Release 0.9.2

- Measure and report scrub I/O statistics in vitastor-cli status
- Make aggregated statistics in vitastor-cli status much smoother
  (first derive, then sum instead of first summing and then deriving)
- Fix an old rare bug leading to journal corruption
  (try to use scrub if you think you're affected...)
- Do not start EC PGs without at least <data chunks> OSDs in each old set
  (prevents spurious read errors with EC during reconnections/restarts)
- Fix failed assert(!scrub_list_op) on OSD restart with pending scrubs
- Fix future planned scrubs not starting because of incorrect time comparison
- Build packages for Debian 12 (Bookworm)

.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'
@@ -323,7 +323,7 @@ jobs:
     steps:
     - name: Run test
       id: test
-      timeout-minutes: 3
+      timeout-minutes: 10
       run: /root/vitastor/tests/test_rebalance_verify.sh
     - name: Print logs
       if: always() && steps.test.outcome == 'failure'
@@ -341,7 +341,7 @@ jobs:
     steps:
     - name: Run test
       id: test
-      timeout-minutes: 3
+      timeout-minutes: 10
       run: IMMEDIATE_COMMIT=1 /root/vitastor/tests/test_rebalance_verify.sh
     - name: Print logs
       if: always() && steps.test.outcome == 'failure'
@@ -359,7 +359,7 @@ jobs:
     steps:
     - name: Run test
       id: test
-      timeout-minutes: 3
+      timeout-minutes: 10
       run: SCHEME=ec /root/vitastor/tests/test_rebalance_verify.sh
     - name: Print logs
       if: always() && steps.test.outcome == 'failure'
@@ -377,7 +377,7 @@ jobs:
     steps:
     - name: Run test
       id: test
-      timeout-minutes: 3
+      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'

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

@@ -9,7 +9,8 @@ for my $line (<>)
         chomp $line;
         my $test_name = $1;
         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;
         }

CMakeLists.txt

@@ -2,6 +2,6 @@ cmake_minimum_required(VERSION 2.8.12)
 
 project(vitastor)
 
-set(VERSION "0.9.1")
+set(VERSION "0.9.2")
 
 add_subdirectory(src)

csi/Makefile

@@ -1,4 +1,4 @@
-VERSION ?= v0.9.1
+VERSION ?= v0.9.2
 
 all: build push
 

csi/deploy/004-csi-nodeplugin.yaml

@@ -49,7 +49,7 @@ spec:
             capabilities:
               add: ["SYS_ADMIN"]
             allowPrivilegeEscalation: true
-          image: vitalif/vitastor-csi:v0.9.1
+          image: vitalif/vitastor-csi:v0.9.2
           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:v0.9.1
+          image: vitalif/vitastor-csi:v0.9.2
           args:
             - "--node=$(NODE_ID)"
             - "--endpoint=$(CSI_ENDPOINT)"

csi/src/config.go

@@ -5,7 +5,7 @@ package vitastor
 
 const (
     vitastorCSIDriverName    = "csi.vitastor.io"
-    vitastorCSIDriverVersion = "0.9.1"
+    vitastorCSIDriverVersion = "0.9.2"
 )
 
 // Config struct fills the parameters of request or user input

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 (0.9.1-1) unstable; urgency=medium
+vitastor (0.9.2-1) unstable; urgency=medium
 
   * Bugfixes
 
  -- Vitaliy Filippov <vitalif@yourcmc.ru>  Fri, 03 Jun 2022 02:09:44 +0300
 
-vitastor (0.9.1-1) unstable; urgency=medium
+vitastor (0.9.2-1) 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 qemu 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; \
@@ -37,23 +39,13 @@ 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+).*!$1!'); \
+    cp /root/vitastor/patches/qemu-$QEMU_VER-vitastor.patch qemu-*/debian/patches; \
+    echo qemu-$QEMU_VER-vitastor.patch >> qemu-*/debian/patches/series; \
     cd /root/packages/qemu-$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; \
     DEBEMAIL="Vitaliy Filippov <vitalif@yourcmc.ru>" dch -D $REL -v $V 'Plug Vitastor block driver'; \

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-$REL; \
-    cp -r /root/vitastor vitastor-0.9.1; \
-    cd vitastor-0.9.1; \
+    cp -r /root/vitastor vitastor-0.9.2; \
+    cd vitastor-0.9.2; \
     ln -s /root/fio-build/fio-*/ ./fio; \
     FIO=$(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-$REL; \
-    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.9.1.orig.tar.xz vitastor-0.9.1; \
-    cd vitastor-0.9.1; \
+    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.9.2.orig.tar.xz vitastor-0.9.2; \
+    cd vitastor-0.9.2; \
     V=$(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; \

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()
@@ -1627,7 +1638,8 @@ class Mon
             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);
@@ -1743,13 +1755,14 @@ 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/cinder-vitastor.py

@@ -50,7 +50,7 @@ from cinder.volume import configuration
 from cinder.volume import driver
 from cinder.volume import volume_utils
 
-VERSION = '0.9.1'
+VERSION = '0.9.2'
 
 LOG = logging.getLogger(__name__)
 

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-0.9.1/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.9.1$(rpm --eval '%dist').tar.gz *
+tar --transform 's#^#vitastor-0.9.2/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.9.2$(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-0.9.1.el7.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.9.2.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:        0.9.1
+Version:        0.9.2
 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-0.9.1.el7.tar.gz
+Source0:        vitastor-0.9.2.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-0.9.1.el8.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.9.2.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:        0.9.1
+Version:        0.9.2
 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-0.9.1.el8.tar.gz
+Source0:        vitastor-0.9.2.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-0.9.1.el9.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.9.2.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:        0.9.1
+Version:        0.9.2
 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-0.9.1.el9.tar.gz
+Source0:        vitastor-0.9.2.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="0.9.1")
+add_definitions(-DVERSION="0.9.2")
 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/blockstore_flush.cpp

@@ -675,53 +675,58 @@ resume_1:
                     return false;
                 }
                 flusher->trimming = true;
-                // First update journal "superblock" and only then update <used_start> in memory
-                await_sqe(12);
-                *((journal_entry_start*)flusher->journal_superblock) = {
-                    .crc32 = 0,
-                    .magic = JOURNAL_MAGIC,
-                    .type = JE_START,
-                    .size = sizeof(journal_entry_start),
-                    .reserved = 0,
-                    .journal_start = new_trim_pos,
-                    .version = JOURNAL_VERSION,
-                };
-                ((journal_entry_start*)flusher->journal_superblock)->crc32 = je_crc32((journal_entry*)flusher->journal_superblock);
-                data->iov = (struct iovec){ flusher->journal_superblock, bs->dsk.journal_block_size };
-                data->callback = simple_callback_w;
-                my_uring_prep_writev(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset);
-                wait_count++;
-            resume_13:
-                if (wait_count > 0)
+                // Recheck the position with the "lock" taken
+                new_trim_pos = bs->journal.get_trim_pos();
+                if (new_trim_pos != bs->journal.used_start)
                 {
-                    wait_state = 13;
-                    return false;
-                }
-                if (!bs->disable_journal_fsync)
-                {
-                    await_sqe(20);
-                    my_uring_prep_fsync(sqe, bs->dsk.journal_fd, IORING_FSYNC_DATASYNC);
-                    data->iov = { 0 };
+                    // First update journal "superblock" and only then update <used_start> in memory
+                    await_sqe(12);
+                    *((journal_entry_start*)flusher->journal_superblock) = {
+                        .crc32 = 0,
+                        .magic = JOURNAL_MAGIC,
+                        .type = JE_START,
+                        .size = sizeof(journal_entry_start),
+                        .reserved = 0,
+                        .journal_start = new_trim_pos,
+                        .version = JOURNAL_VERSION,
+                    };
+                    ((journal_entry_start*)flusher->journal_superblock)->crc32 = je_crc32((journal_entry*)flusher->journal_superblock);
+                    data->iov = (struct iovec){ flusher->journal_superblock, bs->dsk.journal_block_size };
                     data->callback = simple_callback_w;
-                resume_21:
+                    my_uring_prep_writev(sqe, bs->dsk.journal_fd, &data->iov, 1, bs->journal.offset);
+                    wait_count++;
+                resume_13:
                     if (wait_count > 0)
                     {
-                        wait_state = 21;
+                        wait_state = 13;
                         return false;
                     }
-                }
-                bs->journal.used_start = new_trim_pos;
+                    if (!bs->disable_journal_fsync)
+                    {
+                        await_sqe(20);
+                        my_uring_prep_fsync(sqe, bs->dsk.journal_fd, IORING_FSYNC_DATASYNC);
+                        data->iov = { 0 };
+                        data->callback = simple_callback_w;
+                    resume_21:
+                        if (wait_count > 0)
+                        {
+                            wait_state = 21;
+                            return false;
+                        }
+                    }
+                    bs->journal.used_start = new_trim_pos;
 #ifdef BLOCKSTORE_DEBUG
-                printf("Journal trimmed to %08lx (next_free=%08lx)\n", bs->journal.used_start, bs->journal.next_free);
+                    printf("Journal trimmed to %08lx (next_free=%08lx)\n", bs->journal.used_start, bs->journal.next_free);
 #endif
+                    if (bs->journal.flush_journal && !flusher->flush_queue.size())
+                    {
+                        assert(bs->journal.used_start == bs->journal.next_free);
+                        printf("Journal flushed\n");
+                        exit(0);
+                    }
+                }
                 flusher->trimming = false;
             }
-            if (bs->journal.flush_journal && !flusher->flush_queue.size())
-            {
-                assert(bs->journal.used_start == bs->journal.next_free);
-                printf("Journal flushed\n");
-                exit(0);
-            }
         }
         // All done
         flusher->active_flushers--;

src/blockstore_journal.cpp

@@ -236,14 +236,6 @@ 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 +248,26 @@ 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

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_write.cpp

@@ -409,7 +409,23 @@ 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;

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/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_send.cpp

@@ -150,8 +150,10 @@ void osd_messenger_t::measure_exec(osd_op_t *cur_op)
         (cur_op->tv_end.tv_nsec - cur_op->tv_begin.tv_nsec)/1000
     );
     if (cur_op->req.hdr.opcode == OSD_OP_READ ||
-        cur_op->req.hdr.opcode == OSD_OP_WRITE)
+        cur_op->req.hdr.opcode == OSD_OP_WRITE ||
+        cur_op->req.hdr.opcode == OSD_OP_SCRUB)
     {
+        // req.rw.len is internally set to the full object size for scrubs
         stats.op_stat_bytes[cur_op->req.hdr.opcode] += cur_op->req.rw.len;
     }
     else if (cur_op->req.hdr.opcode == OSD_OP_SEC_READ ||

src/osd_peering.cpp

@@ -240,21 +240,20 @@ void osd_t::start_pg_peering(pg_t & pg)
         // (PG history is kept up to the latest active+clean state)
         for (auto & history_set: pg.target_history)
         {
-            bool found = true;
+            int nonzero = 0, found = 0;
             for (auto history_osd: history_set)
             {
                 if (history_osd != 0)
                 {
-                    found = false;
+                    nonzero++;
                     if (history_osd == this->osd_num ||
                         msgr.osd_peer_fds.find(history_osd) != msgr.osd_peer_fds.end())
                     {
-                        found = true;
-                        break;
+                        found++;
                     }
                 }
             }
-            if (!found)
+            if (found < (nonzero >= pg.pg_data_size ? pg.pg_data_size : 1))
             {
                 pg.state = PG_INCOMPLETE;
                 report_pg_state(pg);
@@ -486,6 +485,10 @@ void osd_t::report_pg_state(pg_t & pg)
 {
     pg.print_state();
     this->pg_state_dirty.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
+    if (pg.state & PG_ACTIVE)
+    {
+        plan_scrub(pg, false);
+    }
     if (pg.state == PG_ACTIVE && (pg.target_history.size() > 0 || pg.all_peers.size() > pg.target_set.size()))
     {
         // Clear history of active+clean PGs
@@ -494,7 +497,6 @@ void osd_t::report_pg_state(pg_t & pg)
         pg.all_peers = pg.target_set;
         std::sort(pg.all_peers.begin(), pg.all_peers.end());
         pg.cur_peers = pg.target_set;
-        plan_scrub(pg, false);
         // Change pg_config at the same time, otherwise our PG reconciling loop may try to apply the old metadata
         auto & pg_cfg = st_cli.pool_config[pg.pool_id].pg_config[pg.pg_num];
         pg_cfg.target_history = pg.target_history;
@@ -538,7 +540,6 @@ void osd_t::report_pg_state(pg_t & pg)
                 pg.cur_peers.push_back(pg_osd);
             }
         }
-        plan_scrub(pg, false);
         auto & pg_cfg = st_cli.pool_config[pg.pool_id].pg_config[pg.pg_num];
         pg_cfg.target_history = pg.target_history;
         pg_cfg.all_peers = pg.all_peers;

src/osd_primary_subops.cpp

@@ -87,6 +87,7 @@ void osd_t::finish_op(osd_op_t *cur_op, int retval)
     cur_op->reply.hdr.retval = retval;
     if (cur_op->peer_fd == -1)
     {
+        msgr.measure_exec(cur_op);
         // Copy lambda to be unaffected by `delete op`
         std::function<void(osd_op_t*)>(cur_op->callback)(cur_op);
     }

src/osd_scrub.cpp

@@ -115,6 +115,10 @@ int osd_t::pick_next_scrub(object_id & next_oid)
         }
         return 0;
     }
+    if (scrub_list_op)
+    {
+        return 1;
+    }
     timespec tv_now;
     clock_gettime(CLOCK_REALTIME, &tv_now);
     bool rescan = scrub_last_pg.pool_id != 0 || scrub_last_pg.pg_num != 0;
@@ -127,7 +131,7 @@ int osd_t::pick_next_scrub(object_id & next_oid)
     }
     while (pg_it != pgs.end())
     {
-        if ((pg_it->second.state & PG_ACTIVE) && pg_it->second.next_scrub && pg_it->second.next_scrub < tv_now.tv_sec)
+        if ((pg_it->second.state & PG_ACTIVE) && pg_it->second.next_scrub && pg_it->second.next_scrub <= tv_now.tv_sec)
         {
             // Continue scrubbing from the next object
             if (scrub_last_pg == pg_it->first)
@@ -346,7 +350,7 @@ void osd_t::schedule_scrub(pg_t & pg)
             tfd->clear_timer(scrub_timer_id);
             scrub_timer_id = -1;
         }
-        if (tv_now.tv_sec > scrub_nearest_ts)
+        if (tv_now.tv_sec >= scrub_nearest_ts)
         {
             scrub_nearest_ts = 0;
             peering_state = peering_state | OSD_SCRUBBING;
@@ -376,6 +380,7 @@ void osd_t::continue_primary_scrub(osd_op_t *cur_op)
         goto resume_2;
     {
         auto & pg = pgs.at({ .pool_id = INODE_POOL(op_data->oid.inode), .pg_num = op_data->pg_num });
+        cur_op->req.rw.len = bs_block_size * pg.pg_data_size;
         // Determine version
         auto vo_it = pg.ver_override.find(op_data->oid);
         op_data->target_ver = vo_it != pg.ver_override.end() ? vo_it->second : UINT64_MAX;

src/str_util.cpp

@@ -300,3 +300,11 @@ std::string read_all_fd(int fd)
     res.resize(res_size);
     return res;
 }
+
+std::string str_repeat(const std::string & str, int times)
+{
+    std::string r;
+    for (int i = 0; i < times; i++)
+        r += str;
+    return r;
+}

src/str_util.h

@@ -17,3 +17,4 @@ std::string format_size(uint64_t size, bool nobytes = false);
 void print_help(const char *help_text, std::string exe_name, std::string cmd, bool all);
 uint64_t parse_time(std::string time_str, bool *ok = NULL);
 std::string read_all_fd(int fd);
+std::string str_repeat(const std::string & str, int times);

src/vitastor.pc.in

@@ -6,7 +6,7 @@ includedir=${prefix}/@CMAKE_INSTALL_INCLUDEDIR@
 
 Name: Vitastor
 Description: Vitastor client library
-Version: 0.9.1
+Version: 0.9.2
 Libs: -L${libdir} -lvitastor_client
 Cflags: -I${includedir}
 

tests/test_change_pg_count.sh

@@ -85,7 +85,7 @@ try_change 16
 
 # Monitor should report non-zero overall statistics at least once
 
-if ! (grep /vitastor/stats ./testdata/mon.log | jq -s -e '[ .[] | select((.kv.value.op_stats.primary_write.count | tonumber) > 0) ] | length > 0'); then
+if ! (grep /vitastor/stats ./testdata/mon.log | jq -s -e '[ .[] | select((.kv.value.op_stats.primary_write.count // 0 | tonumber) > 0) ] | length > 0'); then
     format_error "FAILED: monitor doesn't aggregate stats"
 fi