappear.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_rm:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: /root/vitastor/tests/test_rm.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_snapshot_chain:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: /root/vitastor/tests/test_snapshot_chain.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_snapshot_chain_ec:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: SCHEME=ec /root/vitastor/tests/test_snapshot_chain.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_snapshot_down:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: /root/vitastor/tests/test_snapshot_down.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_snapshot_down_ec:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 3
+      run: SCHEME=ec /root/vitastor/tests/test_snapshot_down.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
   test_splitbrain:
     runs-on: ubuntu-latest
     needs: build
@@ -460,6 +460,78 @@ jobs:
           echo ""
         done
 
+  test_rebalance_verify:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: /root/vitastor/tests/test_rebalance_verify.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_rebalance_verify_imm:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: IMMEDIATE_COMMIT=1 /root/vitastor/tests/test_rebalance_verify.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_rebalance_verify_ec:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: SCHEME=ec /root/vitastor/tests/test_rebalance_verify.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_rebalance_verify_ec_imm:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: SCHEME=ec IMMEDIATE_COMMIT=1 /root/vitastor/tests/test_rebalance_verify.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
   test_write:
     runs-on: ubuntu-latest
     needs: build
@@ -550,6 +622,114 @@ jobs:
           echo ""
         done
 
+  test_heal_csum_32k_dmj:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: TEST_NAME=csum_32k_dmj OSD_ARGS="--data_csum_type crc32c --csum_block_size 32k --inmemory_metadata false --inmemory_journal false" OFFSET_ARGS=$OSD_ARGS /root/vitastor/tests/test_heal.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_heal_csum_32k_dj:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: TEST_NAME=csum_32k_dj  OSD_ARGS="--data_csum_type crc32c --csum_block_size 32k --inmemory_journal false" OFFSET_ARGS=$OSD_ARGS /root/vitastor/tests/test_heal.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_heal_csum_32k:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: TEST_NAME=csum_32k     OSD_ARGS="--data_csum_type crc32c --csum_block_size 32k" OFFSET_ARGS=$OSD_ARGS /root/vitastor/tests/test_heal.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_heal_csum_4k_dmj:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: TEST_NAME=csum_4k_dmj  OSD_ARGS="--data_csum_type crc32c --inmemory_metadata false --inmemory_journal false" OFFSET_ARGS=$OSD_ARGS /root/vitastor/tests/test_heal.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_heal_csum_4k_dj:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: TEST_NAME=csum_4k_dj   OSD_ARGS="--data_csum_type crc32c --inmemory_journal false" OFFSET_ARGS=$OSD_ARGS /root/vitastor/tests/test_heal.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
+  test_heal_csum_4k:
+    runs-on: ubuntu-latest
+    needs: build
+    container: ${{env.TEST_IMAGE}}:${{github.sha}}
+    steps:
+    - name: Run test
+      id: test
+      timeout-minutes: 10
+      run: TEST_NAME=csum_4k      OSD_ARGS="--data_csum_type crc32c" OFFSET_ARGS=$OSD_ARGS /root/vitastor/tests/test_heal.sh
+    - name: Print logs
+      if: always() && steps.test.outcome == 'failure'
+      run: |
+        for i in /root/vitastor/testdata/*.log /root/vitastor/testdata/*.txt; do
+          echo "-------- $i --------"
+          cat $i
+          echo ""
+        done
+
   test_scrub:
     runs-on: ubuntu-latest
     needs: build

.gitea/workflows/tests-to-yaml.pl

@@ -7,15 +7,22 @@ for my $line (<>)
     if ($line =~ /\.\/(test_[^\.]+)/s)
     {
         chomp $line;
-        my $test_name = $1;
+        my $base_name = $1;
+        my $test_name = $base_name;
         my $timeout = 3;
-        if ($test_name eq 'test_etcd_fail' || $test_name eq 'test_heal' || $test_name eq 'test_interrupted_rebalance')
+        if ($test_name eq 'test_etcd_fail' || $test_name eq 'test_heal' || $test_name eq 'test_add_osd' ||
+            $test_name eq 'test_interrupted_rebalance' || $test_name eq 'test_rebalance_verify')
         {
             $timeout = 10;
         }
         while ($line =~ /([^\s=]+)=(\S+)/gs)
         {
-            if ($1 eq 'SCHEME' && $2 eq 'ec')
+            if ($1 eq 'TEST_NAME')
+            {
+                $test_name = $base_name.'_'.$2;
+                last;
+            }
+            elsif ($1 eq 'SCHEME' && $2 eq 'ec')
             {
                 $test_name .= '_ec';
             }

CMakeLists.txt

@@ -2,6 +2,6 @@ cmake_minimum_required(VERSION 2.8.12)
 
 project(vitastor)
 
-set(VERSION "0.9.0")
+set(VERSION "0.9.3")
 
 add_subdirectory(src)

README-ru.md

@@ -15,7 +15,7 @@ Vitastor архитектурно похож на Ceph, что означает
 и автоматическое распределение данных по любому числу дисков любого размера с настраиваемыми схемами
 избыточности - репликацией или с произвольными кодами коррекции ошибок.
 
-Vitastor нацелен на SSD и SSD+HDD кластеры с как минимум 10 Гбит/с сетью, поддерживает
+Vitastor нацелен в первую очередь на SSD и SSD+HDD кластеры с как минимум 10 Гбит/с сетью, поддерживает
 TCP и RDMA и на хорошем железе может достигать задержки 4 КБ чтения и записи на уровне ~0.1 мс,
 что примерно в 10 раз быстрее, чем Ceph и другие популярные программные СХД.
 

README.md

@@ -14,8 +14,8 @@ Vitastor is architecturally similar to Ceph which means strong consistency,
 primary-replication, symmetric clustering and automatic data distribution over any
 number of drives of any size with configurable redundancy (replication or erasure codes/XOR).
 
-Vitastor targets SSD and SSD+HDD clusters with at least 10 Gbit/s network, supports
-TCP and RDMA and may achieve 4 KB read and write latency as low as ~0.1 ms
+Vitastor targets primarily SSD and SSD+HDD clusters with at least 10 Gbit/s network,
+supports TCP and RDMA and may achieve 4 KB read and write latency as low as ~0.1 ms
 with proper hardware which is ~10 times faster than other popular SDS's like Ceph
 or internal systems of public clouds.
 

csi/Makefile

@@ -1,4 +1,4 @@
-VERSION ?= v0.9.0
+VERSION ?= v0.9.3
 
 all: build push
 

csi/deploy/004-csi-nodeplugin.yaml

@@ -49,7 +49,7 @@ spec:
             capabilities:
               add: ["SYS_ADMIN"]
             allowPrivilegeEscalation: true
-          image: vitalif/vitastor-csi:v0.9.0
+          image: vitalif/vitastor-csi:v0.9.3
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/deploy/007-csi-provisioner.yaml

@@ -116,7 +116,7 @@ spec:
             privileged: true
             capabilities:
               add: ["SYS_ADMIN"]
-          image: vitalif/vitastor-csi:v0.9.0
+          image: vitalif/vitastor-csi:v0.9.3
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/src/config.go

@@ -5,7 +5,7 @@ package vitastor
 
 const (
     vitastorCSIDriverName    = "csi.vitastor.io"
-    vitastorCSIDriverVersion = "0.9.0"
+    vitastorCSIDriverVersion = "0.9.3"
 )
 
 // Config struct fills the parameters of request or user input

debian/build-pve-qemu.sh

@@ -0,0 +1,58 @@
+exit
+
+git clone https://git.yourcmc.ru/vitalif/pve-qemu .
+
+# bookworm
+
+docker run -it -v `pwd`/pve-qemu:/root/pve-qemu --name pve-qemu-bullseye debian:bullseye bash
+
+perl -i -pe 's/Types: deb$/Types: deb deb-src/' /etc/apt/sources.list.d/debian.sources
+echo 'deb [arch=amd64] http://download.proxmox.com/debian/pve bookworm pve-no-subscription' >> /etc/apt/sources.list
+echo 'deb https://vitastor.io/debian bookworm main' >> /etc/apt/sources.list
+echo 'APT::Install-Recommends false;' >> /etc/apt/apt.conf
+echo 'ru_RU UTF-8' >> /etc/locale.gen
+echo 'en_US UTF-8' >> /etc/locale.gen
+apt-get update
+apt-get install wget ca-certificates
+wget https://enterprise.proxmox.com/debian/proxmox-release-bookworm.gpg -O /etc/apt/trusted.gpg.d/proxmox-release-bookworm.gpg
+wget https://vitastor.io/debian/pubkey.gpg -O /etc/apt/trusted.gpg.d/vitastor.gpg
+apt-get update
+apt-get install git devscripts equivs wget mc libjemalloc-dev vitastor-client-dev lintian locales
+mk-build-deps --install ./control
+
+# bullseye
+
+docker run -it -v `pwd`/pve-qemu:/root/pve-qemu --name pve-qemu-bullseye debian:bullseye bash
+
+grep '^deb ' /etc/apt/sources.list | perl -pe 's/^deb /deb-src /' >> /etc/apt/sources.list
+echo 'deb [arch=amd64] http://download.proxmox.com/debian/pve bullseye pve-no-subscription' >> /etc/apt/sources.list
+echo 'deb https://vitastor.io/debian bullseye main' >> /etc/apt/sources.list
+echo 'APT::Install-Recommends false;' >> /etc/apt/apt.conf
+echo 'ru_RU UTF-8' >> /etc/locale.gen
+echo 'en_US UTF-8' >> /etc/locale.gen
+apt-get update
+apt-get install wget
+wget https://enterprise.proxmox.com/debian/proxmox-release-bullseye.gpg -O /etc/apt/trusted.gpg.d/proxmox-release-bullseye.gpg
+wget https://vitastor.io/debian/pubkey.gpg -O /etc/apt/trusted.gpg.d/vitastor.gpg
+apt-get update
+apt-get install git devscripts equivs wget mc libjemalloc-dev vitastor-client-dev lintian locales
+mk-build-deps --install ./control
+
+# buster
+
+docker run -it -v `pwd`/pve-qemu:/root/pve-qemu --name pve-qemu-buster debian:buster bash
+
+grep '^deb ' /etc/apt/sources.list | perl -pe 's/^deb /deb-src /' >> /etc/apt/sources.list
+echo 'deb [arch=amd64] http://download.proxmox.com/debian/pve buster pve-no-subscription' >> /etc/apt/sources.list
+echo 'deb https://vitastor.io/debian buster main' >> /etc/apt/sources.list
+echo 'deb http://deb.debian.org/debian buster-backports main' >> /etc/apt/sources.list
+echo 'APT::Install-Recommends false;' >> /etc/apt/apt.conf
+echo 'ru_RU UTF-8' >> /etc/locale.gen
+echo 'en_US UTF-8' >> /etc/locale.gen
+apt-get update
+apt-get install wget ca-certificates
+wget http://download.proxmox.com/debian/proxmox-ve-release-6.x.gpg -O /etc/apt/trusted.gpg.d/proxmox-ve-release-6.x.gpg
+wget https://vitastor.io/debian/pubkey.gpg -O /etc/apt/trusted.gpg.d/vitastor.gpg
+apt-get update
+apt-get install git devscripts equivs wget mc libjemalloc-dev vitastor-client-dev lintian locales
+mk-build-deps --install ./control

debian/build-vitastor-bookworm.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+cat < vitastor.Dockerfile > ../Dockerfile
+cd ..
+mkdir -p packages
+sudo podman build --build-arg REL=bookworm -v `pwd`/packages:/root/packages -f Dockerfile .
+rm Dockerfile

debian/changelog

@@ -1,10 +1,10 @@
-vitastor (0.9.0-1) unstable; urgency=medium
+vitastor (0.9.3-1) unstable; urgency=medium
 
   * Bugfixes
 
  -- Vitaliy Filippov <vitalif@yourcmc.ru>  Fri, 03 Jun 2022 02:09:44 +0300
 
-vitastor (0.9.0-1) unstable; urgency=medium
+vitastor (0.9.3-1) unstable; urgency=medium
 
   * Implement NFS proxy
   * Add documentation

debian/patched-qemu.Dockerfile

@@ -1,4 +1,4 @@
-# Build patched QEMU for Debian Buster or Bullseye/Sid inside a container
+# Build patched QEMU for Debian inside a container
 # cd ..; podman build --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f debian/patched-qemu.Dockerfile .
 
 ARG REL=
@@ -15,17 +15,19 @@ RUN if [ "$REL" = "buster" -o "$REL" = "bullseye" ]; then \
         echo 'Pin-Priority: 500' >> /etc/apt/preferences; \
     fi; \
     grep '^deb ' /etc/apt/sources.list | perl -pe 's/^deb/deb-src/' >> /etc/apt/sources.list; \
+    perl -i -pe 's/Types: deb$/Types: deb deb-src/' /etc/apt/sources.list.d/debian.sources || true; \
     echo 'APT::Install-Recommends false;' >> /etc/apt/apt.conf; \
     echo 'APT::Install-Suggests false;' >> /etc/apt/apt.conf
 
 RUN apt-get update
-RUN apt-get -y install qemu fio liburing1 liburing-dev libgoogle-perftools-dev devscripts
+RUN apt-get -y install fio liburing-dev libgoogle-perftools-dev devscripts
 RUN apt-get -y build-dep qemu
 # To build a custom version
 #RUN cp /root/packages/qemu-orig/* /root
 RUN apt-get --download-only source qemu
 
-ADD patches/qemu-5.0-vitastor.patch patches/qemu-5.1-vitastor.patch patches/qemu-6.1-vitastor.patch src/qemu_driver.c /root/vitastor/patches/
+ADD patches /root/vitastor/patches
+ADD src/qemu_driver.c /root/vitastor/src/qemu_driver.c
 RUN set -e; \
     apt-get install -y wget; \
     wget -q -O /etc/apt/trusted.gpg.d/vitastor.gpg https://vitastor.io/debian/pubkey.gpg; \
@@ -37,25 +39,16 @@ RUN set -e; \
     rm -rf /root/packages/qemu-$REL/*; \
     cd /root/packages/qemu-$REL; \
     dpkg-source -x /root/qemu*.dsc; \
-    if ls -d /root/packages/qemu-$REL/qemu-5.0*; then \
-        D=$(ls -d /root/packages/qemu-$REL/qemu-5.0*); \
-        cp /root/vitastor/patches/qemu-5.0-vitastor.patch $D/debian/patches; \
-        echo qemu-5.0-vitastor.patch >> $D/debian/patches/series; \
-    elif ls /root/packages/qemu-$REL/qemu-6.1*; then \
-        D=$(ls -d /root/packages/qemu-$REL/qemu-6.1*); \
-        cp /root/vitastor/patches/qemu-6.1-vitastor.patch $D/debian/patches; \
-        echo qemu-6.1-vitastor.patch >> $D/debian/patches/series; \
-    else \
-        cp /root/vitastor/patches/qemu-5.1-vitastor.patch /root/packages/qemu-$REL/qemu-*/debian/patches; \
-        P=`ls -d /root/packages/qemu-$REL/qemu-*/debian/patches`; \
-        echo qemu-5.1-vitastor.patch >> $P/series; \
-    fi; \
+    QEMU_VER=$(ls -d qemu*/ | perl -pe 's!^.*(\d+\.\d+).*!$1!'); \
+    D=$(ls -d qemu*/); \
+    cp /root/vitastor/patches/qemu-$QEMU_VER-vitastor.patch ./qemu-*/debian/patches; \
+    echo qemu-$QEMU_VER-vitastor.patch >> $D/debian/patches/series; \
     cd /root/packages/qemu-$REL/qemu-*/; \
     quilt push -a; \
     quilt add block/vitastor.c; \
-    cp /root/vitastor/patches/qemu_driver.c block/vitastor.c; \
+    cp /root/vitastor/src/qemu_driver.c block/vitastor.c; \
     quilt refresh; \
-    V=$(head -n1 debian/changelog | perl -pe 's/^.*\((.*?)(~bpo[\d\+]*)?\).*$/$1/')+vitastor1; \
+    V=$(head -n1 debian/changelog | perl -pe 's/^.*\((.*?)(~bpo[\d\+]*)?\).*$/$1/')+vitastor3; \
     DEBEMAIL="Vitaliy Filippov <vitalif@yourcmc.ru>" dch -D $REL -v $V 'Plug Vitastor block driver'; \
     DEB_BUILD_OPTIONS=nocheck dpkg-buildpackage --jobs=auto -sa; \
     rm -rf /root/packages/qemu-$REL/qemu-*/

debian/vitastor.Dockerfile

@@ -1,4 +1,4 @@
-# Build Vitastor packages for Debian Buster or Bullseye/Sid inside a container
+# Build Vitastor packages for Debian inside a container
 # cd ..; podman build --build-arg REL=bullseye -v `pwd`/packages:/root/packages -f debian/vitastor.Dockerfile .
 
 ARG REL=
@@ -15,11 +15,12 @@ RUN if [ "$REL" = "buster" -o "$REL" = "bullseye" ]; then \
         echo 'Pin-Priority: 500' >> /etc/apt/preferences; \
     fi; \
     grep '^deb ' /etc/apt/sources.list | perl -pe 's/^deb/deb-src/' >> /etc/apt/sources.list; \
+    perl -i -pe 's/Types: deb$/Types: deb deb-src/' /etc/apt/sources.list.d/debian.sources || true; \
     echo 'APT::Install-Recommends false;' >> /etc/apt/apt.conf; \
     echo 'APT::Install-Suggests false;' >> /etc/apt/apt.conf
 
 RUN apt-get update
-RUN apt-get -y install fio liburing1 liburing-dev libgoogle-perftools-dev devscripts
+RUN apt-get -y install fio liburing-dev libgoogle-perftools-dev devscripts
 RUN apt-get -y build-dep fio
 RUN apt-get --download-only source fio
 RUN apt-get update && apt-get -y install libjerasure-dev cmake libibverbs-dev libisal-dev
@@ -34,8 +35,8 @@ RUN set -e -x; \
     mkdir -p /root/packages/vitastor-$REL; \
     rm -rf /root/packages/vitastor-$REL/*; \
     cd /root/packages/vitastor-$REL; \
-    cp -r /root/vitastor vitastor-0.9.0; \
-    cd vitastor-0.9.0; \
+    cp -r /root/vitastor vitastor-0.9.3; \
+    cd vitastor-0.9.3; \
     ln -s /root/fio-build/fio-*/ ./fio; \
     FIO=$(head -n1 fio/debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     ls /usr/include/linux/raw.h || cp ./debian/raw.h /usr/include/linux/raw.h; \
@@ -48,8 +49,8 @@ RUN set -e -x; \
     rm -rf a b; \
     echo "dep:fio=$FIO" > debian/fio_version; \
     cd /root/packages/vitastor-$REL; \
-    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.9.0.orig.tar.xz vitastor-0.9.0; \
-    cd vitastor-0.9.0; \
+    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.9.3.orig.tar.xz vitastor-0.9.3; \
+    cd vitastor-0.9.3; \
     V=$(head -n1 debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     DEBFULLNAME="Vitaliy Filippov <vitalif@yourcmc.ru>" dch -D $REL -v "$V""$REL" "Rebuild for $REL"; \
     DEB_BUILD_OPTIONS=nocheck dpkg-buildpackage --jobs=auto -sa; \

docs/config.en.md

@@ -21,7 +21,7 @@ Configuration parameters can be set in 3 places:
   mon, fio and QEMU options, OpenStack/Proxmox/etc configuration. The latter
   doesn't allow to set all variables directly, but it allows to override the
   configuration file and set everything you need inside it.
-- OSD superblocks created by [vitastor-disk](../usage/disk.en.md) contain
+- OSD superblocks created by [vitastor-disk](usage/disk.en.md) contain
   primarily disk layout parameters of specific OSDs. In fact, these parameters
   are automatically passed into the command line of vitastor-osd process, so
   they have the same "status" as command-line parameters.

docs/config.ru.md

@@ -23,7 +23,7 @@
   монитора, опциях fio и QEMU, настроек OpenStack, Proxmox и т.п. Последние,
   как правило, не включают полный набор параметров напрямую, но позволяют
   определить путь к файлу конфигурации и задать любые параметры в нём.
-- В суперблоке OSD, записываемом [vitastor-disk](../usage/disk.ru.md) - параметры,
+- В суперблоке OSD, записываемом [vitastor-disk](usage/disk.ru.md) - параметры,
   связанные с дисковым форматом и с этим конкретным OSD. На самом деле,
   при запуске OSD эти параметры автоматически передаются в командную строку
   процесса vitastor-osd, то есть по "статусу" они эквивалентны параметрам

docs/config/layout-cluster.en.md

@@ -33,12 +33,13 @@ Size of objects (data blocks) into which all physical and virtual drives
 in Vitastor, affects memory usage, write amplification and I/O load
 distribution effectiveness.
 
-Recommended default block size is 128 KB for SSD and 4 MB for HDD. In fact,
-it's possible to use 4 MB for SSD too - it will lower memory usage, but
+Recommended default block size is 128 KB for SSD and 1 MB for HDD. In fact,
+it's possible to use 1 MB for SSD too - it will lower memory usage, but
 may increase average WA and reduce linear performance.
 
 OSD memory usage is roughly (SIZE / BLOCK * 68 bytes) which is roughly
 544 MB per 1 TB of used disk space with the default 128 KB block size.
+With 1 MB it's 8 times lower.
 
 ## bitmap_granularity
 

docs/config/layout-cluster.ru.md

@@ -33,14 +33,14 @@ OSD) могут сосуществовать в одном кластере Vita
 настроек, влияет на потребление памяти, объём избыточной записи (write
 amplification) и эффективность распределения нагрузки по OSD.
 
-Рекомендуемые по умолчанию размеры блока - 128 килобайт для SSD и 4
-мегабайта для HDD. В принципе, для SSD можно тоже использовать 4 мегабайта,
+Рекомендуемые по умолчанию размеры блока - 128 килобайт для SSD и 1 мегабайт
+для HDD. В принципе, для SSD можно тоже использовать блок размером 1 мегабайт,
 это понизит использование памяти, но ухудшит распределение нагрузки и в
 среднем увеличит WA.
 
 Потребление памяти OSD составляет примерно (РАЗМЕР / БЛОК * 68 байт),
 т.е. примерно 544 МБ памяти на 1 ТБ занятого места на диске при
-стандартном 128 КБ блоке.
+стандартном 128 КБ блоке. При 1 МБ блоке памяти нужно в 8 раз меньше.
 
 ## bitmap_granularity
 

docs/config/layout-osd.en.md

@@ -24,6 +24,8 @@ initialization and can't be changed after it without losing data.
 - [disable_journal_fsync](#disable_journal_fsync)
 - [disable_device_lock](#disable_device_lock)
 - [disk_alignment](#disk_alignment)
+- [data_csum_type](#data_csum_type)
+- [csum_block_size](#csum_block_size)
 
 ## data_device
 
@@ -174,3 +176,42 @@ Intel Optane (probably, not tested yet).
 
 Clients don't need to be aware of disk_alignment, so it's not required to
 put a modified value into etcd key /vitastor/config/global.
+
+## data_csum_type
+
+- Type: string
+- Default: none
+
+Data checksum type to use. May be "crc32c" or "none". Set to "crc32c" to
+enable data checksums.
+
+## csum_block_size
+
+- Type: integer
+- Default: 4096
+
+Checksum calculation block size.
+
+Must be equal or a multiple of [bitmap_granularity](layout-cluster.en.md#bitmap_granularity)
+(which is usually 4 KB).
+
+Checksums increase metadata size by 4 bytes per each csum_block_size of data.
+
+Checksums are always a compromise:
+1. You either sacrifice +1 GB RAM per 1 TB of data
+2. Or you raise csum_block_size, for example, to 32k and sacrifice
+   50% random write iops due to checksum read-modify-write
+3. Or you turn off [inmemory_metadata](osd.en.md#inmemory_metadata) and
+   sacrifice 50% random read iops due to checksum reads
+
+Option 1 (default) is recommended for all-flash setups because these usually
+have enough RAM.
+
+Option 2 is recommended for HDD-only setups. HDD-only setups usually do NOT
+have enough RAM for the default 4 KB csum_block_size.
+
+Option 3 is recommended for SSD+HDD setups (because metadata SSDs will handle
+extra reads without any performance drop) and also *maybe* for NVMe all-flash
+setups when you don't have enough RAM (because NVMe drives have plenty
+of read iops to spare). You may also consider enabling
+[cached_read_meta](osd.en.md#cached_read_meta) in this case.

docs/config/layout-osd.ru.md

@@ -25,6 +25,8 @@
 - [disable_journal_fsync](#disable_journal_fsync)
 - [disable_device_lock](#disable_device_lock)
 - [disk_alignment](#disk_alignment)
+- [data_csum_type](#data_csum_type)
+- [csum_block_size](#csum_block_size)
 
 ## data_device
 
@@ -183,3 +185,52 @@ journal_block_size и meta_block_size. Однако единственные SSD
 
 Клиентам не обязательно знать про disk_alignment, так что помещать значение
 этого параметра в etcd в /vitastor/config/global не нужно.
+
+## data_csum_type
+
+- Тип: строка
+- Значение по умолчанию: none
+
+Тип используемых OSD контрольных сумм данных. Может быть "crc32c" или "none".
+Установите в "crc32c", чтобы включить расчёт и проверку контрольных сумм данных.
+
+Следует понимать, что контрольные суммы в зависимости от размера блока их
+расчёта либо увеличивают потребление памяти, либо снижают производительность.
+Подробнее смотрите в описании параметра [csum_block_size](#csum_block_size).
+
+## csum_block_size
+
+- Тип: целое число
+- Значение по умолчанию: 4096
+
+Размер блока расчёта контрольных сумм.
+
+Должен быть равен или кратен [bitmap_granularity](layout-cluster.ru.md#bitmap_granularity)
+(который обычно равен 4 КБ).
+
+Контрольные суммы увеличивают размер метаданных на 4 байта на каждые
+csum_block_size данных.
+
+Контрольные суммы - это всегда компромисс:
+1. Вы либо жертвуете потреблением +1 ГБ памяти на 1 ТБ дискового пространства
+2. Либо вы повышаете csum_block_size до, скажем, 32k и жертвуете 50%
+   скорости случайной записи из-за цикла чтения-изменения-записи для расчёта
+   новых контрольных сумм
+3. Либо вы отключаете [inmemory_metadata](osd.ru.md#inmemory_metadata) и
+   жертвуете 50% скорости случайного чтения из-за чтения контрольных сумм
+   с диска
+
+Вариант 1 (при настройках по умолчанию) рекомендуется для SSD (All-Flash)
+кластеров, потому что памяти в них обычно хватает.
+
+Вариант 2 рекомендуется для кластеров на одних жёстких дисках (без SSD
+под метаданные). На 4 кб блок контрольной суммы памяти в таких кластерах
+обычно НЕ хватает.
+
+Вариант 3 рекомендуется для гибридных кластеров (SSD+HDD), потому что
+скорости SSD под метаданными хватит, чтобы обработать дополнительные чтения
+без снижения производительности. Также вариант 3 *может* рекомендоваться
+для All-Flash кластеров на основе NVMe-дисков, когда памяти НЕ достаточно,
+потому что NVMe-диски имеют огромный запас производительности по чтению.
+В таких случаях, возможно, также имеет смысл включать параметр
+[cached_read_meta](osd.ru.md#cached_read_meta).

docs/config/osd.en.md

@@ -31,6 +31,9 @@ them, even without restarting by updating configuration in etcd.
 - [max_flusher_count](#max_flusher_count)
 - [inmemory_metadata](#inmemory_metadata)
 - [inmemory_journal](#inmemory_journal)
+- [cached_read_data](#cached_read_data)
+- [cached_read_meta](#cached_read_meta)
+- [cached_read_journal](#cached_read_journal)
 - [journal_sector_buffer_count](#journal_sector_buffer_count)
 - [journal_no_same_sector_overwrites](#journal_no_same_sector_overwrites)
 - [throttle_small_writes](#throttle_small_writes)
@@ -255,6 +258,46 @@ is typically very small because it's sufficient to have 16-32 MB journal
 for SSD OSDs. However, in theory it's possible that you'll want to turn it
 off for hybrid (HDD+SSD) OSDs with large journals on quick devices.
 
+## cached_read_data
+
+- Type: boolean
+- Default: false
+
+Read data through Linux page cache, i.e. use a file descriptor opened without
+O_DIRECT for data reads. May improve read performance for frequently accessed
+data if it fits in RAM. Memory in page cache is shared by all processes and
+not accounted in OSD memory consumption.
+
+## cached_read_meta
+
+- Type: boolean
+- Default: false
+
+Read metadata through Linux page cache. May be beneficial when checksums
+are enabled and [inmemory_metadata](#inmemory_metadata) is disabled, because
+in this case metadata blocks are read from disk to verify checksums on every
+read request and caching them may reduce this extra read load.
+
+Absolutely pointless to enable with enabled inmemory_metadata because all
+metadata is kept in memory anyway, and likely pointless without checksums,
+because in that case, metadata blocks are read from disk only during journal
+flushing.
+
+If the same device is used for data and metadata, enabling [cached_read_data](#cached_read_data)
+also enables this parameter, given that it isn't turned off explicitly.
+
+## cached_read_journal
+
+- Type: boolean
+- Default: false
+
+Read buffered data from journal through Linux page cache. Does not have sense
+without disabling [inmemory_journal](#inmemory_journal), which, again, is
+enabled by default.
+
+If the same device is used for metadata and journal, enabling [cached_read_meta](#cached_read_meta)
+also enables this parameter, given that it isn't turned off explicitly.
+
 ## journal_sector_buffer_count
 
 - Type: integer

docs/config/osd.ru.md

@@ -32,6 +32,9 @@
 - [max_flusher_count](#max_flusher_count)
 - [inmemory_metadata](#inmemory_metadata)
 - [inmemory_journal](#inmemory_journal)
+- [cached_read_data](#cached_read_data)
+- [cached_read_meta](#cached_read_meta)
+- [cached_read_journal](#cached_read_journal)
 - [journal_sector_buffer_count](#journal_sector_buffer_count)
 - [journal_no_same_sector_overwrites](#journal_no_same_sector_overwrites)
 - [throttle_small_writes](#throttle_small_writes)
@@ -263,6 +266,51 @@ Flusher - это микро-поток (корутина), которая коп
 параметра может оказаться полезным для гибридных OSD (HDD+SSD) с большими
 журналами, расположенными на быстром по сравнению с HDD устройстве.
 
+## cached_read_data
+
+- Тип: булево (да/нет)
+- Значение по умолчанию: false
+
+Читать данные через системный кэш Linux (page cache), то есть, использовать
+для чтения данных файловый дескриптор, открытый без флага O_DIRECT. Может
+улучшить производительность чтения для часто используемых данных, если они
+помещаются в память. Память кэша разделяется между всеми процессами в
+системе и не учитывается в потреблении памяти процессом OSD.
+
+## cached_read_meta
+
+- Тип: булево (да/нет)
+- Значение по умолчанию: false
+
+Читать метаданные через системный кэш Linux. Может быть полезно, когда
+включены контрольные суммы, а параметр [inmemory_metadata](#inmemory_metadata)
+отключён, так как в этом случае блоки метаданных читаются с диска при каждом
+запросе чтения для проверки контрольных сумм и их кэширование может снизить
+дополнительную нагрузку на диск.
+
+Абсолютно бессмысленно включать данный параметр, если параметр
+inmemory_metadata включён (по умолчанию это так), и также вероятно
+бессмысленно включать его, если не включены контрольные суммы, так как в
+этом случае блоки метаданных читаются с диска только во время сброса
+журнала.
+
+Если одно и то же устройство используется для данных и метаданных, включение
+[cached_read_data](#cached_read_data) также включает данный параметр, при
+условии, что он не отключён явным образом.
+
+## cached_read_journal
+
+- Тип: булево (да/нет)
+- Значение по умолчанию: false
+
+Читать буферизованные в журнале данные через системный кэш Linux. Не имеет
+смысла без отключения параметра [inmemory_journal](#inmemory_journal),
+который, опять же, по умолчанию включён.
+
+Если одно и то же устройство используется для метаданных и журнала,
+включение [cached_read_meta](#cached_read_meta) также включает данный
+параметр, при условии, что он не отключён явным образом.
+
 ## journal_sector_buffer_count
 
 - Тип: целое число

docs/config/src/include.js

@@ -0,0 +1,145 @@
+#!/usr/bin/nodejs
+
+const fsp = require('fs').promises;
+
+run(process.argv).catch(console.error);
+
+async function run(argv)
+{
+    if (argv.length < 3)
+    {
+        console.log('Markdown preprocessor\nUSAGE: ./include.js file.md');
+        return;
+    }
+    const index_file = await fsp.realpath(argv[2]);
+    const re = /(\{\{[\s\S]*?\}\}|\[[^\]]+\]\([^\)]+\)|(?:^|\n)#[^\n]+)/;
+    let text = await fsp.readFile(index_file, { encoding: 'utf-8' });
+    text = text.split(re);
+    let included = {};
+    let heading = 0, heading_name = '', m;
+    for (let i = 0; i < text.length; i++)
+    {
+        if (text[i].substr(0, 2) == '{{')
+        {
+            // Inclusion
+            let incfile = text[i].substr(2, text[i].length-4);
+            let section = null;
+            let indent = heading;
+            incfile = incfile.replace(/\s*\|\s*indent\s*=\s*(-?\d+)\s*$/, (m, m1) => { indent = parseInt(m1); return ''; });
+            incfile = incfile.replace(/\s*#\s*([^#]+)$/, (m, m1) => { section = m1; return ''; });
+            let inc_heading = section;
+            incfile = rel2abs(index_file, incfile);
+            let inc = await fsp.readFile(incfile, { encoding: 'utf-8' });
+            inc = inc.trim().replace(/^[\s\S]+?\n#/, '#'); // remove until the first header
+            inc = inc.split(re);
+            const indent_str = new Array(indent+1).join('#');
+            let section_start = -1, section_end = -1;
+            for (let j = 0; j < inc.length; j++)
+            {
+                if ((m = /^(\n?)(#+\s*)([\s\S]+)$/.exec(inc[j])))
+                {
+                    if (!inc_heading)
+                    {
+                        inc_heading = m[3].trim();
+                    }
+                    if (section)
+                    {
+                        if (m[3].trim() == section)
+                            section_start = j;
+                        else if (section_start >= 0)
+                        {
+                            section_end = j;
+                            break;
+                        }
+                    }
+                    inc[j] = m[1] + indent_str + m[2] + m[3];
+                }
+                else if ((m = /^(\[[^\]]+\]\()([^\)]+)(\))$/.exec(inc[j])) && !/^https?:(\/\/)|^#/.exec(m[2]))
+                {
+                    const abs_m2 = rel2abs(incfile, m[2]);
+                    const rel_m = abs2rel(__filename, abs_m2);
+                    if (rel_m.substr(0, 9) == '../../../') // outside docs
+                        inc[j] = m[1] + 'https://git.yourcmc.ru/vitalif/vitastor/src/branch/master/'+rel2abs('docs/config/src/include.js', rel_m) + m[3];
+                    else
+                        inc[j] = m[1] + abs_m2 + m[3];
+                }
+            }
+            if (section)
+            {
+                inc = section_start >= 0 ? inc.slice(section_start, section_end < 0 ? inc.length : section_end) : [];
+            }
+            if (inc.length)
+            {
+                if (!inc_heading)
+                    inc_heading = heading_name||'';
+                included[incfile+(section ? '#'+section : '')] = '#'+inc_heading.toLowerCase().replace(/\P{L}+/ug, '-').replace(/^-|-$/g, '');
+                inc[0] = inc[0].replace(/^\s+/, '');
+                inc[inc.length-1] = inc[inc.length-1].replace(/\s+$/, '');
+            }
+            text.splice(i, 1, ...inc);
+            i = i + inc.length - 1;
+        }
+        else if ((m = /^\n?(#+)\s*([\s\S]+)$/.exec(text[i])))
+        {
+            // Heading
+            heading = m[1].length;
+            heading_name = m[2].trim();
+        }
+    }
+    for (let i = 0; i < text.length; i++)
+    {
+        if ((m = /^(\[[^\]]+\]\()([^\)]+)(\))$/.exec(text[i])) && !/^https?:(\/\/)|^#/.exec(m[2]))
+        {
+            const p = m[2].indexOf('#');
+            if (included[m[2]])
+            {
+                text[i] = m[1]+included[m[2]]+m[3];
+            }
+            else if (p >= 0 && included[m[2].substr(0, p)])
+            {
+                text[i] = m[1]+m[2].substr(p)+m[3];
+            }
+        }
+    }
+    console.log(text.join(''));
+}
+
+function rel2abs(ref, rel)
+{
+    rel = [ ...ref.replace(/^(.*)\/[^\/]+$/, '$1').split(/\/+/), ...rel.split(/\/+/) ];
+    return killdots(rel).join('/');
+}
+
+function abs2rel(ref, abs)
+{
+    ref = ref.split(/\/+/);
+    abs = abs.split(/\/+/);
+    while (ref.length > 1 && ref[0] == abs[0])
+    {
+        ref.shift();
+        abs.shift();
+    }
+    for (let i = 1; i < ref.length; i++)
+    {
+        abs.unshift('..');
+    }
+    return killdots(abs).join('/');
+}
+
+function killdots(rel)
+{
+    for (let i = 0; i < rel.length; i++)
+    {
+        if (rel[i] == '.')
+        {
+            rel.splice(i, 1);
+            i--;
+        }
+        else if (i >= 1 && rel[i] == '..' && rel[i-1] != '..')
+        {
+            rel.splice(i-1, 2);
+            i -= 2;
+        }
+    }
+    return rel;
+}

docs/config/src/included.en.md

@@ -0,0 +1,65 @@
+# Vitastor
+
+{{../../../README.md#The Idea}}
+
+{{../../../README.md#Talks and presentations}}
+
+{{../../intro/features.en.md}}
+
+{{../../intro/quickstart.en.md}}
+
+{{../../intro/architecture.en.md}}
+
+## Installation
+
+{{../../installation/packages.en.md}}
+
+{{../../installation/proxmox.en.md}}
+
+{{../../installation/openstack.en.md}}
+
+{{../../installation/kubernetes.en.md}}
+
+{{../../installation/source.en.md}}
+
+{{../../config.en.md|indent=1}}
+
+{{../../config/common.en.md|indent=2}}
+
+{{../../config/network.en.md|indent=2}}
+
+{{../../config/layout-cluster.en.md|indent=2}}
+
+{{../../config/layout-osd.en.md|indent=2}}
+
+{{../../config/osd.en.md|indent=2}}
+
+{{../../config/monitor.en.md|indent=2}}
+
+{{../../config/pool.en.md|indent=2}}
+
+{{../../config/inode.en.md|indent=2}}
+
+## Usage
+
+{{../../usage/cli.en.md}}
+
+{{../../usage/disk.en.md}}
+
+{{../../usage/fio.en.md}}
+
+{{../../usage/nbd.en.md}}
+
+{{../../usage/qemu.en.md}}
+
+{{../../usage/nfs.en.md}}
+
+## Performance
+
+{{../../performance/understanding.en.md}}
+
+{{../../performance/theoretical.en.md}}
+
+{{../../performance/comparison1.en.md}}
+
+{{../../intro/author.en.md|indent=1}}

docs/config/src/included.ru.md

@@ -0,0 +1,65 @@
+# Vitastor
+
+{{../../../README-ru.md#Идея|indent=0}}
+
+{{../../../README-ru.md#Презентации и записи докладов|indent=0}}
+
+{{../../intro/features.ru.md}}
+
+{{../../intro/quickstart.ru.md}}
+
+{{../../intro/architecture.ru.md}}
+
+## Установка
+
+{{../../installation/packages.ru.md}}
+
+{{../../installation/proxmox.ru.md}}
+
+{{../../installation/openstack.ru.md}}
+
+{{../../installation/kubernetes.ru.md}}
+
+{{../../installation/source.ru.md}}
+
+{{../../config.ru.md|indent=1}}
+
+{{../../config/common.ru.md|indent=2}}
+
+{{../../config/network.ru.md|indent=2}}
+
+{{../../config/layout-cluster.ru.md|indent=2}}
+
+{{../../config/layout-osd.ru.md|indent=2}}
+
+{{../../config/osd.ru.md|indent=2}}
+
+{{../../config/monitor.ru.md|indent=2}}
+
+{{../../config/pool.ru.md|indent=2}}
+
+{{../../config/inode.ru.md|indent=2}}
+
+## Использование
+
+{{../../usage/cli.ru.md}}
+
+{{../../usage/disk.ru.md}}
+
+{{../../usage/fio.ru.md}}
+
+{{../../usage/nbd.ru.md}}
+
+{{../../usage/qemu.ru.md}}
+
+{{../../usage/nfs.ru.md}}
+
+## Производительность
+
+{{../../performance/understanding.ru.md}}
+
+{{../../performance/theoretical.ru.md}}
+
+{{../../performance/comparison1.ru.md}}
+
+{{../../intro/author.ru.md|indent=1}}

docs/config/src/layout-cluster.yml

@@ -7,26 +7,27 @@
     in Vitastor, affects memory usage, write amplification and I/O load
     distribution effectiveness.
 
-    Recommended default block size is 128 KB for SSD and 4 MB for HDD. In fact,
-    it's possible to use 4 MB for SSD too - it will lower memory usage, but
+    Recommended default block size is 128 KB for SSD and 1 MB for HDD. In fact,
+    it's possible to use 1 MB for SSD too - it will lower memory usage, but
     may increase average WA and reduce linear performance.
 
     OSD memory usage is roughly (SIZE / BLOCK * 68 bytes) which is roughly
     544 MB per 1 TB of used disk space with the default 128 KB block size.
+    With 1 MB it's 8 times lower.
   info_ru: |
     Размер объектов (блоков данных), на которые делятся физические и виртуальные
     диски в Vitastor (в рамках каждого пула). Одна из ключевых на данный момент
     настроек, влияет на потребление памяти, объём избыточной записи (write
     amplification) и эффективность распределения нагрузки по OSD.
 
-    Рекомендуемые по умолчанию размеры блока - 128 килобайт для SSD и 4
-    мегабайта для HDD. В принципе, для SSD можно тоже использовать 4 мегабайта,
+    Рекомендуемые по умолчанию размеры блока - 128 килобайт для SSD и 1 мегабайт
+    для HDD. В принципе, для SSD можно тоже использовать блок размером 1 мегабайт,
     это понизит использование памяти, но ухудшит распределение нагрузки и в
     среднем увеличит WA.
 
     Потребление памяти OSD составляет примерно (РАЗМЕР / БЛОК * 68 байт),
     т.е. примерно 544 МБ памяти на 1 ТБ занятого места на диске при
-    стандартном 128 КБ блоке.
+    стандартном 128 КБ блоке. При 1 МБ блоке памяти нужно в 8 раз меньше.
 - name: bitmap_granularity
   type: int
   default: 4096

docs/config/src/layout-osd.yml

@@ -204,3 +204,77 @@
 
     Клиентам не обязательно знать про disk_alignment, так что помещать значение
     этого параметра в etcd в /vitastor/config/global не нужно.
+- name: data_csum_type
+  type: string
+  default: none
+  info: |
+    Data checksum type to use. May be "crc32c" or "none". Set to "crc32c" to
+    enable data checksums.
+  info_ru: |
+    Тип используемых OSD контрольных сумм данных. Может быть "crc32c" или "none".
+    Установите в "crc32c", чтобы включить расчёт и проверку контрольных сумм данных.
+
+    Следует понимать, что контрольные суммы в зависимости от размера блока их
+    расчёта либо увеличивают потребление памяти, либо снижают производительность.
+    Подробнее смотрите в описании параметра [csum_block_size](#csum_block_size).
+- name: csum_block_size
+  type: int
+  default: 4096
+  info: |
+    Checksum calculation block size.
+
+    Must be equal or a multiple of [bitmap_granularity](layout-cluster.en.md#bitmap_granularity)
+    (which is usually 4 KB).
+
+    Checksums increase metadata size by 4 bytes per each csum_block_size of data.
+
+    Checksums are always a compromise:
+    1. You either sacrifice +1 GB RAM per 1 TB of data
+    2. Or you raise csum_block_size, for example, to 32k and sacrifice
+       50% random write iops due to checksum read-modify-write
+    3. Or you turn off [inmemory_metadata](osd.en.md#inmemory_metadata) and
+       sacrifice 50% random read iops due to checksum reads
+
+    Option 1 (default) is recommended for all-flash setups because these usually
+    have enough RAM.
+
+    Option 2 is recommended for HDD-only setups. HDD-only setups usually do NOT
+    have enough RAM for the default 4 KB csum_block_size.
+
+    Option 3 is recommended for SSD+HDD setups (because metadata SSDs will handle
+    extra reads without any performance drop) and also *maybe* for NVMe all-flash
+    setups when you don't have enough RAM (because NVMe drives have plenty
+    of read iops to spare). You may also consider enabling
+    [cached_read_meta](osd.en.md#cached_read_meta) in this case.
+  info_ru: |
+    Размер блока расчёта контрольных сумм.
+
+    Должен быть равен или кратен [bitmap_granularity](layout-cluster.ru.md#bitmap_granularity)
+    (который обычно равен 4 КБ).
+
+    Контрольные суммы увеличивают размер метаданных на 4 байта на каждые
+    csum_block_size данных.
+
+    Контрольные суммы - это всегда компромисс:
+    1. Вы либо жертвуете потреблением +1 ГБ памяти на 1 ТБ дискового пространства
+    2. Либо вы повышаете csum_block_size до, скажем, 32k и жертвуете 50%
+       скорости случайной записи из-за цикла чтения-изменения-записи для расчёта
+       новых контрольных сумм
+    3. Либо вы отключаете [inmemory_metadata](osd.ru.md#inmemory_metadata) и
+       жертвуете 50% скорости случайного чтения из-за чтения контрольных сумм
+       с диска
+
+    Вариант 1 (при настройках по умолчанию) рекомендуется для SSD (All-Flash)
+    кластеров, потому что памяти в них обычно хватает.
+
+    Вариант 2 рекомендуется для кластеров на одних жёстких дисках (без SSD
+    под метаданные). На 4 кб блок контрольной суммы памяти в таких кластерах
+    обычно НЕ хватает.
+
+    Вариант 3 рекомендуется для гибридных кластеров (SSD+HDD), потому что
+    скорости SSD под метаданными хватит, чтобы обработать дополнительные чтения
+    без снижения производительности. Также вариант 3 *может* рекомендоваться
+    для All-Flash кластеров на основе NVMe-дисков, когда памяти НЕ достаточно,
+    потому что NVMe-диски имеют огромный запас производительности по чтению.
+    В таких случаях, возможно, также имеет смысл включать параметр
+    [cached_read_meta](osd.ru.md#cached_read_meta).

docs/config/src/osd.yml

@@ -260,6 +260,70 @@
     достаточно 16- или 32-мегабайтного журнала. Однако в теории отключение
     параметра может оказаться полезным для гибридных OSD (HDD+SSD) с большими
     журналами, расположенными на быстром по сравнению с HDD устройстве.
+- name: cached_read_data
+  type: bool
+  default: false
+  info: |
+    Read data through Linux page cache, i.e. use a file descriptor opened without
+    O_DIRECT for data reads. May improve read performance for frequently accessed
+    data if it fits in RAM. Memory in page cache is shared by all processes and
+    not accounted in OSD memory consumption.
+  info_ru: |
+    Читать данные через системный кэш Linux (page cache), то есть, использовать
+    для чтения данных файловый дескриптор, открытый без флага O_DIRECT. Может
+    улучшить производительность чтения для часто используемых данных, если они
+    помещаются в память. Память кэша разделяется между всеми процессами в
+    системе и не учитывается в потреблении памяти процессом OSD.
+- name: cached_read_meta
+  type: bool
+  default: false
+  info: |
+    Read metadata through Linux page cache. May be beneficial when checksums
+    are enabled and [inmemory_metadata](#inmemory_metadata) is disabled, because
+    in this case metadata blocks are read from disk to verify checksums on every
+    read request and caching them may reduce this extra read load.
+
+    Absolutely pointless to enable with enabled inmemory_metadata because all
+    metadata is kept in memory anyway, and likely pointless without checksums,
+    because in that case, metadata blocks are read from disk only during journal
+    flushing.
+
+    If the same device is used for data and metadata, enabling [cached_read_data](#cached_read_data)
+    also enables this parameter, given that it isn't turned off explicitly.
+  info_ru: |
+    Читать метаданные через системный кэш Linux. Может быть полезно, когда
+    включены контрольные суммы, а параметр [inmemory_metadata](#inmemory_metadata)
+    отключён, так как в этом случае блоки метаданных читаются с диска при каждом
+    запросе чтения для проверки контрольных сумм и их кэширование может снизить
+    дополнительную нагрузку на диск.
+
+    Абсолютно бессмысленно включать данный параметр, если параметр
+    inmemory_metadata включён (по умолчанию это так), и также вероятно
+    бессмысленно включать его, если не включены контрольные суммы, так как в
+    этом случае блоки метаданных читаются с диска только во время сброса
+    журнала.
+
+    Если одно и то же устройство используется для данных и метаданных, включение
+    [cached_read_data](#cached_read_data) также включает данный параметр, при
+    условии, что он не отключён явным образом.
+- name: cached_read_journal
+  type: bool
+  default: false
+  info: |
+    Read buffered data from journal through Linux page cache. Does not have sense
+    without disabling [inmemory_journal](#inmemory_journal), which, again, is
+    enabled by default.
+
+    If the same device is used for metadata and journal, enabling [cached_read_meta](#cached_read_meta)
+    also enables this parameter, given that it isn't turned off explicitly.
+  info_ru: |
+    Читать буферизованные в журнале данные через системный кэш Linux. Не имеет
+    смысла без отключения параметра [inmemory_journal](#inmemory_journal),
+    который, опять же, по умолчанию включён.
+
+    Если одно и то же устройство используется для метаданных и журнала,
+    включение [cached_read_meta](#cached_read_meta) также включает данный
+    параметр, при условии, что он не отключён явным образом.
 - name: journal_sector_buffer_count
   type: int
   default: 32

docs/installation/kubernetes.ru.md

@@ -8,13 +8,13 @@
 
 У Vitastor есть CSI-плагин для Kubernetes, поддерживающий RWO, а также блочные RWX, тома.
 
-Для установки возьмите манифесты из директории [csi/deploy/](../csi/deploy/), поместите
-вашу конфигурацию подключения к Vitastor в [csi/deploy/001-csi-config-map.yaml](../csi/deploy/001-csi-config-map.yaml),
-настройте StorageClass в [csi/deploy/009-storage-class.yaml](../csi/deploy/009-storage-class.yaml)
+Для установки возьмите манифесты из директории [csi/deploy/](../../csi/deploy/), поместите
+вашу конфигурацию подключения к Vitastor в [csi/deploy/001-csi-config-map.yaml](../../csi/deploy/001-csi-config-map.yaml),
+настройте StorageClass в [csi/deploy/009-storage-class.yaml](../../csi/deploy/009-storage-class.yaml)
 и примените все `NNN-*.yaml` к вашей инсталляции Kubernetes.
 
 ```
 for i in ./???-*.yaml; do kubectl apply -f $i; done
 ```
 
-После этого вы сможете создавать PersistentVolume. Пример смотрите в файле [csi/deploy/example-pvc.yaml](../csi/deploy/example-pvc.yaml).
+После этого вы сможете создавать PersistentVolume. Пример смотрите в файле [csi/deploy/example-pvc.yaml](../../csi/deploy/example-pvc.yaml).

docs/installation/openstack.en.md

@@ -36,5 +36,5 @@ vitastor_pool_id = 1
 image_upload_use_cinder_backend = True
 ```
 
-To put Glance images in Vitastor, use [https://docs.openstack.org/cinder/pike/admin/blockstorage-volume-backed-image.html](volume-backed images),
+To put Glance images in Vitastor, use [volume-backed images](https://docs.openstack.org/cinder/pike/admin/blockstorage-volume-backed-image.html),
 although the support has not been verified yet.

docs/installation/openstack.ru.md

@@ -36,5 +36,5 @@ image_upload_use_cinder_backend = True
 ```
 
 Чтобы помещать в Vitastor Glance-образы, нужно использовать
-[https://docs.openstack.org/cinder/pike/admin/blockstorage-volume-backed-image.html](образы на основе томов Cinder),
+[образы на основе томов Cinder](https://docs.openstack.org/cinder/pike/admin/blockstorage-volume-backed-image.html),
 однако, поддержка этой функции ещё не проверялась.

docs/installation/packages.en.md

@@ -11,7 +11,8 @@
 - Trust Vitastor package signing key:
   `wget https://vitastor.io/debian/pubkey.gpg -O /etc/apt/trusted.gpg.d/vitastor.gpg`
 - Add Vitastor package repository to your /etc/apt/sources.list:
-  - Debian 11 (Bullseye/Sid): `deb https://vitastor.io/debian bullseye main`
+  - Debian 12 (Bookworm/Sid): `deb https://vitastor.io/debian bookworm main`
+  - Debian 11 (Bullseye): `deb https://vitastor.io/debian bullseye main`
   - Debian 10 (Buster): `deb https://vitastor.io/debian buster main`
 - For Debian 10 (Buster) also enable backports repository:
   `deb http://deb.debian.org/debian buster-backports main`

docs/installation/packages.ru.md

@@ -11,7 +11,8 @@
 - Добавьте ключ репозитория Vitastor:
   `wget https://vitastor.io/debian/pubkey.gpg -O /etc/apt/trusted.gpg.d/vitastor.gpg`
 - Добавьте репозиторий Vitastor в /etc/apt/sources.list:
-  - Debian 11 (Bullseye/Sid): `deb https://vitastor.io/debian bullseye main`
+  - Debian 12 (Bookworm/Sid): `deb https://vitastor.io/debian bookworm main`
+  - Debian 11 (Bullseye): `deb https://vitastor.io/debian bullseye main`
   - Debian 10 (Buster): `deb https://vitastor.io/debian buster main`
 - Для Debian 10 (Buster) также включите репозиторий backports:
   `deb http://deb.debian.org/debian buster-backports main`

docs/installation/proxmox.en.md

@@ -6,10 +6,10 @@
 
 # Proxmox VE
 
-To enable Vitastor support in Proxmox Virtual Environment (6.4-7.4 are supported):
+To enable Vitastor support in Proxmox Virtual Environment (6.4-8.0 are supported):
 
 - Add the corresponding Vitastor Debian repository into sources.list on Proxmox hosts:
-  buster for 6.4, bullseye for 7.4, pve7.1 for 7.1, pve7.2 for 7.2, pve7.3 for 7.3
+  bookworm for 8.0, bullseye for 7.4, pve7.3 for 7.3, pve7.2 for 7.2, pve7.1 for 7.1, buster for 6.4
 - Install vitastor-client, pve-qemu-kvm, pve-storage-vitastor (* or see note) packages from Vitastor repository
 - Define storage in `/etc/pve/storage.cfg` (see below)
 - Block network access from VMs to Vitastor network (to OSDs and etcd),
@@ -35,5 +35,5 @@ vitastor: vitastor
     vitastor_nbd 0
 ```
 
-\* Note: you can also manually copy [patches/VitastorPlugin.pm](patches/VitastorPlugin.pm) to Proxmox hosts
+\* Note: you can also manually copy [patches/VitastorPlugin.pm](../../patches/VitastorPlugin.pm) to Proxmox hosts
 as `/usr/share/perl5/PVE/Storage/Custom/VitastorPlugin.pm` instead of installing pve-storage-vitastor.

docs/installation/proxmox.ru.md

@@ -1,15 +1,15 @@
-[Документация](../../README-ru.md#документация) → Установка → Proxmox
+[Документация](../../README-ru.md#документация) → Установка → Proxmox VE
 
 -----
 
 [Read in English](proxmox.en.md)
 
-# Proxmox
+# Proxmox VE
 
-Чтобы подключить Vitastor к Proxmox Virtual Environment (поддерживаются версии 6.4-7.4):
+Чтобы подключить Vitastor к Proxmox Virtual Environment (поддерживаются версии 6.4-8.0):
 
 - Добавьте соответствующий Debian-репозиторий Vitastor в sources.list на хостах Proxmox:
-  buster для 6.4, bullseye для 7.4, pve7.1 для 7.1, pve7.2 для 7.2, pve7.3 для 7.3
+  bookworm для 8.0, bullseye для 7.4, pve7.3 для 7.3, pve7.2 для 7.2, pve7.1 для 7.1, buster для 6.4
 - Установите пакеты vitastor-client, pve-qemu-kvm, pve-storage-vitastor (* или см. сноску) из репозитория Vitastor
 - Определите тип хранилища в `/etc/pve/storage.cfg` (см. ниже)
 - Обязательно заблокируйте доступ от виртуальных машин к сети Vitastor (OSD и etcd), т.к. Vitastor (пока) не поддерживает аутентификацию
@@ -35,5 +35,5 @@ vitastor: vitastor
 ```
 
 \* Примечание: вместо установки пакета pve-storage-vitastor вы можете вручную скопировать файл
-[patches/VitastorPlugin.pm](patches/VitastorPlugin.pm) на хосты Proxmox как
+[patches/VitastorPlugin.pm](../../patches/VitastorPlugin.pm) на хосты Proxmox как
 `/usr/share/perl5/PVE/Storage/Custom/VitastorPlugin.pm`.

docs/intro/architecture.en.md

@@ -44,7 +44,7 @@
   depends linearly on drive capacity and data store block size which is 128 KB by default.
   With 128 KB blocks metadata takes around 512 MB per 1 TB (which is still less than Ceph wants).
   Journal is also kept in memory by default, but in SSD-only clusters it's only 32 MB, and in SSD+HDD
-  clusters, where it's beneficial to increase it, [inmemory_journal](docs/config/osd.en.md#inmemory_journal) can be disabled.
+  clusters, where it's beneficial to increase it, [inmemory_journal](../config/osd.en.md#inmemory_journal) can be disabled.
 - Vitastor storage layer doesn't have internal copy-on-write or redirect-write. I know that maybe
   it's possible to create a good copy-on-write storage, but it's much harder and makes performance
   less deterministic, so CoW isn't used in Vitastor.

docs/intro/architecture.ru.md

@@ -156,7 +156,7 @@
   блока хранилища (block_size, по умолчанию 128 КБ). С 128 КБ блоком потребление памяти
   составляет примерно 512 МБ на 1 ТБ данных. Журналы по умолчанию тоже хранятся в памяти,
   но в SSD-кластерах нужный размер журнала составляет всего 32 МБ, а в гибридных (SSD+HDD)
-  кластерах, в которых есть смысл делать журналы больше, можно отключить [inmemory_journal](../docs/config/osd.ru.md#inmemory_journal).
+  кластерах, в которых есть смысл делать журналы больше, можно отключить [inmemory_journal](../config/osd.ru.md#inmemory_journal).
 - В Vitastor нет внутреннего copy-on-write. Я считаю, что реализация CoW-хранилища гораздо сложнее,
   поэтому сложнее добиться устойчиво хороших результатов. Возможно, в один прекрасный день
   я придумаю красивый алгоритм для CoW-хранилища, но пока нет — внутреннего CoW в Vitastor не будет.

docs/intro/features.en.md

@@ -30,12 +30,13 @@
 - [Write throttling to smooth random write workloads in SSD+HDD configurations](../config/osd.en.md#throttle_small_writes)
 - [RDMA/RoCEv2 support via libibverbs](../config/network.en.md#rdma_device)
 - [Scrubbing without checksums](../config/osd.en.md#auto_scrub) (verification of copies)
+- [Checksums](../config/layout-osd.en.md#data_csum_type)
 
 ## Plugins and tools
 
 - [Debian and CentOS packages](../installation/packages.en.md)
 - [Image management CLI (vitastor-cli)](../usage/cli.en.md)
-- [Disk management CLI (vitastor-disk)](docs/usage/disk.en.md)
+- [Disk management CLI (vitastor-disk)](../usage/disk.en.md)
 - Generic user-space client library
 - [Native QEMU driver](../usage/qemu.en.md)
 - [Loadable fio engine for benchmarks](../usage/fio.en.md)
@@ -55,7 +56,6 @@ The following features are planned for the future:
 - iSCSI proxy
 - Multi-threaded client
 - Faster failover
-- Checksums
 - Tiered storage (SSD caching)
 - NVDIMM support
 - Compression (possibly)

docs/intro/features.ru.md

@@ -13,7 +13,7 @@
 ## Серверные функции
 
 - Базовая часть - надёжное кластерное блочное хранилище без единой точки отказа
-- [Производительность](../comparison1.ru.md) ;-D
+- [Производительность](../performance/comparison1.ru.md) ;-D
 - [Несколько схем отказоустойчивости](../config/pool.ru.md#scheme): репликация, XOR n+1 (1 диск чётности), коды коррекции ошибок
   Рида-Соломона на основе библиотек jerasure и ISA-L с любым числом дисков данных и чётности в группе
 - Конфигурация через простые человекочитаемые JSON-структуры в etcd
@@ -32,12 +32,13 @@
 - [Сглаживание производительности случайной записи в SSD+HDD конфигурациях](../config/osd.ru.md#throttle_small_writes)
 - [Поддержка RDMA/RoCEv2 через libibverbs](../config/network.ru.md#rdma_device)
 - [Фоновая проверка целостности без контрольных сумм](../config/osd.ru.md#auto_scrub) (сверка копий)
+- [Контрольные суммы](../config/layout-osd.ru.md#data_csum_type)
 
 ## Драйверы и инструменты
 
 - [Пакеты для Debian и CentOS](../installation/packages.ru.md)
 - [Консольный интерфейс управления образами (vitastor-cli)](../usage/cli.ru.md)
-- [Инструмент управления дисками (vitastor-disk)](docs/usage/disk.ru.md)
+- [Инструмент управления дисками (vitastor-disk)](../usage/disk.ru.md)
 - Общая пользовательская клиентская библиотека для работы с кластером
 - [Драйвер диска для QEMU](../usage/qemu.ru.md)
 - [Драйвер диска для утилиты тестирования производительности fio](../usage/fio.ru.md)
@@ -55,7 +56,6 @@
 - iSCSI-прокси
 - Многопоточный клиент
 - Более быстрое переключение при отказах
-- Контрольные суммы
 - Поддержка SSD-кэширования (tiered storage)
 - Поддержка NVDIMM
 - Возможно, сжатие

docs/intro/quickstart.en.md

@@ -7,6 +7,7 @@
 # Quick Start
 
 - [Preparation](#preparation)
+- [Recommended drives](#recommended-drives)
 - [Configure monitors](#configure-monitors)
 - [Configure OSDs](#configure-osds)
 - [Create a pool](#create-a-pool)
@@ -19,10 +20,20 @@
 - Get some SATA or NVMe SSDs with capacitors (server-grade drives). You can use desktop SSDs
   with lazy fsync, but prepare for inferior single-thread latency. Read more about capacitors
   [here](../config/layout-cluster.en.md#immediate_commit).
+- If you want to use HDDs, get modern HDDs with Media Cache or SSD Cache: HGST Ultrastar,
+  Toshiba MG08, Seagate EXOS or something similar. If your drives don't have such cache then
+  you also need small SSDs for journal and metadata (even 2 GB per 1 TB of HDD space is enough).
 - Get a fast network (at least 10 Gbit/s). Something like Mellanox ConnectX-4 with RoCEv2 is ideal.
 - Disable CPU powersaving: `cpupower idle-set -D 0 && cpupower frequency-set -g performance`.
 - [Install Vitastor packages](../installation/packages.en.md).
 
+## Recommended drives
+
+- SATA SSD: Micron 5100/5200/5300/5400, Samsung PM863/PM883/PM893, Intel D3-S4510/4520/4610/4620, Kingston DC500M
+- NVMe: Micron 9100/9200/9300/9400, Micron 7300/7450, Samsung PM983/PM9A3, Samsung PM1723/1735/1743,
+  Intel DC-P3700/P4500/P4600, Intel D7-P5500/P5600, Intel Optane, Kingston DC1000B/DC1500M
+- HDD: HGST Ultrastar, Toshiba MG06/MG07/MG08, Seagate EXOS
+
 ## Configure monitors
 
 On the monitor hosts:
@@ -45,9 +56,10 @@ On the monitor hosts:
   }
   ```
 - Initialize OSDs:
-  - SSD-only: `vitastor-disk prepare /dev/sdXXX [/dev/sdYYY ...]`. You can add
-    `--disable_data_fsync off` to leave disk cache enabled if you use desktop
-    SSDs without capacitors.
+  - SSD-only or HDD-only: `vitastor-disk prepare /dev/sdXXX [/dev/sdYYY ...]`.
+    Add `--disable_data_fsync off` to leave disk write cache enabled if you use
+    desktop SSDs without capacitors. Do NOT add `--disable_data_fsync off` if you
+    use HDDs or SSD+HDD.
   - Hybrid, SSD+HDD: `vitastor-disk prepare --hybrid /dev/sdXXX [/dev/sdYYY ...]`.
     Pass all your devices (HDD and SSD) to this script — it will partition disks and initialize journals on its own.
     This script skips HDDs which are already partitioned so if you want to use non-empty disks for

docs/intro/quickstart.ru.md

@@ -7,6 +7,7 @@
 # Быстрый старт
 
 - [Подготовка](#подготовка)
+- [Рекомендуемые диски](#рекомендуемые-диски)
 - [Настройте мониторы](#настройте-мониторы)
 - [Настройте OSD](#настройте-osd)
 - [Создайте пул](#создайте-пул)
@@ -19,10 +20,20 @@
 - Возьмите серверы с SSD (SATA или NVMe), желательно с конденсаторами (серверные SSD). Можно
   использовать и десктопные SSD, включив режим отложенного fsync, но производительность будет хуже.
   О конденсаторах читайте [здесь](../config/layout-cluster.ru.md#immediate_commit).
+- Если хотите использовать HDD, берите современные модели с Media или SSD кэшем - HGST Ultrastar,
+  Toshiba MG08, Seagate EXOS или что-то похожее. Если такого кэша у ваших дисков нет,
+  обязательно возьмите SSD под метаданные и журнал (маленькие, буквально 2 ГБ на 1 ТБ HDD-места).
 - Возьмите быструю сеть, минимум 10 гбит/с. Идеал - что-то вроде Mellanox ConnectX-4 с RoCEv2.
 - Для лучшей производительности отключите энергосбережение CPU: `cpupower idle-set -D 0 && cpupower frequency-set -g performance`.
 - [Установите пакеты Vitastor](../installation/packages.ru.md).
 
+## Рекомендуемые диски
+
+- SATA SSD: Micron 5100/5200/5300/5400, Samsung PM863/PM883/PM893, Intel D3-S4510/4520/4610/4620, Kingston DC500M
+- NVMe: Micron 9100/9200/9300/9400, Micron 7300/7450, Samsung PM983/PM9A3, Samsung PM1723/1735/1743,
+  Intel DC-P3700/P4500/P4600, Intel D7-P5500/P5600, Intel Optane, Kingston DC1000B/DC1500M
+- HDD: HGST Ultrastar, Toshiba MG06/MG07/MG08, Seagate EXOS
+
 ## Настройте мониторы
 
 На хостах, выделенных под мониторы:
@@ -45,9 +56,10 @@
   }
   ```
 - Инициализуйте OSD:
-  - SSD: `vitastor-disk prepare /dev/sdXXX [/dev/sdYYY ...]`. Если вы используете
-    десктопные SSD без конденсаторов, можете оставить кэш включённым, добавив
-    опцию `--disable_data_fsync off`.
+  - Только SSD или только HDD: `vitastor-disk prepare /dev/sdXXX [/dev/sdYYY ...]`.
+    Если вы используете десктопные SSD без конденсаторов, добавьте опцию `--disable_data_fsync off`,
+    чтобы оставить кэш записи диска включённым. НЕ добавляйте эту опцию, если используете
+    жёсткие диски (HDD).
   - Гибридные, SSD+HDD: `vitastor-disk prepare --hybrid /dev/sdXXX [/dev/sdYYY ...]`.
     Передайте все ваши SSD и HDD скрипту в командной строке подряд, скрипт автоматически выделит
     разделы под журналы на SSD и данные на HDD. Скрипт пропускает HDD, на которых уже есть разделы

docs/usage/disk.en.md

@@ -86,6 +86,8 @@ Options (both modes):
 --journal_size 1G/32M      Set journal size (area or partition size)
 --block_size 1M/128k       Set blockstore object size
 --bitmap_granularity 4k    Set bitmap granularity
+--data_csum_type none      Set data checksum type (crc32c or none)
+--csum_block_size 4k       Set data checksum block size
 --data_device_block 4k     Override data device block size
 --meta_device_block 4k     Override metadata device block size
 --journal_device_block 4k  Override journal device block size
@@ -100,8 +102,9 @@ checks the device cache status on start and tries to disable cache for SATA/SAS
 If it doesn't succeed it issues a warning in the system log.
 
 You can also pass other OSD options here as arguments and they'll be persisted
-to the superblock: max_write_iodepth, max_write_iodepth, min_flusher_count,
-max_flusher_count, inmemory_metadata, inmemory_journal, journal_sector_buffer_count,
+in the superblock: cached_read_data, cached_read_meta, cached_read_journal,
+inmemory_metadata, inmemory_journal, max_write_iodepth,
+min_flusher_count, max_flusher_count, journal_sector_buffer_count,
 journal_no_same_sector_overwrites, throttle_small_writes, throttle_target_iops,
 throttle_target_mbs, throttle_target_parallelism, throttle_threshold_us.
 See [Runtime OSD Parameters](../config/osd.en.md) for details.
@@ -249,7 +252,9 @@ Options (see also [Cluster-Wide Disk Layout Parameters](../config/layout-cluster
 ```
 --object_size 128k       Set blockstore block size
 --bitmap_granularity 4k  Set bitmap granularity
---journal_size 32M       Set journal size
+--journal_size 16M       Set journal size
+--data_csum_type none    Set data checksum type (crc32c or none)
+--csum_block_size 4k     Set data checksum block size
 --device_block_size 4k   Set device block size
 --journal_offset 0       Set journal offset
 --device_size 0          Set device size

docs/usage/disk.ru.md

@@ -87,6 +87,8 @@ vitastor-disk - инструмент командной строки для уп
 --journal_size 1G/32M      Задать размер журнала (области или раздела журнала)
 --block_size 1M/128k       Задать размер объекта хранилища
 --bitmap_granularity 4k    Задать гранулярность битовых карт
+--data_csum_type none      Задать тип контрольных сумм (crc32c или none)
+--csum_block_size 4k       Задать размер блока расчёта контрольных сумм
 --data_device_block 4k     Задать размер блока устройства данных
 --meta_device_block 4k     Задать размер блока метаданных
 --journal_device_block 4k  Задать размер блока журнала
@@ -101,8 +103,9 @@ vitastor-disk - инструмент командной строки для уп
 это не удаётся, в системный журнал выводится предупреждение.
 
 Вы можете передать данной команде и некоторые другие опции OSD в качестве аргументов
-и они тоже будут сохранены в суперблок: max_write_iodepth, max_write_iodepth, min_flusher_count,
-max_flusher_count, inmemory_metadata, inmemory_journal, journal_sector_buffer_count,
+и они тоже будут сохранены в суперблок: cached_read_data, cached_read_meta,
+cached_read_journal, inmemory_metadata, inmemory_journal, max_write_iodepth,
+min_flusher_count, max_flusher_count, journal_sector_buffer_count,
 journal_no_same_sector_overwrites, throttle_small_writes, throttle_target_iops,
 throttle_target_mbs, throttle_target_parallelism, throttle_threshold_us.
 Читайте об этих параметрах подробнее в разделе [Изменяемые параметры OSD](../config/osd.ru.md).
@@ -254,7 +257,9 @@ OSD отключены fsync-и.
 ```
 --object_size 128k       Размер блока хранилища
 --bitmap_granularity 4k  Гранулярность битовых карт
---journal_size 32M       Размер журнала
+--journal_size 16M       Размер журнала
+--data_csum_type none    Задать тип контрольных сумм (crc32c или none)
+--csum_block_size 4k     Задать размер блока расчёта контрольных сумм
 --device_block_size 4k   Размер блока устройства
 --journal_offset 0       Смещение журнала
 --device_size 0          Размер устройства

docs/usage/nbd.en.md

@@ -13,6 +13,8 @@ remains decent (see an example [here](../performance/comparison1.en.md#vitastor-
 
 Vitastor Kubernetes CSI driver is based on NBD.
 
+See also [VDUSE](qemu.en.md#vduse).
+
 ## Map image
 
 To create a local block device for a Vitastor image run:

docs/usage/nbd.ru.md

@@ -16,6 +16,8 @@ NBD немного снижает производительность из-за
 
 CSI-драйвер Kubernetes Vitastor основан на NBD.
 
+Смотрите также [VDUSE](qemu.ru.md#vduse).
+
 ## Подключить устройство
 
 Чтобы создать локальное блочное устройство для образа, выполните команду:

docs/usage/nfs.ru.md

@@ -29,7 +29,7 @@ vitastor-nfs [--etcd_address ADDR] [ДРУГИЕ ОПЦИИ]
 --bind <IP>       принимать соединения по адресу <IP> (по умолчанию 0.0.0.0 - на всех)
 --nfspath <PATH>  установить путь NFS-экспорта в <PATH> (по умолчанию /)
 --port <PORT>     использовать порт <PORT> для NFS-сервисов (по умолчанию 2049)
---pool <POOL>     исползовать пул <POOL> для новых образов (обязательно, если пул в кластере не один)
+--pool <POOL>     использовать пул <POOL> для новых образов (обязательно, если пул в кластере не один)
 --foreground 1    не уходить в фон после запуска
 ```
 

docs/usage/qemu.en.md

@@ -83,3 +83,43 @@ qemu-img rebase -u -b '' testimg.qcow2
 This can be used for backups. Just note that exporting an image that is currently being written to
 is of course unsafe and doesn't produce a consistent result, so only export snapshots if you do this
 on a live VM.
+
+## VDUSE
+
+Linux kernel, starting with version 5.15, supports a new interface for attaching virtual disks
+to the host - VDUSE (vDPA Device in Userspace). QEMU, starting with 7.2, has support for
+exporting QEMU block devices over this protocol using qemu-storage-daemon.
+
+VDUSE has the same problem as other FUSE-like interfaces in Linux: if a userspace process hangs,
+for example, if it loses connectivity with Vitastor cluster - active processes doing I/O may
+hang in the D state (uninterruptible sleep) and you won't be able to kill them even with kill -9.
+In this case reboot will be the only way to remove VDUSE devices from system.
+
+On the other hand, VDUSE is faster than [NBD](nbd.en.md), so you may prefer to use it if
+performance is important for you. Approximate performance numbers:
+direct fio benchmark - 115000 iops, NBD - 60000 iops, VDUSE - 90000 iops.
+
+To try VDUSE you need at least Linux 5.15, built with VDUSE support
+(CONFIG_VIRTIO_VDPA=m and CONFIG_VDPA_USER=m). Debian Linux kernels have these options
+disabled by now, so if you want to try it on Debian, use a kernel from Ubuntu
+[kernel-ppa/mainline](https://kernel.ubuntu.com/~kernel-ppa/mainline/) or Proxmox.
+
+Commands to attach Vitastor image as a VDUSE device:
+
+```
+modprobe vduse virtio-vdpa
+qemu-storage-daemon --daemonize --blockdev '{"node-name":"test1","driver":"vitastor",\
+  "etcd-host":"192.168.7.2:2379/v3","image":"testosd1","cache":{"direct":true,"no-flush":false},"discard":"unmap"}' \
+  --export vduse-blk,id=test1,node-name=test1,name=test1,num-queues=16,queue-size=128,writable=true
+vdpa dev add name test1 mgmtdev vduse
+```
+
+After running these commands /dev/vda device will appear in the system and you'll be able to
+use it as a normal disk.
+
+To remove the device:
+
+```
+vdpa dev del test1
+kill <qemu-storage-daemon_process_PID>
+```

docs/usage/qemu.ru.md

@@ -87,3 +87,43 @@ qemu-img rebase -u -b '' testimg.qcow2
 Это можно использовать для резервного копирования. Только помните, что экспортировать образ, в который
 в то же время идёт запись, небезопасно - результат чтения не будет целостным. Так что если вы работаете
 с активными виртуальными машинами, экспортируйте только их снимки, но не сам образ.
+
+## VDUSE
+
+В Linux, начиная с версии ядра 5.15, доступен новый интерфейс для подключения виртуальных дисков
+к системе - VDUSE (vDPA Device in Userspace), а в QEMU, начиная с версии 7.2, есть поддержка
+экспорта блочных устройств QEMU по этому протоколу через qemu-storage-daemon.
+
+VDUSE страдает общей проблемой FUSE-подобных интерфейсов в Linux: если пользовательский процесс
+подвиснет, например, если будет потеряна связь с кластером Vitastor - читающие/пишущие в кластер
+процессы могут "залипнуть" в состоянии D (непрерываемый сон) и их будет невозможно убить даже
+через kill -9. В этом случае удалить из системы устройство можно только перезагрузившись.
+
+С другой стороны, VDUSE быстрее по сравнению с [NBD](nbd.ru.md), поэтому его может
+быть предпочтительно использовать там, где производительность важнее. Порядок показателей:
+прямое тестирование через fio - 115000 iops, NBD - 60000 iops, VDUSE - 90000 iops.
+
+Чтобы использовать VDUSE, вам нужно ядро Linux версии хотя бы 5.15, собранное с поддержкой
+VDUSE (CONFIG_VIRTIO_VDPA=m и CONFIG_VDPA_USER=m). В ядрах в Debian Linux поддержка пока
+отключена - если хотите попробовать эту функцию на Debian, поставьте ядро из Ubuntu
+[kernel-ppa/mainline](https://kernel.ubuntu.com/~kernel-ppa/mainline/) или из Proxmox.
+
+Команды для подключения виртуального диска через VDUSE:
+
+```
+modprobe vduse virtio-vdpa
+qemu-storage-daemon --daemonize --blockdev '{"node-name":"test1","driver":"vitastor",\
+  "etcd-host":"192.168.7.2:2379/v3","image":"testosd1","cache":{"direct":true,"no-flush":false},"discard":"unmap"}' \
+  --export vduse-blk,id=test1,node-name=test1,name=test1,num-queues=16,queue-size=128,writable=true
+vdpa dev add name test1 mgmtdev vduse
+```
+
+После этого в системе появится устройство /dev/vda, которое можно будет использовать как
+обычный диск.
+
+Для удаления устройства из системы:
+
+```
+vdpa dev del test1
+kill <PID_процесса_qemu-storage-daemon>
+```

mon/make-etcd

@@ -63,8 +63,9 @@ Wants=network-online.target local-fs.target time-sync.target
 
 [Service]
 Restart=always
-ExecStart=/usr/local/bin/etcd -name etcd${num} --data-dir /var/lib/etcd${num}.etcd \\
-    --advertise-client-urls http://${etcds[num]}:2379 --listen-client-urls http://${etcds[num]}:2379 \\
+Environment=GOGC=50
+ExecStart=etcd -name etcd${num} --data-dir /var/lib/etcd${num}.etcd \\
+    --snapshot-count 10000 --advertise-client-urls http://${etcds[num]}:2379 --listen-client-urls http://${etcds[num]}:2379 \\
     --initial-advertise-peer-urls http://${etcds[num]}:2380 --listen-peer-urls http://${etcds[num]}:2380 \\
     --initial-cluster-token vitastor-etcd-1 --initial-cluster ${etcd_cluster} \\
     --initial-cluster-state new --max-txn-ops=100000 --max-request-bytes=104857600 \\

mon/mon.js

@@ -391,6 +391,7 @@ class Mon
         this.etcd_start_timeout = (config.etcd_start_timeout || 5) * 1000;
         this.state = JSON.parse(JSON.stringify(this.constructor.etcd_tree));
         this.signals_set = false;
+        this.stat_time = Date.now();
         this.ws = null;
         this.ws_alive = false;
         this.ws_keepalive_timer = null;
@@ -1410,65 +1411,75 @@ class Mon
         }
     }
 
+    derive_osd_stats(st, prev)
+    {
+        const zero_stats = { op: { bps: 0n, iops: 0n, lat: 0n }, subop: { iops: 0n, lat: 0n }, recovery: { bps: 0n, iops: 0n } };
+        const diff = { op_stats: {}, subop_stats: {}, recovery_stats: {} };
+        if (!st || !st.time || prev && (prev.time || this.stat_time/1000) >= st.time)
+        {
+            return diff;
+        }
+        const timediff = BigInt(st.time*1000 - (prev && prev.time*1000 || this.stat_time));
+        for (const op in st.op_stats||{})
+        {
+            const pr = prev && prev.op_stats && prev.op_stats[op];
+            let c = st.op_stats[op];
+            c = { bytes: BigInt(c.bytes||0), usec: BigInt(c.usec||0), count: BigInt(c.count||0) };
+            const b = c.bytes - BigInt(pr && pr.bytes||0);
+            const us = c.usec - BigInt(pr && pr.usec||0);
+            const n = c.count - BigInt(pr && pr.count||0);
+            if (n > 0)
+                diff.op_stats[op] = { ...c, bps: b*1000n/timediff, iops: n*1000n/timediff, lat: us/n };
+        }
+        for (const op in st.subop_stats||{})
+        {
+            const pr = prev && prev.subop_stats && prev.subop_stats[op];
+            let c = st.subop_stats[op];
+            c = { usec: BigInt(c.usec||0), count: BigInt(c.count||0) };
+            const us = c.usec - BigInt(pr && pr.usec||0);
+            const n = c.count - BigInt(pr && pr.count||0);
+            if (n > 0)
+                diff.subop_stats[op] = { ...c, iops: n*1000n/timediff, lat: us/n };
+        }
+        for (const op in st.recovery_stats||{})
+        {
+            const pr = prev && prev.recovery_stats && prev.recovery_stats[op];
+            let c = st.recovery_stats[op];
+            c = { bytes: BigInt(c.bytes||0), count: BigInt(c.count||0) };
+            const b = c.bytes - BigInt(pr && pr.bytes||0);
+            const n = c.count - BigInt(pr && pr.count||0);
+            if (n > 0)
+                diff.recovery_stats[op] = { ...c, bps: b*1000n/timediff, iops: n*1000n/timediff };
+        }
+        return diff;
+    }
+
     sum_op_stats(timestamp, prev_stats)
     {
-        const op_stats = {}, subop_stats = {}, recovery_stats = {};
+        const sum_diff = { op_stats: {}, subop_stats: {}, recovery_stats: {} };
+        if (!prev_stats || prev_stats.timestamp >= timestamp)
+        {
+            return sum_diff;
+        }
+        const tm = BigInt(timestamp - (prev_stats.timestamp || 0));
+        // Sum derived values instead of deriving summed
         for (const osd in this.state.osd.stats)
         {
-            const st = this.state.osd.stats[osd]||{};
-            for (const op in st.op_stats||{})
+            const derived = this.derive_osd_stats(this.state.osd.stats[osd],
+                this.prev_stats && this.prev_stats.osd_stats && this.prev_stats.osd_stats[osd]);
+            for (const type in derived)
             {
-                op_stats[op] = op_stats[op] || { count: 0n, usec: 0n, bytes: 0n };
-                op_stats[op].count += BigInt(st.op_stats[op].count||0);
-                op_stats[op].usec += BigInt(st.op_stats[op].usec||0);
-                op_stats[op].bytes += BigInt(st.op_stats[op].bytes||0);
-            }
-            for (const op in st.subop_stats||{})
-            {
-                subop_stats[op] = subop_stats[op] || { count: 0n, usec: 0n };
-                subop_stats[op].count += BigInt(st.subop_stats[op].count||0);
-                subop_stats[op].usec += BigInt(st.subop_stats[op].usec||0);
-            }
-            for (const op in st.recovery_stats||{})
-            {
-                recovery_stats[op] = recovery_stats[op] || { count: 0n, bytes: 0n };
-                recovery_stats[op].count += BigInt(st.recovery_stats[op].count||0);
-                recovery_stats[op].bytes += BigInt(st.recovery_stats[op].bytes||0);
+                for (const op in derived[type])
+                {
+                    for (const k in derived[type][op])
+                    {
+                        sum_diff[type][op] = sum_diff[type][op] || {};
+                        sum_diff[type][op][k] = (sum_diff[type][op][k] || 0n) + derived[type][op][k];
+                    }
+                }
             }
         }
-        if (prev_stats && prev_stats.timestamp >= timestamp)
-        {
-            prev_stats = null;
-        }
-        const tm = prev_stats ? BigInt(timestamp - prev_stats.timestamp) : 0;
-        for (const op in op_stats)
-        {
-            if (prev_stats && prev_stats.op_stats && prev_stats.op_stats[op])
-            {
-                op_stats[op].bps = (op_stats[op].bytes - prev_stats.op_stats[op].bytes) * 1000n / tm;
-                op_stats[op].iops = (op_stats[op].count - prev_stats.op_stats[op].count) * 1000n / tm;
-                op_stats[op].lat = (op_stats[op].usec - prev_stats.op_stats[op].usec)
-                    / ((op_stats[op].count - prev_stats.op_stats[op].count) || 1n);
-            }
-        }
-        for (const op in subop_stats)
-        {
-            if (prev_stats && prev_stats.subop_stats && prev_stats.subop_stats[op])
-            {
-                subop_stats[op].iops = (subop_stats[op].count - prev_stats.subop_stats[op].count) * 1000n / tm;
-                subop_stats[op].lat = (subop_stats[op].usec - prev_stats.subop_stats[op].usec)
-                    / ((subop_stats[op].count - prev_stats.subop_stats[op].count) || 1n);
-            }
-        }
-        for (const op in recovery_stats)
-        {
-            if (prev_stats && prev_stats.recovery_stats && prev_stats.recovery_stats[op])
-            {
-                recovery_stats[op].bps = (recovery_stats[op].bytes - prev_stats.recovery_stats[op].bytes) * 1000n / tm;
-                recovery_stats[op].iops = (recovery_stats[op].count - prev_stats.recovery_stats[op].count) * 1000n / tm;
-            }
-        }
-        return { op_stats, subop_stats, recovery_stats };
+        return sum_diff;
     }
 
     sum_object_counts()
@@ -1597,7 +1608,7 @@ class Mon
                 }
             }
         }
-        return inode_stats;
+        return { inode_stats, seen_pools };
     }
 
     serialize_bigints(obj)
@@ -1623,11 +1634,12 @@ class Mon
         const timestamp = Date.now();
         const { object_counts, object_bytes } = this.sum_object_counts();
         let stats = this.sum_op_stats(timestamp, this.prev_stats);
-        let inode_stats = this.sum_inode_stats(
+        let { inode_stats, seen_pools } = this.sum_inode_stats(
             this.prev_stats ? this.prev_stats.inode_stats : null,
             timestamp, this.prev_stats ? this.prev_stats.timestamp : null
         );
-        this.prev_stats = { timestamp, ...stats, inode_stats };
+        this.prev_stats = { timestamp, inode_stats, osd_stats: { ...this.state.osd.stats } };
+        this.stat_time = Date.now();
         stats.object_counts = object_counts;
         stats.object_bytes = object_bytes;
         stats = this.serialize_bigints(stats);
@@ -1657,12 +1669,22 @@ class Mon
         }
         for (const pool_id in this.state.pool.stats)
         {
-            const pool_stats = { ...this.state.pool.stats[pool_id] };
-            this.serialize_bigints(pool_stats);
-            txn.push({ requestPut: {
-                key: b64(this.etcd_prefix+'/pool/stats/'+pool_id),
-                value: b64(JSON.stringify(pool_stats)),
-            } });
+            if (!seen_pools[pool_id])
+            {
+                txn.push({ requestDeleteRange: {
+                    key: b64(this.etcd_prefix+'/pool/stats/'+pool_id),
+                } });
+                delete this.state.pool.stats[pool_id];
+            }
+            else
+            {
+                const pool_stats = { ...this.state.pool.stats[pool_id] };
+                this.serialize_bigints(pool_stats);
+                txn.push({ requestPut: {
+                    key: b64(this.etcd_prefix+'/pool/stats/'+pool_id),
+                    value: b64(JSON.stringify(pool_stats)),
+                } });
+            }
         }
         if (txn.length)
         {
@@ -1743,13 +1765,14 @@ class Mon
         else if (key_parts[0] === 'osd' && key_parts[1] === 'stats')
         {
             // Recheck OSD tree on OSD addition/deletion
+            const osd_num = key_parts[2];
             if ((!old) != (!kv.value) || old && kv.value && old.size != kv.value.size)
             {
                 this.schedule_recheck();
             }
             // Recheck PGs <osd_out_time> after last OSD statistics report
             this.schedule_next_recheck_at(
-                !this.state.osd.stats[key[2]] ? 0 : this.state.osd.stats[key[2]].time+this.config.osd_out_time
+                !this.state.osd.stats[osd_num] ? 0 : this.state.osd.stats[osd_num].time+this.config.osd_out_time
             );
         }
     }

patches/VitastorPlugin.pm

@@ -388,8 +388,6 @@ sub unmap_volume
     my ($class, $storeid, $scfg, $volname, $snapname) = @_;
     my $prefix = defined $scfg->{vitastor_prefix} ? $scfg->{vitastor_prefix} : 'pve/';
 
-    return 1 if !$scfg->{vitastor_nbd};
-
     my ($vtype, $name, $vmid) = $class->parse_volname($volname);
     $name .= '@'.$snapname if $snapname;
 
@@ -413,7 +411,7 @@ sub activate_volume
 sub deactivate_volume
 {
     my ($class, $storeid, $scfg, $volname, $snapname, $cache) = @_;
-    $class->unmap_volume($storeid, $scfg, $volname, $snapname);
+    $class->unmap_volume($storeid, $scfg, $volname, $snapname) if $scfg->{vitastor_nbd};
     return 1;
 }
 

patches/cinder-vitastor.py

@@ -50,7 +50,7 @@ from cinder.volume import configuration
 from cinder.volume import driver
 from cinder.volume import volume_utils
 
-VERSION = '0.9.0'
+VERSION = '0.9.3'
 
 LOG = logging.getLogger(__name__)
 

patches/libvirt-9.0-vitastor.diff

@@ -0,0 +1,644 @@
+commit e6f935157944279c2c0634915c3c00feeec748c9
+Author: Vitaliy Filippov <vitalif@yourcmc.ru>
+Date:   Mon Jun 19 00:58:19 2023 +0300
+
+    Add Vitastor support
+
+diff --git a/include/libvirt/libvirt-storage.h b/include/libvirt/libvirt-storage.h
+index aaad4a3..5f5daa8 100644
+--- a/include/libvirt/libvirt-storage.h
++++ b/include/libvirt/libvirt-storage.h
+@@ -326,6 +326,7 @@ typedef enum {
+     VIR_CONNECT_LIST_STORAGE_POOLS_ZFS           = 1 << 17, /* (Since: 1.2.8) */
+     VIR_CONNECT_LIST_STORAGE_POOLS_VSTORAGE      = 1 << 18, /* (Since: 3.1.0) */
+     VIR_CONNECT_LIST_STORAGE_POOLS_ISCSI_DIRECT  = 1 << 19, /* (Since: 5.6.0) */
++    VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR      = 1 << 20, /* (Since: 5.0.0) */
+ } virConnectListAllStoragePoolsFlags;
+ 
+ int                     virConnectListAllStoragePools(virConnectPtr conn,
+diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
+index 45965fa..b7c23d3 100644
+--- a/src/conf/domain_conf.c
++++ b/src/conf/domain_conf.c
+@@ -7103,7 +7103,8 @@ virDomainDiskSourceNetworkParse(xmlNodePtr node,
+     src->configFile = virXPathString("string(./config/@file)", ctxt);
+ 
+     if (src->protocol == VIR_STORAGE_NET_PROTOCOL_HTTP ||
+-        src->protocol == VIR_STORAGE_NET_PROTOCOL_HTTPS)
++        src->protocol == VIR_STORAGE_NET_PROTOCOL_HTTPS ||
++        src->protocol == VIR_STORAGE_NET_PROTOCOL_VITASTOR)
+         src->query = virXMLPropString(node, "query");
+ 
+     if (virDomainStorageNetworkParseHosts(node, ctxt, &src->hosts, &src->nhosts) < 0)
+@@ -30121,6 +30122,7 @@ virDomainStorageSourceTranslateSourcePool(virStorageSource *src,
+ 
+     case VIR_STORAGE_POOL_MPATH:
+     case VIR_STORAGE_POOL_RBD:
++    case VIR_STORAGE_POOL_VITASTOR:
+     case VIR_STORAGE_POOL_SHEEPDOG:
+     case VIR_STORAGE_POOL_GLUSTER:
+     case VIR_STORAGE_POOL_LAST:
+diff --git a/src/conf/domain_validate.c b/src/conf/domain_validate.c
+index 5a9bf20..05058b8 100644
+--- a/src/conf/domain_validate.c
++++ b/src/conf/domain_validate.c
+@@ -494,6 +494,7 @@ virDomainDiskDefValidateSourceChainOne(const virStorageSource *src)
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
+             break;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_NBD:
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+@@ -541,7 +542,7 @@ virDomainDiskDefValidateSourceChainOne(const virStorageSource *src)
+         }
+     }
+ 
+-    /* internal snapshots and config files are currently supported only with rbd: */
++    /* internal snapshots are currently supported only with rbd: */
+     if (virStorageSourceGetActualType(src) != VIR_STORAGE_TYPE_NETWORK &&
+         src->protocol != VIR_STORAGE_NET_PROTOCOL_RBD) {
+         if (src->snapshot) {
+@@ -550,11 +551,15 @@ virDomainDiskDefValidateSourceChainOne(const virStorageSource *src)
+                              "only with 'rbd' disks"));
+             return -1;
+         }
+-
++    }
++    /* config files are currently supported only with rbd and vitastor: */
++    if (virStorageSourceGetActualType(src) != VIR_STORAGE_TYPE_NETWORK &&
++        src->protocol != VIR_STORAGE_NET_PROTOCOL_RBD &&
++        src->protocol != VIR_STORAGE_NET_PROTOCOL_VITASTOR) {
+         if (src->configFile) {
+             virReportError(VIR_ERR_XML_ERROR, "%s",
+                            _("<config> element is currently supported "
+-                             "only with 'rbd' disks"));
++                             "only with 'rbd' and 'vitastor' disks"));
+             return -1;
+         }
+     }
+diff --git a/src/conf/schemas/domaincommon.rng b/src/conf/schemas/domaincommon.rng
+index 6cb0a20..8bf7de9 100644
+--- a/src/conf/schemas/domaincommon.rng
++++ b/src/conf/schemas/domaincommon.rng
+@@ -1972,6 +1972,35 @@
+     </element>
+   </define>
+ 
++  <define name="diskSourceNetworkProtocolVitastor">
++    <element name="source">
++      <interleave>
++        <attribute name="protocol">
++          <value>vitastor</value>
++        </attribute>
++        <ref name="diskSourceCommon"/>
++        <optional>
++          <attribute name="name"/>
++        </optional>
++        <optional>
++          <attribute name="query"/>
++        </optional>
++        <zeroOrMore>
++          <ref name="diskSourceNetworkHost"/>
++        </zeroOrMore>
++        <optional>
++          <element name="config">
++            <attribute name="file">
++              <ref name="absFilePath"/>
++            </attribute>
++            <empty/>
++          </element>
++        </optional>
++        <empty/>
++      </interleave>
++    </element>
++  </define>
++
+   <define name="diskSourceNetworkProtocolISCSI">
+     <element name="source">
+       <attribute name="protocol">
+@@ -2264,6 +2293,7 @@
+       <ref name="diskSourceNetworkProtocolSimple"/>
+       <ref name="diskSourceNetworkProtocolVxHS"/>
+       <ref name="diskSourceNetworkProtocolNFS"/>
++      <ref name="diskSourceNetworkProtocolVitastor"/>
+     </choice>
+   </define>
+ 
+diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c
+index f5a9636..8339bc4 100644
+--- a/src/conf/storage_conf.c
++++ b/src/conf/storage_conf.c
+@@ -56,7 +56,7 @@ VIR_ENUM_IMPL(virStoragePool,
+               "logical", "disk", "iscsi",
+               "iscsi-direct", "scsi", "mpath",
+               "rbd", "sheepdog", "gluster",
+-              "zfs", "vstorage",
++              "zfs", "vstorage", "vitastor",
+ );
+ 
+ VIR_ENUM_IMPL(virStoragePoolFormatFileSystem,
+@@ -242,6 +242,18 @@ static virStoragePoolTypeInfo poolTypeInfo[] = {
+           .formatToString = virStorageFileFormatTypeToString,
+       }
+     },
++    {.poolType = VIR_STORAGE_POOL_VITASTOR,
++     .poolOptions = {
++         .flags = (VIR_STORAGE_POOL_SOURCE_HOST |
++                   VIR_STORAGE_POOL_SOURCE_NETWORK |
++                   VIR_STORAGE_POOL_SOURCE_NAME),
++      },
++      .volOptions = {
++          .defaultFormat = VIR_STORAGE_FILE_RAW,
++          .formatFromString = virStorageVolumeFormatFromString,
++          .formatToString = virStorageFileFormatTypeToString,
++      }
++    },
+     {.poolType = VIR_STORAGE_POOL_SHEEPDOG,
+      .poolOptions = {
+          .flags = (VIR_STORAGE_POOL_SOURCE_HOST |
+@@ -542,6 +554,11 @@ virStoragePoolDefParseSource(xmlXPathContextPtr ctxt,
+                        _("element 'name' is mandatory for RBD pool"));
+         return -1;
+     }
++    if (pool_type == VIR_STORAGE_POOL_VITASTOR && source->name == NULL) {
++        virReportError(VIR_ERR_XML_ERROR, "%s",
++                       _("element 'name' is mandatory for Vitastor pool"));
++        return -1;
++    }
+ 
+     if (options->formatFromString) {
+         g_autofree char *format = NULL;
+@@ -1132,6 +1149,7 @@ virStoragePoolDefFormatBuf(virBuffer *buf,
+     /* RBD, Sheepdog, Gluster and Iscsi-direct devices are not local block devs nor
+      * files, so they don't have a target */
+     if (def->type != VIR_STORAGE_POOL_RBD &&
++        def->type != VIR_STORAGE_POOL_VITASTOR &&
+         def->type != VIR_STORAGE_POOL_SHEEPDOG &&
+         def->type != VIR_STORAGE_POOL_GLUSTER &&
+         def->type != VIR_STORAGE_POOL_ISCSI_DIRECT) {
+diff --git a/src/conf/storage_conf.h b/src/conf/storage_conf.h
+index fc67957..720c07e 100644
+--- a/src/conf/storage_conf.h
++++ b/src/conf/storage_conf.h
+@@ -103,6 +103,7 @@ typedef enum {
+     VIR_STORAGE_POOL_GLUSTER,  /* Gluster device */
+     VIR_STORAGE_POOL_ZFS,      /* ZFS */
+     VIR_STORAGE_POOL_VSTORAGE, /* Virtuozzo Storage */
++    VIR_STORAGE_POOL_VITASTOR, /* Vitastor */
+ 
+     VIR_STORAGE_POOL_LAST,
+ } virStoragePoolType;
+@@ -454,6 +455,7 @@ VIR_ENUM_DECL(virStoragePartedFs);
+                  VIR_CONNECT_LIST_STORAGE_POOLS_SCSI     | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_MPATH    | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_RBD      | \
++                 VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_SHEEPDOG | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_GLUSTER  | \
+                  VIR_CONNECT_LIST_STORAGE_POOLS_ZFS      | \
+diff --git a/src/conf/storage_source_conf.c b/src/conf/storage_source_conf.c
+index cecd7e8..d7b79a4 100644
+--- a/src/conf/storage_source_conf.c
++++ b/src/conf/storage_source_conf.c
+@@ -87,6 +87,7 @@ VIR_ENUM_IMPL(virStorageNetProtocol,
+               "ssh",
+               "vxhs",
+               "nfs",
++              "vitastor",
+ );
+ 
+ 
+@@ -1286,6 +1287,7 @@ virStorageSourceNetworkDefaultPort(virStorageNetProtocol protocol)
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+             return 24007;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
+             /* we don't provide a default for RBD */
+             return 0;
+diff --git a/src/conf/storage_source_conf.h b/src/conf/storage_source_conf.h
+index 14a6825..eb4acac 100644
+--- a/src/conf/storage_source_conf.h
++++ b/src/conf/storage_source_conf.h
+@@ -128,6 +128,7 @@ typedef enum {
+     VIR_STORAGE_NET_PROTOCOL_SSH,
+     VIR_STORAGE_NET_PROTOCOL_VXHS,
+     VIR_STORAGE_NET_PROTOCOL_NFS,
++    VIR_STORAGE_NET_PROTOCOL_VITASTOR,
+ 
+     VIR_STORAGE_NET_PROTOCOL_LAST
+ } virStorageNetProtocol;
+diff --git a/src/conf/virstorageobj.c b/src/conf/virstorageobj.c
+index e6c187e..035b423 100644
+--- a/src/conf/virstorageobj.c
++++ b/src/conf/virstorageobj.c
+@@ -1433,6 +1433,7 @@ virStoragePoolObjSourceFindDuplicateCb(const void *payload,
+             return 1;
+         break;
+ 
++    case VIR_STORAGE_POOL_VITASTOR:
+     case VIR_STORAGE_POOL_ISCSI_DIRECT:
+     case VIR_STORAGE_POOL_RBD:
+     case VIR_STORAGE_POOL_LAST:
+@@ -1918,6 +1919,8 @@ virStoragePoolObjMatch(virStoragePoolObj *obj,
+                (obj->def->type == VIR_STORAGE_POOL_MPATH))   ||
+               (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_RBD) &&
+                (obj->def->type == VIR_STORAGE_POOL_RBD))     ||
++              (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR) &&
++               (obj->def->type == VIR_STORAGE_POOL_VITASTOR)) ||
+               (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_SHEEPDOG) &&
+                (obj->def->type == VIR_STORAGE_POOL_SHEEPDOG)) ||
+               (MATCH(VIR_CONNECT_LIST_STORAGE_POOLS_GLUSTER) &&
+diff --git a/src/libvirt-storage.c b/src/libvirt-storage.c
+index 8490034..ab2cdaa 100644
+--- a/src/libvirt-storage.c
++++ b/src/libvirt-storage.c
+@@ -94,6 +94,7 @@ virStoragePoolGetConnect(virStoragePoolPtr pool)
+  * VIR_CONNECT_LIST_STORAGE_POOLS_SCSI
+  * VIR_CONNECT_LIST_STORAGE_POOLS_MPATH
+  * VIR_CONNECT_LIST_STORAGE_POOLS_RBD
++ * VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR
+  * VIR_CONNECT_LIST_STORAGE_POOLS_SHEEPDOG
+  * VIR_CONNECT_LIST_STORAGE_POOLS_GLUSTER
+  * VIR_CONNECT_LIST_STORAGE_POOLS_ZFS
+diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c
+index 17ac880..59711b5 100644
+--- a/src/libxl/libxl_conf.c
++++ b/src/libxl/libxl_conf.c
+@@ -970,6 +970,7 @@ libxlMakeNetworkDiskSrcStr(virStorageSource *src,
+     case VIR_STORAGE_NET_PROTOCOL_SSH:
+     case VIR_STORAGE_NET_PROTOCOL_VXHS:
+     case VIR_STORAGE_NET_PROTOCOL_NFS:
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+     case VIR_STORAGE_NET_PROTOCOL_LAST:
+     case VIR_STORAGE_NET_PROTOCOL_NONE:
+         virReportError(VIR_ERR_NO_SUPPORT,
+diff --git a/src/libxl/xen_xl.c b/src/libxl/xen_xl.c
+index 6919325..55ffc32 100644
+--- a/src/libxl/xen_xl.c
++++ b/src/libxl/xen_xl.c
+@@ -1445,6 +1445,7 @@ xenFormatXLDiskSrcNet(virStorageSource *src)
+     case VIR_STORAGE_NET_PROTOCOL_SSH:
+     case VIR_STORAGE_NET_PROTOCOL_VXHS:
+     case VIR_STORAGE_NET_PROTOCOL_NFS:
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+     case VIR_STORAGE_NET_PROTOCOL_LAST:
+     case VIR_STORAGE_NET_PROTOCOL_NONE:
+         virReportError(VIR_ERR_NO_SUPPORT,
+diff --git a/src/qemu/qemu_block.c b/src/qemu/qemu_block.c
+index e865aa1..40162af 100644
+--- a/src/qemu/qemu_block.c
++++ b/src/qemu/qemu_block.c
+@@ -604,6 +604,38 @@ qemuBlockStorageSourceGetRBDProps(virStorageSource *src,
+ }
+ 
+ 
++static virJSONValue *
++qemuBlockStorageSourceGetVitastorProps(virStorageSource *src)
++{
++    virJSONValue *ret = NULL;
++    virStorageNetHostDef *host;
++    size_t i;
++    g_auto(virBuffer) buf = VIR_BUFFER_INITIALIZER;
++    g_autofree char *etcd = NULL;
++
++    for (i = 0; i < src->nhosts; i++) {
++        host = src->hosts + i;
++        if ((virStorageNetHostTransport)host->transport != VIR_STORAGE_NET_HOST_TRANS_TCP) {
++            return NULL;
++        }
++        virBufferAsprintf(&buf, i > 0 ? ",%s:%u" : "%s:%u", host->name, host->port);
++    }
++    if (src->nhosts > 0) {
++        etcd = virBufferContentAndReset(&buf);
++    }
++
++    if (virJSONValueObjectAdd(&ret,
++                              "S:etcd-host", etcd,
++                              "S:etcd-prefix", src->query,
++                              "S:config-path", src->configFile,
++                              "s:image", src->path,
++                              NULL) < 0)
++        return NULL;
++
++    return ret;
++}
++
++
+ static virJSONValue *
+ qemuBlockStorageSourceGetSheepdogProps(virStorageSource *src)
+ {
+@@ -917,6 +949,12 @@ qemuBlockStorageSourceGetBackendProps(virStorageSource *src,
+                 return NULL;
+             break;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++            driver = "vitastor";
++            if (!(fileprops = qemuBlockStorageSourceGetVitastorProps(src)))
++                return NULL;
++            break;
++
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+             driver = "sheepdog";
+             if (!(fileprops = qemuBlockStorageSourceGetSheepdogProps(src)))
+@@ -1860,6 +1898,7 @@ qemuBlockGetBackingStoreString(virStorageSource *src,
+ 
+             case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+             case VIR_STORAGE_NET_PROTOCOL_RBD:
++            case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+             case VIR_STORAGE_NET_PROTOCOL_VXHS:
+             case VIR_STORAGE_NET_PROTOCOL_NFS:
+             case VIR_STORAGE_NET_PROTOCOL_SSH:
+@@ -2242,6 +2281,12 @@ qemuBlockStorageSourceCreateGetStorageProps(virStorageSource *src,
+                 return -1;
+             break;
+ 
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++            driver = "vitastor";
++            if (!(location = qemuBlockStorageSourceGetVitastorProps(src)))
++                return -1;
++            break;
++
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+             driver = "sheepdog";
+             if (!(location = qemuBlockStorageSourceGetSheepdogProps(src)))
+diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c
+index 2eb5653..60ee82d 100644
+--- a/src/qemu/qemu_domain.c
++++ b/src/qemu/qemu_domain.c
+@@ -4958,7 +4958,8 @@ qemuDomainValidateStorageSource(virStorageSource *src,
+     if (src->query &&
+         (actualType != VIR_STORAGE_TYPE_NETWORK ||
+          (src->protocol != VIR_STORAGE_NET_PROTOCOL_HTTPS &&
+-          src->protocol != VIR_STORAGE_NET_PROTOCOL_HTTP))) {
++          src->protocol != VIR_STORAGE_NET_PROTOCOL_HTTP &&
++          src->protocol != VIR_STORAGE_NET_PROTOCOL_VITASTOR))) {
+         virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s",
+                        _("query is supported only with HTTP(S) protocols"));
+         return -1;
+@@ -10129,6 +10130,7 @@ qemuDomainPrepareStorageSourceTLS(virStorageSource *src,
+         break;
+ 
+     case VIR_STORAGE_NET_PROTOCOL_RBD:
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+     case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+     case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+     case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+diff --git a/src/qemu/qemu_snapshot.c b/src/qemu/qemu_snapshot.c
+index b841680..a6be771 100644
+--- a/src/qemu/qemu_snapshot.c
++++ b/src/qemu/qemu_snapshot.c
+@@ -373,6 +373,7 @@ qemuSnapshotPrepareDiskExternalInactive(virDomainSnapshotDiskDef *snapdisk,
+         case VIR_STORAGE_NET_PROTOCOL_NONE:
+         case VIR_STORAGE_NET_PROTOCOL_NBD:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+         case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+@@ -578,6 +579,7 @@ qemuSnapshotPrepareDiskInternal(virDomainDiskDef *disk,
+         case VIR_STORAGE_NET_PROTOCOL_NONE:
+         case VIR_STORAGE_NET_PROTOCOL_NBD:
+         case VIR_STORAGE_NET_PROTOCOL_RBD:
++        case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
+         case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+         case VIR_STORAGE_NET_PROTOCOL_GLUSTER:
+         case VIR_STORAGE_NET_PROTOCOL_ISCSI:
+diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
+index d90c1c9..e853457 100644
+--- a/src/storage/storage_driver.c
++++ b/src/storage/storage_driver.c
+@@ -1627,6 +1627,7 @@ storageVolLookupByPathCallback(virStoragePoolObj *obj,
+ 
+         case VIR_STORAGE_POOL_GLUSTER:
+         case VIR_STORAGE_POOL_RBD:
++        case VIR_STORAGE_POOL_VITASTOR:
+         case VIR_STORAGE_POOL_SHEEPDOG:
+         case VIR_STORAGE_POOL_ZFS:
+         case VIR_STORAGE_POOL_LAST:
+diff --git a/src/storage_file/storage_source_backingstore.c b/src/storage_file/storage_source_backingstore.c
+index e48ae72..2017ccc 100644
+--- a/src/storage_file/storage_source_backingstore.c
++++ b/src/storage_file/storage_source_backingstore.c
+@@ -284,6 +284,75 @@ virStorageSourceParseRBDColonString(const char *rbdstr,
+ }
+ 
+ 
++static int
++virStorageSourceParseVitastorColonString(const char *colonstr,
++                                         virStorageSource *src)
++{
++    char *p, *e, *next;
++    g_autofree char *options = NULL;
++
++    /* optionally skip the "vitastor:" prefix if provided */
++    if (STRPREFIX(colonstr, "vitastor:"))
++        colonstr += strlen("vitastor:");
++
++    options = g_strdup(colonstr);
++
++    p = options;
++    while (*p) {
++        /* find : delimiter or end of string */
++        for (e = p; *e && *e != ':'; ++e) {
++            if (*e == '\\') {
++                e++;
++                if (*e == '\0')
++                    break;
++            }
++        }
++        if (*e == '\0') {
++            next = e;    /* last kv pair */
++        } else {
++            next = e + 1;
++            *e = '\0';
++        }
++
++        if (STRPREFIX(p, "image=")) {
++            src->path = g_strdup(p + strlen("image="));
++        } else if (STRPREFIX(p, "etcd-prefix=")) {
++            src->query = g_strdup(p + strlen("etcd-prefix="));
++        } else if (STRPREFIX(p, "config-path=")) {
++            src->configFile = g_strdup(p + strlen("config-path="));
++        } else if (STRPREFIX(p, "etcd-host=")) {
++            char *h, *sep;
++
++            h = p + strlen("etcd-host=");
++            while (h < e) {
++                for (sep = h; sep < e; ++sep) {
++                    if (*sep == '\\' && (sep[1] == ',' ||
++                                         sep[1] == ';' ||
++                                         sep[1] == ' ')) {
++                        *sep = '\0';
++                        sep += 2;
++                        break;
++                    }
++                }
++
++                if (virStorageSourceRBDAddHost(src, h) < 0)
++                    return -1;
++
++                h = sep;
++            }
++        }
++
++        p = next;
++    }
++
++    if (!src->path) {
++        return -1;
++    }
++
++    return 0;
++}
++
++
+ static int
+ virStorageSourceParseNBDColonString(const char *nbdstr,
+                                     virStorageSource *src)
+@@ -396,6 +465,11 @@ virStorageSourceParseBackingColon(virStorageSource *src,
+             return -1;
+         break;
+ 
++    case VIR_STORAGE_NET_PROTOCOL_VITASTOR:
++        if (virStorageSourceParseVitastorColonString(path, src) < 0)
++            return -1;
++        break;
++
+     case VIR_STORAGE_NET_PROTOCOL_SHEEPDOG:
+     case VIR_STORAGE_NET_PROTOCOL_LAST:
+     case VIR_STORAGE_NET_PROTOCOL_NONE:
+@@ -984,6 +1058,54 @@ virStorageSourceParseBackingJSONRBD(virStorageSource *src,
+     return 0;
+ }
+ 
++static int
++virStorageSourceParseBackingJSONVitastor(virStorageSource *src,
++                                         virJSONValue *json,
++                                         const char *jsonstr G_GNUC_UNUSED,
++                                         int opaque G_GNUC_UNUSED)
++{
++    const char *filename;
++    const char *image = virJSONValueObjectGetString(json, "image");
++    const char *conf = virJSONValueObjectGetString(json, "config-path");
++    const char *etcd_prefix = virJSONValueObjectGetString(json, "etcd-prefix");
++    virJSONValue *servers = virJSONValueObjectGetArray(json, "server");
++    size_t nservers;
++    size_t i;
++
++    src->type = VIR_STORAGE_TYPE_NETWORK;
++    src->protocol = VIR_STORAGE_NET_PROTOCOL_VITASTOR;
++
++    /* legacy syntax passed via 'filename' option */
++    if ((filename = virJSONValueObjectGetString(json, "filename")))
++        return virStorageSourceParseVitastorColonString(filename, src);
++
++    if (!image) {
++        virReportError(VIR_ERR_INVALID_ARG, "%s",
++                       _("missing image name in Vitastor backing volume "
++                         "JSON specification"));
++        return -1;
++    }
++
++    src->path = g_strdup(image);
++    src->configFile = g_strdup(conf);
++    src->query = g_strdup(etcd_prefix);
++
++    if (servers) {
++        nservers = virJSONValueArraySize(servers);
++
++        src->hosts = g_new0(virStorageNetHostDef, nservers);
++        src->nhosts = nservers;
++
++        for (i = 0; i < nservers; i++) {
++            if (virStorageSourceParseBackingJSONInetSocketAddress(src->hosts + i,
++                                                                  virJSONValueArrayGet(servers, i)) < 0)
++                return -1;
++        }
++    }
++
++    return 0;
++}
++
+ static int
+ virStorageSourceParseBackingJSONRaw(virStorageSource *src,
+                                     virJSONValue *json,
+@@ -1162,6 +1284,7 @@ static const struct virStorageSourceJSONDriverParser jsonParsers[] = {
+     {"sheepdog", false, virStorageSourceParseBackingJSONSheepdog, 0},
+     {"ssh", false, virStorageSourceParseBackingJSONSSH, 0},
+     {"rbd", false, virStorageSourceParseBackingJSONRBD, 0},
++    {"vitastor", false, virStorageSourceParseBackingJSONVitastor, 0},
+     {"raw", true, virStorageSourceParseBackingJSONRaw, 0},
+     {"nfs", false, virStorageSourceParseBackingJSONNFS, 0},
+     {"vxhs", false, virStorageSourceParseBackingJSONVxHS, 0},
+diff --git a/src/test/test_driver.c b/src/test/test_driver.c
+index bd6f063..cce34e1 100644
+--- a/src/test/test_driver.c
++++ b/src/test/test_driver.c
+@@ -7338,6 +7338,7 @@ testStorageVolumeTypeForPool(int pooltype)
+     case VIR_STORAGE_POOL_ISCSI_DIRECT:
+     case VIR_STORAGE_POOL_GLUSTER:
+     case VIR_STORAGE_POOL_RBD:
++    case VIR_STORAGE_POOL_VITASTOR:
+         return VIR_STORAGE_VOL_NETWORK;
+     case VIR_STORAGE_POOL_LOGICAL:
+     case VIR_STORAGE_POOL_DISK:
+diff --git a/tests/storagepoolcapsschemadata/poolcaps-fs.xml b/tests/storagepoolcapsschemadata/poolcaps-fs.xml
+index eee75af..8bd0a57 100644
+--- a/tests/storagepoolcapsschemadata/poolcaps-fs.xml
++++ b/tests/storagepoolcapsschemadata/poolcaps-fs.xml
+@@ -204,4 +204,11 @@
+       </enum>
+     </volOptions>
+   </pool>
++  <pool type='vitastor' supported='no'>
++    <volOptions>
++      <defaultFormat type='raw'/>
++      <enum name='targetFormatType'>
++      </enum>
++    </volOptions>
++  </pool>
+ </storagepoolCapabilities>
+diff --git a/tests/storagepoolcapsschemadata/poolcaps-full.xml b/tests/storagepoolcapsschemadata/poolcaps-full.xml
+index 805950a..852df0d 100644
+--- a/tests/storagepoolcapsschemadata/poolcaps-full.xml
++++ b/tests/storagepoolcapsschemadata/poolcaps-full.xml
+@@ -204,4 +204,11 @@
+       </enum>
+     </volOptions>
+   </pool>
++  <pool type='vitastor' supported='yes'>
++    <volOptions>
++      <defaultFormat type='raw'/>
++      <enum name='targetFormatType'>
++      </enum>
++    </volOptions>
++  </pool>
+ </storagepoolCapabilities>
+diff --git a/tests/storagepoolxml2argvtest.c b/tests/storagepoolxml2argvtest.c
+index e8e40d6..db55fe5 100644
+--- a/tests/storagepoolxml2argvtest.c
++++ b/tests/storagepoolxml2argvtest.c
+@@ -65,6 +65,7 @@ testCompareXMLToArgvFiles(bool shouldFail,
+     case VIR_STORAGE_POOL_GLUSTER:
+     case VIR_STORAGE_POOL_ZFS:
+     case VIR_STORAGE_POOL_VSTORAGE:
++    case VIR_STORAGE_POOL_VITASTOR:
+     case VIR_STORAGE_POOL_LAST:
+     default:
+         VIR_TEST_DEBUG("pool type '%s' has no xml2argv test", defTypeStr);
+diff --git a/tools/virsh-pool.c b/tools/virsh-pool.c
+index 8a98c6a..4b1bbd4 100644
+--- a/tools/virsh-pool.c
++++ b/tools/virsh-pool.c
+@@ -1221,6 +1221,9 @@ cmdPoolList(vshControl *ctl, const vshCmd *cmd G_GNUC_UNUSED)
+             case VIR_STORAGE_POOL_VSTORAGE:
+                 flags |= VIR_CONNECT_LIST_STORAGE_POOLS_VSTORAGE;
+                 break;
++            case VIR_STORAGE_POOL_VITASTOR:
++                flags |= VIR_CONNECT_LIST_STORAGE_POOLS_VITASTOR;
++                break;
+             case VIR_STORAGE_POOL_LAST:
+                 break;
+             }

patches/pve-qemu-8.0-vitastor.patch

@@ -0,0 +1,190 @@
+diff --git a/block/meson.build b/block/meson.build
+index 382bec0e7d..af6207dbce 100644
+--- a/block/meson.build
++++ b/block/meson.build
+@@ -114,6 +114,7 @@ foreach m : [
+   [libnfs, 'nfs', files('nfs.c')],
+   [libssh, 'ssh', files('ssh.c')],
+   [rbd, 'rbd', files('rbd.c')],
++  [vitastor, 'vitastor', files('vitastor.c')],
+ ]
+   if m[0].found()
+     module_ss = ss.source_set()
+diff --git a/meson.build b/meson.build
+index c44d05a13f..ebedb42843 100644
+--- a/meson.build
++++ b/meson.build
+@@ -1028,6 +1028,26 @@ if not get_option('rbd').auto() or have_block
+   endif
+ endif
+ 
++vitastor = not_found
++if not get_option('vitastor').auto() or have_block
++  libvitastor_client = cc.find_library('vitastor_client', has_headers: ['vitastor_c.h'],
++    required: get_option('vitastor'), kwargs: static_kwargs)
++  if libvitastor_client.found()
++    if cc.links('''
++      #include <vitastor_c.h>
++      int main(void) {
++        vitastor_c_create_qemu(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0);
++        return 0;
++      }''', dependencies: libvitastor_client)
++      vitastor = declare_dependency(dependencies: libvitastor_client)
++    elif get_option('vitastor').enabled()
++      error('could not link libvitastor_client')
++    else
++      warning('could not link libvitastor_client, disabling')
++    endif
++  endif
++endif
++
+ glusterfs = not_found
+ glusterfs_ftruncate_has_stat = false
+ glusterfs_iocb_has_stat = false
+@@ -1882,6 +1902,7 @@ endif
+ config_host_data.set('CONFIG_OPENGL', opengl.found())
+ config_host_data.set('CONFIG_PROFILER', get_option('profiler'))
+ config_host_data.set('CONFIG_RBD', rbd.found())
++config_host_data.set('CONFIG_VITASTOR', vitastor.found())
+ config_host_data.set('CONFIG_RDMA', rdma.found())
+ config_host_data.set('CONFIG_SDL', sdl.found())
+ config_host_data.set('CONFIG_SDL_IMAGE', sdl_image.found())
+@@ -4020,6 +4041,7 @@ if spice_protocol.found()
+   summary_info += {'  spice server support': spice}
+ endif
+ summary_info += {'rbd support':       rbd}
++summary_info += {'vitastor support':  vitastor}
+ summary_info += {'smartcard support': cacard}
+ summary_info += {'U2F support':       u2f}
+ summary_info += {'libusb':            libusb}
+diff --git a/meson_options.txt b/meson_options.txt
+index fc9447d267..c4ac55c283 100644
+--- a/meson_options.txt
++++ b/meson_options.txt
+@@ -173,6 +173,8 @@ option('lzo', type : 'feature', value : 'auto',
+        description: 'lzo compression support')
+ option('rbd', type : 'feature', value : 'auto',
+        description: 'Ceph block device driver')
++option('vitastor', type : 'feature', value : 'auto',
++       description: 'Vitastor block device driver')
+ option('opengl', type : 'feature', value : 'auto',
+        description: 'OpenGL support')
+ option('rdma', type : 'feature', value : 'auto',
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index c05ad0c07e..f5eb701604 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -3308,7 +3308,7 @@
+             'raw', 'rbd',
+             { 'name': 'replication', 'if': 'CONFIG_REPLICATION' },
+             'pbs',
+-            'ssh', 'throttle', 'vdi', 'vhdx',
++            'ssh', 'throttle', 'vdi', 'vhdx', 'vitastor',
+             { 'name': 'virtio-blk-vfio-pci', 'if': 'CONFIG_BLKIO' },
+             { 'name': 'virtio-blk-vhost-user', 'if': 'CONFIG_BLKIO' },
+             { 'name': 'virtio-blk-vhost-vdpa', 'if': 'CONFIG_BLKIO' },
+@@ -4338,6 +4338,28 @@
+             '*key-secret': 'str',
+             '*server': ['InetSocketAddressBase'] } }
+ 
++##
++# @BlockdevOptionsVitastor:
++#
++# Driver specific block device options for vitastor
++#
++# @image:       Image name
++# @inode:       Inode number
++# @pool:        Pool ID
++# @size:        Desired image size in bytes
++# @config-path: Path to Vitastor configuration
++# @etcd-host:   etcd connection address(es)
++# @etcd-prefix: etcd key/value prefix
++##
++{ 'struct': 'BlockdevOptionsVitastor',
++  'data': { '*inode': 'uint64',
++            '*pool': 'uint64',
++            '*size': 'uint64',
++            '*image': 'str',
++            '*config-path': 'str',
++            '*etcd-host': 'str',
++            '*etcd-prefix': 'str' } }
++
+ ##
+ # @ReplicationMode:
+ #
+@@ -4787,6 +4809,7 @@
+       'throttle':   'BlockdevOptionsThrottle',
+       'vdi':        'BlockdevOptionsGenericFormat',
+       'vhdx':       'BlockdevOptionsGenericFormat',
++      'vitastor':   'BlockdevOptionsVitastor',
+       'virtio-blk-vfio-pci':
+                     { 'type': 'BlockdevOptionsVirtioBlkVfioPci',
+                       'if': 'CONFIG_BLKIO' },
+@@ -5187,6 +5210,17 @@
+             '*cluster-size' :   'size',
+             '*encrypt' :        'RbdEncryptionCreateOptions' } }
+ 
++##
++# @BlockdevCreateOptionsVitastor:
++#
++# Driver specific image creation options for Vitastor.
++#
++# @size: Size of the virtual disk in bytes
++##
++{ 'struct': 'BlockdevCreateOptionsVitastor',
++  'data': { 'location':         'BlockdevOptionsVitastor',
++            'size':             'size' } }
++
+ ##
+ # @BlockdevVmdkSubformat:
+ #
+@@ -5385,6 +5419,7 @@
+       'ssh':            'BlockdevCreateOptionsSsh',
+       'vdi':            'BlockdevCreateOptionsVdi',
+       'vhdx':           'BlockdevCreateOptionsVhdx',
++      'vitastor':       'BlockdevCreateOptionsVitastor',
+       'vmdk':           'BlockdevCreateOptionsVmdk',
+       'vpc':            'BlockdevCreateOptionsVpc'
+   } }
+diff --git a/scripts/ci/org.centos/stream/8/x86_64/configure b/scripts/ci/org.centos/stream/8/x86_64/configure
+index 6e8983f39c..1b0b9fcf3e 100755
+--- a/scripts/ci/org.centos/stream/8/x86_64/configure
++++ b/scripts/ci/org.centos/stream/8/x86_64/configure
+@@ -32,7 +32,7 @@
+ --with-git=meson \
+ --with-git-submodules=update \
+ --target-list="x86_64-softmmu" \
+---block-drv-rw-whitelist="qcow2,raw,file,host_device,nbd,iscsi,rbd,blkdebug,luks,null-co,nvme,copy-on-read,throttle,gluster" \
++--block-drv-rw-whitelist="qcow2,raw,file,host_device,nbd,iscsi,rbd,vitastor,blkdebug,luks,null-co,nvme,copy-on-read,throttle,gluster" \
+ --audio-drv-list="" \
+ --block-drv-ro-whitelist="vmdk,vhdx,vpc,https,ssh" \
+ --with-coroutine=ucontext \
+@@ -179,6 +179,7 @@
+ --enable-opengl \
+ --enable-pie \
+ --enable-rbd \
++--enable-vitastor \
+ --enable-rdma \
+ --enable-seccomp \
+ --enable-snappy \
+diff --git a/scripts/meson-buildoptions.sh b/scripts/meson-buildoptions.sh
+index 009fab1515..95914e6ebc 100644
+--- a/scripts/meson-buildoptions.sh
++++ b/scripts/meson-buildoptions.sh
+@@ -144,6 +144,7 @@ meson_options_help() {
+   printf "%s\n" '  qed             qed image format support'
+   printf "%s\n" '  qga-vss         build QGA VSS support (broken with MinGW)'
+   printf "%s\n" '  rbd             Ceph block device driver'
++  printf "%s\n" '  vitastor        Vitastor block device driver'
+   printf "%s\n" '  rdma            Enable RDMA-based migration'
+   printf "%s\n" '  replication     replication support'
+   printf "%s\n" '  sdl             SDL user interface'
+@@ -392,6 +393,8 @@ _meson_option_parse() {
+     --disable-qom-cast-debug) printf "%s" -Dqom_cast_debug=false ;;
+     --enable-rbd) printf "%s" -Drbd=enabled ;;
+     --disable-rbd) printf "%s" -Drbd=disabled ;;
++    --enable-vitastor) printf "%s" -Dvitastor=enabled ;;
++    --disable-vitastor) printf "%s" -Dvitastor=disabled ;;
+     --enable-rdma) printf "%s" -Drdma=enabled ;;
+     --disable-rdma) printf "%s" -Drdma=disabled ;;
+     --enable-replication) printf "%s" -Dreplication=enabled ;;

rpm/build-tarball.sh

@@ -24,4 +24,4 @@ rm fio
 mv fio-copy fio
 FIO=`rpm -qi fio | perl -e 'while(<>) { /^Epoch[\s:]+(\S+)/ && print "$1:"; /^Version[\s:]+(\S+)/ && print $1; /^Release[\s:]+(\S+)/ && print "-$1"; }'`
 perl -i -pe 's/(Requires:\s*fio)([^\n]+)?/$1 = '$FIO'/' $VITASTOR/rpm/vitastor-el$EL.spec
-tar --transform 's#^#vitastor-0.9.0/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.9.0$(rpm --eval '%dist').tar.gz *
+tar --transform 's#^#vitastor-0.9.3/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.9.3$(rpm --eval '%dist').tar.gz *

rpm/vitastor-el7.Dockerfile

@@ -35,7 +35,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-0.9.0.el7.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.9.3.el7.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el7.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el7.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        0.9.0
+Version:        0.9.3
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-0.9.0.el7.tar.gz
+Source0:        vitastor-0.9.3.el7.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

rpm/vitastor-el8.Dockerfile

@@ -35,7 +35,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-0.9.0.el8.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.9.3.el8.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el8.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el8.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        0.9.0
+Version:        0.9.3
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-0.9.0.el8.tar.gz
+Source0:        vitastor-0.9.3.el8.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

rpm/vitastor-el9.Dockerfile

@@ -18,7 +18,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-0.9.0.el9.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.9.3.el9.tar.gz ~/rpmbuild/SOURCES; \
     cp vitastor-el9.spec ~/rpmbuild/SPECS/vitastor.spec; \
     cd ~/rpmbuild/SPECS/; \
     rpmbuild -ba vitastor.spec; \

rpm/vitastor-el9.spec

@@ -1,11 +1,11 @@
 Name:           vitastor
-Version:        0.9.0
+Version:        0.9.3
 Release:        1%{?dist}
 Summary:        Vitastor, a fast software-defined clustered block storage
 
 License:        Vitastor Network Public License 1.1
 URL:            https://vitastor.io/
-Source0:        vitastor-0.9.0.el9.tar.gz
+Source0:        vitastor-0.9.3.el9.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

src/CMakeLists.txt

@@ -16,7 +16,7 @@ if("${CMAKE_INSTALL_PREFIX}" MATCHES "^/usr/local/?$")
 	set(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}")
 endif()
 
-add_definitions(-DVERSION="0.9.0")
+add_definitions(-DVERSION="0.9.3")
 add_definitions(-Wall -Wno-sign-compare -Wno-comment -Wno-parentheses -Wno-pointer-arith -fdiagnostics-color=always -I ${CMAKE_SOURCE_DIR}/src)
 if (${WITH_ASAN})
 	add_definitions(-fsanitize=address -fno-omit-frame-pointer)

src/allocator.cpp

@@ -143,34 +143,83 @@ uint64_t allocator::get_free_count()
     return free;
 }
 
+// FIXME: Move to utils?
 void bitmap_set(void *bitmap, uint64_t start, uint64_t len, uint64_t bitmap_granularity)
 {
-    if (start == 0)
+    if (start == 0 && len == 32*bitmap_granularity)
+        *((uint32_t*)bitmap) = UINT32_MAX;
+    else if (start == 0 && len == 64*bitmap_granularity)
+        *((uint64_t*)bitmap) = UINT64_MAX;
+    else
     {
-        if (len == 32*bitmap_granularity)
+        unsigned bit_start = start / bitmap_granularity;
+        unsigned bit_end = ((start + len) + bitmap_granularity - 1) / bitmap_granularity;
+        while (bit_start < bit_end)
         {
-            *((uint32_t*)bitmap) = UINT32_MAX;
-            return;
-        }
-        else if (len == 64*bitmap_granularity)
-        {
-            *((uint64_t*)bitmap) = UINT64_MAX;
-            return;
-        }
-    }
-    unsigned bit_start = start / bitmap_granularity;
-    unsigned bit_end = ((start + len) + bitmap_granularity - 1) / bitmap_granularity;
-    while (bit_start < bit_end)
-    {
-        if (!(bit_start & 7) && bit_end >= bit_start+8)
-        {
-            ((uint8_t*)bitmap)[bit_start / 8] = UINT8_MAX;
-            bit_start += 8;
-        }
-        else
-        {
-            ((uint8_t*)bitmap)[bit_start / 8] |= 1 << (bit_start % 8);
-            bit_start++;
+            if (!(bit_start & 7) && bit_end >= bit_start+8)
+            {
+                ((uint8_t*)bitmap)[bit_start / 8] = UINT8_MAX;
+                bit_start += 8;
+            }
+            else
+            {
+                ((uint8_t*)bitmap)[bit_start / 8] |= 1 << (bit_start % 8);
+                bit_start++;
+            }
         }
     }
 }
+
+void bitmap_clear(void *bitmap, uint64_t start, uint64_t len, uint64_t bitmap_granularity)
+{
+    if (start == 0 && len == 32*bitmap_granularity)
+        *((uint32_t*)bitmap) = 0;
+    else if (start == 0 && len == 64*bitmap_granularity)
+        *((uint64_t*)bitmap) = 0;
+    else
+    {
+        unsigned bit_start = start / bitmap_granularity;
+        unsigned bit_end = ((start + len) + bitmap_granularity - 1) / bitmap_granularity;
+        while (bit_start < bit_end)
+        {
+            if (!(bit_start & 7) && bit_end >= bit_start+8)
+            {
+                ((uint8_t*)bitmap)[bit_start / 8] = 0;
+                bit_start += 8;
+            }
+            else
+            {
+                ((uint8_t*)bitmap)[bit_start / 8] &= (0xFF ^ (1 << (bit_start % 8)));
+                bit_start++;
+            }
+        }
+    }
+}
+
+bool bitmap_check(void *bitmap, uint64_t start, uint64_t len, uint64_t bitmap_granularity)
+{
+    bool r = false;
+    if (start == 0 && len == 32*bitmap_granularity)
+        r = !!*((uint32_t*)bitmap);
+    else if (start == 0 && len == 64*bitmap_granularity)
+        r = !!*((uint64_t*)bitmap);
+    else
+    {
+        unsigned bit_start = start / bitmap_granularity;
+        unsigned bit_end = ((start + len) + bitmap_granularity - 1) / bitmap_granularity;
+        while (bit_start < bit_end)
+        {
+            if (!(bit_start & 7) && bit_end >= bit_start+8)
+            {
+                r = r || !!((uint8_t*)bitmap)[bit_start / 8];
+                bit_start += 8;
+            }
+            else
+            {
+                r = r || (((uint8_t*)bitmap)[bit_start / 8] & (1 << (bit_start % 8)));
+                bit_start++;
+            }
+        }
+    }
+    return r;
+}

src/allocator.h

@@ -23,3 +23,5 @@ public:
 };
 
 void bitmap_set(void *bitmap, uint64_t start, uint64_t len, uint64_t bitmap_granularity);
+void bitmap_clear(void *bitmap, uint64_t start, uint64_t len, uint64_t bitmap_granularity);
+bool bitmap_check(void *bitmap, uint64_t start, uint64_t len, uint64_t bitmap_granularity);

src/blockstore.h

@@ -77,6 +77,7 @@ Output:
   -EINVAL = invalid input parameters
   -ENOENT = requested object/version does not exist for reads
   -ENOSPC = no space left in the store for writes
+  -EDOM = checksum error.
 - version = the version actually read or written
 
 ## BS_OP_DELETE

src/blockstore_disk.cpp

@@ -40,10 +40,31 @@ void blockstore_disk_t::parse_config(std::map<std::string, std::string> & config
     data_block_size = parse_size(config["block_size"]);
     journal_device = config["journal_device"];
     journal_offset = parse_size(config["journal_offset"]);
-    disk_alignment = strtoull(config["disk_alignment"].c_str(), NULL, 10);
-    journal_block_size = strtoull(config["journal_block_size"].c_str(), NULL, 10);
-    meta_block_size = strtoull(config["meta_block_size"].c_str(), NULL, 10);
-    bitmap_granularity = strtoull(config["bitmap_granularity"].c_str(), NULL, 10);
+    disk_alignment = parse_size(config["disk_alignment"]);
+    journal_block_size = parse_size(config["journal_block_size"]);
+    meta_block_size = parse_size(config["meta_block_size"]);
+    bitmap_granularity = parse_size(config["bitmap_granularity"]);
+    meta_format = stoull_full(config["meta_format"]);
+    cached_read_data = config["cached_read_data"] == "true" || config["cached_read_data"] == "yes" || config["cached_read_data"] == "1";
+    cached_read_meta = cached_read_data && (meta_device == data_device || meta_device == "") &&
+        config.find("cached_read_meta") == config.end() ||
+        config["cached_read_meta"] == "true" || config["cached_read_meta"] == "yes" || config["cached_read_meta"] == "1";
+    cached_read_journal = cached_read_meta && (journal_device == meta_device || journal_device == "") &&
+        config.find("cached_read_journal") == config.end() ||
+        config["cached_read_journal"] == "true" || config["cached_read_journal"] == "yes" || config["cached_read_journal"] == "1";
+    if (config["data_csum_type"] == "crc32c")
+    {
+        data_csum_type = BLOCKSTORE_CSUM_CRC32C;
+    }
+    else if (config["data_csum_type"] == "" || config["data_csum_type"] == "none")
+    {
+        data_csum_type = BLOCKSTORE_CSUM_NONE;
+    }
+    else
+    {
+        throw std::runtime_error("data_csum_type="+config["data_csum_type"]+" is unsupported, only \"crc32c\" and \"none\" are supported");
+    }
+    csum_block_size = parse_size(config["csum_block_size"]);
     // Validate
     if (!data_block_size)
     {
@@ -91,7 +112,23 @@ void blockstore_disk_t::parse_config(std::map<std::string, std::string> & config
     }
     if (data_block_size % bitmap_granularity)
     {
-        throw std::runtime_error("Block size must be a multiple of sparse write tracking granularity");
+        throw std::runtime_error("Data block size must be a multiple of sparse write tracking granularity");
+    }
+    if (!data_csum_type)
+    {
+        csum_block_size = 0;
+    }
+    else if (!csum_block_size)
+    {
+        csum_block_size = bitmap_granularity;
+    }
+    if (csum_block_size && (csum_block_size % bitmap_granularity))
+    {
+        throw std::runtime_error("Checksum block size must be a multiple of sparse write tracking granularity");
+    }
+    if (csum_block_size && (data_block_size % csum_block_size))
+    {
+        throw std::runtime_error("Checksum block size must be a divisor of data block size");
     }
     if (meta_device == "")
     {
@@ -110,7 +147,9 @@ void blockstore_disk_t::parse_config(std::map<std::string, std::string> & config
         throw std::runtime_error("journal_offset must be a multiple of journal_block_size = "+std::to_string(journal_block_size));
     }
     clean_entry_bitmap_size = data_block_size / bitmap_granularity / 8;
-    clean_entry_size = sizeof(clean_disk_entry) + 2*clean_entry_bitmap_size;
+    clean_dyn_size = clean_entry_bitmap_size*2 + (csum_block_size
+        ? data_block_size/csum_block_size*(data_csum_type & 0xFF) : 0);
+    clean_entry_size = sizeof(clean_disk_entry) + clean_dyn_size + 4 /*entry_csum*/;
 }
 
 void blockstore_disk_t::calc_lengths(bool skip_meta_check)
@@ -160,6 +199,25 @@ void blockstore_disk_t::calc_lengths(bool skip_meta_check)
     // required metadata size
     block_count = data_len / data_block_size;
     meta_len = (1 + (block_count - 1 + meta_block_size / clean_entry_size) / (meta_block_size / clean_entry_size)) * meta_block_size;
+    if (meta_format == BLOCKSTORE_META_FORMAT_V1 ||
+        !meta_format && !skip_meta_check && meta_area_size < meta_len && !data_csum_type)
+    {
+        uint64_t clean_entry_v0_size = sizeof(clean_disk_entry) + 2*clean_entry_bitmap_size;
+        uint64_t meta_v0_len = (1 + (block_count - 1 + meta_block_size / clean_entry_v0_size)
+            / (meta_block_size / clean_entry_v0_size)) * meta_block_size;
+        if (meta_format == BLOCKSTORE_META_FORMAT_V1 || meta_area_size >= meta_v0_len)
+        {
+            // Old metadata fits.
+            printf("Warning: Using old metadata format without checksums because the new format doesn't fit into provided area\n");
+            clean_entry_size = clean_entry_v0_size;
+            meta_len = meta_v0_len;
+            meta_format = BLOCKSTORE_META_FORMAT_V1;
+        }
+        else
+            meta_format = BLOCKSTORE_META_FORMAT_V2;
+    }
+    else
+        meta_format = BLOCKSTORE_META_FORMAT_V2;
     if (!skip_meta_check && meta_area_size < meta_len)
     {
         throw std::runtime_error("Metadata area is too small, need at least "+std::to_string(meta_len)+" bytes");
@@ -237,6 +295,18 @@ void blockstore_disk_t::open_data()
     {
         throw std::runtime_error(std::string("Failed to lock data device: ") + strerror(errno));
     }
+    if (cached_read_data)
+    {
+        read_data_fd = open(data_device.c_str(), O_RDWR);
+        if (read_data_fd == -1)
+        {
+            throw std::runtime_error("Failed to open data device "+data_device+": "+std::string(strerror(errno)));
+        }
+    }
+    else
+    {
+        read_data_fd = data_fd;
+    }
 }
 
 void blockstore_disk_t::open_meta()
@@ -257,6 +327,18 @@ void blockstore_disk_t::open_meta()
         {
             throw std::runtime_error(std::string("Failed to lock metadata device: ") + strerror(errno));
         }
+        if (cached_read_meta)
+        {
+            read_meta_fd = open(meta_device.c_str(), O_RDWR);
+            if (read_meta_fd == -1)
+            {
+                throw std::runtime_error("Failed to open metadata device "+meta_device+": "+std::string(strerror(errno)));
+            }
+        }
+        else
+        {
+            read_meta_fd = meta_fd;
+        }
     }
     else
     {
@@ -275,6 +357,22 @@ void blockstore_disk_t::open_meta()
             ") is not a multiple of data device sector size ("+std::to_string(meta_device_sect)+")"
         );
     }
+    if (!cached_read_meta)
+    {
+        read_meta_fd = meta_fd;
+    }
+    else if (meta_device == data_device && cached_read_data)
+    {
+        read_meta_fd = read_data_fd;
+    }
+    else
+    {
+        read_meta_fd = open(meta_device.c_str(), O_RDWR);
+        if (read_meta_fd == -1)
+        {
+            throw std::runtime_error("Failed to open metadata device "+meta_device+": "+std::string(strerror(errno)));
+        }
+    }
 }
 
 void blockstore_disk_t::open_journal()
@@ -309,6 +407,26 @@ void blockstore_disk_t::open_journal()
             ") is not a multiple of journal device sector size ("+std::to_string(journal_device_sect)+")"
         );
     }
+    if (!cached_read_journal)
+    {
+        read_journal_fd = journal_fd;
+    }
+    else if (journal_device == meta_device && cached_read_meta)
+    {
+        read_journal_fd = read_meta_fd;
+    }
+    else if (journal_device == data_device && cached_read_data)
+    {
+        read_journal_fd = read_data_fd;
+    }
+    else
+    {
+        read_journal_fd = open(journal_device.c_str(), O_RDWR);
+        if (read_journal_fd == -1)
+        {
+            throw std::runtime_error("Failed to open journal device "+journal_device+": "+std::string(strerror(errno)));
+        }
+    }
 }
 
 void blockstore_disk_t::close_all()
@@ -319,5 +437,12 @@ void blockstore_disk_t::close_all()
         close(meta_fd);
     if (journal_fd >= 0 && journal_fd != meta_fd)
         close(journal_fd);
+    if (read_data_fd >= 0 && read_data_fd != data_fd)
+        close(read_data_fd);
+    if (read_meta_fd >= 0 && read_meta_fd != meta_fd)
+        close(read_meta_fd);
+    if (read_journal_fd >= 0 && read_journal_fd != journal_fd)
+        close(read_journal_fd);
     data_fd = meta_fd = journal_fd = -1;
+    read_data_fd = read_meta_fd = read_journal_fd = -1;
 }

src/blockstore_disk.h

@@ -8,6 +8,10 @@
 #include <string>
 #include <map>
 
+#define BLOCKSTORE_CSUM_NONE 0
+// Lower byte of checksum type is its length
+#define BLOCKSTORE_CSUM_CRC32C 0x104
+
 struct blockstore_disk_t
 {
     std::string data_device, meta_device, journal_device;
@@ -21,17 +25,24 @@ struct blockstore_disk_t
     uint64_t meta_block_size = 4096;
     // Sparse write tracking granularity. 4 KB is a good choice. Must be a multiple of disk_alignment
     uint64_t bitmap_granularity = 4096;
+    // Data checksum type, BLOCKSTORE_CSUM_NONE or BLOCKSTORE_CSUM_CRC32C
+    uint32_t data_csum_type = BLOCKSTORE_CSUM_NONE;
+    // Checksum block size, must be a multiple of bitmap_granularity
+    uint32_t csum_block_size = 4096;
     // By default, Blockstore locks all opened devices exclusively. This option can be used to disable locking
     bool disable_flock = false;
+    // Use linux page cache for reads. If enabled, separate buffered FDs will be opened for reading
+    bool cached_read_data = false, cached_read_meta = false, cached_read_journal = false;
 
     int meta_fd = -1, data_fd = -1, journal_fd = -1;
-    uint64_t meta_offset, meta_device_sect, meta_device_size, meta_len;
+    int read_meta_fd = -1, read_data_fd = -1, read_journal_fd = -1;
+    uint64_t meta_offset, meta_device_sect, meta_device_size, meta_len, meta_format = 0;
     uint64_t data_offset, data_device_sect, data_device_size, data_len;
     uint64_t journal_offset, journal_device_sect, journal_device_size, journal_len;
 
     uint32_t block_order;
     uint64_t block_count;
-    uint32_t clean_entry_bitmap_size = 0, clean_entry_size = 0;
+    uint32_t clean_entry_bitmap_size = 0, clean_entry_size = 0, clean_dyn_size = 0;
 
     void parse_config(std::map<std::string, std::string> & config);
     void open_data();
@@ -39,4 +50,13 @@ struct blockstore_disk_t
     void open_journal();
     void calc_lengths(bool skip_meta_check = false);
     void close_all();
+
+    inline uint64_t dirty_dyn_size(uint64_t offset, uint64_t len)
+    {
+        // Checksums may be partial if write is not aligned with csum_block_size
+        return clean_entry_bitmap_size + (csum_block_size && len > 0
+            ? ((offset+len+csum_block_size-1)/csum_block_size - offset/csum_block_size)
+                * (data_csum_type & 0xFF)
+            : 0);
+    }
 };

src/blockstore_flush.h

@@ -1,10 +1,22 @@
 // Copyright (c) Vitaliy Filippov, 2019+
 // License: VNPL-1.1 (see README.md for details)
 
+#define COPY_BUF_JOURNAL 1
+#define COPY_BUF_DATA 2
+#define COPY_BUF_ZERO 4
+#define COPY_BUF_CSUM_FILL 8
+#define COPY_BUF_COALESCED 16
+#define COPY_BUF_META_BLOCK 32
+#define COPY_BUF_JOURNALED_BIG 64
+
 struct copy_buffer_t
 {
-    uint64_t offset, len;
+    int copy_flags;
+    uint64_t offset, len, disk_offset;
+    uint64_t journal_sector; // only for reads: sector+1 if used and !journal.inmemory, otherwise 0
     void *buf;
+    uint8_t *csum_buf;
+    int *dyn_data;
 };
 
 struct meta_sector_t
@@ -37,7 +49,7 @@ class journal_flusher_co
 {
     blockstore_impl_t *bs;
     journal_flusher_t *flusher;
-    int wait_state, wait_count;
+    int wait_state, wait_count, wait_journal_count;
     struct io_uring_sqe *sqe;
     struct ring_data_t *data;
 
@@ -46,28 +58,39 @@ class journal_flusher_co
     obj_ver_id cur;
     std::map<obj_ver_id, dirty_entry>::iterator dirty_it, dirty_start, dirty_end;
     std::map<object_id, uint64_t>::iterator repeat_it;
-    std::function<void(ring_data_t*)> simple_callback_r, simple_callback_w;
+    std::function<void(ring_data_t*)> simple_callback_r, simple_callback_rj, simple_callback_w;
 
     bool skip_copy, has_delete, has_writes;
     std::vector<copy_buffer_t> v;
     std::vector<copy_buffer_t>::iterator it;
+    int i;
+    bool fill_incomplete, cleared_incomplete;
+    int read_to_fill_incomplete;
     int copy_count;
-    uint64_t clean_loc, old_clean_loc;
+    uint64_t clean_loc, clean_ver, old_clean_loc, old_clean_ver;
     flusher_meta_write_t meta_old, meta_new;
     bool clean_init_bitmap;
     uint64_t clean_bitmap_offset, clean_bitmap_len;
-    void *new_clean_bitmap;
+    uint8_t *clean_init_dyn_ptr;
+    uint8_t *new_clean_bitmap;
 
     uint64_t new_trim_pos;
 
-    // local: scan_dirty()
-    uint64_t offset, end_offset, submit_offset, submit_len;
-
     friend class journal_flusher_t;
-    bool scan_dirty(int wait_base);
+    void scan_dirty();
+    bool read_dirty(int wait_base);
+    bool modify_meta_do_reads(int wait_base);
+    bool wait_meta_reads(int wait_base);
     bool modify_meta_read(uint64_t meta_loc, flusher_meta_write_t &wr, int wait_base);
+    bool clear_incomplete_csum_block_bits(int wait_base);
+    void calc_block_checksums(uint32_t *new_data_csums, bool skip_overwrites);
+    void update_metadata_entry();
+    bool write_meta_block(flusher_meta_write_t & meta_block, int wait_base);
     void update_clean_db();
+    void free_data_blocks();
     bool fsync_batch(bool fsync_meta, int wait_base);
+    bool trim_journal(int wait_base);
+    void free_buffers();
 public:
     journal_flusher_co();
     bool loop();
@@ -95,14 +118,16 @@ class journal_flusher_t
 
     std::map<uint64_t, meta_sector_t> meta_sectors;
     std::deque<object_id> flush_queue;
-    std::map<object_id, uint64_t> flush_versions;
+    std::map<object_id, uint64_t> flush_versions; // FIXME: consider unordered_map?
 
     bool try_find_older(std::map<obj_ver_id, dirty_entry>::iterator & dirty_end, obj_ver_id & cur);
+    bool try_find_other(std::map<obj_ver_id, dirty_entry>::iterator & dirty_end, obj_ver_id & cur);
 
 public:
     journal_flusher_t(blockstore_impl_t *bs);
     ~journal_flusher_t();
     void loop();
+    bool is_trim_wanted() { return trim_wanted; }
     bool is_active();
     void mark_trim_possible();
     void request_trim();
@@ -111,4 +136,5 @@ public:
     void unshift_flush(obj_ver_id oid, bool force);
     void remove_flush(object_id oid);
     void dump_diagnostics();
+    bool is_mutated(uint64_t clean_loc);
 };

src/blockstore_impl.cpp

@@ -13,6 +13,7 @@ blockstore_impl_t::blockstore_impl_t(blockstore_config_t & config, ring_loop_t *
     initialized = 0;
     parse_config(config, true);
     zero_object = (uint8_t*)memalign_or_die(MEM_ALIGNMENT, dsk.data_block_size);
+    alloc_dyn_data = dsk.clean_dyn_size > sizeof(void*) || dsk.csum_block_size > 0;
     try
     {
         dsk.open_data();
@@ -38,8 +39,8 @@ blockstore_impl_t::~blockstore_impl_t()
     dsk.close_all();
     if (metadata_buffer)
         free(metadata_buffer);
-    if (clean_bitmap)
-        free(clean_bitmap);
+    if (clean_bitmaps)
+        free(clean_bitmaps);
 }
 
 bool blockstore_impl_t::is_started()

src/blockstore_impl.h

@@ -93,11 +93,10 @@
 
 // "VITAstor"
 #define BLOCKSTORE_META_MAGIC_V1 0x726F747341544956l
-#define BLOCKSTORE_META_VERSION_V1 1
+#define BLOCKSTORE_META_FORMAT_V1 1
+#define BLOCKSTORE_META_FORMAT_V2 2
 
 // metadata header (superblock)
-// FIXME: After adding the OSD superblock, add a key to metadata
-// and journal headers to check if they belong to the same OSD
 struct __attribute__((__packed__)) blockstore_meta_header_v1_t
 {
     uint64_t zero;
@@ -108,14 +107,29 @@ struct __attribute__((__packed__)) blockstore_meta_header_v1_t
     uint32_t bitmap_granularity;
 };
 
+struct __attribute__((__packed__)) blockstore_meta_header_v2_t
+{
+    uint64_t zero;
+    uint64_t magic;
+    uint64_t version;
+    uint32_t meta_block_size;
+    uint32_t data_block_size;
+    uint32_t bitmap_granularity;
+    uint32_t data_csum_type;
+    uint32_t csum_block_size;
+    uint32_t header_csum;
+};
+
 // 32 bytes = 24 bytes + block bitmap (4 bytes by default) + external attributes (also bitmap, 4 bytes by default)
 // per "clean" entry on disk with fixed metadata tables
-// FIXME: maybe add crc32's to metadata
 struct __attribute__((__packed__)) clean_disk_entry
 {
     object_id oid;
     uint64_t version;
     uint8_t bitmap[];
+    // Two more fields come after bitmap in metadata version 2:
+    // uint32_t data_csum[];
+    // uint32_t entry_csum;
 };
 
 // 32 = 16 + 16 bytes per "clean" entry in memory (object_id => clean_entry)
@@ -125,7 +139,7 @@ struct __attribute__((__packed__)) clean_entry
     uint64_t location;
 };
 
-// 64 = 24 + 40 bytes per dirty entry in memory (obj_ver_id => dirty_entry)
+// 64 = 24 + 40 bytes per dirty entry in memory (obj_ver_id => dirty_entry). Plus checksums
 struct __attribute__((__packed__)) dirty_entry
 {
     uint32_t state;
@@ -134,7 +148,7 @@ struct __attribute__((__packed__)) dirty_entry
     uint32_t offset;   // data offset within object (stripe)
     uint32_t len;      // data length
     uint64_t journal_sector; // journal sector used for this entry
-    void* bitmap;   // either external bitmap itself when it fits, or a pointer to it when it doesn't
+    void* dyn_data;    // dynamic data: external bitmap and data block checksums. may be a pointer to the in-memory journal
 };
 
 // - Sync must be submitted after previous writes/deletes (not before!)
@@ -163,12 +177,23 @@ struct __attribute__((__packed__)) dirty_entry
 // Suspend operation until there is some free space on the data device
 #define WAIT_FREE 5
 
-struct fulfill_read_t
+struct used_clean_obj_t
 {
-    uint64_t offset, len;
-    uint64_t journal_sector; // sector+1 if used and !journal.inmemory, otherwise 0
+    int refs;
+    bool was_freed; // was freed by a parallel flush?
+    bool was_changed; // was changed by a parallel flush?
 };
 
+// https://github.com/algorithm-ninja/cpp-btree
+// https://github.com/greg7mdp/sparsepp/ was used previously, but it was TERRIBLY slow after resizing
+// with sparsepp, random reads dropped to ~700 iops very fast with just as much as ~32k objects in the DB
+typedef btree::btree_map<object_id, clean_entry> blockstore_clean_db_t;
+typedef std::map<obj_ver_id, dirty_entry> blockstore_dirty_db_t;
+
+#include "blockstore_init.h"
+
+#include "blockstore_flush.h"
+
 #define PRIV(op) ((blockstore_op_private_t*)(op)->private_data)
 #define FINISH_OP(op) PRIV(op)->~blockstore_op_private_t(); std::function<void (blockstore_op_t*)>(op->callback)(op)
 
@@ -181,7 +206,8 @@ struct blockstore_op_private_t
     int op_state;
 
     // Read
-    std::vector<fulfill_read_t> read_vec;
+    uint64_t clean_block_used;
+    std::vector<copy_buffer_t> read_vec;
 
     // Sync, write
     int min_flushed_journal_sector, max_flushed_journal_sector;
@@ -197,16 +223,6 @@ struct blockstore_op_private_t
     int sync_small_checked, sync_big_checked;
 };
 
-// https://github.com/algorithm-ninja/cpp-btree
-// https://github.com/greg7mdp/sparsepp/ was used previously, but it was TERRIBLY slow after resizing
-// with sparsepp, random reads dropped to ~700 iops very fast with just as much as ~32k objects in the DB
-typedef btree::btree_map<object_id, clean_entry> blockstore_clean_db_t;
-typedef std::map<obj_ver_id, dirty_entry> blockstore_dirty_db_t;
-
-#include "blockstore_init.h"
-
-#include "blockstore_flush.h"
-
 typedef uint32_t pool_id_t;
 typedef uint64_t pool_pg_id_t;
 
@@ -253,7 +269,7 @@ class blockstore_impl_t
 
     std::map<pool_id_t, pool_shard_settings_t> clean_db_settings;
     std::map<pool_pg_id_t, blockstore_clean_db_t> clean_db_shards;
-    uint8_t *clean_bitmap = NULL;
+    uint8_t *clean_bitmaps = NULL;
     blockstore_dirty_db_t dirty_db;
     std::vector<blockstore_op_t*> submit_queue;
     std::vector<obj_ver_id> unsynced_big_writes, unsynced_small_writes;
@@ -267,6 +283,10 @@ class blockstore_impl_t
     journal_flusher_t *flusher;
     int big_to_flush = 0;
     int write_iodepth = 0;
+    bool alloc_dyn_data = false;
+
+    // clean data blocks referenced by read operations
+    std::map<uint64_t, used_clean_obj_t> used_clean_objects;
 
     bool live = false, queue_stall = false;
     ring_loop_t *ringloop;
@@ -310,8 +330,30 @@ class blockstore_impl_t
 
     // Read
     int dequeue_read(blockstore_op_t *read_op);
-    int fulfill_read(blockstore_op_t *read_op, uint64_t &fulfilled, uint32_t item_start, uint32_t item_end,
-        uint32_t item_state, uint64_t item_version, uint64_t item_location, uint64_t journal_sector);
+    void find_holes(std::vector<copy_buffer_t> & read_vec, uint32_t item_start, uint32_t item_end,
+        std::function<int(int, bool, uint32_t, uint32_t)> callback);
+    int fulfill_read(blockstore_op_t *read_op,
+        uint64_t &fulfilled, uint32_t item_start, uint32_t item_end,
+        uint32_t item_state, uint64_t item_version, uint64_t item_location,
+        uint64_t journal_sector, uint8_t *csum, int *dyn_data);
+    bool fulfill_clean_read(blockstore_op_t *read_op, uint64_t & fulfilled,
+        uint8_t *clean_entry_bitmap, int *dyn_data,
+        uint32_t item_start, uint32_t item_end, uint64_t clean_loc, uint64_t clean_ver);
+    int fill_partial_checksum_blocks(std::vector<copy_buffer_t> & rv, uint64_t & fulfilled,
+        uint8_t *clean_entry_bitmap, int *dyn_data, bool from_journal, uint8_t *read_buf, uint64_t read_offset, uint64_t read_end);
+    int pad_journal_read(std::vector<copy_buffer_t> & rv, copy_buffer_t & cp,
+        uint64_t dirty_offset, uint64_t dirty_end, uint64_t dirty_loc, uint8_t *csum_ptr, int *dyn_data,
+        uint64_t offset, uint64_t submit_len, uint64_t & blk_begin, uint64_t & blk_end, uint8_t* & blk_buf);
+    bool read_range_fulfilled(std::vector<copy_buffer_t> & rv, uint64_t & fulfilled, uint8_t *read_buf,
+        uint8_t *clean_entry_bitmap, uint32_t item_start, uint32_t item_end);
+    bool read_checksum_block(blockstore_op_t *op, int rv_pos, uint64_t &fulfilled, uint64_t clean_loc);
+    uint8_t* read_clean_meta_block(blockstore_op_t *read_op, uint64_t clean_loc, int rv_pos);
+    bool verify_padded_checksums(uint8_t *clean_entry_bitmap, uint8_t *csum_buf, uint32_t offset,
+        iovec *iov, int n_iov, std::function<void(uint32_t, uint32_t, uint32_t)> bad_block_cb);
+    bool verify_journal_checksums(uint8_t *csums, uint32_t offset,
+        iovec *iov, int n_iov, std::function<void(uint32_t, uint32_t, uint32_t)> bad_block_cb);
+    bool verify_clean_padded_checksums(blockstore_op_t *op, uint64_t clean_loc, uint8_t *dyn_data, bool from_journal,
+        iovec *iov, int n_iov, std::function<void(uint32_t, uint32_t, uint32_t)> bad_block_cb);
     int fulfill_read_push(blockstore_op_t *op, void *buf, uint64_t offset, uint64_t len,
         uint32_t item_state, uint64_t item_version);
     void handle_read_event(ring_data_t *data, blockstore_op_t *op);
@@ -342,6 +384,7 @@ class blockstore_impl_t
     int continue_rollback(blockstore_op_t *op);
     void mark_rolled_back(const obj_ver_id & ov);
     void erase_dirty(blockstore_dirty_db_t::iterator dirty_start, blockstore_dirty_db_t::iterator dirty_end, uint64_t clean_loc);
+    void free_dirty_dyn_data(dirty_entry & e);
 
     // List
     void process_list(blockstore_op_t *op);

src/blockstore_init.cpp

@@ -65,7 +65,7 @@ int blockstore_init_meta::loop()
     GET_SQE();
     data->iov = { metadata_buffer, bs->dsk.meta_block_size };
     data->callback = [this](ring_data_t *data) { handle_event(data, -1); };
-    my_uring_prep_readv(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset);
+    my_uring_prep_readv(sqe, bs->dsk.read_meta_fd, &data->iov, 1, bs->dsk.meta_offset);
     bs->ringloop->submit();
     submitted++;
 resume_1:
@@ -77,13 +77,20 @@ resume_1:
     if (iszero((uint64_t*)metadata_buffer, bs->dsk.meta_block_size / sizeof(uint64_t)))
     {
         {
-            blockstore_meta_header_v1_t *hdr = (blockstore_meta_header_v1_t *)metadata_buffer;
+            blockstore_meta_header_v2_t *hdr = (blockstore_meta_header_v2_t *)metadata_buffer;
             hdr->zero = 0;
             hdr->magic = BLOCKSTORE_META_MAGIC_V1;
-            hdr->version = BLOCKSTORE_META_VERSION_V1;
+            hdr->version = bs->dsk.meta_format;
             hdr->meta_block_size = bs->dsk.meta_block_size;
             hdr->data_block_size = bs->dsk.data_block_size;
             hdr->bitmap_granularity = bs->dsk.bitmap_granularity;
+            if (bs->dsk.meta_format >= BLOCKSTORE_META_FORMAT_V2)
+            {
+                hdr->data_csum_type = bs->dsk.data_csum_type;
+                hdr->csum_block_size = bs->dsk.csum_block_size;
+                hdr->header_csum = 0;
+                hdr->header_csum = crc32c(0, hdr, sizeof(*hdr));
+            }
         }
         if (bs->readonly)
         {
@@ -109,28 +116,62 @@ resume_1:
     }
     else
     {
-        blockstore_meta_header_v1_t *hdr = (blockstore_meta_header_v1_t *)metadata_buffer;
-        if (hdr->zero != 0 ||
-            hdr->magic != BLOCKSTORE_META_MAGIC_V1 ||
-            hdr->version != BLOCKSTORE_META_VERSION_V1)
+        blockstore_meta_header_v2_t *hdr = (blockstore_meta_header_v2_t *)metadata_buffer;
+        if (hdr->zero != 0 || hdr->magic != BLOCKSTORE_META_MAGIC_V1 || hdr->version < BLOCKSTORE_META_FORMAT_V1)
         {
             printf(
-                "Metadata is corrupt or old version.\n"
-                " If this is a new OSD please zero out the metadata area before starting it.\n"
-                " If you need to upgrade from 0.5.x please request it via the issue tracker.\n"
+                "Metadata is corrupt or too old (pre-0.6.x).\n"
+                " If this is a new OSD, please zero out the metadata area before starting it.\n"
+                " If you need to upgrade from 0.5.x, convert metadata with vitastor-disk.\n"
+            );
+            exit(1);
+        }
+        if (hdr->version == BLOCKSTORE_META_FORMAT_V2)
+        {
+            uint32_t csum = hdr->header_csum;
+            hdr->header_csum = 0;
+            if (crc32c(0, hdr, sizeof(*hdr)) != csum)
+            {
+                printf("Metadata header is corrupt (checksum mismatch).\n");
+                exit(1);
+            }
+            hdr->header_csum = csum;
+            bs->dsk.meta_format = BLOCKSTORE_META_FORMAT_V2;
+        }
+        else if (hdr->version == BLOCKSTORE_META_FORMAT_V1)
+        {
+            hdr->data_csum_type = 0;
+            hdr->csum_block_size = 0;
+            hdr->header_csum = 0;
+            // Enable compatibility mode - entries without checksums
+            bs->dsk.clean_entry_size = sizeof(clean_disk_entry) + bs->dsk.clean_entry_bitmap_size*2;
+            bs->dsk.meta_len = (1 + (bs->dsk.block_count - 1 + bs->dsk.meta_block_size / bs->dsk.clean_entry_size)
+                / (bs->dsk.meta_block_size / bs->dsk.clean_entry_size)) * bs->dsk.meta_block_size;
+            bs->dsk.meta_format = BLOCKSTORE_META_FORMAT_V1;
+            printf("Warning: Starting with metadata in the old format without checksums, as stored on disk\n");
+        }
+        else if (hdr->version > BLOCKSTORE_META_FORMAT_V2)
+        {
+            printf(
+                "Metadata format is too new for me (stored version is %lu, max supported %u).\n",
+                hdr->version, BLOCKSTORE_META_FORMAT_V2
             );
             exit(1);
         }
         if (hdr->meta_block_size != bs->dsk.meta_block_size ||
             hdr->data_block_size != bs->dsk.data_block_size ||
-            hdr->bitmap_granularity != bs->dsk.bitmap_granularity)
+            hdr->bitmap_granularity != bs->dsk.bitmap_granularity ||
+            hdr->data_csum_type != bs->dsk.data_csum_type ||
+            hdr->csum_block_size != bs->dsk.csum_block_size)
         {
             printf(
                 "Configuration stored in metadata superblock"
-                " (meta_block_size=%u, data_block_size=%u, bitmap_granularity=%u)"
-                " differs from OSD configuration (%lu/%u/%lu).\n",
+                " (meta_block_size=%u, data_block_size=%u, bitmap_granularity=%u, data_csum_type=%u, csum_block_size=%u)"
+                " differs from OSD configuration (%lu/%u/%lu, %u/%u).\n",
                 hdr->meta_block_size, hdr->data_block_size, hdr->bitmap_granularity,
-                bs->dsk.meta_block_size, bs->dsk.data_block_size, bs->dsk.bitmap_granularity
+                hdr->data_csum_type, hdr->csum_block_size,
+                bs->dsk.meta_block_size, bs->dsk.data_block_size, bs->dsk.bitmap_granularity,
+                bs->dsk.data_csum_type, bs->dsk.csum_block_size
             );
             exit(1);
         }
@@ -161,7 +202,7 @@ resume_2:
                 data->iov = { bufs[i].buf, bufs[i].size };
                 data->callback = [this, i](ring_data_t *data) { handle_event(data, i); };
                 if (!zero_on_init)
-                    my_uring_prep_readv(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + bufs[i].offset);
+                    my_uring_prep_readv(sqe, bs->dsk.read_meta_fd, &data->iov, 1, bs->dsk.meta_offset + bufs[i].offset);
                 else
                 {
                     // Fill metadata with zeroes
@@ -218,7 +259,7 @@ resume_2:
             GET_SQE();
             data->iov = { metadata_buffer, bs->dsk.meta_block_size };
             data->callback = [this](ring_data_t *data) { handle_event(data, -1); };
-            my_uring_prep_readv(sqe, bs->dsk.meta_fd, &data->iov, 1, bs->dsk.meta_offset + (1+next_offset)*bs->dsk.meta_block_size);
+            my_uring_prep_readv(sqe, bs->dsk.read_meta_fd, &data->iov, 1, bs->dsk.meta_offset + (1+next_offset)*bs->dsk.meta_block_size);
             submitted++;
 resume_5:
             if (submitted > 0)
@@ -279,12 +320,22 @@ bool blockstore_init_meta::handle_meta_block(uint8_t *buf, uint64_t entries_per_
     for (uint64_t i = 0; i < max_i; i++)
     {
         clean_disk_entry *entry = (clean_disk_entry*)(buf + i*bs->dsk.clean_entry_size);
-        if (!bs->inmemory_meta && bs->dsk.clean_entry_bitmap_size)
-        {
-            memcpy(bs->clean_bitmap + (done_cnt+i)*2*bs->dsk.clean_entry_bitmap_size, &entry->bitmap, 2*bs->dsk.clean_entry_bitmap_size);
-        }
         if (entry->oid.inode > 0)
         {
+            if (bs->dsk.meta_format >= BLOCKSTORE_META_FORMAT_V2)
+            {
+                // Check entry crc32
+                uint32_t *entry_csum = (uint32_t*)((uint8_t*)entry + bs->dsk.clean_entry_size - 4);
+                if (*entry_csum != crc32c(0, entry, bs->dsk.clean_entry_size - 4))
+                {
+                    printf("Metadata entry %lu is corrupt (checksum mismatch), skipping\n", done_cnt+i);
+                    continue;
+                }
+            }
+            if (!bs->inmemory_meta && bs->dsk.clean_entry_bitmap_size)
+            {
+                memcpy(bs->clean_bitmaps + (done_cnt+i) * 2 * bs->dsk.clean_entry_bitmap_size, &entry->bitmap, 2 * bs->dsk.clean_entry_bitmap_size);
+            }
             auto & clean_db = bs->clean_db_shard(entry->oid);
             auto clean_it = clean_db.find(entry->oid);
             if (clean_it == clean_db.end() || clean_it->second.version < entry->version)
@@ -416,7 +467,7 @@ int blockstore_init_journal::loop()
     data = ((ring_data_t*)sqe->user_data);
     data->iov = { submitted_buf, bs->journal.block_size };
     data->callback = simple_callback;
-    my_uring_prep_readv(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset);
+    my_uring_prep_readv(sqe, bs->dsk.read_journal_fd, &data->iov, 1, bs->journal.offset);
     bs->ringloop->submit();
     wait_count = 1;
 resume_1:
@@ -440,7 +491,9 @@ resume_1:
             .size = sizeof(journal_entry_start),
             .reserved = 0,
             .journal_start = bs->journal.block_size,
-            .version = JOURNAL_VERSION,
+            .version = JOURNAL_VERSION_V2,
+            .data_csum_type = bs->dsk.data_csum_type,
+            .csum_block_size = bs->dsk.csum_block_size,
         };
         ((journal_entry_start*)submitted_buf)->crc32 = je_crc32((journal_entry*)submitted_buf);
         if (bs->readonly)
@@ -492,18 +545,36 @@ resume_1:
         if (je_start->magic != JOURNAL_MAGIC ||
             je_start->type != JE_START ||
             je_crc32((journal_entry*)je_start) != je_start->crc32 ||
-            je_start->size != sizeof(journal_entry_start) && je_start->size != JE_START_LEGACY_SIZE)
+            je_start->size != JE_START_V0_SIZE && je_start->size != JE_START_V1_SIZE && je_start->size != JE_START_V2_SIZE)
         {
             // Entry is corrupt
-            fprintf(stderr, "First entry of the journal is corrupt\n");
+            fprintf(stderr, "First entry of the journal is corrupt or unsupported\n");
             exit(1);
         }
-        if (je_start->size == JE_START_LEGACY_SIZE || je_start->version != JOURNAL_VERSION)
+        if (je_start->size == JE_START_V0_SIZE ||
+            (je_start->version != JOURNAL_VERSION_V1 || je_start->size != JE_START_V1_SIZE) &&
+            (je_start->version != JOURNAL_VERSION_V2 || je_start->size != JE_START_V2_SIZE))
         {
             fprintf(
-                stderr, "The code only supports journal version %d, but it is %lu on disk."
-                    " Please use the previous version to flush the journal before upgrading OSD\n",
-                JOURNAL_VERSION, je_start->size == JE_START_LEGACY_SIZE ? 0 : je_start->version
+                stderr, "The code only supports journal versions 2 and 1, but it is %lu on disk."
+                    " Please use vitastor-disk to rewrite the journal\n",
+                je_start->size == JE_START_V0_SIZE ? 0 : je_start->version
+            );
+            exit(1);
+        }
+        if (je_start->version == JOURNAL_VERSION_V1)
+        {
+            je_start->data_csum_type = 0;
+            je_start->csum_block_size = 0;
+        }
+        if (je_start->data_csum_type != bs->dsk.data_csum_type ||
+            je_start->csum_block_size != bs->dsk.csum_block_size)
+        {
+            printf(
+                "Configuration stored in journal superblock (data_csum_type=%u, csum_block_size=%u)"
+                " differs from OSD configuration (%u/%u).\n",
+                je_start->data_csum_type, je_start->csum_block_size,
+                bs->dsk.data_csum_type, bs->dsk.csum_block_size
             );
             exit(1);
         }
@@ -536,7 +607,7 @@ resume_1:
                     end - journal_pos < JOURNAL_BUFFER_SIZE ? end - journal_pos : JOURNAL_BUFFER_SIZE,
                 };
                 data->callback = [this](ring_data_t *data1) { handle_event(data1); };
-                my_uring_prep_readv(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset + journal_pos);
+                my_uring_prep_readv(sqe, bs->dsk.read_journal_fd, &data->iov, 1, bs->journal.offset + journal_pos);
                 bs->ringloop->submit();
             }
             while (done.size() > 0)
@@ -705,11 +776,14 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     snprintf(err, 1024, "BUG: calculated journal data offset (%08lx) != stored journal data offset (%08lx)", location, je->small_write.data_offset);
                     throw std::runtime_error(err);
                 }
-                uint32_t data_crc32 = 0;
+                small_write_data.clear();
                 if (location >= done_pos && location+je->small_write.len <= done_pos+len)
                 {
                     // data is within this buffer
-                    data_crc32 = crc32c(0, (uint8_t*)buf + location - done_pos, je->small_write.len);
+                    small_write_data.push_back((iovec){
+                        .iov_base = (uint8_t*)buf + location - done_pos,
+                        .iov_len = je->small_write.len,
+                    });
                 }
                 else
                 {
@@ -724,7 +798,10 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                                 ? location+je->small_write.len : done[i].pos+done[i].len);
                             uint64_t part_begin = (location < done[i].pos ? done[i].pos : location);
                             covered += part_end - part_begin;
-                            data_crc32 = crc32c(data_crc32, (uint8_t*)done[i].buf + part_begin - done[i].pos, part_end - part_begin);
+                            small_write_data.push_back((iovec){
+                                .iov_base = (uint8_t*)done[i].buf + part_begin - done[i].pos,
+                                .iov_len = part_end - part_begin,
+                            });
                         }
                     }
                     if (covered < je->small_write.len)
@@ -734,12 +811,102 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         return 2;
                     }
                 }
-                if (data_crc32 != je->small_write.crc32_data)
+                bool data_csum_valid = true;
+                if (!bs->dsk.csum_block_size)
+                {
+                    uint32_t data_crc32 = 0;
+                    for (auto & sd: small_write_data)
+                    {
+                        data_crc32 = crc32c(data_crc32, sd.iov_base, sd.iov_len);
+                    }
+                    data_csum_valid = data_crc32 == je->small_write.crc32_data;
+                    if (!data_csum_valid)
+                    {
+                        printf(
+                            "Journal entry data is corrupt for small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u - data crc32 %x != %x\n",
+                            je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
+                            je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
+                            je->small_write.offset, je->small_write.len,
+                            data_crc32, je->small_write.crc32_data
+                        );
+                    }
+                }
+                else if (je->small_write.len > 0)
+                {
+                    // FIXME: deduplicate with disk_tool_journal.cpp
+                    // like in enqueue_write()
+                    uint32_t start = je->small_write.offset / bs->dsk.csum_block_size;
+                    uint32_t end = (je->small_write.offset+je->small_write.len-1) / bs->dsk.csum_block_size;
+                    uint32_t data_csum_size = (end-start+1) * (bs->dsk.data_csum_type & 0xFF);
+                    uint32_t required_size = sizeof(journal_entry_small_write) + bs->dsk.clean_entry_bitmap_size + data_csum_size;
+                    if (je->size != required_size)
+                    {
+                        printf(
+                            "Journal entry data has invalid size for small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u - should be %u bytes but is %u bytes\n",
+                            je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
+                            je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
+                            je->small_write.offset, je->small_write.len,
+                            required_size, je->size
+                        );
+                        data_csum_valid = false;
+                    }
+                    else
+                    {
+                        int sd_num = 0;
+                        size_t sd_pos = 0;
+                        uint32_t *block_csums = (uint32_t*)((uint8_t*)je + sizeof(journal_entry_small_write) + bs->dsk.clean_entry_bitmap_size);
+                        for (uint32_t pos = start; pos <= end; pos++, block_csums++)
+                        {
+                            size_t block_left = (pos == start
+                                ? (start == end
+                                    ? je->small_write.len
+                                    : bs->dsk.csum_block_size - je->small_write.offset%bs->dsk.csum_block_size)
+                                : (pos < end
+                                    ? bs->dsk.csum_block_size
+                                    : (je->small_write.offset + je->small_write.len)%bs->dsk.csum_block_size));
+                            if (pos > start && pos == end && block_left == 0)
+                            {
+                                // full last block
+                                block_left = bs->dsk.csum_block_size;
+                            }
+                            uint32_t block_crc32 = 0;
+                            while (block_left > 0)
+                            {
+                                assert(sd_num < small_write_data.size());
+                                if (small_write_data[sd_num].iov_len >= sd_pos+block_left)
+                                {
+                                    block_crc32 = crc32c(block_crc32, (uint8_t*)small_write_data[sd_num].iov_base+sd_pos, block_left);
+                                    sd_pos += block_left;
+                                    break;
+                                }
+                                else
+                                {
+                                    block_crc32 = crc32c(block_crc32, (uint8_t*)small_write_data[sd_num].iov_base+sd_pos, small_write_data[sd_num].iov_len-sd_pos);
+                                    block_left -= (small_write_data[sd_num].iov_len-sd_pos);
+                                    sd_pos = 0;
+                                    sd_num++;
+                                }
+                            }
+                            if (block_crc32 != *block_csums)
+                            {
+                                printf(
+                                    "Journal entry data is corrupt for small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u - block %u crc32 %x != %x\n",
+                                    je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
+                                    je->small_write.oid.inode, je->small_write.oid.stripe, je->small_write.version,
+                                    je->small_write.offset, je->small_write.len,
+                                    pos, block_crc32, *block_csums
+                                );
+                                data_csum_valid = false;
+                                break;
+                            }
+                        }
+                    }
+                }
+                if (!data_csum_valid)
                 {
                     // journal entry is corrupt, stop here
                     // interesting thing is that we must clear the corrupt entry if we're not readonly,
                     // because we don't write next entries in the same journal block
-                    printf("Journal entry data is corrupt (data crc32 %x != %x)\n", data_crc32, je->small_write.crc32_data);
                     memset((uint8_t*)buf + proc_pos - done_pos + pos, 0, bs->journal.block_size - pos);
                     bs->journal.next_free = prev_free;
                     init_write_buf = (uint8_t*)buf + proc_pos - done_pos;
@@ -755,11 +922,14 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         .oid = je->small_write.oid,
                         .version = je->small_write.version,
                     };
-                    void *bmp = NULL;
-                    void *bmp_from = (uint8_t*)je + sizeof(journal_entry_small_write);
-                    if (bs->dsk.clean_entry_bitmap_size <= sizeof(void*))
+                    uint64_t dyn_size = bs->dsk.dirty_dyn_size(je->small_write.offset, je->small_write.len);
+                    void *dyn = NULL;
+                    void *dyn_from = (uint8_t*)je + sizeof(journal_entry_small_write);
+                    if (!bs->alloc_dyn_data)
                     {
-                        memcpy(&bmp, bmp_from, bs->dsk.clean_entry_bitmap_size);
+                        // Bitmap without checksum is only 4 bytes for 128k objects, save it inline
+                        // It can even contain 4 byte bitmap + 4 byte CRC32 for 4 kb writes :)
+                        memcpy(&dyn, dyn_from, dyn_size);
                     }
                     else
                     {
@@ -767,8 +937,9 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         // allocations for entry bitmaps. This can only be fixed by using
                         // a patched map with dynamic entry size, but not the btree_map,
                         // because it doesn't keep iterators valid all the time.
-                        bmp = malloc_or_die(bs->dsk.clean_entry_bitmap_size);
-                        memcpy(bmp, bmp_from, bs->dsk.clean_entry_bitmap_size);
+                        dyn = malloc_or_die(dyn_size+sizeof(int));
+                        *((int*)dyn) = 1;
+                        memcpy((uint8_t*)dyn+sizeof(int), dyn_from, dyn_size);
                     }
                     bs->dirty_db.emplace(ov, (dirty_entry){
                         .state = (BS_ST_SMALL_WRITE | BS_ST_SYNCED),
@@ -777,7 +948,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         .offset = je->small_write.offset,
                         .len = je->small_write.len,
                         .journal_sector = proc_pos,
-                        .bitmap = bmp,
+                        .dyn_data = dyn,
                     });
                     bs->journal.used_sectors[proc_pos]++;
 #ifdef BLOCKSTORE_DEBUG
@@ -836,11 +1007,13 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         .oid = je->big_write.oid,
                         .version = je->big_write.version,
                     };
-                    void *bmp = NULL;
-                    void *bmp_from = (uint8_t*)je + sizeof(journal_entry_big_write);
-                    if (bs->dsk.clean_entry_bitmap_size <= sizeof(void*))
+                    uint64_t dyn_size = bs->dsk.dirty_dyn_size(je->big_write.offset, je->big_write.len);
+                    void *dyn = NULL;
+                    void *dyn_from = (uint8_t*)je + sizeof(journal_entry_big_write);
+                    if (!bs->alloc_dyn_data)
                     {
-                        memcpy(&bmp, bmp_from, bs->dsk.clean_entry_bitmap_size);
+                        // Bitmap without checksum is only 4 bytes for 128k objects, save it inline
+                        memcpy(&dyn, dyn_from, dyn_size);
                     }
                     else
                     {
@@ -848,8 +1021,9 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         // allocations for entry bitmaps. This can only be fixed by using
                         // a patched map with dynamic entry size, but not the btree_map,
                         // because it doesn't keep iterators valid all the time.
-                        bmp = malloc_or_die(bs->dsk.clean_entry_bitmap_size);
-                        memcpy(bmp, bmp_from, bs->dsk.clean_entry_bitmap_size);
+                        dyn = malloc_or_die(dyn_size+sizeof(int));
+                        *((int*)dyn) = 1;
+                        memcpy((uint8_t*)dyn+sizeof(int), dyn_from, dyn_size);
                     }
                     auto dirty_it = bs->dirty_db.emplace(ov, (dirty_entry){
                         .state = (BS_ST_BIG_WRITE | BS_ST_SYNCED),
@@ -858,7 +1032,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         .offset = je->big_write.offset,
                         .len = je->big_write.len,
                         .journal_sector = proc_pos,
-                        .bitmap = bmp,
+                        .dyn_data = dyn,
                     }).first;
                     if (bs->data_alloc->get(je->big_write.location >> bs->dsk.block_order))
                     {

src/blockstore_init.h

@@ -50,6 +50,7 @@ class blockstore_init_journal
     uint64_t next_free;
     std::vector<bs_init_journal_done> done;
     std::vector<obj_ver_id> double_allocs;
+    std::vector<iovec> small_write_data;
     uint64_t journal_pos = 0;
     uint64_t continue_pos = 0;
     void *init_write_buf = NULL;

src/blockstore_journal.cpp

@@ -17,6 +17,7 @@ blockstore_journal_check_t::blockstore_journal_check_t(blockstore_impl_t *bs)
 // Check if we can write <required> entries of <size> bytes and <data_after> data bytes after them to the journal
 int blockstore_journal_check_t::check_available(blockstore_op_t *op, int entries_required, int size, int data_after)
 {
+    uint64_t prev_next = next_sector;
     int required = entries_required;
     while (1)
     {
@@ -35,11 +36,19 @@ int blockstore_journal_check_t::check_available(blockstore_op_t *op, int entries
             }
             required -= fits;
             next_in_pos += fits * size;
-            sectors_to_write++;
+            if (next_sector != prev_next || !sectors_to_write)
+            {
+                // Except the previous call to this function
+                sectors_to_write++;
+            }
         }
         else if (bs->journal.sector_info[next_sector].dirty)
         {
-            sectors_to_write++;
+            if (next_sector != prev_next || !sectors_to_write)
+            {
+                // Except the previous call to this function
+                sectors_to_write++;
+            }
         }
         if (required <= 0)
         {
@@ -236,14 +245,6 @@ journal_t::~journal_t()
 uint64_t journal_t::get_trim_pos()
 {
     auto journal_used_it = used_sectors.lower_bound(used_start);
-#ifdef BLOCKSTORE_DEBUG
-    printf(
-        "Trimming journal (used_start=%08lx, next_free=%08lx, dirty_start=%08lx, new_start=%08lx, new_refcount=%ld)\n",
-        used_start, next_free, dirty_start,
-        journal_used_it == used_sectors.end() ? 0 : journal_used_it->first,
-        journal_used_it == used_sectors.end() ? 0 : journal_used_it->second
-    );
-#endif
     if (journal_used_it == used_sectors.end())
     {
         // Journal is cleared to its end, restart from the beginning
@@ -256,12 +257,26 @@ uint64_t journal_t::get_trim_pos()
         else
         {
             // next_free does not need updating during trim
+#ifdef BLOCKSTORE_DEBUG
+            printf(
+                "Trimming journal (used_start=%08lx, next_free=%08lx, dirty_start=%08lx, new_start=%08lx, new_refcount=%ld)\n",
+                used_start, next_free, dirty_start,
+                journal_used_it->first, journal_used_it->second
+            );
+#endif
             return journal_used_it->first;
         }
     }
     else if (journal_used_it->first > used_start)
     {
         // Journal is cleared up to <journal_used_it>
+#ifdef BLOCKSTORE_DEBUG
+        printf(
+            "Trimming journal (used_start=%08lx, next_free=%08lx, dirty_start=%08lx, new_start=%08lx, new_refcount=%ld)\n",
+            used_start, next_free, dirty_start,
+            journal_used_it->first, journal_used_it->second
+        );
+#endif
         return journal_used_it->first;
     }
     // Can't trim journal
@@ -283,3 +298,31 @@ void journal_t::dump_diagnostics()
         journal_used_it == used_sectors.end() ? 0 : journal_used_it->second
     );
 }
+
+static uint64_t zero_page[4096];
+
+uint32_t crc32c_pad(uint32_t prev_crc, const void *buf, size_t len, size_t left_pad, size_t right_pad)
+{
+    uint32_t r = prev_crc;
+    while (left_pad >= 4096)
+    {
+        r = crc32c(r, zero_page, 4096);
+        left_pad -= 4096;
+    }
+    if (left_pad > 0)
+        r = crc32c(r, zero_page, left_pad);
+    r = crc32c(r, buf, len);
+    while (right_pad >= 4096)
+    {
+        r = crc32c(r, zero_page, 4096);
+        right_pad -= 4096;
+    }
+    if (left_pad > 0)
+        r = crc32c(r, zero_page, right_pad);
+    return r;
+}
+
+uint32_t crc32c_nopad(uint32_t prev_crc, const void *buf, size_t len, size_t left_pad, size_t right_pad)
+{
+    return crc32c(0, buf, len);
+}

src/blockstore_journal.h

@@ -8,7 +8,8 @@
 
 #define MIN_JOURNAL_SIZE 4*1024*1024
 #define JOURNAL_MAGIC 0x4A33
-#define JOURNAL_VERSION 1
+#define JOURNAL_VERSION_V1 1
+#define JOURNAL_VERSION_V2 2
 #define JOURNAL_BUFFER_SIZE 4*1024*1024
 #define JOURNAL_ENTRY_HEADER_SIZE 16
 
@@ -32,7 +33,7 @@
 #define JE_BIG_WRITE_INSTANT   0x08
 #define JE_MAX         0x08
 
-// crc32c comes first to ease calculation and is equal to crc32()
+// crc32c comes first to ease calculation
 struct __attribute__((__packed__)) journal_entry_start
 {
     uint32_t crc32;
@@ -42,8 +43,12 @@ struct __attribute__((__packed__)) journal_entry_start
     uint32_t reserved;
     uint64_t journal_start;
     uint64_t version;
+    uint32_t data_csum_type;
+    uint32_t csum_block_size;
 };
-#define JE_START_LEGACY_SIZE 24
+#define JE_START_V0_SIZE 24
+#define JE_START_V1_SIZE 32
+#define JE_START_V2_SIZE 40
 
 struct __attribute__((__packed__)) journal_entry_small_write
 {
@@ -59,10 +64,12 @@ struct __attribute__((__packed__)) journal_entry_small_write
     // small_write entries contain <len> bytes of data which is stored in next sectors
     // data_offset is its offset within journal
     uint64_t data_offset;
-    uint32_t crc32_data;
+    uint32_t crc32_data; // zero when data_csum_type != 0
     // small_write and big_write entries are followed by the "external" bitmap
     // its size is dynamic and included in journal entry's <size> field
     uint8_t bitmap[];
+    // and then data checksums if data_csum_type != 0
+    // uint32_t data_crc32c[];
 };
 
 struct __attribute__((__packed__)) journal_entry_big_write
@@ -80,6 +87,8 @@ struct __attribute__((__packed__)) journal_entry_big_write
     // small_write and big_write entries are followed by the "external" bitmap
     // its size is dynamic and included in journal entry's <size> field
     uint8_t bitmap[];
+    // and then data checksums if data_csum_type != 0
+    // uint32_t data_crc32c[];
 };
 
 struct __attribute__((__packed__)) journal_entry_stable
@@ -218,3 +227,6 @@ struct blockstore_journal_check_t
 };
 
 journal_entry* prefill_single_journal_entry(journal_t & journal, uint16_t type, uint32_t size);
+
+uint32_t crc32c_pad(uint32_t prev_crc, const void *buf, size_t len, size_t left_pad, size_t right_pad);
+uint32_t crc32c_nopad(uint32_t prev_crc, const void *buf, size_t len, size_t left_pad, size_t right_pad);

src/blockstore_open.cpp

@@ -133,19 +133,24 @@ void blockstore_impl_t::calc_lengths()
     {
         metadata_buffer = memalign(MEM_ALIGNMENT, dsk.meta_len);
         if (!metadata_buffer)
-            throw std::runtime_error("Failed to allocate memory for the metadata");
+            throw std::runtime_error("Failed to allocate memory for the metadata ("+std::to_string(dsk.meta_len/1024/1024)+" MB)");
     }
-    else if (dsk.clean_entry_bitmap_size)
+    else if (dsk.clean_entry_bitmap_size || dsk.data_csum_type)
     {
-        clean_bitmap = (uint8_t*)malloc(dsk.block_count * 2*dsk.clean_entry_bitmap_size);
-        if (!clean_bitmap)
-            throw std::runtime_error("Failed to allocate memory for the metadata sparse write bitmap");
+        clean_bitmaps = (uint8_t*)malloc(dsk.block_count * 2 * dsk.clean_entry_bitmap_size);
+        if (!clean_bitmaps)
+        {
+            throw std::runtime_error(
+                "Failed to allocate memory for the metadata sparse write bitmap ("+
+                std::to_string(dsk.block_count * 2 * dsk.clean_entry_bitmap_size / 1024 / 1024)+" MB)"
+            );
+        }
     }
     if (journal.inmemory)
     {
         journal.buffer = memalign(MEM_ALIGNMENT, journal.len);
         if (!journal.buffer)
-            throw std::runtime_error("Failed to allocate memory for journal");
+            throw std::runtime_error("Failed to allocate memory for journal ("+std::to_string(journal.len/1024/1024)+" MB)");
     }
     else
     {

src/blockstore_rollback.cpp

@@ -218,7 +218,7 @@ void blockstore_impl_t::erase_dirty(blockstore_dirty_db_t::iterator dirty_start,
         auto used = --journal.used_sectors[dirty_it->second.journal_sector];
 #ifdef BLOCKSTORE_DEBUG
         printf(
-            "remove usage of journal offset %08lx by %lx:%lx v%lu (%d refs)\n", dirty_it->second.journal_sector,
+            "remove usage of journal offset %08lx by %lx:%lx v%lu (%lu refs)\n", dirty_it->second.journal_sector,
             dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version, used
         );
 #endif
@@ -227,11 +227,7 @@ void blockstore_impl_t::erase_dirty(blockstore_dirty_db_t::iterator dirty_start,
             journal.used_sectors.erase(dirty_it->second.journal_sector);
             flusher->mark_trim_possible();
         }
-        if (dsk.clean_entry_bitmap_size > sizeof(void*))
-        {
-            free(dirty_it->second.bitmap);
-            dirty_it->second.bitmap = NULL;
-        }
+        free_dirty_dyn_data(dirty_it->second);
         if (dirty_it == dirty_start)
         {
             break;
@@ -240,3 +236,18 @@ void blockstore_impl_t::erase_dirty(blockstore_dirty_db_t::iterator dirty_start,
     }
     dirty_db.erase(dirty_start, dirty_end);
 }
+
+void blockstore_impl_t::free_dirty_dyn_data(dirty_entry & e)
+{
+    if (e.dyn_data)
+    {
+        if (alloc_dyn_data &&
+            --*((int*)e.dyn_data) == 0) // refcount
+        {
+            // dyn_data contains the bitmap and checksums
+            // free it if it doesn't refer to the in-memory journal
+            free(e.dyn_data);
+        }
+        e.dyn_data = NULL;
+    }
+}

src/blockstore_stable.cpp

@@ -458,6 +458,16 @@ void blockstore_impl_t::mark_stable(const obj_ver_id & v, bool forget_dirty)
                     big_to_flush++;
                 }
             }
+            else if (IS_IN_FLIGHT(dirty_it->second.state))
+            {
+                // mark_stable should never be called for in-flight or submitted writes
+                printf(
+                    "BUG: Attempt to mark_stable object %lx:%lx v%lu state of which is %x\n",
+                    dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version,
+                    dirty_it->second.state
+                );
+                exit(1);
+            }
             if (forget_dirty && (IS_BIG_WRITE(dirty_it->second.state) ||
                 IS_DELETE(dirty_it->second.state)))
             {

src/blockstore_sync.cpp

@@ -78,7 +78,23 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
         // 2nd step: Data device is synced, prepare & write journal entries
         // Check space in the journal and journal memory buffers
         blockstore_journal_check_t space_check(this);
-        if (!space_check.check_available(op, PRIV(op)->sync_big_writes.size(),
+        if (dsk.csum_block_size)
+        {
+            // More complex check because all journal entries have different lengths
+            int left = PRIV(op)->sync_big_writes.size();
+            for (auto & sbw: PRIV(op)->sync_big_writes)
+            {
+                left--;
+                auto & dirty_entry = dirty_db.at(sbw);
+                uint64_t dyn_size = dsk.dirty_dyn_size(dirty_entry.offset, dirty_entry.len);
+                if (!space_check.check_available(op, 1, sizeof(journal_entry_big_write) + dyn_size,
+                    left == 0 ? JOURNAL_STABILIZE_RESERVATION : 0))
+                {
+                    return 0;
+                }
+            }
+        }
+        else if (!space_check.check_available(op, PRIV(op)->sync_big_writes.size(),
             sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size, JOURNAL_STABILIZE_RESERVATION))
         {
             return 0;
@@ -90,16 +106,17 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
         int s = 0;
         while (it != PRIV(op)->sync_big_writes.end())
         {
-            if (!journal.entry_fits(sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size) &&
+            auto & dirty_entry = dirty_db.at(*it);
+            uint64_t dyn_size = dsk.dirty_dyn_size(dirty_entry.offset, dirty_entry.len);
+            if (!journal.entry_fits(sizeof(journal_entry_big_write) + dyn_size) &&
                 journal.sector_info[journal.cur_sector].dirty)
             {
                 prepare_journal_sector_write(journal.cur_sector, op);
                 s++;
             }
-            auto & dirty_entry = dirty_db.at(*it);
             journal_entry_big_write *je = (journal_entry_big_write*)prefill_single_journal_entry(
                 journal, (dirty_entry.state & BS_ST_INSTANT) ? JE_BIG_WRITE_INSTANT : JE_BIG_WRITE,
-                sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size
+                sizeof(journal_entry_big_write) + dyn_size
             );
             dirty_entry.journal_sector = journal.sector_info[journal.cur_sector].offset;
             journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
@@ -115,8 +132,8 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
             je->offset = dirty_entry.offset;
             je->len = dirty_entry.len;
             je->location = dirty_entry.location;
-            memcpy((void*)(je+1), (dsk.clean_entry_bitmap_size > sizeof(void*)
-                ? dirty_entry.bitmap : &dirty_entry.bitmap), dsk.clean_entry_bitmap_size);
+            memcpy((void*)(je+1), (alloc_dyn_data
+                ? (uint8_t*)dirty_entry.dyn_data+sizeof(int) : (uint8_t*)&dirty_entry.dyn_data), dyn_size);
             je->crc32 = je_crc32((journal_entry*)je);
             journal.crc32_last = je->crc32;
             it++;

src/blockstore_write.cpp

@@ -8,12 +8,21 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
     // Check or assign version number
     bool found = false, deleted = false, unsynced = false, is_del = (op->opcode == BS_OP_DELETE);
     bool wait_big = false, wait_del = false;
-    void *bmp = NULL;
-    uint64_t version = 1;
-    if (!is_del && dsk.clean_entry_bitmap_size > sizeof(void*))
+    void *dyn = NULL;
+    if (is_del)
     {
-        bmp = calloc_or_die(1, dsk.clean_entry_bitmap_size);
+        op->len = 0;
     }
+    size_t dyn_size = dsk.dirty_dyn_size(op->offset, op->len);
+    if (!is_del && alloc_dyn_data)
+    {
+        // FIXME: Working with `dyn_data` has to be refactored somehow but I first have to decide how :)
+        // +sizeof(int) = refcount
+        dyn = calloc_or_die(1, dyn_size+sizeof(int));
+        *((int*)dyn) = 1;
+    }
+    uint8_t *dyn_ptr = (uint8_t*)(alloc_dyn_data ? dyn+sizeof(int) : &dyn);
+    uint64_t version = 1;
     if (dirty_db.size() > 0)
     {
         auto dirty_it = dirty_db.upper_bound((obj_ver_id){
@@ -33,10 +42,9 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
                 : ((dirty_it->second.state & BS_ST_WORKFLOW_MASK) == BS_ST_WAIT_BIG);
             if (!is_del && !deleted)
             {
-                if (dsk.clean_entry_bitmap_size > sizeof(void*))
-                    memcpy(bmp, dirty_it->second.bitmap, dsk.clean_entry_bitmap_size);
-                else
-                    bmp = dirty_it->second.bitmap;
+                void *dyn_from = alloc_dyn_data
+                    ? (uint8_t*)dirty_it->second.dyn_data + sizeof(int) : (uint8_t*)&dirty_it->second.dyn_data;
+                memcpy(dyn_ptr, dyn_from, dsk.clean_entry_bitmap_size);
             }
         }
     }
@@ -50,7 +58,7 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
             if (!is_del)
             {
                 void *bmp_ptr = get_clean_entry_bitmap(clean_it->second.location, dsk.clean_entry_bitmap_size);
-                memcpy((dsk.clean_entry_bitmap_size > sizeof(void*) ? bmp : &bmp), bmp_ptr, dsk.clean_entry_bitmap_size);
+                memcpy(dyn_ptr, bmp_ptr, dsk.clean_entry_bitmap_size);
             }
         }
         else
@@ -112,9 +120,9 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
             printf("Write %lx:%lx v%lu requested, but we already have v%lu\n", op->oid.inode, op->oid.stripe, op->version, version);
 #endif
             op->retval = -EEXIST;
-            if (!is_del && dsk.clean_entry_bitmap_size > sizeof(void*))
+            if (!is_del && alloc_dyn_data)
             {
-                free(bmp);
+                free(dyn);
             }
             return false;
         }
@@ -158,26 +166,50 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
         if (op->bitmap)
         {
             // Only allow to overwrite part of the object bitmap respective to the write's offset/len
-            uint8_t *bmp_ptr = (uint8_t*)(dsk.clean_entry_bitmap_size > sizeof(void*) ? bmp : &bmp);
             uint32_t bit = op->offset/dsk.bitmap_granularity;
             uint32_t bits_left = op->len/dsk.bitmap_granularity;
             while (!(bit % 8) && bits_left >= 8)
             {
                 // Copy bytes
-                bmp_ptr[bit/8] = ((uint8_t*)op->bitmap)[bit/8];
+                dyn_ptr[bit/8] = ((uint8_t*)op->bitmap)[bit/8];
                 bit += 8;
                 bits_left -= 8;
             }
             while (bits_left > 0)
             {
                 // Copy bits
-                bmp_ptr[bit/8] = (bmp_ptr[bit/8] & ~(1 << (bit%8)))
+                dyn_ptr[bit/8] = (dyn_ptr[bit/8] & ~(1 << (bit%8)))
                     | (((uint8_t*)op->bitmap)[bit/8] & (1 << bit%8));
                 bit++;
                 bits_left--;
             }
         }
     }
+    // Calculate checksums
+    // FIXME: Allow to receive checksums from outside?
+    if (!is_del && dsk.data_csum_type && op->len > 0)
+    {
+        uint32_t *data_csums = (uint32_t*)(dyn_ptr + dsk.clean_entry_bitmap_size);
+        uint32_t start = op->offset / dsk.csum_block_size;
+        uint32_t end = (op->offset+op->len-1) / dsk.csum_block_size;
+        auto fn = state & BS_ST_BIG_WRITE ? crc32c_pad : crc32c_nopad;
+        if (start == end)
+            data_csums[0] = fn(0, op->buf, op->len, op->offset - start*dsk.csum_block_size, end*dsk.csum_block_size - (op->offset+op->len));
+        else
+        {
+            // First block
+            data_csums[0] = fn(0, op->buf, dsk.csum_block_size*(start+1)-op->offset, op->offset - start*dsk.csum_block_size, 0);
+            // Intermediate blocks
+            for (uint32_t i = start+1; i < end; i++)
+                data_csums[i-start] = crc32c(0, (uint8_t*)op->buf + dsk.csum_block_size*i-op->offset, dsk.csum_block_size);
+            // Last block
+            data_csums[end-start] = fn(
+                0, (uint8_t*)op->buf + end*dsk.csum_block_size - op->offset,
+                op->offset+op->len - end*dsk.csum_block_size,
+                0, (end+1)*dsk.csum_block_size - (op->offset+op->len)
+            );
+        }
+    }
     dirty_db.emplace((obj_ver_id){
         .oid = op->oid,
         .version = op->version,
@@ -188,7 +220,7 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
         .offset = is_del ? 0 : op->offset,
         .len = is_del ? 0 : op->len,
         .journal_sector = 0,
-        .bitmap = bmp,
+        .dyn_data = dyn,
     });
     return true;
 }
@@ -197,8 +229,7 @@ void blockstore_impl_t::cancel_all_writes(blockstore_op_t *op, blockstore_dirty_
 {
     while (dirty_it != dirty_db.end() && dirty_it->first.oid == op->oid)
     {
-        if (dsk.clean_entry_bitmap_size > sizeof(void*))
-            free(dirty_it->second.bitmap);
+        free_dirty_dyn_data(dirty_it->second);
         dirty_db.erase(dirty_it++);
     }
     bool found = false;
@@ -280,7 +311,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
     {
         blockstore_journal_check_t space_check(this);
         if (!space_check.check_available(op, unsynced_big_write_count + 1,
-            sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size,
+            sizeof(journal_entry_big_write) + dsk.clean_dyn_size,
             (dirty_it->second.state & BS_ST_INSTANT) ? JOURNAL_INSTANT_RESERVATION : JOURNAL_STABILIZE_RESERVATION))
         {
             return 0;
@@ -363,12 +394,13 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
     {
         // Small (journaled) write
         // First check if the journal has sufficient space
+        uint64_t dyn_size = dsk.dirty_dyn_size(op->offset, op->len);
         blockstore_journal_check_t space_check(this);
         if (unsynced_big_write_count &&
             !space_check.check_available(op, unsynced_big_write_count,
-                sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size, 0)
+                sizeof(journal_entry_big_write) + dsk.clean_dyn_size, 0)
             || !space_check.check_available(op, 1,
-                sizeof(journal_entry_small_write) + dsk.clean_entry_bitmap_size,
+                sizeof(journal_entry_small_write) + dyn_size,
                 op->len + ((dirty_it->second.state & BS_ST_INSTANT) ? JOURNAL_INSTANT_RESERVATION : JOURNAL_STABILIZE_RESERVATION)))
         {
             return 0;
@@ -377,7 +409,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         BS_SUBMIT_CHECK_SQES(
             // Write current journal sector only if it's dirty and full, or in the immediate_commit mode
             (immediate_commit != IMMEDIATE_NONE ||
-                !journal.entry_fits(sizeof(journal_entry_small_write) + dsk.clean_entry_bitmap_size) ? 1 : 0) +
+                !journal.entry_fits(sizeof(journal_entry_small_write) + dyn_size) ? 1 : 0) +
             (op->len > 0 ? 1 : 0)
         );
         write_iodepth++;
@@ -385,7 +417,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         auto cb = [this, op](ring_data_t *data) { handle_write_event(data, op); };
         if (immediate_commit == IMMEDIATE_NONE)
         {
-            if (!journal.entry_fits(sizeof(journal_entry_small_write) + dsk.clean_entry_bitmap_size))
+            if (!journal.entry_fits(sizeof(journal_entry_small_write) + dyn_size))
             {
                 prepare_journal_sector_write(journal.cur_sector, op);
             }
@@ -397,7 +429,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         // Then pre-fill journal entry
         journal_entry_small_write *je = (journal_entry_small_write*)prefill_single_journal_entry(
             journal, op->opcode == BS_OP_WRITE_STABLE ? JE_SMALL_WRITE_INSTANT : JE_SMALL_WRITE,
-            sizeof(journal_entry_small_write) + dsk.clean_entry_bitmap_size
+            sizeof(journal_entry_small_write) + dyn_size
         );
         dirty_it->second.journal_sector = journal.sector_info[journal.cur_sector].offset;
         journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
@@ -409,14 +441,31 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         );
 #endif
         // Figure out where data will be
-        journal.next_free = (journal.next_free + op->len) <= journal.len ? journal.next_free : dsk.journal_block_size;
+        auto next_next_free = (journal.next_free + op->len) <= journal.len ? journal.next_free : dsk.journal_block_size;
+        if (op->len > 0)
+        {
+            auto journal_used_it = journal.used_sectors.lower_bound(next_next_free);
+            if (journal_used_it != journal.used_sectors.end() &&
+                journal_used_it->first < next_next_free + op->len)
+            {
+                printf(
+                    "BUG: Attempt to overwrite used offset (%lx, %lu refs) of the journal with the object %lx:%lx v%lu: data at %lx, len %x!"
+                    " Journal used_start=%08lx (%lu refs), next_free=%08lx, dirty_start=%08lx\n",
+                    journal_used_it->first, journal_used_it->second, op->oid.inode, op->oid.stripe, op->version, next_next_free, op->len,
+                    journal.used_start, journal.used_sectors[journal.used_start], journal.next_free, journal.dirty_start
+                );
+                exit(1);
+            }
+        }
+        journal.next_free = next_next_free;
         je->oid = op->oid;
         je->version = op->version;
         je->offset = op->offset;
         je->len = op->len;
         je->data_offset = journal.next_free;
-        je->crc32_data = crc32c(0, op->buf, op->len);
-        memcpy((void*)(je+1), (dsk.clean_entry_bitmap_size > sizeof(void*) ? dirty_it->second.bitmap : &dirty_it->second.bitmap), dsk.clean_entry_bitmap_size);
+        je->crc32_data = dsk.csum_block_size ? 0 : crc32c(0, op->buf, op->len);
+        memcpy((void*)(je+1), (alloc_dyn_data
+            ? (uint8_t*)dirty_it->second.dyn_data+sizeof(int) : (uint8_t*)&dirty_it->second.dyn_data), dyn_size);
         je->crc32 = je_crc32((journal_entry*)je);
         journal.crc32_last = je->crc32;
         if (immediate_commit != IMMEDIATE_NONE)
@@ -485,9 +534,9 @@ resume_2:
             .version = op->version,
         });
         assert(dirty_it != dirty_db.end());
+        uint64_t dyn_size = dsk.dirty_dyn_size(op->offset, op->len);
         blockstore_journal_check_t space_check(this);
-        if (!space_check.check_available(op, 1,
-            sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size,
+        if (!space_check.check_available(op, 1, sizeof(journal_entry_big_write) + dyn_size,
             ((dirty_it->second.state & BS_ST_INSTANT) ? JOURNAL_INSTANT_RESERVATION : JOURNAL_STABILIZE_RESERVATION)))
         {
             return 0;
@@ -495,7 +544,7 @@ resume_2:
         BS_SUBMIT_CHECK_SQES(1);
         journal_entry_big_write *je = (journal_entry_big_write*)prefill_single_journal_entry(
             journal, op->opcode == BS_OP_WRITE_STABLE ? JE_BIG_WRITE_INSTANT : JE_BIG_WRITE,
-            sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size
+            sizeof(journal_entry_big_write) + dyn_size
         );
         dirty_it->second.journal_sector = journal.sector_info[journal.cur_sector].offset;
         journal.used_sectors[journal.sector_info[journal.cur_sector].offset]++;
@@ -511,7 +560,8 @@ resume_2:
         je->offset = op->offset;
         je->len = op->len;
         je->location = dirty_it->second.location;
-        memcpy((void*)(je+1), (dsk.clean_entry_bitmap_size > sizeof(void*) ? dirty_it->second.bitmap : &dirty_it->second.bitmap), dsk.clean_entry_bitmap_size);
+        memcpy((void*)(je+1), (alloc_dyn_data
+            ? (uint8_t*)dirty_it->second.dyn_data+sizeof(int) : (uint8_t*)&dirty_it->second.dyn_data), dyn_size);
         je->crc32 = je_crc32((journal_entry*)je);
         journal.crc32_last = je->crc32;
         prepare_journal_sector_write(journal.cur_sector, op);
@@ -645,8 +695,13 @@ void blockstore_impl_t::release_journal_sectors(blockstore_op_t *op)
         uint64_t s = PRIV(op)->min_flushed_journal_sector;
         while (1)
         {
-            if (s != (1+journal.cur_sector) && journal.sector_info[s-1].flush_count == 0)
+            if (!journal.sector_info[s-1].dirty && journal.sector_info[s-1].flush_count == 0)
             {
+                if (s == (1+journal.cur_sector))
+                {
+                    // Forcibly move to the next sector and move dirty position
+                    journal.in_sector_pos = journal.block_size;
+                }
                 // We know for sure that we won't write into this sector anymore
                 uint64_t new_ds = journal.sector_info[s-1].offset + journal.block_size;
                 if (new_ds >= journal.len)

src/cli_merge.cpp

@@ -41,7 +41,7 @@ struct snap_merger_t
     int fsync_interval = 128;
 
     // -- STATE --
-    inode_t target;
+    inode_t target, to_num;
     int target_rank;
     bool inside_continue = false;
     int state = 0;
@@ -98,6 +98,7 @@ struct snap_merger_t
             state = 100;
             return;
         }
+        to_num = to_cfg->num;
         // Check that to_cfg is actually a child of from_cfg and target_cfg is somewhere between them
         std::vector<inode_t> chain_list;
         inode_config_t *cur = to_cfg;
@@ -451,7 +452,7 @@ struct snap_merger_t
     {
         cluster_op_t *op = &rwo->op;
         op->opcode = OSD_OP_READ;
-        op->inode = target;
+        op->inode = to_num;
         op->offset = rwo->offset;
         op->len = target_block_size;
         op->iov.push_back(rwo->buf, target_block_size);
@@ -483,7 +484,7 @@ struct snap_merger_t
                 {
                     // write start->end
                     rwo->todo++;
-                    write_subop(rwo, rwo->start*gran, rwo->end*gran, use_cas ? 1+rwo->op.version : 0);
+                    write_subop(rwo, rwo->start*gran, rwo->end*gran, use_cas && to_num == target ? 1+rwo->op.version : 0);
                     rwo->start = rwo->end;
                     if (use_cas)
                     {
@@ -502,7 +503,7 @@ struct snap_merger_t
         {
             // write start->end
             rwo->todo++;
-            write_subop(rwo, rwo->start*gran, rwo->end*gran, use_cas ? 1+rwo->op.version : 0);
+            write_subop(rwo, rwo->start*gran, rwo->end*gran, use_cas && to_num == target ? 1+rwo->op.version : 0);
             rwo->start = rwo->end;
             if (use_cas)
             {
@@ -532,7 +533,7 @@ struct snap_merger_t
                 if (use_cas && subop->retval == -EINTR)
                 {
                     // CAS failure - reread and repeat optimistically
-                    rwo->start = subop->offset - rwo->offset;
+                    rwo->start = rwo->end = 0;
                     rwo_read(rwo);
                     delete subop;
                     return;
@@ -542,7 +543,7 @@ struct snap_merger_t
                 rwo->error_read = false;
             }
             // Increment CAS version
-            rwo->op.version++;
+            rwo->op.version = subop->version;
             if (use_cas)
                 next_write(rwo);
             else

src/cli_rm.cpp

@@ -65,6 +65,9 @@ struct snap_remover_t
     int current_child = 0;
     std::function<bool(cli_result_t &)> cb;
 
+    std::vector<std::string> rebased_images, deleted_images;
+    std::vector<uint64_t> deleted_ids;
+    std::string inverse_child_name, inverse_parent_name;
     cli_result_t result;
 
     bool is_done()
@@ -122,6 +125,7 @@ resume_1:
         {
             if (merge_children[current_child] == inverse_child)
                 continue;
+            rebased_images.push_back(parent->cli->st_cli.inode_config.at(merge_children[current_child]).name);
             start_merge_child(merge_children[current_child], merge_children[current_child]);
             if (state == 100)
                 return;
@@ -134,9 +138,12 @@ resume_2:
             cb = NULL;
             if (result.err)
             {
+                result.data = my_result(result.data);
                 state = 100;
                 return;
             }
+            else if (parent->progress)
+                printf("%s\n", result.text.c_str());
             parent->change_parent(merge_children[current_child], new_parent, &result);
             state = 3;
 resume_3:
@@ -144,6 +151,7 @@ resume_3:
                 return;
             if (result.err)
             {
+                result.data = my_result(result.data);
                 state = 100;
                 return;
             }
@@ -165,9 +173,12 @@ resume_4:
             cb = NULL;
             if (result.err)
             {
+                result.data = my_result(result.data);
                 state = 100;
                 return;
             }
+            else if (parent->progress)
+                printf("%s\n", result.text.c_str());
             // Delete "inverse" child data
             start_delete_source(inverse_child);
             if (state == 100)
@@ -181,9 +192,12 @@ resume_5:
             cb = NULL;
             if (result.err)
             {
+                result.data = my_result(result.data);
                 state = 100;
                 return;
             }
+            else if (parent->progress)
+                printf("%s\n", result.text.c_str());
             // Delete "inverse" child metadata, rename parent over it,
             // and also change parent links of the previous "inverse" child
             rename_inverse_parent();
@@ -199,6 +213,12 @@ resume_6:
         {
             if (chain_list[current_child] == inverse_parent)
                 continue;
+            {
+                auto parent_it = parent->cli->st_cli.inode_config.find(chain_list[current_child]);
+                if (parent_it != parent->cli->st_cli.inode_config.end())
+                    deleted_images.push_back(parent_it->second.name);
+                deleted_ids.push_back(chain_list[current_child]);
+            }
             start_delete_source(chain_list[current_child]);
 resume_7:
             while (!cb(result))
@@ -209,9 +229,12 @@ resume_7:
             cb = NULL;
             if (result.err)
             {
+                result.data = my_result(result.data);
                 state = 100;
                 return;
             }
+            else if (parent->progress)
+                printf("%s\n", result.text.c_str());
             delete_inode_config(chain_list[current_child]);
             if (state == 100)
                 return;
@@ -221,11 +244,26 @@ resume_8:
                 return;
         }
         state = 100;
+        result = (cli_result_t){
+            .text = "",
+            .data = my_result(result.data),
+        };
 resume_100:
         // Done
         return;
     }
 
+    json11::Json my_result(json11::Json src)
+    {
+        auto obj = src.object_items();
+        obj["deleted_ids"] = deleted_ids;
+        obj["deleted_images"] = deleted_images;
+        obj["rebased_images"] = rebased_images;
+        obj["renamed_from"] = inverse_parent_name;
+        obj["renamed_to"] = inverse_child_name;
+        return obj;
+    }
+
     void get_merge_children()
     {
         // Get all children of from..to
@@ -338,7 +376,11 @@ resume_100:
             }
             for (auto inode_result: data["responses"].array_items())
             {
-                auto kv = parent->cli->st_cli.parse_etcd_kv(inode_result["kvs"][0]);
+                if (inode_result["response_range"]["kvs"].array_items().size() == 0)
+                {
+                    continue;
+                }
+                auto kv = parent->cli->st_cli.parse_etcd_kv(inode_result["response_range"]["kvs"][0]);
                 pool_id_t pool_id = 0;
                 inode_t inode = 0;
                 char null_byte = 0;
@@ -377,7 +419,7 @@ resume_100:
             inode_t child = cp.first;
             uint64_t child_used = inode_used[child];
             int rank = cp.second;
-            for (int i = chain_list.size()-rank; i < chain_list.size(); i++)
+            for (int i = chain_list.size()-1-rank; i < chain_list.size(); i++)
             {
                 inode_t parent = chain_list[i];
                 uint64_t parent_used = inode_used[parent];
@@ -413,8 +455,8 @@ resume_100:
         }
         inode_config_t *child_cfg = &child_it->second;
         inode_config_t *target_cfg = &target_it->second;
-        std::string child_name = child_cfg->name;
-        std::string target_name = target_cfg->name;
+        inverse_child_name = child_cfg->name;
+        inverse_parent_name = target_cfg->name;
         std::string child_cfg_key = base64_encode(
             parent->cli->st_cli.etcd_prefix+
             "/config/inode/"+std::to_string(INODE_POOL(inverse_child))+
@@ -425,6 +467,9 @@ resume_100:
             "/config/inode/"+std::to_string(INODE_POOL(inverse_parent))+
             "/"+std::to_string(INODE_NO_POOL(inverse_parent))
         );
+        std::string target_idx_key = base64_encode(
+            parent->cli->st_cli.etcd_prefix+"/index/image/"+inverse_parent_name
+        );
         // Fill new configuration
         inode_config_t new_cfg = *child_cfg;
         new_cfg.num = target_cfg->num;
@@ -449,6 +494,11 @@ resume_100:
                     { "key", child_cfg_key },
                 } },
             },
+            json11::Json::object {
+                { "request_delete_range", json11::Json::object {
+                    { "key", target_idx_key },
+                } },
+            },
             json11::Json::object {
                 { "request_put", json11::Json::object {
                     { "key", target_cfg_key },
@@ -495,12 +545,12 @@ resume_100:
         parent->cli->st_cli.etcd_txn_slow(json11::Json::object {
             { "compare", cmp },
             { "success", txn },
-        }, [this, target_name, child_name](std::string err, json11::Json res)
+        }, [this](std::string err, json11::Json res)
         {
             parent->waiting--;
             if (err != "")
             {
-                result = (cli_result_t){ .err = EIO, .text = "Error renaming "+target_name+" to "+child_name+": "+err };
+                result = (cli_result_t){ .err = EIO, .text = "Error renaming "+inverse_parent_name+" to "+inverse_child_name+": "+err };
                 state = 100;
                 return;
             }
@@ -508,14 +558,14 @@ resume_100:
             {
                 result = (cli_result_t){
                     .err = EAGAIN,
-                    .text = "Parent ("+target_name+"), child ("+child_name+"), or one of its children"
+                    .text = "Parent ("+inverse_parent_name+"), child ("+inverse_child_name+"), or one of its children"
                         " configuration was modified during rename",
                 };
                 state = 100;
                 return;
             }
             if (parent->progress)
-                printf("Layer %s renamed to %s\n", target_name.c_str(), child_name.c_str());
+                printf("Layer %s renamed to %s\n", inverse_parent_name.c_str(), inverse_child_name.c_str());
             parent->ringloop->wakeup();
         });
     }

src/cli_rm_data.cpp

@@ -28,6 +28,7 @@ struct rm_inode_t
     cli_tool_t *parent = NULL;
     inode_list_t *lister = NULL;
     std::vector<rm_pg_t*> lists;
+    std::vector<osd_num_t> inactive_osds;
     uint64_t total_count = 0, total_done = 0, total_prev_pct = 0;
     uint64_t pgs_to_list = 0;
     bool lists_done = false;
@@ -86,6 +87,16 @@ struct rm_inode_t
             state = 100;
             return;
         }
+        inactive_osds = parent->cli->list_inode_get_inactive_osds(lister);
+        if (inactive_osds.size() && !parent->json_output)
+        {
+            fprintf(stderr, "Some data may remain after delete on OSDs which are currently down: ");
+            for (int i = 0; i < inactive_osds.size(); i++)
+            {
+                fprintf(stderr, i > 0 ? ", %lu" : "%lu", inactive_osds[i]);
+            }
+            fprintf(stderr, "\n");
+        }
         pgs_to_list = parent->cli->list_pg_count(lister);
         parent->cli->list_inode_next(lister, parent->parallel_osds);
     }
@@ -167,16 +178,33 @@ struct rm_inode_t
         }
         if (parent->progress && total_count > 0 && total_done*1000/total_count != total_prev_pct)
         {
-            printf("\rRemoved %lu/%lu objects, %lu more PGs to list...", total_done, total_count, pgs_to_list);
+            fprintf(stderr, "\rRemoved %lu/%lu objects, %lu more PGs to list...", total_done, total_count, pgs_to_list);
             total_prev_pct = total_done*1000/total_count;
         }
         if (lists_done && !lists.size())
         {
+            if (parent->progress && total_count > 0)
+            {
+                fprintf(stderr, "\n");
+            }
+            if (parent->progress && (total_done < total_count || inactive_osds.size() > 0))
+            {
+                fprintf(
+                    stderr, "Warning: Pool:%u,ID:%lu inode data may not have been fully removed.\n"
+                    " Use `vitastor-cli rm-data --pool %u --inode %lu` if you encounter it in listings.\n",
+                    pool_id, INODE_NO_POOL(inode), pool_id, INODE_NO_POOL(inode)
+                );
+            }
             result = (cli_result_t){
                 .err = error_count > 0 ? EIO : 0,
                 .text = error_count > 0 ? "Some blocks were not removed" : (
                     "Done, inode "+std::to_string(INODE_NO_POOL(inode))+" from pool "+
                     std::to_string(pool_id)+" removed"),
+                .data = json11::Json::object {
+                    { "removed_objects", total_done },
+                    { "total_objects", total_count },
+                    { "inactive_osds", inactive_osds },
+                },
             };
             state = 100;
         }

src/cli_status.cpp

@@ -12,7 +12,7 @@ static const char *obj_states[] = { "clean", "misplaced", "degraded", "incomplet
 // Print cluster status:
 // etcd, mon, osd states
 // raw/used space, object states, pool states, pg states
-// client io, recovery io, rebalance io
+// client io, recovery io, rebalance io, scrub io
 struct status_printer_t
 {
     cli_tool_t *parent;
@@ -252,18 +252,25 @@ resume_2:
         }
         more_states.resize(more_states.size()-2);
         std::string recovery_io;
+        int io_indent = 0;
         {
             uint64_t deg_bps = agg_stats["recovery_stats"]["degraded"]["bps"].uint64_value();
             uint64_t deg_iops = agg_stats["recovery_stats"]["degraded"]["iops"].uint64_value();
             uint64_t misp_bps = agg_stats["recovery_stats"]["misplaced"]["bps"].uint64_value();
             uint64_t misp_iops = agg_stats["recovery_stats"]["misplaced"]["iops"].uint64_value();
+            uint64_t scrub_bps = agg_stats["op_stats"]["scrub"]["bps"].uint64_value();
+            uint64_t scrub_iops = agg_stats["op_stats"]["scrub"]["iops"].uint64_value();
+            if (misp_iops > 0 || misp_bps > 0 || no_rebalance)
+                io_indent = 3;
+            else if (deg_iops > 0 || deg_bps > 0 || no_recovery)
+                io_indent = 2;
             if (deg_iops > 0 || deg_bps > 0)
             {
-                recovery_io += "    recovery:  "+std::string(no_recovery ? "disabled, " : "")+
+                recovery_io += "    recovery: "+str_repeat(" ", io_indent-2)+std::string(no_recovery ? "disabled, " : "")+
                     format_size(deg_bps)+"/s, "+format_size(deg_iops, true)+" op/s\n";
             }
             else if (no_recovery)
-                recovery_io += "    recovery:  disabled\n";
+                recovery_io += "    recovery: disabled\n";
             if (misp_iops > 0 || misp_bps > 0)
             {
                 recovery_io += "    rebalance: "+std::string(no_rebalance ? "disabled, " : "")+
@@ -271,6 +278,13 @@ resume_2:
             }
             else if (no_rebalance)
                 recovery_io += "    rebalance: disabled\n";
+            if (scrub_iops > 0 || scrub_bps > 0)
+            {
+                recovery_io += "    scrub: "+str_repeat(" ", io_indent+1)+std::string(no_scrub ? "disabled, " : "")+
+                    format_size(scrub_bps)+"/s, "+format_size(scrub_iops, true)+" op/s\n";
+            }
+            else if (no_scrub)
+                recovery_io += "    scrub: "+str_repeat(" ", io_indent+1)+"disabled\n";
         }
         printf(
             "  cluster:\n"
@@ -298,7 +312,7 @@ resume_2:
             pools_active, pool_count,
             pgs_by_state_str.c_str(),
             readonly ? " (read-only mode)" : "",
-            recovery_io.size() > 0 ? "   " : "",
+            str_repeat(" ", io_indent).c_str(),
             format_size(agg_stats["op_stats"]["primary_read"]["bps"].uint64_value()).c_str(),
             format_size(agg_stats["op_stats"]["primary_read"]["iops"].uint64_value(), true).c_str(),
             format_size(agg_stats["op_stats"]["primary_write"]["bps"].uint64_value()).c_str(),

src/cluster_client.cpp

@@ -1209,6 +1209,10 @@ void cluster_client_t::handle_op_part(cluster_op_part_t *part)
             copy_part_bitmap(op, part);
             op->version = op->parts.size() == 1 ? part->op.reply.rw.version : 0;
         }
+        else if (op->opcode == OSD_OP_WRITE)
+        {
+            op->version = op->parts.size() == 1 ? part->op.reply.rw.version : 0;
+        }
         if (op->inflight_count == 0)
         {
             if (op->opcode == OSD_OP_SYNC)

src/cluster_client.h

@@ -130,6 +130,7 @@ public:
     inode_list_t *list_inode_start(inode_t inode,
         std::function<void(inode_list_t* lst, std::set<object_id>&& objects, pg_num_t pg_num, osd_num_t primary_osd, int status)> callback);
     int list_pg_count(inode_list_t *lst);
+    const std::vector<osd_num_t> & list_inode_get_inactive_osds(inode_list_t *lst);
     void list_inode_next(inode_list_t *lst, int next_pgs);
     //inline uint32_t get_bs_bitmap_granularity() { return st_cli.global_bitmap_granularity; }
     //inline uint64_t get_bs_block_size() { return st_cli.global_block_size; }

src/cluster_client_list.cpp

@@ -36,6 +36,7 @@ struct inode_list_t
     inode_t inode = 0;
     int done_pgs = 0;
     int want = 0;
+    std::vector<osd_num_t> inactive_osds;
     std::vector<inode_list_pg_t*> pgs;
     std::function<void(inode_list_t* lst, std::set<object_id>&& objects, pg_num_t pg_num, osd_num_t primary_osd, int status)> callback;
 };
@@ -60,6 +61,7 @@ inode_list_t* cluster_client_t::list_inode_start(inode_t inode,
     lst->inode = inode;
     lst->callback = callback;
     auto pool_cfg = st_cli.pool_config[pool_id];
+    std::set<osd_num_t> inactive_osd_set;
     for (auto & pg_item: pool_cfg.pg_config)
     {
         auto & pg = pg_item.second;
@@ -106,11 +108,18 @@ inode_list_t* cluster_client_t::list_inode_start(inode_t inode,
             }
             for (osd_num_t peer_osd: all_peers)
             {
-                r->list_osds.push_back((inode_list_osd_t){
-                    .pg = r,
-                    .osd_num = peer_osd,
-                    .sent = false,
-                });
+                if (st_cli.peer_states.find(peer_osd) != st_cli.peer_states.end())
+                {
+                    r->list_osds.push_back((inode_list_osd_t){
+                        .pg = r,
+                        .osd_num = peer_osd,
+                        .sent = false,
+                    });
+                }
+                else
+                {
+                    inactive_osd_set.insert(peer_osd);
+                }
             }
         }
         else
@@ -132,6 +141,7 @@ inode_list_t* cluster_client_t::list_inode_start(inode_t inode,
     {
         lst->pgs[i]->pos = i;
     }
+    lst->inactive_osds.insert(lst->inactive_osds.end(), inactive_osd_set.begin(), inactive_osd_set.end());
     lists.push_back(lst);
     return lst;
 }
@@ -141,6 +151,11 @@ int cluster_client_t::list_pg_count(inode_list_t *lst)
     return lst->pgs.size();
 }
 
+const std::vector<osd_num_t> & cluster_client_t::list_inode_get_inactive_osds(inode_list_t *lst)
+{
+    return lst->inactive_osds;
+}
+
 void cluster_client_t::list_inode_next(inode_list_t *lst, int next_pgs)
 {
     if (next_pgs >= 0)

src/disk_simple_offsets.cpp

@@ -10,6 +10,7 @@
 #include "json11/json11.hpp"
 #include "str_util.h"
 #include "blockstore.h"
+#include "blockstore_disk.h"
 
 // Calculate offsets for a block device and print OSD command line parameters
 void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
@@ -20,23 +21,39 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
         fprintf(stderr, "Device path is missing\n");
         exit(1);
     }
-    uint64_t object_size = parse_size(cfg["object_size"].string_value());
+    uint64_t data_block_size = parse_size(cfg["object_size"].string_value());
     uint64_t bitmap_granularity = parse_size(cfg["bitmap_granularity"].string_value());
     uint64_t journal_size = parse_size(cfg["journal_size"].string_value());
     uint64_t device_block_size = parse_size(cfg["device_block_size"].string_value());
     uint64_t journal_offset = parse_size(cfg["journal_offset"].string_value());
     uint64_t device_size = parse_size(cfg["device_size"].string_value());
+    uint32_t csum_block_size = parse_size(cfg["csum_block_size"].string_value());
+    uint32_t data_csum_type = BLOCKSTORE_CSUM_NONE;
+    if (cfg["data_csum_type"] == "crc32c")
+        data_csum_type = BLOCKSTORE_CSUM_CRC32C;
+    else if (cfg["data_csum_type"].string_value() != "" && cfg["data_csum_type"].string_value() != "none")
+    {
+        fprintf(
+            stderr, "data_csum_type=%s is unsupported, only \"crc32c\" and \"none\" are supported",
+            cfg["data_csum_type"].string_value().c_str()
+        );
+        exit(1);
+    }
     std::string format = cfg["format"].string_value();
     if (json_output)
         format = "json";
-    if (!object_size)
-        object_size = 1 << DEFAULT_DATA_BLOCK_ORDER;
+    if (!data_block_size)
+        data_block_size = 1 << DEFAULT_DATA_BLOCK_ORDER;
     if (!bitmap_granularity)
         bitmap_granularity = DEFAULT_BITMAP_GRANULARITY;
     if (!journal_size)
         journal_size = 16*1024*1024;
     if (!device_block_size)
         device_block_size = 4096;
+    if (!data_csum_type)
+        csum_block_size = 0;
+    else if (!csum_block_size)
+        csum_block_size = bitmap_granularity;
     uint64_t orig_device_size = device_size;
     if (!device_size)
     {
@@ -85,22 +102,30 @@ void disk_tool_simple_offsets(json11::Json cfg, bool json_output)
         fprintf(stderr, "Invalid device block size specified: %lu\n", device_block_size);
         exit(1);
     }
-    if (object_size < device_block_size || object_size > MAX_DATA_BLOCK_SIZE ||
-        object_size & (object_size-1) != 0)
+    if (data_block_size < device_block_size || data_block_size > MAX_DATA_BLOCK_SIZE ||
+        data_block_size & (data_block_size-1) != 0)
     {
-        fprintf(stderr, "Invalid object size specified: %lu\n", object_size);
+        fprintf(stderr, "Invalid object size specified: %lu\n", data_block_size);
         exit(1);
     }
-    if (bitmap_granularity < device_block_size || bitmap_granularity > object_size ||
+    if (bitmap_granularity < device_block_size || bitmap_granularity > data_block_size ||
         bitmap_granularity & (bitmap_granularity-1) != 0)
     {
         fprintf(stderr, "Invalid bitmap granularity specified: %lu\n", bitmap_granularity);
         exit(1);
     }
+    if (csum_block_size && (data_block_size % csum_block_size))
+    {
+        fprintf(stderr, "csum_block_size must be a divisor of data_block_size\n");
+        exit(1);
+    }
     journal_offset = ((journal_offset+device_block_size-1)/device_block_size)*device_block_size;
     uint64_t meta_offset = journal_offset + ((journal_size+device_block_size-1)/device_block_size)*device_block_size;
-    uint64_t entries_per_block = (device_block_size / (24 + 2*object_size/bitmap_granularity/8));
-    uint64_t object_count = ((device_size-meta_offset)/object_size);
+    uint64_t data_csum_size = (data_csum_type ? data_block_size/csum_block_size*(data_csum_type & 0xFF) : 0);
+    uint64_t clean_entry_bitmap_size = data_block_size/bitmap_granularity/8;
+    uint64_t clean_entry_size = 24 /*sizeof(clean_disk_entry)*/ + 2*clean_entry_bitmap_size + data_csum_size + 4 /*entry_csum*/;
+    uint64_t entries_per_block = device_block_size / clean_entry_size;
+    uint64_t object_count = ((device_size-meta_offset)/data_block_size);
     uint64_t meta_size = (1 + (object_count+entries_per_block-1)/entries_per_block) * device_block_size;
     uint64_t data_offset = meta_offset + meta_size;
     if (format == "json")

src/disk_tool.cpp

@@ -59,6 +59,8 @@ static const char *help_text =
     "    --journal_size 32M/1G      Set journal size (area or partition size)\n"
     "    --block_size 128k/1M       Set blockstore object size\n"
     "    --bitmap_granularity 4k    Set bitmap granularity\n"
+    "    --data_csum_type none      Set data checksum type (crc32c or none)\n"
+    "    --csum_block_size 4k       Set data checksum block size\n"
     "    --data_device_block 4k     Override data device block size\n"
     "    --meta_device_block 4k     Override metadata device block size\n"
     "    --journal_device_block 4k  Override journal device block size\n"
@@ -72,8 +74,9 @@ static const char *help_text =
     "  If it doesn't succeed it issues a warning in the system log.\n"
     "  \n"
     "  You can also pass other OSD options here as arguments and they'll be persisted\n"
-    "  to the superblock: max_write_iodepth, max_write_iodepth, min_flusher_count,\n"
-    "  max_flusher_count, inmemory_metadata, inmemory_journal, journal_sector_buffer_count,\n"
+    "  in the superblock: cached_read_data, cached_read_meta, cached_read_journal,\n"
+    "  inmemory_metadata, inmemory_journal, max_write_iodepth,\n"
+    "  min_flusher_count, max_flusher_count, journal_sector_buffer_count,\n"
     "  journal_no_same_sector_overwrites, throttle_small_writes, throttle_target_iops,\n"
     "  throttle_target_mbs, throttle_target_parallelism, throttle_threshold_us.\n"
     "\n"
@@ -161,6 +164,8 @@ static const char *help_text =
     "    --object_size 128k       Set blockstore block size\n"
     "    --bitmap_granularity 4k  Set bitmap granularity\n"
     "    --journal_size 16M       Set journal size\n"
+    "    --data_csum_type none    Set data checksum type (crc32c or none)\n"
+    "    --csum_block_size 4k     Set data checksum block size\n"
     "    --device_block_size 4k   Set device block size\n"
     "    --journal_offset 0       Set journal offset\n"
     "    --device_size 0          Set device size\n"
@@ -270,6 +275,19 @@ int main(int argc, char *argv[])
             fprintf(stderr, "Invalid JSON: %s\n", json_err.c_str());
             return 1;
         }
+        if (entries[0]["type"] == "start")
+        {
+            self.dsk.data_csum_type = csum_type_from_str(entries[0]["data_csum_type"].string_value());
+            self.dsk.csum_block_size = entries[0]["csum_block_size"].uint64_value();
+        }
+        if (self.options["data_csum_type"] != "")
+        {
+            self.dsk.data_csum_type = csum_type_from_str(self.options["data_csum_type"]);
+        }
+        if (self.options["csum_block_size"] != "")
+        {
+            self.dsk.csum_block_size = stoull_full(self.options["csum_block_size"], 0);
+        }
         return self.write_json_journal(entries);
     }
     else if (!strcmp(cmd[0], "dump-meta"))

src/disk_tool.h

@@ -64,17 +64,19 @@ struct disk_tool_t
     ring_loop_t *ringloop;
     ring_consumer_t ring_consumer;
     int remap_active;
+    journal_entry_start je_start;
     uint8_t *new_journal_buf, *new_meta_buf, *new_journal_ptr, *new_journal_data;
     uint64_t new_journal_in_pos;
     int64_t data_idx_diff;
     uint64_t total_blocks, free_first, free_last;
-    uint64_t new_clean_entry_bitmap_size, new_clean_entry_size, new_entries_per_block;
+    uint64_t new_clean_entry_bitmap_size, new_data_csum_size, new_clean_entry_size, new_entries_per_block;
     int new_journal_fd, new_meta_fd;
     resizer_data_moving_t *moving_blocks;
 
     bool started;
     void *small_write_data;
     uint32_t data_crc32;
+    bool data_csum_valid;
     uint32_t crc32_last;
     uint32_t new_crc32_prev;
 
@@ -84,11 +86,11 @@ struct disk_tool_t
     void dump_journal_entry(int num, journal_entry *je, bool json);
     int process_journal(std::function<int(void*)> block_fn);
     int process_journal_block(void *buf, std::function<void(int, journal_entry*)> iter_fn);
-    int process_meta(std::function<void(blockstore_meta_header_v1_t *)> hdr_fn,
+    int process_meta(std::function<void(blockstore_meta_header_v2_t *)> hdr_fn,
         std::function<void(uint64_t, clean_disk_entry*, uint8_t*)> record_fn);
 
     int dump_meta();
-    void dump_meta_header(blockstore_meta_header_v1_t *hdr);
+    void dump_meta_header(blockstore_meta_header_v2_t *hdr);
     void dump_meta_entry(uint64_t block_num, clean_disk_entry *entry, uint8_t *bitmap);
 
     int write_json_journal(json11::Json entries);
@@ -96,7 +98,7 @@ struct disk_tool_t
 
     int resize_data();
     int resize_parse_params();
-    void resize_init(blockstore_meta_header_v1_t *hdr);
+    void resize_init(blockstore_meta_header_v2_t *hdr);
     int resize_remap_blocks();
     int resize_copy_data();
     int resize_rewrite_journal();
@@ -141,3 +143,5 @@ json11::Json read_parttable(std::string dev);
 uint64_t dev_size_from_parttable(json11::Json pt);
 uint64_t free_from_parttable(json11::Json pt);
 int fix_partition_type(std::string dev_by_uuid);
+std::string csum_type_str(uint32_t data_csum_type);
+uint32_t csum_type_from_str(std::string data_csum_type);

src/disk_tool_journal.cpp

@@ -55,6 +55,23 @@ int disk_tool_t::dump_journal()
                     printf("offset %08lx:\n", journal_pos);
                 else
                     printf(",\"entries\":[\n");
+                if (journal_pos == 0)
+                {
+                    // Fill journal header to know checksum type & size
+                    journal_entry *je = (journal_entry*)journal_buf;
+                    if (je->magic == JOURNAL_MAGIC && je->type == JE_START &&
+                        (je->start.version == JOURNAL_VERSION_V1 || je->start.version == JOURNAL_VERSION_V2))
+                    {
+                        memcpy(&je_start, je, sizeof(je_start));
+                        if (je_start.size == JE_START_V0_SIZE)
+                            je_start.version = 0;
+                        if (je_start.version < JOURNAL_VERSION_V2)
+                        {
+                            je_start.data_csum_type = 0;
+                            je_start.csum_block_size = 0;
+                        }
+                    }
+                }
                 first_entry = true;
                 process_journal_block(journal_buf, [this](int num, journal_entry *je) { dump_journal_entry(num, je, json); });
                 if (json)
@@ -120,8 +137,22 @@ int disk_tool_t::process_journal(std::function<int(void*)> block_fn)
         fprintf(stderr, "offset %08lx: journal superblock is invalid\n", journal_pos);
         r = 1;
     }
+    else if (je->start.size != JE_START_V0_SIZE && je->start.version != JOURNAL_VERSION_V1 && je->start.version != JOURNAL_VERSION_V2)
+    {
+        fprintf(stderr, "offset %08lx: journal superblock contains version %lu, but I only understand 0, 1 and 2\n",
+            journal_pos, je->start.size == JE_START_V0_SIZE ? 0 : je->start.version);
+        r = 1;
+    }
     else
     {
+        memcpy(&je_start, je, sizeof(je_start));
+        if (je_start.size == JE_START_V0_SIZE)
+            je_start.version = 0;
+        if (je_start.version < JOURNAL_VERSION_V2)
+        {
+            je_start.data_csum_type = 0;
+            je_start.csum_block_size = 0;
+        }
         started = false;
         crc32_last = 0;
         block_fn(data);
@@ -183,7 +214,49 @@ int disk_tool_t::process_journal_block(void *buf, std::function<void(int, journa
             }
             small_write_data = memalign_or_die(MEM_ALIGNMENT, je->small_write.len);
             assert(pread(dsk.journal_fd, small_write_data, je->small_write.len, dsk.journal_offset+je->small_write.data_offset) == je->small_write.len);
-            data_crc32 = crc32c(0, small_write_data, je->small_write.len);
+            data_crc32 = je_start.csum_block_size ? 0 : crc32c(0, small_write_data, je->small_write.len);
+            data_csum_valid = (data_crc32 == je->small_write.crc32_data);
+            if (je_start.csum_block_size && je->small_write.len > 0)
+            {
+                // like in enqueue_write()
+                uint32_t start = je->small_write.offset / je_start.csum_block_size;
+                uint32_t end = (je->small_write.offset+je->small_write.len-1) / je_start.csum_block_size;
+                uint32_t data_csum_size = (end-start+1) * (je_start.data_csum_type & 0xFF);
+                if (je->size < sizeof(journal_entry_small_write) + data_csum_size)
+                {
+                    data_csum_valid = false;
+                }
+                else
+                {
+                    uint32_t calc_csum = 0;
+                    uint32_t *block_csums = (uint32_t*)((uint8_t*)je + je->size - data_csum_size);
+                    if (start == end)
+                    {
+                        calc_csum = crc32c(0, (uint8_t*)small_write_data, je->small_write.len);
+                        data_csum_valid = data_csum_valid && (calc_csum == *block_csums++);
+                    }
+                    else
+                    {
+                        // First block
+                        calc_csum = crc32c(0, (uint8_t*)small_write_data,
+                            je_start.csum_block_size*(start+1)-je->small_write.offset);
+                        data_csum_valid = data_csum_valid && (calc_csum == *block_csums++);
+                        // Intermediate blocks
+                        for (uint32_t i = start+1; i < end; i++)
+                        {
+                            calc_csum = crc32c(0, (uint8_t*)small_write_data +
+                                je_start.csum_block_size*i-je->small_write.offset, je_start.csum_block_size);
+                            data_csum_valid = data_csum_valid && (calc_csum == *block_csums++);
+                        }
+                        // Last block
+                        calc_csum = crc32c(
+                            0, (uint8_t*)small_write_data + end*je_start.csum_block_size - je->small_write.offset,
+                            je->small_write.offset+je->small_write.len - end*je_start.csum_block_size
+                        );
+                        data_csum_valid = data_csum_valid && (calc_csum == *block_csums++);
+                    }
+                }
+            }
         }
         iter_fn(entry, je);
         if (je->type == JE_SMALL_WRITE || je->type == JE_SMALL_WRITE_INSTANT)
@@ -223,29 +296,40 @@ void disk_tool_t::dump_journal_entry(int num, journal_entry *je, bool json)
     if (je->type == JE_START)
     {
         printf(
-            json ? ",\"type\":\"start\",\"start\":\"0x%lx\"}" : "je_start start=%08lx\n",
+            json ? ",\"type\":\"start\",\"start\":\"0x%lx\"" : "je_start start=%08lx",
             je->start.journal_start
         );
+        if (je->start.data_csum_type)
+        {
+            printf(
+                json ? ",\"data_csum_type\":\"%s\",\"csum_block_size\":%u" : " data_csum_type=%s csum_block_size=%u",
+                csum_type_str(je->start.data_csum_type).c_str(), je->start.csum_block_size
+            );
+        }
+        printf(json ? "}" : "\n");
     }
     else if (je->type == JE_SMALL_WRITE || je->type == JE_SMALL_WRITE_INSTANT)
     {
+        auto & sw = je->small_write;
         printf(
             json ? ",\"type\":\"small_write%s\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\",\"offset\":%u,\"len\":%u,\"loc\":\"0x%lx\""
                 : "je_small_write%s oid=%lx:%lx ver=%lu offset=%u len=%u loc=%08lx",
             je->type == JE_SMALL_WRITE_INSTANT ? "_instant" : "",
-            je->small_write.oid.inode, je->small_write.oid.stripe,
-            je->small_write.version, je->small_write.offset, je->small_write.len,
-            je->small_write.data_offset
+            sw.oid.inode, sw.oid.stripe, sw.version, sw.offset, sw.len, sw.data_offset
         );
-        if (journal_calc_data_pos != je->small_write.data_offset)
+        if (journal_calc_data_pos != sw.data_offset)
         {
             printf(json ? ",\"bad_loc\":true,\"calc_loc\":\"0x%lx\""
                 : " (mismatched, calculated = %lu)", journal_pos);
         }
-        if (je->small_write.size > sizeof(journal_entry_small_write))
+        uint32_t data_csum_size = (!je_start.csum_block_size
+            ? 0
+            : ((sw.offset + sw.len - 1)/je_start.csum_block_size - sw.offset/je_start.csum_block_size + 1)
+                *(je_start.data_csum_type & 0xFF));
+        if (je->size > sizeof(journal_entry_small_write) + data_csum_size)
         {
             printf(json ? ",\"bitmap\":\"" : " (bitmap: ");
-            for (int i = sizeof(journal_entry_small_write); i < je->small_write.size; i++)
+            for (int i = sizeof(journal_entry_small_write); i < je->size - data_csum_size; i++)
             {
                 printf("%02x", ((uint8_t*)je)[i]);
             }
@@ -254,34 +338,56 @@ void disk_tool_t::dump_journal_entry(int num, journal_entry *je, bool json)
         if (dump_with_data)
         {
             printf(json ? ",\"data\":\"" : " (data: ");
-            for (int i = 0; i < je->small_write.len; i++)
+            for (int i = 0; i < sw.len; i++)
             {
                 printf("%02x", ((uint8_t*)small_write_data)[i]);
             }
             printf(json ? "\"" : ")");
         }
+        if (data_csum_size > 0 && je->size >= sizeof(journal_entry_small_write) + data_csum_size)
+        {
+            printf(json ? ",\"block_csums\":\"" : " block_csums=");
+            uint8_t *block_csums = (uint8_t*)je + je->size - data_csum_size;
+            for (int i = 0; i < data_csum_size; i++)
+                printf("%02x", block_csums[i]);
+            printf(json ? "\"" : "");
+        }
+        else
+        {
+            printf(json ? ",\"data_crc32\":\"%08x\"" : " data_crc32=%08x", sw.crc32_data);
+        }
         printf(
-            json ? ",\"data_crc32\":\"%08x\",\"data_valid\":%s}" : " data_crc32=%08x%s\n",
-            je->small_write.crc32_data,
-            (data_crc32 != je->small_write.crc32_data
-                ? (json ? "false" : " (invalid)")
-                : (json ? "true" : " (valid)"))
+            json ? ",\"data_valid\":%s}" : "%s\n",
+            (data_csum_valid
+                ? (json ? "true" : " (valid)")
+                : (json ? "false" : " (invalid)"))
         );
     }
     else if (je->type == JE_BIG_WRITE || je->type == JE_BIG_WRITE_INSTANT)
     {
+        auto & bw = je->big_write;
         printf(
             json ? ",\"type\":\"big_write%s\",\"inode\":\"0x%lx\",\"stripe\":\"0x%lx\",\"ver\":\"%lu\",\"offset\":%u,\"len\":%u,\"loc\":\"0x%lx\""
                 : "je_big_write%s oid=%lx:%lx ver=%lu offset=%u len=%u loc=%08lx",
             je->type == JE_BIG_WRITE_INSTANT ? "_instant" : "",
-            je->big_write.oid.inode, je->big_write.oid.stripe,
-            je->big_write.version, je->big_write.offset, je->big_write.len,
-            je->big_write.location
+            bw.oid.inode, bw.oid.stripe, bw.version, bw.offset, bw.len, bw.location
         );
-        if (je->big_write.size > sizeof(journal_entry_big_write))
+        uint32_t data_csum_size = (!je_start.csum_block_size
+            ? 0
+            : ((bw.offset + bw.len - 1)/je_start.csum_block_size - bw.offset/je_start.csum_block_size + 1)
+                *(je_start.data_csum_type & 0xFF));
+        if (data_csum_size > 0 && je->size >= sizeof(journal_entry_big_write) + data_csum_size)
+        {
+            printf(json ? ",\"block_csums\":\"" : " block_csums=");
+            uint8_t *block_csums = (uint8_t*)je + je->size - data_csum_size;
+            for (int i = 0; i < data_csum_size; i++)
+                printf("%02x", block_csums[i]);
+            printf(json ? "\"" : "");
+        }
+        if (bw.size > sizeof(journal_entry_big_write) + data_csum_size)
         {
             printf(json ? ",\"bitmap\":\"" : " (bitmap: ");
-            for (int i = sizeof(journal_entry_big_write); i < je->big_write.size; i++)
+            for (int i = sizeof(journal_entry_big_write); i < bw.size - data_csum_size; i++)
             {
                 printf("%02x", ((uint8_t*)je)[i]);
             }
@@ -338,7 +444,9 @@ int disk_tool_t::write_json_journal(json11::Json entries)
         .type = JE_START,
         .size = sizeof(journal_entry_start),
         .journal_start = dsk.journal_block_size,
-        .version = JOURNAL_VERSION,
+        .version = JOURNAL_VERSION_V2,
+        .data_csum_type = dsk.data_csum_type,
+        .csum_block_size = dsk.csum_block_size,
     };
     ((journal_entry*)new_journal_buf)->crc32 = je_crc32((journal_entry*)new_journal_buf);
     new_journal_ptr += dsk.journal_block_size;
@@ -358,9 +466,11 @@ int disk_tool_t::write_json_journal(json11::Json entries)
         uint32_t entry_size = (type == JE_START
             ? sizeof(journal_entry_start)
             : (type == JE_SMALL_WRITE || type == JE_SMALL_WRITE_INSTANT
-                ? sizeof(journal_entry_small_write) + dsk.clean_entry_bitmap_size
+                ? sizeof(journal_entry_small_write) + dsk.clean_entry_bitmap_size +
+                    (dsk.data_csum_type ? rec["len"].uint64_value()/dsk.csum_block_size*(dsk.data_csum_type & 0xFF) : 0)
                 : (type == JE_BIG_WRITE || type == JE_BIG_WRITE_INSTANT
-                    ? sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size
+                    ? sizeof(journal_entry_big_write) + dsk.clean_entry_bitmap_size +
+                        (dsk.data_csum_type ? rec["len"].uint64_value()/dsk.csum_block_size*(dsk.data_csum_type & 0xFF) : 0)
                     : sizeof(journal_entry_del))));
         if (dsk.journal_block_size < new_journal_in_pos + entry_size)
         {
@@ -402,12 +512,24 @@ int disk_tool_t::write_json_journal(json11::Json entries)
                 .offset = (uint32_t)rec["offset"].uint64_value(),
                 .len = (uint32_t)rec["len"].uint64_value(),
                 .data_offset = (uint64_t)(new_journal_data-new_journal_buf),
-                .crc32_data = (uint32_t)sscanf_json("%x", rec["data_crc32"]),
+                .crc32_data = !dsk.data_csum_type ? 0 : (uint32_t)sscanf_json("%x", rec["data_crc32"]),
             };
-            fromhexstr(rec["bitmap"].string_value(), dsk.clean_entry_bitmap_size, ((uint8_t*)ne) + sizeof(journal_entry_small_write));
+            uint32_t data_csum_size = !dsk.data_csum_type ? 0 : ne->small_write.len/dsk.csum_block_size*(dsk.data_csum_type & 0xFF);
+            fromhexstr(rec["bitmap"].string_value(), dsk.clean_entry_bitmap_size, ((uint8_t*)ne) + sizeof(journal_entry_small_write) + data_csum_size);
             fromhexstr(rec["data"].string_value(), ne->small_write.len, new_journal_data);
+            if (dsk.data_csum_type)
+                fromhexstr(rec["block_csums"].string_value(), data_csum_size, ((uint8_t*)ne) + sizeof(journal_entry_small_write));
             if (rec["data"].is_string())
-                ne->small_write.crc32_data = crc32c(0, new_journal_data, ne->small_write.len);
+            {
+                if (!dsk.data_csum_type)
+                    ne->small_write.crc32_data = crc32c(0, new_journal_data, ne->small_write.len);
+                else if (dsk.data_csum_type == BLOCKSTORE_CSUM_CRC32C)
+                {
+                    uint32_t *block_csums = (uint32_t*)(((uint8_t*)ne) + sizeof(journal_entry_small_write));
+                    for (uint32_t i = 0; i < ne->small_write.len; i += dsk.csum_block_size, block_csums++)
+                        *block_csums = crc32c(0, new_journal_data+i, dsk.csum_block_size);
+                }
+            }
             new_journal_data += ne->small_write.len;
         }
         else if (type == JE_BIG_WRITE || type == JE_BIG_WRITE_INSTANT)
@@ -426,7 +548,10 @@ int disk_tool_t::write_json_journal(json11::Json entries)
                 .len = (uint32_t)rec["len"].uint64_value(),
                 .location = sscanf_json(NULL, rec["loc"]),
             };
-            fromhexstr(rec["bitmap"].string_value(), dsk.clean_entry_bitmap_size, ((uint8_t*)ne) + sizeof(journal_entry_big_write));
+            uint32_t data_csum_size = !dsk.data_csum_type ? 0 : ne->big_write.len/dsk.csum_block_size*(dsk.data_csum_type & 0xFF);
+            fromhexstr(rec["bitmap"].string_value(), dsk.clean_entry_bitmap_size, ((uint8_t*)ne) + sizeof(journal_entry_big_write) + data_csum_size);
+            if (dsk.data_csum_type)
+                fromhexstr(rec["block_csums"].string_value(), data_csum_size, ((uint8_t*)ne) + sizeof(journal_entry_big_write));
         }
         else if (type == JE_STABLE || type == JE_ROLLBACK || type == JE_DELETE)
         {

src/disk_tool_meta.cpp

@@ -5,7 +5,7 @@
 #include "rw_blocking.h"
 #include "osd_id.h"
 
-int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v1_t *)> hdr_fn,
+int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v2_t *)> hdr_fn,
     std::function<void(uint64_t, clean_disk_entry*, uint8_t*)> record_fn)
 {
     if (dsk.meta_block_size % DIRECT_IO_ALIGNMENT)
@@ -28,12 +28,38 @@ int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v1_t *)>
     lseek64(dsk.meta_fd, dsk.meta_offset, 0);
     read_blocking(dsk.meta_fd, data, dsk.meta_block_size);
     // Check superblock
-    blockstore_meta_header_v1_t *hdr = (blockstore_meta_header_v1_t *)data;
-    if (hdr->zero == 0 &&
-        hdr->magic == BLOCKSTORE_META_MAGIC_V1 &&
-        hdr->version == BLOCKSTORE_META_VERSION_V1)
+    blockstore_meta_header_v2_t *hdr = (blockstore_meta_header_v2_t *)data;
+    if (hdr->zero == 0 && hdr->magic == BLOCKSTORE_META_MAGIC_V1)
     {
-        // Vitastor 0.6-0.7 - static array of clean_disk_entry with bitmaps
+        if (hdr->version == BLOCKSTORE_META_FORMAT_V1)
+        {
+            // Vitastor 0.6-0.8 - static array of clean_disk_entry with bitmaps
+            hdr->data_csum_type = 0;
+            hdr->csum_block_size = 0;
+            hdr->header_csum = 0;
+        }
+        else if (hdr->version == BLOCKSTORE_META_FORMAT_V2)
+        {
+            // Vitastor 0.9 - static array of clean_disk_entry with bitmaps and checksums
+            if (hdr->data_csum_type != 0 &&
+                hdr->data_csum_type != BLOCKSTORE_CSUM_CRC32C)
+            {
+                fprintf(stderr, "I don't know checksum format %u, the only supported format is crc32c = %u.\n", hdr->data_csum_type, BLOCKSTORE_CSUM_CRC32C);
+                free(data);
+                close(dsk.meta_fd);
+                dsk.meta_fd = -1;
+                return 1;
+            }
+        }
+        else
+        {
+            // Unsupported version
+            fprintf(stderr, "Metadata format is too new for me (stored version is %lu, max supported %u).\n", hdr->version, BLOCKSTORE_META_FORMAT_V2);
+            free(data);
+            close(dsk.meta_fd);
+            dsk.meta_fd = -1;
+            return 1;
+        }
         if (hdr->meta_block_size != dsk.meta_block_size)
         {
             fprintf(stderr, "Using block size of %u bytes based on information from the superblock\n", hdr->meta_block_size);
@@ -45,14 +71,24 @@ int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v1_t *)>
                 memcpy(new_data, data, dsk.meta_block_size);
                 free(data);
                 data = new_data;
-                hdr = (blockstore_meta_header_v1_t *)data;
+                hdr = (blockstore_meta_header_v2_t *)data;
             }
         }
+        dsk.meta_format = hdr->version;
+        dsk.data_block_size = hdr->data_block_size;
+        dsk.csum_block_size = hdr->csum_block_size;
+        dsk.data_csum_type = hdr->data_csum_type;
         dsk.bitmap_granularity = hdr->bitmap_granularity;
-        dsk.clean_entry_bitmap_size = hdr->data_block_size / hdr->bitmap_granularity / 8;
-        dsk.clean_entry_size = sizeof(clean_disk_entry) + 2*dsk.clean_entry_bitmap_size;
+        dsk.clean_entry_bitmap_size = (hdr->data_block_size / hdr->bitmap_granularity + 7) / 8;
+        dsk.clean_entry_size = sizeof(clean_disk_entry) + 2*dsk.clean_entry_bitmap_size
+            + (hdr->data_csum_type
+                ? ((hdr->data_block_size+hdr->csum_block_size-1)/hdr->csum_block_size
+                    *(hdr->data_csum_type & 0xff))
+                : 0)
+            + (dsk.meta_format == BLOCKSTORE_META_FORMAT_V2 ? 4 /*entry_csum*/ : 0);
         uint64_t block_num = 0;
         hdr_fn(hdr);
+        hdr = NULL;
         meta_pos = dsk.meta_block_size;
         lseek64(dsk.meta_fd, dsk.meta_offset+meta_pos, 0);
         while (meta_pos < dsk.meta_len)
@@ -67,6 +103,15 @@ int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v1_t *)>
                     clean_disk_entry *entry = (clean_disk_entry*)((uint8_t*)data + blk + ioff);
                     if (entry->oid.inode)
                     {
+                        if (dsk.data_csum_type)
+                        {
+                            uint32_t *entry_csum = (uint32_t*)((uint8_t*)entry + dsk.clean_entry_size - 4);
+                            if (*entry_csum != crc32c(0, entry, dsk.clean_entry_size - 4))
+                            {
+                                fprintf(stderr, "Metadata entry %lu is corrupt (checksum mismatch), skipping\n", block_num);
+                                continue;
+                            }
+                        }
                         record_fn(block_num, entry, entry->bitmap);
                     }
                 }
@@ -107,21 +152,35 @@ int disk_tool_t::process_meta(std::function<void(blockstore_meta_header_v1_t *)>
 int disk_tool_t::dump_meta()
 {
     int r = process_meta(
-        [this](blockstore_meta_header_v1_t *hdr) { dump_meta_header(hdr); },
+        [this](blockstore_meta_header_v2_t *hdr) { dump_meta_header(hdr); },
         [this](uint64_t block_num, clean_disk_entry *entry, uint8_t *bitmap) { dump_meta_entry(block_num, entry, bitmap); }
     );
-    printf("\n]}\n");
+    if (r == 0)
+        printf("\n]}\n");
     return r;
 }
 
-void disk_tool_t::dump_meta_header(blockstore_meta_header_v1_t *hdr)
+void disk_tool_t::dump_meta_header(blockstore_meta_header_v2_t *hdr)
 {
     if (hdr)
     {
-        printf(
-            "{\"version\":\"0.6\",\"meta_block_size\":%u,\"data_block_size\":%u,\"bitmap_granularity\":%u,\"entries\":[\n",
-            hdr->meta_block_size, hdr->data_block_size, hdr->bitmap_granularity
-        );
+        if (hdr->version == BLOCKSTORE_META_FORMAT_V1)
+        {
+            printf(
+                "{\"version\":\"0.6\",\"meta_block_size\":%u,\"data_block_size\":%u,\"bitmap_granularity\":%u,"
+                "\"entries\":[\n",
+                hdr->meta_block_size, hdr->data_block_size, hdr->bitmap_granularity
+            );
+        }
+        else if (hdr->version == BLOCKSTORE_META_FORMAT_V2)
+        {
+            printf(
+                "{\"version\":\"0.9\",\"meta_block_size\":%u,\"data_block_size\":%u,\"bitmap_granularity\":%u,"
+                "\"data_csum_type\":%s,\"csum_block_size\":%u,\"entries\":[\n",
+                hdr->meta_block_size, hdr->data_block_size, hdr->bitmap_granularity,
+                csum_type_str(hdr->data_csum_type).c_str(), hdr->csum_block_size
+            );
+        }
     }
     else
     {
@@ -151,6 +210,15 @@ void disk_tool_t::dump_meta_entry(uint64_t block_num, clean_disk_entry *entry, u
         {
             printf("%02x", bitmap[dsk.clean_entry_bitmap_size + i]);
         }
+        if (dsk.csum_block_size && dsk.data_csum_type)
+        {
+            uint8_t *csums = bitmap + dsk.clean_entry_bitmap_size*2;
+            printf("\",\"block_csums\":\"");
+            for (uint64_t i = 0; i < (dsk.data_block_size+dsk.csum_block_size-1)/dsk.csum_block_size*(dsk.data_csum_type & 0xFF); i++)
+            {
+                printf("%02x", csums[i]);
+            }
+        }
         printf("\"}");
     }
     else
@@ -164,18 +232,30 @@ int disk_tool_t::write_json_meta(json11::Json meta)
 {
     new_meta_buf = (uint8_t*)memalign_or_die(MEM_ALIGNMENT, new_meta_len);
     memset(new_meta_buf, 0, new_meta_len);
-    blockstore_meta_header_v1_t *new_hdr = (blockstore_meta_header_v1_t *)new_meta_buf;
+    blockstore_meta_header_v2_t *new_hdr = (blockstore_meta_header_v2_t *)new_meta_buf;
     new_hdr->zero = 0;
     new_hdr->magic = BLOCKSTORE_META_MAGIC_V1;
-    new_hdr->version = BLOCKSTORE_META_VERSION_V1;
+    new_hdr->version = meta["version"].uint64_value() == BLOCKSTORE_META_FORMAT_V1
+        ? BLOCKSTORE_META_FORMAT_V1 : BLOCKSTORE_META_FORMAT_V2;
     new_hdr->meta_block_size = meta["meta_block_size"].uint64_value()
         ? meta["meta_block_size"].uint64_value() : 4096;
     new_hdr->data_block_size = meta["data_block_size"].uint64_value()
         ? meta["data_block_size"].uint64_value() : 131072;
     new_hdr->bitmap_granularity = meta["bitmap_granularity"].uint64_value()
         ? meta["bitmap_granularity"].uint64_value() : 4096;
-    new_clean_entry_bitmap_size = new_hdr->data_block_size / new_hdr->bitmap_granularity / 8;
-    new_clean_entry_size = sizeof(clean_disk_entry) + 2*new_clean_entry_bitmap_size;
+    new_hdr->data_csum_type = meta["data_csum_type"].is_number()
+        ? meta["data_csum_type"].uint64_value()
+        : (meta["data_csum_type"].string_value() == "crc32c"
+            ? BLOCKSTORE_CSUM_CRC32C
+            : BLOCKSTORE_CSUM_NONE);
+    new_hdr->csum_block_size = meta["csum_block_size"].uint64_value();
+    uint32_t new_clean_entry_header_size = (new_hdr->version == BLOCKSTORE_META_FORMAT_V1
+        ? sizeof(clean_disk_entry) : sizeof(clean_disk_entry) + 4 /*entry_csum*/);
+    new_clean_entry_bitmap_size = (new_hdr->data_block_size / new_hdr->bitmap_granularity + 7) / 8;
+    new_data_csum_size = (new_hdr->data_csum_type
+        ? ((new_hdr->data_block_size+new_hdr->csum_block_size-1)/new_hdr->csum_block_size*(new_hdr->data_csum_type & 0xFF))
+        : 0);
+    new_clean_entry_size = new_clean_entry_header_size + 2*new_clean_entry_bitmap_size + new_data_csum_size;
     new_entries_per_block = new_hdr->meta_block_size / new_clean_entry_size;
     for (const auto & e: meta["entries"].array_items())
     {
@@ -194,8 +274,21 @@ int disk_tool_t::write_json_meta(json11::Json meta)
         new_entry->oid.inode = (sscanf_json(NULL, e["pool"]) << (64-POOL_ID_BITS)) | sscanf_json(NULL, e["inode"]);
         new_entry->oid.stripe = sscanf_json(NULL, e["stripe"]);
         new_entry->version = sscanf_json(NULL, e["version"]);
-        fromhexstr(e["bitmap"].string_value(), new_clean_entry_bitmap_size, ((uint8_t*)new_entry) + sizeof(clean_disk_entry));
-        fromhexstr(e["ext_bitmap"].string_value(), new_clean_entry_bitmap_size, ((uint8_t*)new_entry) + sizeof(clean_disk_entry) + new_clean_entry_bitmap_size);
+        fromhexstr(e["bitmap"].string_value(), new_clean_entry_bitmap_size,
+            ((uint8_t*)new_entry) + sizeof(clean_disk_entry));
+        fromhexstr(e["ext_bitmap"].string_value(), new_clean_entry_bitmap_size,
+            ((uint8_t*)new_entry) + sizeof(clean_disk_entry) + new_clean_entry_bitmap_size);
+        if (new_hdr->version == BLOCKSTORE_META_FORMAT_V2)
+        {
+            if (new_hdr->data_csum_type != 0)
+            {
+                fromhexstr(e["data_csum"].string_value(), new_data_csum_size,
+                    ((uint8_t*)new_entry) + sizeof(clean_disk_entry) + 2*new_clean_entry_bitmap_size);
+            }
+            uint32_t *new_entry_csum = (uint32_t*)(((uint8_t*)new_entry) + sizeof(clean_disk_entry) +
+                2*new_clean_entry_bitmap_size + new_data_csum_size);
+            *new_entry_csum = crc32c(0, new_entry, new_clean_entry_size - 4);
+        }
     }
     int r = resize_write_new_meta();
     free(new_meta_buf);

src/disk_tool_prepare.cpp

@@ -8,6 +8,9 @@
 int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_hdd)
 {
     static const char *allow_additional_params[] = {
+        "cached_read_data",
+        "cached_read_meta",
+        "cached_read_journal",
         "max_write_iodepth",
         "max_write_iodepth",
         "min_flusher_count",
@@ -99,15 +102,16 @@ int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_
     if (options["journal_size"] == "")
     {
         if (options["journal_device"] == "")
-            options["journal_size"] = "32M";
+            options["journal_size"] = is_hdd ? "128M" : "32M";
         else if (is_hdd)
             options["journal_size"] = DEFAULT_HYBRID_JOURNAL;
     }
+    bool is_hybrid = is_hdd && options["journal_device"] != "" && options["journal_device"] != options["data_device"];
     if (is_hdd)
     {
         if (options["block_size"] == "")
             options["block_size"] = "1M";
-        if (options["throttle_small_writes"] == "")
+        if (is_hybrid && options["throttle_small_writes"] == "")
             options["throttle_small_writes"] = "1";
     }
     json11::Json::object sb;
@@ -115,6 +119,7 @@ int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_
     try
     {
         dsk.parse_config(options);
+        dsk.cached_read_data = dsk.cached_read_meta = dsk.cached_read_journal = false;
         dsk.open_data();
         dsk.open_meta();
         dsk.open_journal();
@@ -134,7 +139,7 @@ int disk_tool_t::prepare_one(std::map<std::string, std::string> options, int is_
             { "meta_offset", 4096 + (dsk.meta_device == dsk.journal_device ? dsk.journal_len : 0) },
             { "data_offset", 4096 + (dsk.data_device == dsk.meta_device ? dsk.meta_len : 0) +
                 (dsk.data_device == dsk.journal_device ? dsk.journal_len : 0) },
-            { "journal_no_same_sector_overwrites", true },
+            { "journal_no_same_sector_overwrites", !is_hdd || is_hybrid },
             { "journal_sector_buffer_count", 1024 },
             { "disable_data_fsync", json_is_true(options["disable_data_fsync"]) },
             { "disable_meta_fsync", json_is_true(options["disable_meta_fsync"]) },
@@ -478,6 +483,7 @@ int disk_tool_t::get_meta_partition(std::vector<vitastor_dev_info_t> & ssds, std
     {
         blockstore_disk_t dsk;
         dsk.parse_config(options);
+        dsk.cached_read_data = dsk.cached_read_meta = dsk.cached_read_journal = false;
         dsk.open_data();
         dsk.open_meta();
         dsk.open_journal();

src/disk_tool_resize.cpp

@@ -29,7 +29,7 @@ int disk_tool_t::resize_data()
     fprintf(stderr, "Reading metadata\n");
     data_alloc = new allocator((new_data_len < dsk.data_len ? dsk.data_len : new_data_len) / dsk.data_block_size);
     r = process_meta(
-        [this](blockstore_meta_header_v1_t *hdr)
+        [this](blockstore_meta_header_v2_t *hdr)
         {
             resize_init(hdr);
         },
@@ -91,6 +91,7 @@ int disk_tool_t::resize_parse_params()
     try
     {
         dsk.parse_config(options);
+        dsk.cached_read_data = dsk.cached_read_meta = dsk.cached_read_journal = false;
         dsk.open_data();
         dsk.open_meta();
         dsk.open_journal();
@@ -139,7 +140,7 @@ int disk_tool_t::resize_parse_params()
     return 0;
 }
 
-void disk_tool_t::resize_init(blockstore_meta_header_v1_t *hdr)
+void disk_tool_t::resize_init(blockstore_meta_header_v2_t *hdr)
 {
     if (hdr && dsk.data_block_size != hdr->data_block_size)
     {
@@ -149,6 +150,15 @@ void disk_tool_t::resize_init(blockstore_meta_header_v1_t *hdr)
         }
         dsk.data_block_size = hdr->data_block_size;
     }
+    if (hdr && (dsk.data_csum_type != hdr->data_csum_type || dsk.csum_block_size != hdr->csum_block_size))
+    {
+        if (dsk.data_csum_type)
+        {
+            fprintf(stderr, "Using data checksum type %s from metadata superblock\n", csum_type_str(hdr->data_csum_type).c_str());
+        }
+        dsk.data_csum_type = hdr->data_csum_type;
+        dsk.csum_block_size = hdr->csum_block_size;
+    }
     if (((new_data_len-dsk.data_len) % dsk.data_block_size) ||
         ((new_data_offset-dsk.data_offset) % dsk.data_block_size))
     {
@@ -160,8 +170,12 @@ void disk_tool_t::resize_init(blockstore_meta_header_v1_t *hdr)
     free_last = (new_data_offset+new_data_len < dsk.data_offset+dsk.data_len)
         ? (dsk.data_offset+dsk.data_len-new_data_offset-new_data_len) / dsk.data_block_size
         : 0;
+    uint32_t new_clean_entry_header_size = sizeof(clean_disk_entry) + 4 /*entry_csum*/;
     new_clean_entry_bitmap_size = dsk.data_block_size / (hdr ? hdr->bitmap_granularity : 4096) / 8;
-    new_clean_entry_size = sizeof(clean_disk_entry) + 2 * new_clean_entry_bitmap_size;
+    new_data_csum_size = (dsk.data_csum_type
+        ? ((dsk.data_block_size+dsk.csum_block_size-1)/dsk.csum_block_size*(dsk.data_csum_type & 0xFF))
+        : 0);
+    new_clean_entry_size = new_clean_entry_header_size + 2*new_clean_entry_bitmap_size + new_data_csum_size;
     new_entries_per_block = dsk.meta_block_size/new_clean_entry_size;
     uint64_t new_meta_blocks = 1 + (new_data_len/dsk.data_block_size + new_entries_per_block-1) / new_entries_per_block;
     if (!new_meta_len)
@@ -349,13 +363,25 @@ int disk_tool_t::resize_rewrite_journal()
         {
             if (je->type == JE_START)
             {
+                if (je_start.data_csum_type != dsk.data_csum_type ||
+                    je_start.csum_block_size != dsk.csum_block_size)
+                {
+                    fprintf(
+                        stderr, "Error: journal header has different checksum parameters: %s/%u vs %s/%u\n",
+                        csum_type_str(je_start.data_csum_type).c_str(), je_start.csum_block_size,
+                        csum_type_str(dsk.data_csum_type).c_str(), dsk.csum_block_size
+                    );
+                    exit(1);
+                }
                 journal_entry *ne = (journal_entry*)(new_journal_ptr + new_journal_in_pos);
                 *((journal_entry_start*)ne) = (journal_entry_start){
                     .magic = JOURNAL_MAGIC,
                     .type = JE_START,
                     .size = sizeof(journal_entry_start),
                     .journal_start = dsk.journal_block_size,
-                    .version = JOURNAL_VERSION,
+                    .version = JOURNAL_VERSION_V2,
+                    .data_csum_type = dsk.data_csum_type,
+                    .csum_block_size = dsk.csum_block_size,
                 };
                 ne->crc32 = je_crc32(ne);
                 new_journal_ptr += dsk.journal_block_size;
@@ -436,15 +462,17 @@ int disk_tool_t::resize_rewrite_meta()
     new_meta_buf = (uint8_t*)memalign_or_die(MEM_ALIGNMENT, new_meta_len);
     memset(new_meta_buf, 0, new_meta_len);
     int r = process_meta(
-        [this](blockstore_meta_header_v1_t *hdr)
+        [this](blockstore_meta_header_v2_t *hdr)
         {
-            blockstore_meta_header_v1_t *new_hdr = (blockstore_meta_header_v1_t *)new_meta_buf;
+            blockstore_meta_header_v2_t *new_hdr = (blockstore_meta_header_v2_t *)new_meta_buf;
             new_hdr->zero = 0;
             new_hdr->magic = BLOCKSTORE_META_MAGIC_V1;
-            new_hdr->version = BLOCKSTORE_META_VERSION_V1;
+            new_hdr->version = BLOCKSTORE_META_FORMAT_V1;
             new_hdr->meta_block_size = dsk.meta_block_size;
             new_hdr->data_block_size = dsk.data_block_size;
             new_hdr->bitmap_granularity = dsk.bitmap_granularity ? dsk.bitmap_granularity : 4096;
+            new_hdr->data_csum_type = dsk.data_csum_type;
+            new_hdr->csum_block_size = dsk.csum_block_size;
         },
         [this](uint64_t block_num, clean_disk_entry *entry, uint8_t *bitmap)
         {
@@ -463,7 +491,7 @@ int disk_tool_t::resize_rewrite_meta()
             new_entry->oid = entry->oid;
             new_entry->version = entry->version;
             if (bitmap)
-                memcpy(new_entry->bitmap, bitmap, 2*new_clean_entry_bitmap_size);
+                memcpy(new_entry->bitmap, bitmap, 2*new_clean_entry_bitmap_size + new_data_csum_size);
             else
                 memset(new_entry->bitmap, 0xff, 2*new_clean_entry_bitmap_size);
         }

src/disk_tool_utils.cpp

@@ -373,3 +373,22 @@ int fix_partition_type(std::string dev_by_uuid)
     std::string out;
     return shell_exec({ "sfdisk", "--no-reread", "--force", "/dev/"+parent_dev }, script, &out, NULL);
 }
+
+std::string csum_type_str(uint32_t data_csum_type)
+{
+    std::string csum_type;
+    if (data_csum_type == BLOCKSTORE_CSUM_NONE)
+        csum_type = "none";
+    else if (data_csum_type == BLOCKSTORE_CSUM_CRC32C)
+        csum_type = "crc32c";
+    else
+        csum_type = std::to_string(data_csum_type);
+    return csum_type;
+}
+
+uint32_t csum_type_from_str(std::string data_csum_type)
+{
+    if (data_csum_type == "crc32c")
+        return BLOCKSTORE_CSUM_CRC32C;
+    return stoull_full(data_csum_type, 0);
+}

src/etcd_state_client.cpp

@@ -187,22 +187,30 @@ void etcd_state_client_t::add_etcd_url(std::string addr)
             check_addr = addr;
         if (pos == std::string::npos)
             addr += "/v3";
+        bool local = false;
         int i;
         for (i = 0; i < local_ips.size(); i++)
         {
             if (local_ips[i] == check_addr)
             {
-                this->etcd_local.push_back(addr);
+                local = true;
                 break;
             }
         }
-        if (i >= local_ips.size())
-            this->etcd_addresses.push_back(addr);
+        auto & to = local ? this->etcd_local : this->etcd_addresses;
+        for (i = 0; i < to.size(); i++)
+        {
+            if (to[i] == addr)
+                break;
+        }
+        if (i >= to.size())
+            to.push_back(addr);
     }
 }
 
 void etcd_state_client_t::parse_config(const json11::Json & config)
 {
+    this->etcd_local.clear();
     this->etcd_addresses.clear();
     if (config["etcd_address"].is_string())
     {
@@ -349,7 +357,7 @@ void etcd_state_client_t::start_etcd_watcher()
                         watch_id == ETCD_OSD_STATE_WATCH_ID)
                         etcd_watches_initialised++;
                     if (etcd_watches_initialised == 4 && this->log_level > 0)
-                        fprintf(stderr, "Successfully subscribed to etcd at %s\n", selected_etcd_address.c_str());
+                        fprintf(stderr, "Successfully subscribed to etcd at %s\n", cur_addr.c_str());
                 }
                 if (data["result"]["canceled"].bool_value())
                 {
@@ -360,15 +368,17 @@ void etcd_state_client_t::start_etcd_watcher()
                         // so we should restart from the beginning if we can
                         if (on_reload_hook != NULL)
                         {
-                            fprintf(stderr, "Revisions before %lu were compacted by etcd, reloading state\n",
-                                data["result"]["compact_revision"].uint64_value());
-                            if (etcd_watch_ws)
+                            // check to not trigger on_reload_hook multiple times
+                            if (etcd_watch_ws != NULL)
                             {
+                                fprintf(stderr, "Revisions before %lu were compacted by etcd, reloading state\n",
+                                    data["result"]["compact_revision"].uint64_value());
                                 http_close(etcd_watch_ws);
                                 etcd_watch_ws = NULL;
+                                etcd_watch_revision = 0;
+                                on_reload_hook();
                             }
-                            etcd_watch_revision = 0;
-                            on_reload_hook();
+                            return;
                         }
                         else
                         {
@@ -415,13 +425,9 @@ void etcd_state_client_t::start_etcd_watcher()
         }
         if (msg->eof)
         {
+            fprintf(stderr, "Disconnected from etcd %s\n", cur_addr.c_str());
             if (cur_addr == selected_etcd_address)
-            {
-                fprintf(stderr, "Disconnected from etcd %s\n", selected_etcd_address.c_str());
                 selected_etcd_address = "";
-            }
-            else
-                fprintf(stderr, "Disconnected from etcd\n");
             if (etcd_watch_ws)
             {
                 http_close(etcd_watch_ws);
@@ -438,6 +444,7 @@ void etcd_state_client_t::start_etcd_watcher()
             else if (etcd_watches_initialised > 0)
             {
                 // Connection was live, retry immediately
+                etcd_watches_initialised = 0;
                 start_etcd_watcher();
             }
         }

src/messenger.h

@@ -173,6 +173,8 @@ public:
     bool connect_rdma(int peer_fd, std::string rdma_address, uint64_t client_max_msg);
 #endif
 
+    void measure_exec(osd_op_t *cur_op);
+
 protected:
     void try_connect_peer(uint64_t osd_num);
     void try_connect_peer_addr(osd_num_t peer_osd, const char *peer_host, int peer_port);
@@ -184,7 +186,6 @@ protected:
     void cancel_op(osd_op_t *op);
 
     bool try_send(osd_client_t *cl);
-    void measure_exec(osd_op_t *cur_op);
     void handle_send(int result, osd_client_t *cl);
 
     bool handle_read(int result, osd_client_t *cl);

src/msgr_op.cpp

@@ -9,6 +9,10 @@ osd_op_t::~osd_op_t()
 {
     assert(!bs_op);
     assert(!op_data);
+    if (bitmap_buf)
+    {
+        free(bitmap_buf);
+    }
     if (rmw_buf)
     {
         free(rmw_buf);