Release 0.5.12

Fix qemu driver broken in 0.5.11 :)

.gitignore

@@ -0,0 +1,18 @@
+*.o
+*.so
+package-lock.json
+fio
+qemu
+osd
+stub_osd
+stub_uring_osd
+stub_bench
+osd_test
+osd_peering_pg_test
+dump_journal
+nbd_proxy
+rm_inode
+test_allocator
+test_blockstore
+test_shit
+osd_rmw_test

debian/build-vitastor-bullseye.sh

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-sed 's/$REL/bullseye/' < vitastor.Dockerfile > ../Dockerfile
+sed 's/$REL/bullseye/g' < vitastor.Dockerfile > ../Dockerfile
 cd ..
 mkdir -p packages
 sudo podman build -v `pwd`/packages:/root/packages -f Dockerfile .

debian/build-vitastor-buster.sh

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-sed 's/$REL/buster/' < vitastor.Dockerfile > ../Dockerfile
+sed 's/$REL/buster/g' < vitastor.Dockerfile > ../Dockerfile
 cd ..
 mkdir -p packages
 sudo podman build -v `pwd`/packages:/root/packages -f Dockerfile .

debian/changelog

@@ -1,4 +1,4 @@
-vitastor (0.5.6-1) unstable; urgency=medium
+vitastor (0.5.12-1) unstable; urgency=medium
 
   * Bugfixes
 

debian/vitastor.Dockerfile

@@ -40,10 +40,10 @@ RUN set -e -x; \
     mkdir -p /root/packages/vitastor-$REL; \
     rm -rf /root/packages/vitastor-$REL/*; \
     cd /root/packages/vitastor-$REL; \
-    cp -r /root/vitastor vitastor-0.5.6; \
-    ln -s /root/packages/qemu-$REL/qemu-*/ vitastor-0.5.6/qemu; \
-    ln -s /root/fio-build/fio-*/ vitastor-0.5.6/fio; \
-    cd vitastor-0.5.6; \
+    cp -r /root/vitastor vitastor-0.5.12; \
+    ln -s /root/packages/qemu-$REL/qemu-*/ vitastor-0.5.12/qemu; \
+    ln -s /root/fio-build/fio-*/ vitastor-0.5.12/fio; \
+    cd vitastor-0.5.12; \
     FIO=$(head -n1 fio/debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     QEMU=$(head -n1 qemu/debian/changelog | perl -pe 's/^.*\((.*?)\).*$/$1/'); \
     sh copy-qemu-includes.sh; \
@@ -59,8 +59,8 @@ RUN set -e -x; \
     echo "dep:fio=$FIO" > debian/substvars; \
     echo "dep:qemu=$QEMU" >> debian/substvars; \
     cd /root/packages/vitastor-$REL; \
-    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.5.6.orig.tar.xz vitastor-0.5.6; \
-    cd vitastor-0.5.6; \
+    tar --sort=name --mtime='2020-01-01' --owner=0 --group=0 --exclude=debian -cJf vitastor_0.5.12.orig.tar.xz vitastor-0.5.12; \
+    cd vitastor-0.5.12; \
     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; \

lambda_size.cpp

@@ -1,51 +0,0 @@
-// Copyright (c) Vitaliy Filippov, 2019+
-// License: VNPL-1.1 (see README.md for details)
-
-#include <iostream>
-#include <functional>
-#include <array>
-#include <cstdlib> // for malloc() and free()
-using namespace std;
-
-// replace operator new and delete to log allocations
-void* operator new(std::size_t n)
-{
-    cout << "Allocating " << n << " bytes" << endl;
-    return malloc(n);
-}
-
-void operator delete(void* p) throw()
-{
-    free(p);
-}
-
-class test
-{
-public:
-    std::string s;
-    void a(std::function<void()> & f, const char *str)
-    {
-        auto l = [this, str]() { cout << str << " ? " << s << " from this\n"; };
-        cout << "Assigning lambda3 of size " << sizeof(l) << endl;
-        f = l;
-    }
-};
-
-int main()
-{
-    std::array<char, 16> arr1;
-    auto lambda1 = [arr1](){};
-    cout << "Assigning lambda1 of size " << sizeof(lambda1) << endl;
-    std::function<void()> f1 = lambda1;
-
-    std::array<char, 17> arr2;
-    auto lambda2 = [arr2](){};
-    cout << "Assigning lambda2 of size " << sizeof(lambda2) << endl;
-    std::function<void()> f2 = lambda2;
-
-    test t;
-    std::function<void()> f3;
-    t.s = "str";
-    t.a(f3, "huyambda");
-    f3();
-}

mon/PGUtil.js

@@ -5,18 +5,55 @@ module.exports = {
     scale_pg_count,
 };
 
+function add_pg_history(new_pg_history, new_pg, prev_pgs, prev_pg_history, old_pg)
+{
+    if (!new_pg_history[new_pg])
+    {
+        new_pg_history[new_pg] = {
+            osd_sets: {},
+            all_peers: {},
+            epoch: 0,
+        };
+    }
+    const nh = new_pg_history[new_pg], oh = prev_pg_history[old_pg];
+    nh.osd_sets[prev_pgs[old_pg].join(' ')] = prev_pgs[old_pg];
+    if (oh && oh.osd_sets && oh.osd_sets.length)
+    {
+        for (const pg of oh.osd_sets)
+        {
+            nh.osd_sets[pg.join(' ')] = pg;
+        }
+    }
+    if (oh && oh.all_peers && oh.all_peers.length)
+    {
+        for (const osd_num of oh.all_peers)
+        {
+            nh.all_peers[osd_num] = Number(osd_num);
+        }
+    }
+    if (oh && oh.epoch)
+    {
+        nh.epoch = nh.epoch < oh.epoch ? oh.epoch : nh.epoch;
+    }
+}
+
+function finish_pg_history(merged_history)
+{
+    merged_history.osd_sets = Object.values(merged_history.osd_sets);
+    merged_history.all_peers = Object.values(merged_history.all_peers);
+}
+
 function scale_pg_count(prev_pgs, prev_pg_history, new_pg_history, new_pg_count)
 {
     const old_pg_count = prev_pgs.length;
     // Add all possibly intersecting PGs to the history of new PGs
     if (!(new_pg_count % old_pg_count))
     {
-        // New PG count is a multiple of the old PG count
-        const mul = (new_pg_count / old_pg_count);
+        // New PG count is a multiple of old PG count
         for (let i = 0; i < new_pg_count; i++)
         {
-            const old_i = Math.floor(new_pg_count / mul);
-            new_pg_history[i] = prev_pg_history[old_i] ? JSON.parse(JSON.stringify(prev_pg_history[old_i])) : undefined;
+            add_pg_history(new_pg_history, i, prev_pgs, prev_pg_history, i % old_pg_count);
+            finish_pg_history(new_pg_history[i]);
         }
     }
     else if (!(old_pg_count % new_pg_count))
@@ -25,68 +62,26 @@ function scale_pg_count(prev_pgs, prev_pg_history, new_pg_history, new_pg_count)
         const mul = (old_pg_count / new_pg_count);
         for (let i = 0; i < new_pg_count; i++)
         {
-            new_pg_history[i] = {
-                osd_sets: [],
-                all_peers: [],
-                epoch: 0,
-            };
             for (let j = 0; j < mul; j++)
             {
-                new_pg_history[i].osd_sets.push(prev_pgs[i*mul]);
-                const hist = prev_pg_history[1+i*mul+j];
-                if (hist && hist.osd_sets && hist.osd_sets.length)
-                {
-                    Array.prototype.push.apply(new_pg_history[i].osd_sets, hist.osd_sets);
-                }
-                if (hist && hist.all_peers && hist.all_peers.length)
-                {
-                    Array.prototype.push.apply(new_pg_history[i].all_peers, hist.all_peers);
-                }
-                if (hist && hist.epoch)
-                {
-                    new_pg_history[i].epoch = new_pg_history[i].epoch < hist.epoch ? hist.epoch : new_pg_history[i].epoch;
-                }
+                add_pg_history(new_pg_history, i, prev_pgs, prev_pg_history, i+j*new_pg_count);
             }
+            finish_pg_history(new_pg_history[i]);
         }
     }
     else
     {
         // Any PG may intersect with any PG after non-multiple PG count change
         // So, merge ALL PGs history
-        let all_sets = {};
-        let all_peers = {};
-        let max_epoch = 0;
-        for (const pg of prev_pgs)
+        let merged_history = {};
+        for (let i = 0; i < old_pg_count; i++)
         {
-            all_sets[pg.join(' ')] = pg;
+            add_pg_history(merged_history, 1, prev_pgs, prev_pg_history, i);
         }
-        for (const pg in prev_pg_history)
-        {
-            const hist = prev_pg_history[pg];
-            if (hist && hist.osd_sets)
-            {
-                for (const pg of hist.osd_sets)
-                {
-                    all_sets[pg.join(' ')] = pg;
-                }
-            }
-            if (hist && hist.all_peers)
-            {
-                for (const osd_num of hist.all_peers)
-                {
-                    all_peers[osd_num] = Number(osd_num);
-                }
-            }
-            if (hist && hist.epoch)
-            {
-                max_epoch = max_epoch < hist.epoch ? hist.epoch : max_epoch;
-            }
-        }
-        all_sets = Object.values(all_sets);
-        all_peers = Object.values(all_peers);
+        finish_pg_history(merged_history[1]);
         for (let i = 0; i < new_pg_count; i++)
         {
-            new_pg_history[i] = { osd_sets: all_sets, all_peers, epoch: max_epoch };
+            new_pg_history[i] = { ...merged_history[1] };
         }
     }
     // Mark history keys for removed PGs as removed
@@ -94,19 +89,16 @@ function scale_pg_count(prev_pgs, prev_pg_history, new_pg_history, new_pg_count)
     {
         new_pg_history[i] = null;
     }
+    // Just for the lp_solve optimizer - pick a "previous" PG for each "new" one
     if (old_pg_count < new_pg_count)
     {
-        for (let i = new_pg_count-1; i >= 0; i--)
+        for (let i = old_pg_count; i < new_pg_count; i++)
         {
-            prev_pgs[i] = prev_pgs[Math.floor(i/new_pg_count*old_pg_count)];
+            prev_pgs[i] = prev_pgs[i % old_pg_count];
         }
     }
     else if (old_pg_count > new_pg_count)
     {
-        for (let i = 0; i < new_pg_count; i++)
-        {
-            prev_pgs[i] = prev_pgs[Math.round(i/new_pg_count*old_pg_count)];
-        }
         prev_pgs.splice(new_pg_count, old_pg_count-new_pg_count);
     }
 }

mon/lp-optimizer.js

@@ -104,6 +104,17 @@ async function optimize_initial({ osd_tree, pg_count, pg_size = 3, pg_minsize =
     return res;
 }
 
+function shuffle(array)
+{
+    for (let i = array.length - 1, j, x; i > 0; i--)
+    {
+        j = Math.floor(Math.random() * (i + 1));
+        x = array[i];
+        array[i] = array[j];
+        array[j] = x;
+    }
+}
+
 function make_int_pgs(weights, pg_count)
 {
     const total_weight = Object.values(weights).reduce((a, c) => Number(a) + Number(c), 0);
@@ -120,6 +131,7 @@ function make_int_pgs(weights, pg_count)
         weight_left -= weights[pg_name];
         pg_left -= n;
     }
+    shuffle(int_pgs);
     return int_pgs;
 }
 

mon/make-osd.sh

@@ -53,7 +53,6 @@ ExecStart=/usr/bin/vitastor-osd \\
     --osd_num $OSD_NUM \\
     --disable_data_fsync 1 \\
     --immediate_commit all \\
-    --flusher_count 256 \\
     --disk_alignment 4096 --journal_block_size 4096 --meta_block_size 4096 \\
     --journal_no_same_sector_overwrites true \\
     --journal_sector_buffer_count 1024 \\

mon/make-units.sh

@@ -32,7 +32,8 @@ ExecStart=/usr/local/bin/etcd -name etcd$ETCD_NUM --data-dir /var/lib/etcd$ETCD_
     --advertise-client-urls http://$IP:2379 --listen-client-urls http://$IP:2379 \\
     --initial-advertise-peer-urls http://$IP:2380 --listen-peer-urls http://$IP:2380 \\
     --initial-cluster-token vitastor-etcd-1 --initial-cluster $ETCD_HOSTS \\
-    --initial-cluster-state new --max-txn-ops=100000 --auto-compaction-retention=10 --auto-compaction-mode=revision
+    --initial-cluster-state new --max-txn-ops=100000 --max-request-bytes=104857600 \\
+    --auto-compaction-retention=10 --auto-compaction-mode=revision
 WorkingDirectory=/var/lib/etcd$ETCD_NUM.etcd
 ExecStartPre=+chown -R etcd /var/lib/etcd$ETCD_NUM.etcd
 User=etcd

mon/mon.js

@@ -10,6 +10,14 @@ const stableStringify = require('./stable-stringify.js');
 const PGUtil = require('./PGUtil.js');
 
 // FIXME document all etcd keys and config variables in the form of JSON schema or similar
+const etcd_nonempty_keys = {
+    'config/global': 1,
+    'config/node_placement': 1,
+    'config/pools': 1,
+    'config/pgs': 1,
+    'history/last_clean_pgs': 1,
+    'stats': 1,
+};
 const etcd_allow = new RegExp('^'+[
     'config/global',
     'config/node_placement',
@@ -22,6 +30,7 @@ const etcd_allow = new RegExp('^'+[
     'pg/state/[1-9]\\d*/[1-9]\\d*',
     'pg/stats/[1-9]\\d*/[1-9]\\d*',
     'pg/history/[1-9]\\d*/[1-9]\\d*',
+    'history/last_clean_pgs',
     'stats',
 ].join('$|^')+'$');
 
@@ -34,7 +43,7 @@ const etcd_tree = {
             etcd_mon_retries: 5, // min: 0
             mon_change_timeout: 1000, // ms. min: 100
             mon_stats_timeout: 1000, // ms. min: 100
-            osd_out_time: 1800, // seconds. min: 0
+            osd_out_time: 600, // seconds. min: 0
             placement_levels: { datacenter: 1, rack: 2, host: 3, osd: 4, ... },
             // client and osd
             use_sync_send_recv: false,
@@ -46,6 +55,8 @@ const etcd_tree = {
             client_dirty_limit: 33554432,
             peer_connect_interval: 5, // seconds. min: 1
             peer_connect_timeout: 5, // seconds. min: 1
+            osd_idle_timeout: 5, // seconds. min: 1
+            osd_ping_timeout: 5, // seconds. min: 1
             up_wait_retry_interval: 500, // ms. min: 50
             // osd
             etcd_report_interval: 30, // min: 10
@@ -55,8 +66,12 @@ const etcd_tree = {
             autosync_interval: 5,
             client_queue_depth: 128, // unused
             recovery_queue_depth: 4,
+            recovery_sync_batch: 16,
             readonly: false,
+            no_recovery: false,
+            no_rebalance: false,
             print_stats_interval: 3,
+            slow_log_interval: 10,
             // blockstore - fixed in superblock
             block_size,
             disk_alignment,
@@ -76,7 +91,9 @@ const etcd_tree = {
             disable_meta_fsync,
             disable_device_lock,
             // blockstore - configurable
-            flusher_count,
+            max_write_iodepth,
+            min_flusher_count: 1,
+            max_flusher_count: 256,
             inmemory_metadata,
             inmemory_journal,
             journal_sector_buffer_count,
@@ -166,7 +183,7 @@ const etcd_tree = {
             /* <pool_id>: {
                 <pg_id>: {
                     primary: osd_num_t,
-                    state: ("starting"|"peering"|"incomplete"|"active"|"stopping"|"offline"|
+                    state: ("starting"|"peering"|"incomplete"|"active"|"repeering"|"stopping"|"offline"|
                         "degraded"|"has_incomplete"|"has_degraded"|"has_misplaced"|"has_unclean"|
                         "has_invalid"|"left_on_dead")[],
                 }
@@ -213,6 +230,9 @@ const etcd_tree = {
             incomplete: uint64_t,
         }, */
     },
+    history: {
+        last_clean_pgs: {},
+    },
 };
 
 // FIXME Split into several files
@@ -291,7 +311,7 @@ class Mon
         this.config.osd_out_time = Number(this.config.osd_out_time) || 0;
         if (!this.config.osd_out_time)
         {
-            this.config.osd_out_time = 30*60; // 30 minutes by default
+            this.config.osd_out_time = 600; // 10 minutes by default
         }
     }
 
@@ -313,8 +333,14 @@ class Mon
                     ok(false);
                 }, this.config.etcd_mon_timeout);
                 this.ws = new WebSocket(base+'/watch');
+                const fail = () =>
+                {
+                    ok(false);
+                };
+                this.ws.on('error', fail);
                 this.ws.on('open', () =>
                 {
+                    this.ws.removeListener('error', fail);
                     if (timer_id)
                         clearTimeout(timer_id);
                     ok(true);
@@ -359,7 +385,7 @@ class Mon
             }
             else
             {
-                let stats_changed = false, changed = false;
+                let stats_changed = false, changed = false, pg_states_changed = false;
                 if (this.verbose)
                 {
                     console.log('Revision '+data.result.header.revision+' events: ');
@@ -373,6 +399,10 @@ class Mon
                     {
                         stats_changed = true;
                     }
+                    else if (key.substr(0, 10) == '/pg/state/')
+                    {
+                        pg_states_changed = true;
+                    }
                     else if (key != '/stats')
                     {
                         changed = true;
@@ -382,6 +412,10 @@ class Mon
                         console.log(JSON.stringify(e));
                     }
                 }
+                if (pg_states_changed)
+                {
+                    this.save_last_clean().catch(console.error);
+                }
                 if (stats_changed)
                 {
                     this.schedule_update_stats();
@@ -394,10 +428,46 @@ class Mon
         });
     }
 
+    async save_last_clean()
+    {
+        // last_clean_pgs is used to avoid extra data move when observing a series of changes in the cluster
+        for (const pool_id in this.state.config.pools)
+        {
+            const pool_cfg = this.state.config.pools[pool_id];
+            if (!this.validate_pool_cfg(pool_id, pool_cfg, false))
+            {
+                continue;
+            }
+            for (let pg_num = 1; pg_num <= pool_cfg.pg_count; pg_num++)
+            {
+                if (!this.state.pg.state[pool_id] ||
+                    !this.state.pg.state[pool_id][pg_num] ||
+                    !(this.state.pg.state[pool_id][pg_num].state instanceof Array))
+                {
+                    // Unclean
+                    return;
+                }
+                let st = this.state.pg.state[pool_id][pg_num].state.join(',');
+                if (st != 'active' && st != 'active,left_on_dead' && st != 'left_on_dead,active')
+                {
+                    // Unclean
+                    return;
+                }
+            }
+        }
+        this.state.history.last_clean_pgs = JSON.parse(JSON.stringify(this.state.config.pgs));
+        await this.etcd_call('/kv/txn', {
+            success: [ { requestPut: {
+                key: b64(this.etcd_prefix+'/history/last_clean_pgs'),
+                value: b64(JSON.stringify(this.state.history.last_clean_pgs))
+            } } ],
+        }, this.etcd_start_timeout, 0);
+    }
+
     async get_lease()
     {
         const max_ttl = this.config.etcd_mon_ttl + this.config.etcd_mon_timeout/1000*this.config.etcd_mon_retries;
-        const res = await this.etcd_call('/lease/grant', { TTL: max_ttl }, this.config.etcd_mon_timeout, this.config.etcd_mon_retries);
+        const res = await this.etcd_call('/lease/grant', { TTL: max_ttl }, this.config.etcd_mon_timeout, -1);
         this.etcd_lease_id = res.ID;
         setInterval(async () =>
         {
@@ -472,7 +542,7 @@ class Mon
         for (const osd_num of this.all_osds().sort((a, b) => a - b))
         {
             const stat = this.state.osd.stats[osd_num];
-            if (stat.size && (this.state.osd.state[osd_num] || Number(stat.time) >= down_time))
+            if (stat && stat.size && (this.state.osd.state[osd_num] || Number(stat.time) >= down_time))
             {
                 // Numeric IDs are reserved for OSDs
                 const osd_cfg = this.state.config.osd[osd_num];
@@ -573,34 +643,61 @@ class Mon
         return !has_online;
     }
 
+    reset_rng()
+    {
+        this.seed = 0x5f020e43;
+    }
+
+    rng()
+    {
+        this.seed ^= this.seed << 13;
+        this.seed ^= this.seed >> 17;
+        this.seed ^= this.seed << 5;
+        return this.seed + 2147483648;
+    }
+
+    pick_primary(pool_id, osd_set, up_osds)
+    {
+        let alive_set;
+        if (this.state.config.pools[pool_id].scheme === 'replicated')
+            alive_set = osd_set.filter(osd_num => osd_num && up_osds[osd_num]);
+        else
+        {
+            // Prefer data OSDs for EC because they can actually read something without an additional network hop
+            const pg_data_size = (this.state.config.pools[pool_id].pg_size||0) -
+                (this.state.config.pools[pool_id].parity_chunks||0);
+            alive_set = osd_set.slice(0, pg_data_size).filter(osd_num => osd_num && up_osds[osd_num]);
+            if (!alive_set.length)
+                alive_set = osd_set.filter(osd_num => osd_num && up_osds[osd_num]);
+        }
+        if (!alive_set.length)
+            return 0;
+        return alive_set[this.rng() % alive_set.length];
+    }
+
     save_new_pgs_txn(request, pool_id, up_osds, prev_pgs, new_pgs, pg_history)
     {
-        const replicated = new_pgs.length && this.state.config.pools[pool_id].scheme === 'replicated';
-        const pg_minsize = new_pgs.length && this.state.config.pools[pool_id].pg_minsize;
         const pg_items = {};
+        this.reset_rng();
         new_pgs.map((osd_set, i) =>
         {
             osd_set = osd_set.map(osd_num => osd_num === LPOptimizer.NO_OSD ? 0 : osd_num);
-            let alive_set;
-            if (replicated)
-                alive_set = osd_set.filter(osd_num => osd_num && up_osds[osd_num]);
-            else
-            {
-                // Prefer data OSDs for EC because they can actually read something without an additional network hop
-                alive_set = osd_set.slice(0, pg_minsize).filter(osd_num => osd_num && up_osds[osd_num]);
-                if (!alive_set.length)
-                    alive_set = osd_set.filter(osd_num => osd_num && up_osds[osd_num]);
-            }
             pg_items[i+1] = {
                 osd_set,
-                primary: alive_set.length ? alive_set[Math.floor(Math.random()*alive_set.length)] : 0,
+                primary: this.pick_primary(pool_id, osd_set, up_osds),
             };
-            if (prev_pgs[i] && prev_pgs[i].join(' ') != osd_set.join(' '))
+            if (prev_pgs[i] && prev_pgs[i].join(' ') != osd_set.join(' ') &&
+                prev_pgs[i].filter(osd_num => osd_num).length > 0)
             {
                 pg_history[i] = pg_history[i] || {};
                 pg_history[i].osd_sets = pg_history[i].osd_sets || [];
                 pg_history[i].osd_sets.push(prev_pgs[i]);
             }
+            if (pg_history[i] && pg_history[i].osd_sets)
+            {
+                pg_history[i].osd_sets = Object.values(pg_history[i].osd_sets
+                    .reduce((a, c) => { a[c.join(' ')] = c; return a; }, {}));
+            }
         });
         for (let i = 0; i < new_pgs.length || i < prev_pgs.length; i++)
         {
@@ -751,7 +848,7 @@ class Mon
     {
         // Take configuration and state, check it against the stored configuration hash
         // Recalculate PGs and save them to etcd if the configuration is changed
-        // FIXME: Also do not change anything if the distribution is good enough and no PGs are degraded
+        // FIXME: Do not change anything if the distribution is good and random enough and no PGs are degraded
         const { up_osds, levels, osd_tree } = this.get_osd_tree();
         const tree_cfg = {
             osd_tree,
@@ -791,13 +888,33 @@ class Mon
                 pool_tree = pool_tree ? pool_tree.children : [];
                 pool_tree = LPOptimizer.flatten_tree(pool_tree, levels, pool_cfg.failure_domain, 'osd');
                 this.filter_osds_by_tags(osd_tree, pool_tree, pool_cfg.osd_tags);
-                const prev_pgs = [];
-                for (const pg in ((this.state.config.pgs.items||{})[pool_id]||{})||{})
+                // These are for the purpose of building history.osd_sets
+                const real_prev_pgs = [];
+                let pg_history = [];
+                for (const pg in ((this.state.config.pgs.items||{})[pool_id]||{}))
                 {
-                    prev_pgs[pg-1] = this.state.config.pgs.items[pool_id][pg].osd_set;
+                    real_prev_pgs[pg-1] = this.state.config.pgs.items[pool_id][pg].osd_set;
+                    if (this.state.pg.history[pool_id] &&
+                        this.state.pg.history[pool_id][pg])
+                    {
+                        pg_history[pg-1] = this.state.pg.history[pool_id][pg];
+                    }
                 }
-                const pg_history = [];
-                const old_pg_count = prev_pgs.length;
+                // And these are for the purpose of minimizing data movement
+                let prev_pgs = [];
+                for (const pg in ((this.state.history.last_clean_pgs.items||{})[pool_id]||{}))
+                {
+                    prev_pgs[pg-1] = this.state.history.last_clean_pgs.items[pool_id][pg].osd_set;
+                }
+                prev_pgs = JSON.parse(JSON.stringify(prev_pgs.length ? prev_pgs : real_prev_pgs));
+                const old_pg_count = real_prev_pgs.length;
+                const optimize_cfg = {
+                    osd_tree: pool_tree,
+                    pg_count: pool_cfg.pg_count,
+                    pg_size: pool_cfg.pg_size,
+                    pg_minsize: pool_cfg.pg_minsize,
+                    max_combinations: pool_cfg.max_osd_combinations,
+                };
                 let optimize_result;
                 if (old_pg_count > 0)
                 {
@@ -809,7 +926,9 @@ class Mon
                             this.schedule_recheck();
                             return;
                         }
-                        PGUtil.scale_pg_count(prev_pgs, this.state.pg.history[pool_id]||{}, pg_history, pool_cfg.pg_count);
+                        const new_pg_history = [];
+                        PGUtil.scale_pg_count(prev_pgs, pg_history, new_pg_history, pool_cfg.pg_count);
+                        pg_history = new_pg_history;
                     }
                     for (const pg of prev_pgs)
                     {
@@ -822,23 +941,22 @@ class Mon
                             pg.pop();
                         }
                     }
-                    optimize_result = await LPOptimizer.optimize_change({
-                        prev_pgs,
-                        osd_tree: pool_tree,
-                        pg_size: pool_cfg.pg_size,
-                        pg_minsize: pool_cfg.pg_minsize,
-                        max_combinations: pool_cfg.max_osd_combinations,
-                    });
+                    if (!this.state.config.pgs.hash)
+                    {
+                        // Re-shuffle PGs
+                        optimize_result = await LPOptimizer.optimize_initial(optimize_cfg);
+                    }
+                    else
+                    {
+                        optimize_result = await LPOptimizer.optimize_change({
+                            prev_pgs,
+                            ...optimize_cfg,
+                        });
+                    }
                 }
                 else
                 {
-                    optimize_result = await LPOptimizer.optimize_initial({
-                        osd_tree: pool_tree,
-                        pg_count: pool_cfg.pg_count,
-                        pg_size: pool_cfg.pg_size,
-                        pg_minsize: pool_cfg.pg_minsize,
-                        max_combinations: pool_cfg.max_osd_combinations,
-                    });
+                    optimize_result = await LPOptimizer.optimize_initial(optimize_cfg);
                 }
                 if (old_pg_count != optimize_result.int_pgs.length)
                 {
@@ -846,16 +964,21 @@ class Mon
                         `PG count for pool ${pool_id} (${pool_cfg.name || 'unnamed'})`+
                         ` changed from: ${old_pg_count} to ${optimize_result.int_pgs.length}`
                     );
+                    // Drop stats
+                    etcd_request.success.push({ requestDeleteRange: {
+                        key: b64(this.etcd_prefix+'/pg/stats/'+pool_id+'/'),
+                        range_end: b64(this.etcd_prefix+'/pg/stats/'+pool_id+'0'),
+                    } });
                 }
                 LPOptimizer.print_change_stats(optimize_result);
-                this.save_new_pgs_txn(etcd_request, pool_id, up_osds, prev_pgs, optimize_result.int_pgs, pg_history);
+                this.save_new_pgs_txn(etcd_request, pool_id, up_osds, real_prev_pgs, optimize_result.int_pgs, pg_history);
             }
             this.state.config.pgs.hash = tree_hash;
             await this.save_pg_config(etcd_request);
         }
         else
         {
-            // Nothing changed, but we still want to check for down OSDs
+            // Nothing changed, but we still want to recheck the distribution of primaries
             let changed = false;
             for (const pool_id in this.state.config.pools)
             {
@@ -865,22 +988,13 @@ class Mon
                     continue;
                 }
                 const replicated = pool_cfg.scheme === 'replicated';
-                for (const pg_num in ((this.state.config.pgs.items||{})[pool_id]||{})||{})
+                this.reset_rng();
+                for (let pg_num = 1; pg_num <= pool_cfg.pg_count; pg_num++)
                 {
                     const pg_cfg = this.state.config.pgs.items[pool_id][pg_num];
-                    if (!Number(pg_cfg.primary) || !up_osds[pg_cfg.primary])
+                    if (pg_cfg)
                     {
-                        let alive_set;
-                        if (replicated)
-                            alive_set = pg_cfg.osd_set.filter(osd_num => osd_num && up_osds[osd_num]);
-                        else
-                        {
-                            // Prefer data OSDs for EC because they can actually read something without an additional network hop
-                            alive_set = pg_cfg.osd_set.slice(0, pool_cfg.pg_minsize).filter(osd_num => osd_num && up_osds[osd_num]);
-                            if (!alive_set.length)
-                                alive_set = pg_cfg.osd_set.filter(osd_num => osd_num && up_osds[osd_num]);
-                        }
-                        const new_primary = alive_set.length ? alive_set[Math.floor(Math.random()*alive_set.length)] : 0;
+                        const new_primary = this.pick_primary(pool_id, pg_cfg.osd_set, up_osds);
                         if (pg_cfg.primary != new_primary)
                         {
                             console.log(
@@ -970,7 +1084,7 @@ class Mon
         const op_stats = {}, subop_stats = {}, recovery_stats = {};
         for (const osd in this.state.osd.stats)
         {
-            const st = this.state.osd.stats[osd];
+            const st = this.state.osd.stats[osd]||{};
             for (const op in st.op_stats||{})
             {
                 op_stats[op] = op_stats[op] || { count: 0n, usec: 0n, bytes: 0n };
@@ -1043,11 +1157,14 @@ class Mon
             for (const pg_num in this.state.pg.stats[pool_id])
             {
                 const st = this.state.pg.stats[pool_id][pg_num];
-                for (const k in object_counts)
+                if (st)
                 {
-                    if (st[k+'_count'])
+                    for (const k in object_counts)
                     {
-                        object_counts[k] += BigInt(st[k+'_count']);
+                        if (st[k+'_count'])
+                        {
+                            object_counts[k] += BigInt(st[k+'_count']);
+                        }
                     }
                 }
             }
@@ -1122,16 +1239,20 @@ class Mon
             console.log('Bad value in etcd: '+kv.key+' = '+kv.value);
             return;
         }
-        key = key.split('/');
+        let key_parts = key.split('/');
         let cur = this.state;
-        for (let i = 0; i < key.length-1; i++)
+        for (let i = 0; i < key_parts.length-1; i++)
         {
-            cur = (cur[key[i]] = cur[key[i]] || {});
+            cur = (cur[key_parts[i]] = cur[key_parts[i]] || {});
         }
-        cur[key[key.length-1]] = kv.value;
-        if (key.join('/') === 'config/global')
+        if (etcd_nonempty_keys[key])
+        {
+            // Do not clear these to null
+            kv.value = kv.value || {};
+        }
+        cur[key_parts[key_parts.length-1]] = kv.value;
+        if (key === 'config/global')
         {
-            this.state.config.global = this.state.config.global || {};
             this.config = this.state.config.global;
             this.check_config();
             for (const osd_num in this.state.osd.stats)
@@ -1142,7 +1263,7 @@ class Mon
                 );
             }
         }
-        else if (key.join('/') === 'config/pools')
+        else if (key === 'config/pools')
         {
             for (const pool_id in this.state.config.pools)
             {
@@ -1151,7 +1272,7 @@ class Mon
                 this.validate_pool_cfg(pool_id, pool_cfg, true);
             }
         }
-        else if (key[0] === 'osd' && key[1] === 'stats')
+        else if (key_parts[0] === 'osd' && key_parts[1] === 'stats')
         {
             // Recheck PGs <osd_out_time> later
             this.schedule_next_recheck_at(
@@ -1183,6 +1304,11 @@ class Mon
                     console.error('etcd returned error: '+res.json.error);
                     break;
                 }
+                if (this.etcd_urls.length > 1)
+                {
+                    // Stick to the same etcd for the rest of calls
+                    this.etcd_urls = [ base ];
+                }
                 return res.json;
             }
             retry++;

rpm/build-tarball.sh

@@ -48,4 +48,4 @@ FIO=`rpm -qi fio | perl -e 'while(<>) { /^Epoch[\s:]+(\S+)/ && print "$1:"; /^Ve
 QEMU=`rpm -qi qemu qemu-kvm | 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
 perl -i -pe 's/(Requires:\s*qemu(?:-kvm)?)([^\n]+)?/$1 = '$QEMU'/' $VITASTOR/rpm/vitastor-el$EL.spec
-tar --transform 's#^#vitastor-0.5.6/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.5.6$(rpm --eval '%dist').tar.gz *
+tar --transform 's#^#vitastor-0.5.12/#' --exclude 'rpm/*.rpm' -czf $VITASTOR/../vitastor-0.5.12$(rpm --eval '%dist').tar.gz *

rpm/vitastor-el7.Dockerfile

@@ -37,7 +37,7 @@ ADD . /root/vitastor
 RUN set -e; \
     cd /root/vitastor/rpm; \
     sh build-tarball.sh; \
-    cp /root/vitastor-0.5.6.el7.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.5.12.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:        0.5.6
+Version:        0.5.12
 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-0.5.6.el7.tar.gz
+Source0:        vitastor-0.5.12.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-0.5.6.el8.tar.gz ~/rpmbuild/SOURCES; \
+    cp /root/vitastor-0.5.12.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:        0.5.6
+Version:        0.5.12
 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-0.5.6.el8.tar.gz
+Source0:        vitastor-0.5.12.el8.tar.gz
 
 BuildRequires:  liburing-devel >= 0.6
 BuildRequires:  gperftools-devel

src/CMakeLists.txt

@@ -14,7 +14,7 @@ if("${CMAKE_INSTALL_PREFIX}" MATCHES "^/usr/local/?$")
 endif()
 
 add_definitions(-DVERSION="0.6-dev")
-add_definitions(-Wall -Wno-sign-compare -Wno-comment -Wno-parentheses -Wno-pointer-arith)
+add_definitions(-Wall -Wno-sign-compare -Wno-comment -Wno-parentheses -Wno-pointer-arith -I ${CMAKE_SOURCE_DIR}/src)
 if (${WITH_ASAN})
 	add_definitions(-fsanitize=address -fno-omit-frame-pointer)
 	add_link_options(-fsanitize=address -fno-omit-frame-pointer)
@@ -66,7 +66,8 @@ target_link_libraries(fio_vitastor_blk
 # vitastor-osd
 add_executable(vitastor-osd
 	osd_main.cpp osd.cpp osd_secondary.cpp msgr_receive.cpp msgr_send.cpp osd_peering.cpp osd_flush.cpp osd_peering_pg.cpp
-	osd_primary.cpp osd_primary_subops.cpp etcd_state_client.cpp messenger.cpp osd_cluster.cpp http_client.cpp osd_ops.cpp pg_states.cpp
+	osd_primary.cpp osd_primary_sync.cpp osd_primary_write.cpp osd_primary_subops.cpp
+	etcd_state_client.cpp messenger.cpp msgr_stop.cpp msgr_op.cpp osd_cluster.cpp http_client.cpp osd_ops.cpp pg_states.cpp
 	osd_rmw.cpp base64.cpp timerfd_manager.cpp epoll_manager.cpp ../json11/json11.cpp
 )
 target_link_libraries(vitastor-osd
@@ -86,7 +87,7 @@ target_link_libraries(fio_vitastor_sec
 # libvitastor_client.so
 add_library(vitastor_client SHARED
 	cluster_client.cpp epoll_manager.cpp etcd_state_client.cpp
-	messenger.cpp msgr_send.cpp msgr_receive.cpp ringloop.cpp ../json11/json11.cpp
+	messenger.cpp msgr_stop.cpp msgr_op.cpp msgr_send.cpp msgr_receive.cpp ringloop.cpp ../json11/json11.cpp
 	http_client.cpp osd_ops.cpp pg_states.cpp timerfd_manager.cpp base64.cpp
 )
 target_link_libraries(vitastor_client
@@ -161,7 +162,8 @@ target_link_libraries(osd_rmw_test Jerasure tcmalloc_minimal)
 
 # stub_uring_osd
 add_executable(stub_uring_osd
-	stub_uring_osd.cpp epoll_manager.cpp messenger.cpp msgr_send.cpp msgr_receive.cpp ringloop.cpp timerfd_manager.cpp ../json11/json11.cpp
+	stub_uring_osd.cpp epoll_manager.cpp messenger.cpp msgr_stop.cpp msgr_op.cpp
+	msgr_send.cpp msgr_receive.cpp ringloop.cpp timerfd_manager.cpp ../json11/json11.cpp
 )
 target_link_libraries(stub_uring_osd
 	${LIBURING_LIBRARIES}
@@ -175,6 +177,15 @@ target_link_libraries(osd_peering_pg_test tcmalloc_minimal)
 # test_allocator
 add_executable(test_allocator test_allocator.cpp allocator.cpp)
 
+# test_cluster_client
+add_executable(test_cluster_client
+	test_cluster_client.cpp
+	pg_states.cpp osd_ops.cpp cluster_client.cpp msgr_op.cpp mock/messenger.cpp msgr_stop.cpp
+	etcd_state_client.cpp timerfd_manager.cpp ../json11/json11.cpp
+)
+target_compile_definitions(test_cluster_client PUBLIC -D__MOCK__)
+target_include_directories(test_cluster_client PUBLIC ${CMAKE_SOURCE_DIR}/src/mock)
+
 ## test_blockstore, test_shit
 #add_executable(test_blockstore test_blockstore.cpp timerfd_interval.cpp)
 #target_link_libraries(test_blockstore blockstore)

src/allocator.cpp

@@ -13,19 +13,19 @@ allocator::allocator(uint64_t blocks)
     {
         throw std::invalid_argument("blocks");
     }
-    uint64_t p2 = 1, total = 1;
+    uint64_t p2 = 1;
+    total = 0;
     while (p2 * 64 < blocks)
     {
-        p2 = p2 * 64;
         total += p2;
+        p2 = p2 * 64;
     }
-    total -= p2;
     total += (blocks+63) / 64;
-    mask = new uint64_t[2 + total];
+    mask = new uint64_t[total];
     size = free = blocks;
     last_one_mask = (blocks % 64) == 0
         ? UINT64_MAX
-        : ~(UINT64_MAX << (64 - blocks % 64));
+        : ((1l << (blocks % 64)) - 1);
     for (uint64_t i = 0; i < total; i++)
     {
         mask[i] = 0;
@@ -37,6 +37,21 @@ allocator::~allocator()
     delete[] mask;
 }
 
+bool allocator::get(uint64_t addr)
+{
+    if (addr >= size)
+    {
+        return false;
+    }
+    uint64_t p2 = 1, offset = 0;
+    while (p2 * 64 < size)
+    {
+        offset += p2;
+        p2 = p2 * 64;
+    }
+    return ((mask[offset + addr/64] >> (addr % 64)) & 1);
+}
+
 void allocator::set(uint64_t addr, bool value)
 {
     if (addr >= size)
@@ -99,6 +114,10 @@ uint64_t allocator::find_free()
     uint64_t p2 = 1, offset = 0, addr = 0, f, i;
     while (p2 < size)
     {
+        if (offset+addr >= total)
+        {
+            return UINT64_MAX;
+        }
         uint64_t m = mask[offset + addr];
         for (i = 0, f = 1; i < 64; i++, f <<= 1)
         {
@@ -113,11 +132,6 @@ uint64_t allocator::find_free()
             return UINT64_MAX;
         }
         addr = (addr * 64) | i;
-        if (addr >= size)
-        {
-            // No space
-            return UINT64_MAX;
-        }
         offset += p2;
         p2 = p2 * 64;
     }

src/allocator.h

@@ -8,6 +8,7 @@
 // Hierarchical bitmap allocator
 class allocator
 {
+    uint64_t total;
     uint64_t size;
     uint64_t free;
     uint64_t last_one_mask;
@@ -15,6 +16,7 @@ class allocator
 public:
     allocator(uint64_t blocks);
     ~allocator();
+    bool get(uint64_t addr);
     void set(uint64_t addr, bool value);
     uint64_t find_free();
     uint64_t get_free_count();

src/blockstore.cpp

@@ -35,12 +35,7 @@ bool blockstore_t::is_safe_to_stop()
 
 void blockstore_t::enqueue_op(blockstore_op_t *op)
 {
-    impl->enqueue_op(op, false);
-}
-
-void blockstore_t::enqueue_op_first(blockstore_op_t *op)
-{
-    impl->enqueue_op(op, true);
+    impl->enqueue_op(op);
 }
 
 std::unordered_map<object_id, uint64_t> & blockstore_t::get_unstable_writes()
@@ -63,7 +58,7 @@ uint64_t blockstore_t::get_free_block_count()
     return impl->get_free_block_count();
 }
 
-uint32_t blockstore_t::get_disk_alignment()
+uint32_t blockstore_t::get_bitmap_granularity()
 {
-    return impl->get_disk_alignment();
+    return impl->get_bitmap_granularity();
 }

src/blockstore.h

@@ -175,10 +175,6 @@ public:
     // Submission
     void enqueue_op(blockstore_op_t *op);
 
-    // Insert operation into the beginning of the queue
-    // Intended for the OSD syncer "thread" to be able to stabilize something when the journal is full
-    void enqueue_op_first(blockstore_op_t *op);
-
     // Unstable writes are added here (map of object_id -> version)
     std::unordered_map<object_id, uint64_t> & get_unstable_writes();
 
@@ -187,5 +183,5 @@ public:
     uint64_t get_block_count();
     uint64_t get_free_block_count();
 
-    uint32_t get_disk_alignment();
+    uint32_t get_bitmap_granularity();
 };

src/blockstore_flush.cpp

@@ -3,12 +3,13 @@
 
 #include "blockstore_impl.h"
 
-journal_flusher_t::journal_flusher_t(int flusher_count, blockstore_impl_t *bs)
+journal_flusher_t::journal_flusher_t(blockstore_impl_t *bs)
 {
     this->bs = bs;
-    this->flusher_count = flusher_count;
-    this->cur_flusher_count = 1;
-    this->target_flusher_count = 1;
+    this->max_flusher_count = bs->max_flusher_count;
+    this->min_flusher_count = bs->min_flusher_count;
+    this->cur_flusher_count = bs->min_flusher_count;
+    this->target_flusher_count = bs->min_flusher_count;
     dequeuing = false;
     trimming = false;
     active_flushers = 0;
@@ -19,8 +20,8 @@ journal_flusher_t::journal_flusher_t(int flusher_count, blockstore_impl_t *bs)
     journal_trim_counter = 0;
     trim_wanted = 0;
     journal_superblock = bs->journal.inmemory ? bs->journal.buffer : memalign_or_die(MEM_ALIGNMENT, bs->journal_block_size);
-    co = new journal_flusher_co[flusher_count];
-    for (int i = 0; i < flusher_count; i++)
+    co = new journal_flusher_co[max_flusher_count];
+    for (int i = 0; i < max_flusher_count; i++)
     {
         co[i].bs = bs;
         co[i].flusher = this;
@@ -71,10 +72,10 @@ bool journal_flusher_t::is_active()
 void journal_flusher_t::loop()
 {
     target_flusher_count = bs->write_iodepth*2;
-    if (target_flusher_count <= 0)
-        target_flusher_count = 1;
-    else if (target_flusher_count > flusher_count)
-        target_flusher_count = flusher_count;
+    if (target_flusher_count < min_flusher_count)
+        target_flusher_count = min_flusher_count;
+    else if (target_flusher_count > max_flusher_count)
+        target_flusher_count = max_flusher_count;
     if (target_flusher_count > cur_flusher_count)
         cur_flusher_count = target_flusher_count;
     else if (target_flusher_count < cur_flusher_count)
@@ -237,7 +238,8 @@ bool journal_flusher_co::loop()
     else if (wait_state == 21)
         goto resume_21;
 resume_0:
-    if (!flusher->flush_queue.size() || !flusher->dequeuing)
+    if (flusher->flush_queue.size() < flusher->min_flusher_count && !flusher->trim_wanted ||
+        !flusher->flush_queue.size() || !flusher->dequeuing)
     {
 stop_flusher:
         if (flusher->trim_wanted > 0 && flusher->journal_trim_counter > 0)
@@ -482,6 +484,14 @@ resume_1:
         }
         if (has_delete)
         {
+            clean_disk_entry *new_entry = (clean_disk_entry*)(meta_new.buf + meta_new.pos*bs->clean_entry_size);
+            if (new_entry->oid.inode != 0 && new_entry->oid != cur.oid)
+            {
+                printf("Fatal error (metadata corruption or bug): tried to delete metadata entry %lu (%lx:%lx) while deleting %lx:%lx\n",
+                    clean_loc >> bs->block_order, new_entry->oid.inode, new_entry->oid.stripe, cur.oid.inode, cur.oid.stripe);
+                exit(1);
+            }
+            // zero out new metadata entry
             memset(meta_new.buf + meta_new.pos*bs->clean_entry_size, 0, bs->clean_entry_size);
         }
         else
@@ -646,7 +656,7 @@ bool journal_flusher_co::scan_dirty(int wait_base)
         {
             char err[1024];
             snprintf(
-                err, 1024, "BUG: Unexpected dirty_entry %lx:%lx v%lu unstable state during flush: %d",
+                err, 1024, "BUG: Unexpected dirty_entry %lx:%lx v%lu unstable state during flush: 0x%x",
                 dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version, dirty_it->second.state
             );
             throw std::runtime_error(err);
@@ -775,7 +785,10 @@ void journal_flusher_co::update_clean_db()
     if (old_clean_loc != UINT64_MAX && old_clean_loc != clean_loc)
     {
 #ifdef BLOCKSTORE_DEBUG
-        printf("Free block %lu (new location is %lu)\n", old_clean_loc >> bs->block_order, clean_loc >> bs->block_order);
+        printf("Free block %lu from %lx:%lx v%lu (new location is %lu)\n",
+            old_clean_loc >> bs->block_order,
+            cur.oid.inode, cur.oid.stripe, cur.version,
+            clean_loc >> bs->block_order);
 #endif
         bs->data_alloc->set(old_clean_loc >> bs->block_order, false);
     }
@@ -783,6 +796,11 @@ void journal_flusher_co::update_clean_db()
     {
         auto clean_it = bs->clean_db.find(cur.oid);
         bs->clean_db.erase(clean_it);
+#ifdef BLOCKSTORE_DEBUG
+        printf("Free block %lu from %lx:%lx v%lu (delete)\n",
+            clean_loc >> bs->block_order,
+            cur.oid.inode, cur.oid.stripe, cur.version);
+#endif
         bs->data_alloc->set(clean_loc >> bs->block_order, false);
         clean_loc = UINT64_MAX;
     }
@@ -804,7 +822,7 @@ bool journal_flusher_co::fsync_batch(bool fsync_meta, int wait_base)
         goto resume_1;
     else if (wait_state == wait_base+2)
         goto resume_2;
-    if (!(fsync_meta ? bs->disable_meta_fsync : bs->disable_journal_fsync))
+    if (!(fsync_meta ? bs->disable_meta_fsync : bs->disable_data_fsync))
     {
         cur_sync = flusher->syncs.end();
         while (cur_sync != flusher->syncs.begin())
@@ -823,31 +841,34 @@ bool journal_flusher_co::fsync_batch(bool fsync_meta, int wait_base)
     sync_found:
         cur_sync->ready_count++;
         flusher->syncing_flushers++;
-        if (flusher->syncing_flushers >= flusher->flusher_count || !flusher->flush_queue.size())
+    resume_1:
+        if (!cur_sync->state)
         {
-            // Sync batch is ready. Do it.
-            await_sqe(0);
-            data->iov = { 0 };
-            data->callback = simple_callback_w;
-            my_uring_prep_fsync(sqe, fsync_meta ? bs->meta_fd : bs->data_fd, IORING_FSYNC_DATASYNC);
-            cur_sync->state = 1;
-            wait_count++;
-        resume_1:
-            if (wait_count > 0)
+            if (flusher->syncing_flushers >= flusher->cur_flusher_count || !flusher->flush_queue.size())
             {
+                // Sync batch is ready. Do it.
+                await_sqe(0);
+                data->iov = { 0 };
+                data->callback = simple_callback_w;
+                my_uring_prep_fsync(sqe, fsync_meta ? bs->meta_fd : bs->data_fd, IORING_FSYNC_DATASYNC);
+                cur_sync->state = 1;
+                wait_count++;
+            resume_2:
+                if (wait_count > 0)
+                {
+                    wait_state = 2;
+                    return false;
+                }
+                // Sync completed. All previous coroutines waiting for it must be resumed
+                cur_sync->state = 2;
+                bs->ringloop->wakeup();
+            }
+            else
+            {
+                // Wait until someone else sends and completes a sync.
                 wait_state = 1;
                 return false;
             }
-            // Sync completed. All previous coroutines waiting for it must be resumed
-            cur_sync->state = 2;
-            bs->ringloop->wakeup();
-        }
-        // Wait until someone else sends and completes a sync.
-    resume_2:
-        if (!cur_sync->state)
-        {
-            wait_state = 2;
-            return false;
         }
         flusher->syncing_flushers--;
         cur_sync->ready_count--;

src/blockstore_flush.h

@@ -80,7 +80,7 @@ class journal_flusher_t
 {
     int trim_wanted = 0;
     bool dequeuing;
-    int flusher_count, cur_flusher_count, target_flusher_count;
+    int min_flusher_count, max_flusher_count, cur_flusher_count, target_flusher_count;
     int flusher_start_threshold;
     journal_flusher_co *co;
     blockstore_impl_t *bs;
@@ -99,7 +99,7 @@ class journal_flusher_t
     std::deque<object_id> flush_queue;
     std::map<object_id, uint64_t> flush_versions;
 public:
-    journal_flusher_t(int flusher_count, blockstore_impl_t *bs);
+    journal_flusher_t(blockstore_impl_t *bs);
     ~journal_flusher_t();
     void loop();
     bool is_active();

src/blockstore_impl.cpp

@@ -10,9 +10,9 @@ blockstore_impl_t::blockstore_impl_t(blockstore_config_t & config, ring_loop_t *
     ring_consumer.loop = [this]() { loop(); };
     ringloop->register_consumer(&ring_consumer);
     initialized = 0;
-    zero_object = (uint8_t*)memalign_or_die(MEM_ALIGNMENT, block_size);
     data_fd = meta_fd = journal.fd = -1;
     parse_config(config);
+    zero_object = (uint8_t*)memalign_or_die(MEM_ALIGNMENT, block_size);
     try
     {
         open_data();
@@ -31,7 +31,7 @@ blockstore_impl_t::blockstore_impl_t(blockstore_config_t & config, ring_loop_t *
             close(journal.fd);
         throw;
     }
-    flusher = new journal_flusher_t(flusher_count, this);
+    flusher = new journal_flusher_t(this);
 }
 
 blockstore_impl_t::~blockstore_impl_t()
@@ -101,26 +101,14 @@ void blockstore_impl_t::loop()
     {
         // try to submit ops
         unsigned initial_ring_space = ringloop->space_left();
-        // FIXME: rework this "sync polling"
-        auto cur_sync = in_progress_syncs.begin();
-        while (cur_sync != in_progress_syncs.end())
+        // has_writes == 0 - no writes before the current queue item
+        // has_writes == 1 - some writes in progress
+        // has_writes == 2 - tried to submit some writes, but failed
+        int has_writes = 0, op_idx = 0, new_idx = 0;
+        for (; op_idx < submit_queue.size(); op_idx++, new_idx++)
         {
-            if (continue_sync(*cur_sync) != 2)
-            {
-                // List is unmodified
-                cur_sync++;
-            }
-            else
-            {
-                cur_sync = in_progress_syncs.begin();
-            }
-        }
-        auto cur = submit_queue.begin();
-        int has_writes = 0;
-        while (cur != submit_queue.end())
-        {
-            auto op_ptr = cur;
-            auto op = *(cur++);
+            auto op = submit_queue[op_idx];
+            submit_queue[new_idx] = op;
             // FIXME: This needs some simplification
             // Writes should not block reads if the ring is not full and reads don't depend on them
             // In all other cases we should stop submission
@@ -142,10 +130,13 @@ void blockstore_impl_t::loop()
             }
             unsigned ring_space = ringloop->space_left();
             unsigned prev_sqe_pos = ringloop->save();
-            bool dequeue_op = false;
+            // 0 = can't submit
+            // 1 = in progress
+            // 2 = can be removed from queue
+            int wr_st = 0;
             if (op->opcode == BS_OP_READ)
             {
-                dequeue_op = dequeue_read(op);
+                wr_st = dequeue_read(op);
             }
             else if (op->opcode == BS_OP_WRITE || op->opcode == BS_OP_WRITE_STABLE)
             {
@@ -154,8 +145,8 @@ void blockstore_impl_t::loop()
                     // Some writes already could not be submitted
                     continue;
                 }
-                dequeue_op = dequeue_write(op);
-                has_writes = dequeue_op ? 1 : 2;
+                wr_st = dequeue_write(op);
+                has_writes = wr_st > 0 ? 1 : 2;
             }
             else if (op->opcode == BS_OP_DELETE)
             {
@@ -164,8 +155,8 @@ void blockstore_impl_t::loop()
                     // Some writes already could not be submitted
                     continue;
                 }
-                dequeue_op = dequeue_del(op);
-                has_writes = dequeue_op ? 1 : 2;
+                wr_st = dequeue_del(op);
+                has_writes = wr_st > 0 ? 1 : 2;
             }
             else if (op->opcode == BS_OP_SYNC)
             {
@@ -178,29 +169,31 @@ void blockstore_impl_t::loop()
                     // Can't submit SYNC before previous writes
                     continue;
                 }
-                dequeue_op = dequeue_sync(op);
+                wr_st = continue_sync(op, false);
+                if (wr_st != 2)
+                {
+                    has_writes = wr_st > 0 ? 1 : 2;
+                }
             }
             else if (op->opcode == BS_OP_STABLE)
             {
-                dequeue_op = dequeue_stable(op);
+                wr_st = dequeue_stable(op);
             }
             else if (op->opcode == BS_OP_ROLLBACK)
             {
-                dequeue_op = dequeue_rollback(op);
+                wr_st = dequeue_rollback(op);
             }
             else if (op->opcode == BS_OP_LIST)
             {
-                // LIST doesn't need to be blocked by previous modifications,
-                // it only needs to include all in-progress writes as they're guaranteed
-                // to be readable and stabilizable/rollbackable by subsequent operations
+                // LIST doesn't need to be blocked by previous modifications
                 process_list(op);
-                dequeue_op = true;
+                wr_st = 2;
             }
-            if (dequeue_op)
+            if (wr_st == 2)
             {
-                submit_queue.erase(op_ptr);
+                new_idx--;
             }
-            else
+            if (wr_st == 0)
             {
                 ringloop->restore(prev_sqe_pos);
                 if (PRIV(op)->wait_for == WAIT_SQE)
@@ -211,6 +204,14 @@ void blockstore_impl_t::loop()
                 }
             }
         }
+        if (op_idx != new_idx)
+        {
+            while (op_idx < submit_queue.size())
+            {
+                submit_queue[new_idx++] = submit_queue[op_idx++];
+            }
+            submit_queue.resize(new_idx);
+        }
         if (!readonly)
         {
             flusher->loop();
@@ -233,7 +234,7 @@ bool blockstore_impl_t::is_safe_to_stop()
 {
     // It's safe to stop blockstore when there are no in-flight operations,
     // no in-progress syncs and flusher isn't doing anything
-    if (submit_queue.size() > 0 || in_progress_syncs.size() > 0 || !readonly && flusher->is_active())
+    if (submit_queue.size() > 0 || !readonly && flusher->is_active())
     {
         return false;
     }
@@ -300,7 +301,7 @@ void blockstore_impl_t::check_wait(blockstore_op_t *op)
     }
     else if (PRIV(op)->wait_for == WAIT_FREE)
     {
-        if (!data_alloc->get_free_count() && !flusher->is_active())
+        if (!data_alloc->get_free_count() && flusher->is_active())
         {
 #ifdef BLOCKSTORE_DEBUG
             printf("Still waiting for free space on the data device\n");
@@ -315,7 +316,7 @@ void blockstore_impl_t::check_wait(blockstore_op_t *op)
     }
 }
 
-void blockstore_impl_t::enqueue_op(blockstore_op_t *op, bool first)
+void blockstore_impl_t::enqueue_op(blockstore_op_t *op)
 {
     if (op->opcode < BS_OP_MIN || op->opcode > BS_OP_MAX ||
         ((op->opcode == BS_OP_READ || op->opcode == BS_OP_WRITE || op->opcode == BS_OP_WRITE_STABLE) && (
@@ -323,8 +324,7 @@ void blockstore_impl_t::enqueue_op(blockstore_op_t *op, bool first)
             op->len > block_size-op->offset ||
             (op->len % disk_alignment)
         )) ||
-        readonly && op->opcode != BS_OP_READ && op->opcode != BS_OP_LIST ||
-        first && (op->opcode == BS_OP_WRITE || op->opcode == BS_OP_WRITE_STABLE))
+        readonly && op->opcode != BS_OP_READ && op->opcode != BS_OP_LIST)
     {
         // Basic verification not passed
         op->retval = -EINVAL;
@@ -374,25 +374,12 @@ void blockstore_impl_t::enqueue_op(blockstore_op_t *op, bool first)
         std::function<void (blockstore_op_t*)>(op->callback)(op);
         return;
     }
-    if (op->opcode == BS_OP_SYNC && immediate_commit == IMMEDIATE_ALL)
-    {
-        op->retval = 0;
-        std::function<void (blockstore_op_t*)>(op->callback)(op);
-        return;
-    }
     // Call constructor without allocating memory. We'll call destructor before returning op back
     new ((void*)op->private_data) blockstore_op_private_t;
     PRIV(op)->wait_for = 0;
     PRIV(op)->op_state = 0;
     PRIV(op)->pending_ops = 0;
-    if (!first)
-    {
-        submit_queue.push_back(op);
-    }
-    else
-    {
-        submit_queue.push_front(op);
-    }
+    submit_queue.push_back(op);
     ringloop->wakeup();
 }
 

src/blockstore_impl.h

@@ -160,8 +160,6 @@ struct blockstore_op_private_t
     // Sync
     std::vector<obj_ver_id> sync_big_writes, sync_small_writes;
     int sync_small_checked, sync_big_checked;
-    std::list<blockstore_op_t*>::iterator in_progress_ptr;
-    int prev_sync_count;
 };
 
 // https://github.com/algorithm-ninja/cpp-btree
@@ -199,8 +197,8 @@ class blockstore_impl_t
     // Suitable only for server SSDs with capacitors, requires disabled data and journal fsyncs
     int immediate_commit = IMMEDIATE_NONE;
     bool inmemory_meta = false;
-    // Maximum flusher count
-    unsigned flusher_count;
+    // Maximum and minimum flusher count
+    unsigned max_flusher_count, min_flusher_count;
     // Maximum queue depth
     unsigned max_write_iodepth = 128;
     /******* END OF OPTIONS *******/
@@ -210,9 +208,9 @@ class blockstore_impl_t
     blockstore_clean_db_t clean_db;
     uint8_t *clean_bitmap = NULL;
     blockstore_dirty_db_t dirty_db;
-    std::list<blockstore_op_t*> submit_queue; // FIXME: funny thing is that vector is better here
+    std::vector<blockstore_op_t*> submit_queue;
     std::vector<obj_ver_id> unsynced_big_writes, unsynced_small_writes;
-    std::list<blockstore_op_t*> in_progress_syncs; // ...and probably here, too
+    int unsynced_big_write_count = 0;
     allocator *data_alloc = NULL;
     uint8_t *zero_object;
 
@@ -271,6 +269,7 @@ class blockstore_impl_t
 
     // Write
     bool enqueue_write(blockstore_op_t *op);
+    void cancel_all_writes(blockstore_op_t *op, blockstore_dirty_db_t::iterator dirty_it, int retval);
     int dequeue_write(blockstore_op_t *op);
     int dequeue_del(blockstore_op_t *op);
     int continue_write(blockstore_op_t *op);
@@ -278,16 +277,14 @@ class blockstore_impl_t
     void handle_write_event(ring_data_t *data, blockstore_op_t *op);
 
     // Sync
-    int dequeue_sync(blockstore_op_t *op);
+    int continue_sync(blockstore_op_t *op, bool queue_has_in_progress_sync);
     void handle_sync_event(ring_data_t *data, blockstore_op_t *op);
-    int continue_sync(blockstore_op_t *op);
-    void ack_one_sync(blockstore_op_t *op);
-    int ack_sync(blockstore_op_t *op);
+    void ack_sync(blockstore_op_t *op);
 
     // Stabilize
     int dequeue_stable(blockstore_op_t *op);
     int continue_stable(blockstore_op_t *op);
-    void mark_stable(const obj_ver_id & ov);
+    void mark_stable(const obj_ver_id & ov, bool forget_dirty = false);
     void handle_stable_event(ring_data_t *data, blockstore_op_t *op);
     void stabilize_object(object_id oid, uint64_t max_ver);
 
@@ -322,7 +319,7 @@ public:
     bool is_stalled();
 
     // Submission
-    void enqueue_op(blockstore_op_t *op, bool first = false);
+    void enqueue_op(blockstore_op_t *op);
 
     // Unstable writes are added here (map of object_id -> version)
     std::unordered_map<object_id, uint64_t> unstable_writes;
@@ -330,5 +327,5 @@ public:
     inline uint32_t get_block_size() { return block_size; }
     inline uint64_t get_block_count() { return block_count; }
     inline uint64_t get_free_block_count() { return data_alloc->get_free_count(); }
-    inline uint32_t get_disk_alignment() { return disk_alignment; }
+    inline uint32_t get_bitmap_granularity() { return disk_alignment; }
 };

src/blockstore_init.cpp

@@ -111,7 +111,10 @@ void blockstore_init_meta::handle_entries(void* entries, unsigned count, int blo
                 {
                     // free the previous block
 #ifdef BLOCKSTORE_DEBUG
-                    printf("Free block %lu (new location is %lu)\n", clean_it->second.location >> block_order, done_cnt+i);
+                    printf("Free block %lu from %lx:%lx v%lu (new location is %lu)\n",
+                        clean_it->second.location >> block_order,
+                        clean_it->first.inode, clean_it->first.stripe, clean_it->second.version,
+                        done_cnt+i);
 #endif
                     bs->data_alloc->set(clean_it->second.location >> block_order, false);
                 }
@@ -399,6 +402,18 @@ resume_1:
             }
         }
     }
+    for (auto ov: double_allocs)
+    {
+        auto dirty_it = bs->dirty_db.find(ov);
+        if (dirty_it != bs->dirty_db.end() &&
+            IS_BIG_WRITE(dirty_it->second.state) &&
+            dirty_it->second.location == UINT64_MAX)
+        {
+            printf("Fatal error (bug): %lx:%lx v%lu big_write journal_entry was allocated over another object\n",
+                dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version);
+            exit(1);
+        }
+    }
     bs->flusher->mark_trim_possible();
     bs->journal.dirty_start = bs->journal.next_free;
     printf(
@@ -549,7 +564,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     unstab = unstab < ov.version ? ov.version : unstab;
                     if (je->type == JE_SMALL_WRITE_INSTANT)
                     {
-                        bs->mark_stable(ov);
+                        bs->mark_stable(ov, true);
                     }
                 }
             }
@@ -579,32 +594,10 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         // its data and metadata are already flushed.
                         // We don't know if newer versions are flushed, but
                         // the previous delete definitely is.
-                        // So we flush previous dirty entries, but retain the clean one.
+                        // So we forget previous dirty entries, but retain the clean one.
                         // This feature is required for writes happening shortly
                         // after deletes.
-                        auto dirty_end = dirty_it;
-                        dirty_end++;
-                        while (1)
-                        {
-                            if (dirty_it == bs->dirty_db.begin())
-                            {
-                                break;
-                            }
-                            dirty_it--;
-                            if (dirty_it->first.oid != je->big_write.oid)
-                            {
-                                dirty_it++;
-                                break;
-                            }
-                        }
-                        auto clean_it = bs->clean_db.find(je->big_write.oid);
-                        bs->erase_dirty(
-                            dirty_it, dirty_end,
-                            clean_it != bs->clean_db.end() ? clean_it->second.location : UINT64_MAX
-                        );
-                        // Remove it from the flusher's queue, too
-                        // Otherwise it may end up referring to a small unstable write after reading the rest of the journal
-                        bs->flusher->remove_flush(je->big_write.oid);
+                        erase_dirty_object(dirty_it);
                     }
                 }
                 auto clean_it = bs->clean_db.find(je->big_write.oid);
@@ -616,18 +609,33 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                         .oid = je->big_write.oid,
                         .version = je->big_write.version,
                     };
-                    bs->dirty_db.emplace(ov, (dirty_entry){
+                    auto dirty_it = bs->dirty_db.emplace(ov, (dirty_entry){
                         .state = (BS_ST_BIG_WRITE | BS_ST_SYNCED),
                         .flags = 0,
                         .location = je->big_write.location,
                         .offset = je->big_write.offset,
                         .len = je->big_write.len,
                         .journal_sector = proc_pos,
-                    });
+                    }).first;
+                    if (bs->data_alloc->get(je->big_write.location >> bs->block_order))
+                    {
+                        // This is probably a big_write that's already flushed and freed, but it may
+                        // also indicate a bug. So we remember such entries and recheck them afterwards.
+                        // If it's not a bug they won't be present after reading the whole journal.
+                        dirty_it->second.location = UINT64_MAX;
+                        double_allocs.push_back(ov);
+                    }
+                    else
+                    {
 #ifdef BLOCKSTORE_DEBUG
-                    printf("Allocate block %lu\n", je->big_write.location >> bs->block_order);
+                        printf(
+                            "Allocate block (journal) %lu: %lx:%lx v%lu\n",
+                            je->big_write.location >> bs->block_order,
+                            ov.oid.inode, ov.oid.stripe, ov.version
+                        );
 #endif
-                    bs->data_alloc->set(je->big_write.location >> bs->block_order, true);
+                        bs->data_alloc->set(je->big_write.location >> bs->block_order, true);
+                    }
                     bs->journal.used_sectors[proc_pos]++;
 #ifdef BLOCKSTORE_DEBUG
                     printf(
@@ -639,7 +647,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     unstab = unstab < ov.version ? ov.version : unstab;
                     if (je->type == JE_BIG_WRITE_INSTANT)
                     {
-                        bs->mark_stable(ov);
+                        bs->mark_stable(ov, true);
                     }
                 }
             }
@@ -653,7 +661,7 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     .oid = je->stable.oid,
                     .version = je->stable.version,
                 };
-                bs->mark_stable(ov);
+                bs->mark_stable(ov, true);
             }
             else if (je->type == JE_ROLLBACK)
             {
@@ -672,9 +680,26 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
 #ifdef BLOCKSTORE_DEBUG
                 printf("je_delete oid=%lx:%lx ver=%lu\n", je->del.oid.inode, je->del.oid.stripe, je->del.version);
 #endif
+                bool dirty_exists = false;
+                auto dirty_it = bs->dirty_db.upper_bound((obj_ver_id){
+                    .oid = je->del.oid,
+                    .version = UINT64_MAX,
+                });
+                if (dirty_it != bs->dirty_db.begin())
+                {
+                    dirty_it--;
+                    dirty_exists = dirty_it->first.oid == je->del.oid;
+                }
                 auto clean_it = bs->clean_db.find(je->del.oid);
-                if (clean_it == bs->clean_db.end() ||
-                    clean_it->second.version < je->del.version)
+                bool clean_exists = (clean_it != bs->clean_db.end() &&
+                    clean_it->second.version < je->del.version);
+                if (!clean_exists && dirty_exists)
+                {
+                    // Clean entry doesn't exist. This means that the delete is already flushed.
+                    // So we must not flush this object anymore.
+                    erase_dirty_object(dirty_it);
+                }
+                else if (clean_exists || dirty_exists)
                 {
                     // oid, version
                     obj_ver_id ov = {
@@ -692,8 +717,9 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
                     bs->journal.used_sectors[proc_pos]++;
                     // Deletions are treated as immediately stable, because
                     // "2-phase commit" (write->stabilize) isn't sufficient for them anyway
-                    bs->mark_stable(ov);
+                    bs->mark_stable(ov, true);
                 }
+                // Ignore delete if neither preceding dirty entries nor the clean one are present
             }
             started = true;
             pos += je->size;
@@ -704,3 +730,30 @@ int blockstore_init_journal::handle_journal_part(void *buf, uint64_t done_pos, u
     bs->journal.next_free = next_free;
     return 1;
 }
+
+void blockstore_init_journal::erase_dirty_object(blockstore_dirty_db_t::iterator dirty_it)
+{
+    auto oid = dirty_it->first.oid;
+    auto dirty_end = dirty_it;
+    dirty_end++;
+    while (1)
+    {
+        if (dirty_it == bs->dirty_db.begin())
+        {
+            break;
+        }
+        dirty_it--;
+        if (dirty_it->first.oid != oid)
+        {
+            dirty_it++;
+            break;
+        }
+    }
+    auto clean_it = bs->clean_db.find(oid);
+    uint64_t clean_loc = clean_it != bs->clean_db.end()
+        ? clean_it->second.location : UINT64_MAX;
+    bs->erase_dirty(dirty_it, dirty_end, clean_loc);
+    // Remove it from the flusher's queue, too
+    // Otherwise it may end up referring to a small unstable write after reading the rest of the journal
+    bs->flusher->remove_flush(oid);
+}

src/blockstore_init.h

@@ -36,6 +36,7 @@ class blockstore_init_journal
     bool started = false;
     uint64_t next_free;
     std::vector<bs_init_journal_done> done;
+    std::vector<obj_ver_id> double_allocs;
     uint64_t journal_pos = 0;
     uint64_t continue_pos = 0;
     void *init_write_buf = NULL;
@@ -48,6 +49,7 @@ class blockstore_init_journal
     std::function<void(ring_data_t*)> simple_callback;
     int handle_journal_part(void *buf, uint64_t done_pos, uint64_t len);
     void handle_event(ring_data_t *data);
+    void erase_dirty_object(blockstore_dirty_db_t::iterator dirty_it);
 public:
     blockstore_init_journal(blockstore_impl_t* bs);
     int loop();

src/blockstore_open.cpp

@@ -69,7 +69,10 @@ void blockstore_impl_t::parse_config(blockstore_config_t & config)
     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);
-    flusher_count = strtoull(config["flusher_count"].c_str(), NULL, 10);
+    max_flusher_count = strtoull(config["max_flusher_count"].c_str(), NULL, 10);
+    if (!max_flusher_count)
+        max_flusher_count = strtoull(config["flusher_count"].c_str(), NULL, 10);
+    min_flusher_count = strtoull(config["min_flusher_count"].c_str(), NULL, 10);
     max_write_iodepth = strtoull(config["max_write_iodepth"].c_str(), NULL, 10);
     // Validate
     if (!block_size)
@@ -80,9 +83,13 @@ void blockstore_impl_t::parse_config(blockstore_config_t & config)
     {
         throw std::runtime_error("Bad block size");
     }
-    if (!flusher_count)
+    if (!max_flusher_count)
     {
-        flusher_count = 32;
+        max_flusher_count = 256;
+    }
+    if (!min_flusher_count)
+    {
+        min_flusher_count = 1;
     }
     if (!max_write_iodepth)
     {

src/blockstore_read.cpp

@@ -112,7 +112,7 @@ int blockstore_impl_t::dequeue_read(blockstore_op_t *read_op)
         read_op->version = 0;
         read_op->retval = read_op->len;
         FINISH_OP(read_op);
-        return 1;
+        return 2;
     }
     uint64_t fulfilled = 0;
     PRIV(read_op)->pending_ops = 0;
@@ -191,8 +191,8 @@ int blockstore_impl_t::dequeue_read(blockstore_op_t *read_op)
                     if (bmp_end > bmp_start)
                     {
                         // fill with zeroes
-                        fulfill_read(read_op, fulfilled, bmp_start * bitmap_granularity,
-                            bmp_end * bitmap_granularity, (BS_ST_DELETE | BS_ST_STABLE), 0, 0);
+                        assert(fulfill_read(read_op, fulfilled, bmp_start * bitmap_granularity,
+                            bmp_end * bitmap_granularity, (BS_ST_DELETE | BS_ST_STABLE), 0, 0));
                     }
                     bmp_start = bmp_end;
                     while (clean_entry_bitmap[bmp_end >> 3] & (1 << (bmp_end & 0x7)) && bmp_end < bmp_size)
@@ -218,7 +218,7 @@ int blockstore_impl_t::dequeue_read(blockstore_op_t *read_op)
     else if (fulfilled < read_op->len)
     {
         // fill remaining parts with zeroes
-        fulfill_read(read_op, fulfilled, 0, block_size, (BS_ST_DELETE | BS_ST_STABLE), 0, 0);
+        assert(fulfill_read(read_op, fulfilled, 0, block_size, (BS_ST_DELETE | BS_ST_STABLE), 0, 0));
     }
     assert(fulfilled == read_op->len);
     read_op->version = result_version;
@@ -232,10 +232,10 @@ int blockstore_impl_t::dequeue_read(blockstore_op_t *read_op)
         }
         read_op->retval = read_op->len;
         FINISH_OP(read_op);
-        return 1;
+        return 2;
     }
     read_op->retval = 0;
-    return 1;
+    return 2;
 }
 
 void blockstore_impl_t::handle_read_event(ring_data_t *data, blockstore_op_t *op)

src/blockstore_rollback.cpp

@@ -50,7 +50,7 @@ skip_ov:
                 {
                     op->retval = -EBUSY;
                     FINISH_OP(op);
-                    return 1;
+                    return 2;
                 }
                 if (dirty_it == dirty_db.begin())
                 {
@@ -66,7 +66,7 @@ skip_ov:
         // Already rolled back
         op->retval = 0;
         FINISH_OP(op);
-        return 1;
+        return 2;
     }
     // Check journal space
     blockstore_journal_check_t space_check(this);
@@ -126,11 +126,8 @@ resume_2:
 resume_3:
     if (!disable_journal_fsync)
     {
-        io_uring_sqe *sqe = get_sqe();
-        if (!sqe)
-        {
-            return 0;
-        }
+        io_uring_sqe *sqe;
+        BS_SUBMIT_GET_SQE_DECL(sqe);
         ring_data_t *data = ((ring_data_t*)sqe->user_data);
         my_uring_prep_fsync(sqe, journal.fd, IORING_FSYNC_DATASYNC);
         data->iov = { 0 };
@@ -151,7 +148,7 @@ resume_5:
     // Acknowledge op
     op->retval = 0;
     FINISH_OP(op);
-    return 1;
+    return 2;
 }
 
 void blockstore_impl_t::mark_rolled_back(const obj_ver_id & ov)
@@ -166,10 +163,7 @@ void blockstore_impl_t::mark_rolled_back(const obj_ver_id & ov)
         auto rm_start = it;
         auto rm_end = it;
         it--;
-        while (it->first.oid == ov.oid &&
-            it->first.version > ov.version &&
-            !IS_IN_FLIGHT(it->second.state) &&
-            !IS_STABLE(it->second.state))
+        while (1)
         {
             if (it->first.oid != ov.oid)
                 break;
@@ -179,7 +173,7 @@ void blockstore_impl_t::mark_rolled_back(const obj_ver_id & ov)
                     max_unstable = it->first.version;
                 break;
             }
-            else if (IS_STABLE(it->second.state))
+            else if (IS_IN_FLIGHT(it->second.state) || IS_STABLE(it->second.state))
                 break;
             // Remove entry
             rm_start = it;
@@ -190,14 +184,14 @@ void blockstore_impl_t::mark_rolled_back(const obj_ver_id & ov)
         if (rm_start != rm_end)
         {
             erase_dirty(rm_start, rm_end, UINT64_MAX);
-        }
-        auto unstab_it = unstable_writes.find(ov.oid);
-        if (unstab_it != unstable_writes.end())
-        {
-            if (max_unstable == 0)
-                unstable_writes.erase(unstab_it);
-            else
-                unstab_it->second = max_unstable;
+            auto unstab_it = unstable_writes.find(ov.oid);
+            if (unstab_it != unstable_writes.end())
+            {
+                if (max_unstable == 0)
+                    unstable_writes.erase(unstab_it);
+                else
+                    unstab_it->second = max_unstable;
+            }
         }
     }
 }
@@ -216,10 +210,7 @@ void blockstore_impl_t::handle_rollback_event(ring_data_t *data, blockstore_op_t
     if (PRIV(op)->pending_ops == 0)
     {
         PRIV(op)->op_state++;
-        if (!continue_rollback(op))
-        {
-            submit_queue.push_front(op);
-        }
+        ringloop->wakeup();
     }
 }
 
@@ -257,10 +248,12 @@ void blockstore_impl_t::erase_dirty(blockstore_dirty_db_t::iterator dirty_start,
     }
     while (1)
     {
-        if (IS_BIG_WRITE(dirty_it->second.state) && dirty_it->second.location != clean_loc)
+        if (IS_BIG_WRITE(dirty_it->second.state) && dirty_it->second.location != clean_loc &&
+            dirty_it->second.location != UINT64_MAX)
         {
 #ifdef BLOCKSTORE_DEBUG
-            printf("Free block %lu\n", dirty_it->second.location >> block_order);
+            printf("Free block %lu from %lx:%lx v%lu\n", dirty_it->second.location >> block_order,
+                dirty_it->first.oid.inode, dirty_it->first.oid.stripe, dirty_it->first.version);
 #endif
             data_alloc->set(dirty_it->second.location >> block_order, false);
         }

src/blockstore_stable.cpp

@@ -60,7 +60,7 @@ int blockstore_impl_t::dequeue_stable(blockstore_op_t *op)
                 // No such object version
                 op->retval = -ENOENT;
                 FINISH_OP(op);
-                return 1;
+                return 2;
             }
             else
             {
@@ -77,7 +77,7 @@ int blockstore_impl_t::dequeue_stable(blockstore_op_t *op)
             // Object not synced yet. Caller must sync it first
             op->retval = -EBUSY;
             FINISH_OP(op);
-            return 1;
+            return 2;
         }
         else if (!IS_STABLE(dirty_it->second.state))
         {
@@ -89,7 +89,7 @@ int blockstore_impl_t::dequeue_stable(blockstore_op_t *op)
         // Already stable
         op->retval = 0;
         FINISH_OP(op);
-        return 1;
+        return 2;
     }
     // Check journal space
     blockstore_journal_check_t space_check(this);
@@ -150,11 +150,8 @@ resume_2:
 resume_3:
     if (!disable_journal_fsync)
     {
-        io_uring_sqe *sqe = get_sqe();
-        if (!sqe)
-        {
-            return 0;
-        }
+        io_uring_sqe *sqe;
+        BS_SUBMIT_GET_SQE_DECL(sqe);
         ring_data_t *data = ((ring_data_t*)sqe->user_data);
         my_uring_prep_fsync(sqe, journal.fd, IORING_FSYNC_DATASYNC);
         data->iov = { 0 };
@@ -171,30 +168,50 @@ resume_5:
     for (i = 0, v = (obj_ver_id*)op->buf; i < op->len; i++, v++)
     {
         // Mark all dirty_db entries up to op->version as stable
+#ifdef BLOCKSTORE_DEBUG
+        printf("Stabilize %lx:%lx v%lu\n", v->oid.inode, v->oid.stripe, v->version);
+#endif
         mark_stable(*v);
     }
     // Acknowledge op
     op->retval = 0;
     FINISH_OP(op);
-    return 1;
+    return 2;
 }
 
-void blockstore_impl_t::mark_stable(const obj_ver_id & v)
+void blockstore_impl_t::mark_stable(const obj_ver_id & v, bool forget_dirty)
 {
     auto dirty_it = dirty_db.find(v);
     if (dirty_it != dirty_db.end())
     {
         while (1)
         {
+            bool was_stable = IS_STABLE(dirty_it->second.state);
             if ((dirty_it->second.state & BS_ST_WORKFLOW_MASK) == BS_ST_SYNCED)
             {
                 dirty_it->second.state = (dirty_it->second.state & ~BS_ST_WORKFLOW_MASK) | BS_ST_STABLE;
             }
-            else if (IS_STABLE(dirty_it->second.state))
+            if (forget_dirty && (IS_BIG_WRITE(dirty_it->second.state) ||
+                IS_DELETE(dirty_it->second.state)))
             {
+                // Big write overrides all previous dirty entries
+                auto erase_end = dirty_it;
+                while (dirty_it != dirty_db.begin())
+                {
+                    dirty_it--;
+                    if (dirty_it->first.oid != v.oid)
+                    {
+                        dirty_it++;
+                        break;
+                    }
+                }
+                auto clean_it = clean_db.find(v.oid);
+                uint64_t clean_loc = clean_it != clean_db.end()
+                    ? clean_it->second.location : UINT64_MAX;
+                erase_dirty(dirty_it, erase_end, clean_loc);
                 break;
             }
-            if (dirty_it == dirty_db.begin())
+            if (was_stable || dirty_it == dirty_db.begin())
             {
                 break;
             }
@@ -228,9 +245,6 @@ void blockstore_impl_t::handle_stable_event(ring_data_t *data, blockstore_op_t *
     if (PRIV(op)->pending_ops == 0)
     {
         PRIV(op)->op_state++;
-        if (!continue_stable(op))
-        {
-            submit_queue.push_front(op);
-        }
+        ringloop->wakeup();
     }
 }

src/blockstore_sync.cpp

@@ -12,11 +12,19 @@
 #define SYNC_JOURNAL_SYNC_SENT 7
 #define SYNC_DONE 8
 
-int blockstore_impl_t::dequeue_sync(blockstore_op_t *op)
+int blockstore_impl_t::continue_sync(blockstore_op_t *op, bool queue_has_in_progress_sync)
 {
+    if (immediate_commit == IMMEDIATE_ALL)
+    {
+        // We can return immediately because sync is only dequeued after all previous writes
+        op->retval = 0;
+        FINISH_OP(op);
+        return 2;
+    }
     if (PRIV(op)->op_state == 0)
     {
         stop_sync_submitted = false;
+        unsynced_big_write_count -= unsynced_big_writes.size();
         PRIV(op)->sync_big_writes.swap(unsynced_big_writes);
         PRIV(op)->sync_small_writes.swap(unsynced_small_writes);
         PRIV(op)->sync_small_checked = 0;
@@ -29,34 +37,15 @@ int blockstore_impl_t::dequeue_sync(blockstore_op_t *op)
             PRIV(op)->op_state = SYNC_HAS_SMALL;
         else
             PRIV(op)->op_state = SYNC_DONE;
-        // Always add sync to in_progress_syncs because we clear unsynced_big_writes and unsynced_small_writes
-        PRIV(op)->prev_sync_count = in_progress_syncs.size();
-        PRIV(op)->in_progress_ptr = in_progress_syncs.insert(in_progress_syncs.end(), op);
     }
-    continue_sync(op);
-    // Always dequeue because we always add syncs to in_progress_syncs
-    return 1;
-}
-
-int blockstore_impl_t::continue_sync(blockstore_op_t *op)
-{
-    auto cb = [this, op](ring_data_t *data) { handle_sync_event(data, op); };
     if (PRIV(op)->op_state == SYNC_HAS_SMALL)
     {
         // No big writes, just fsync the journal
-        for (; PRIV(op)->sync_small_checked < PRIV(op)->sync_small_writes.size(); PRIV(op)->sync_small_checked++)
-        {
-            if (IS_IN_FLIGHT(dirty_db[PRIV(op)->sync_small_writes[PRIV(op)->sync_small_checked]].state))
-            {
-                // Wait for small inflight writes to complete
-                return 0;
-            }
-        }
         if (journal.sector_info[journal.cur_sector].dirty)
         {
             // Write out the last journal sector if it happens to be dirty
             BS_SUBMIT_GET_ONLY_SQE(sqe);
-            prepare_journal_sector_write(journal, journal.cur_sector, sqe, cb);
+            prepare_journal_sector_write(journal, journal.cur_sector, sqe, [this, op](ring_data_t *data) { handle_sync_event(data, op); });
             PRIV(op)->min_flushed_journal_sector = PRIV(op)->max_flushed_journal_sector = 1 + journal.cur_sector;
             PRIV(op)->pending_ops = 1;
             PRIV(op)->op_state = SYNC_JOURNAL_WRITE_SENT;
@@ -69,21 +58,13 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
     }
     if (PRIV(op)->op_state == SYNC_HAS_BIG)
     {
-        for (; PRIV(op)->sync_big_checked < PRIV(op)->sync_big_writes.size(); PRIV(op)->sync_big_checked++)
-        {
-            if (IS_IN_FLIGHT(dirty_db[PRIV(op)->sync_big_writes[PRIV(op)->sync_big_checked]].state))
-            {
-                // Wait for big inflight writes to complete
-                return 0;
-            }
-        }
         // 1st step: fsync data
         if (!disable_data_fsync)
         {
             BS_SUBMIT_GET_SQE(sqe, data);
             my_uring_prep_fsync(sqe, data_fd, IORING_FSYNC_DATASYNC);
             data->iov = { 0 };
-            data->callback = cb;
+            data->callback = [this, op](ring_data_t *data) { handle_sync_event(data, op); };
             PRIV(op)->min_flushed_journal_sector = PRIV(op)->max_flushed_journal_sector = 0;
             PRIV(op)->pending_ops = 1;
             PRIV(op)->op_state = SYNC_DATA_SYNC_SENT;
@@ -96,14 +77,6 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
     }
     if (PRIV(op)->op_state == SYNC_DATA_SYNC_DONE)
     {
-        for (; PRIV(op)->sync_small_checked < PRIV(op)->sync_small_writes.size(); PRIV(op)->sync_small_checked++)
-        {
-            if (IS_IN_FLIGHT(dirty_db[PRIV(op)->sync_small_writes[PRIV(op)->sync_small_checked]].state))
-            {
-                // Wait for small inflight writes to complete
-                return 0;
-            }
-        }
         // 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);
@@ -127,7 +100,7 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
             {
                 if (cur_sector == -1)
                     PRIV(op)->min_flushed_journal_sector = 1 + journal.cur_sector;
-                prepare_journal_sector_write(journal, journal.cur_sector, sqe[s++], cb);
+                prepare_journal_sector_write(journal, journal.cur_sector, sqe[s++], [this, op](ring_data_t *data) { handle_sync_event(data, op); });
                 cur_sector = journal.cur_sector;
             }
             journal_entry_big_write *je = (journal_entry_big_write*)prefill_single_journal_entry(
@@ -152,7 +125,7 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
             journal.crc32_last = je->crc32;
             it++;
         }
-        prepare_journal_sector_write(journal, journal.cur_sector, sqe[s++], cb);
+        prepare_journal_sector_write(journal, journal.cur_sector, sqe[s++], [this, op](ring_data_t *data) { handle_sync_event(data, op); });
         assert(s == space_check.sectors_to_write);
         if (cur_sector == -1)
             PRIV(op)->min_flushed_journal_sector = 1 + journal.cur_sector;
@@ -168,7 +141,7 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
             BS_SUBMIT_GET_SQE(sqe, data);
             my_uring_prep_fsync(sqe, journal.fd, IORING_FSYNC_DATASYNC);
             data->iov = { 0 };
-            data->callback = cb;
+            data->callback = [this, op](ring_data_t *data) { handle_sync_event(data, op); };
             PRIV(op)->pending_ops = 1;
             PRIV(op)->op_state = SYNC_JOURNAL_SYNC_SENT;
             return 1;
@@ -178,9 +151,10 @@ int blockstore_impl_t::continue_sync(blockstore_op_t *op)
             PRIV(op)->op_state = SYNC_DONE;
         }
     }
-    if (PRIV(op)->op_state == SYNC_DONE)
+    if (PRIV(op)->op_state == SYNC_DONE && !queue_has_in_progress_sync)
     {
-        return ack_sync(op);
+        ack_sync(op);
+        return 2;
     }
     return 1;
 }
@@ -212,42 +186,16 @@ void blockstore_impl_t::handle_sync_event(ring_data_t *data, blockstore_op_t *op
         else if (PRIV(op)->op_state == SYNC_JOURNAL_SYNC_SENT)
         {
             PRIV(op)->op_state = SYNC_DONE;
-            ack_sync(op);
         }
         else
         {
             throw std::runtime_error("BUG: unexpected sync op state");
         }
+        ringloop->wakeup();
     }
 }
 
-int blockstore_impl_t::ack_sync(blockstore_op_t *op)
-{
-    if (PRIV(op)->op_state == SYNC_DONE && PRIV(op)->prev_sync_count == 0)
-    {
-        // Remove dependency of subsequent syncs
-        auto it = PRIV(op)->in_progress_ptr;
-        int done_syncs = 1;
-        ++it;
-        // Acknowledge sync
-        ack_one_sync(op);
-        while (it != in_progress_syncs.end())
-        {
-            auto & next_sync = *it++;
-            PRIV(next_sync)->prev_sync_count -= done_syncs;
-            if (PRIV(next_sync)->prev_sync_count == 0 && PRIV(next_sync)->op_state == SYNC_DONE)
-            {
-                done_syncs++;
-                // Acknowledge next_sync
-                ack_one_sync(next_sync);
-            }
-        }
-        return 2;
-    }
-    return 0;
-}
-
-void blockstore_impl_t::ack_one_sync(blockstore_op_t *op)
+void blockstore_impl_t::ack_sync(blockstore_op_t *op)
 {
     // Handle states
     for (auto it = PRIV(op)->sync_big_writes.begin(); it != PRIV(op)->sync_big_writes.end(); it++)
@@ -295,7 +243,6 @@ void blockstore_impl_t::ack_one_sync(blockstore_op_t *op)
             }
         }
     }
-    in_progress_syncs.erase(PRIV(op)->in_progress_ptr);
     op->retval = 0;
     FINISH_OP(op);
 }

src/blockstore_write.cpp

@@ -124,6 +124,29 @@ bool blockstore_impl_t::enqueue_write(blockstore_op_t *op)
     return true;
 }
 
+void blockstore_impl_t::cancel_all_writes(blockstore_op_t *op, blockstore_dirty_db_t::iterator dirty_it, int retval)
+{
+    while (dirty_it != dirty_db.end() && dirty_it->first.oid == op->oid)
+    {
+        dirty_db.erase(dirty_it++);
+    }
+    bool found = false;
+    for (auto other_op: submit_queue)
+    {
+        if (!found && other_op == op)
+            found = true;
+        else if (found && other_op->oid == op->oid &&
+            (other_op->opcode == BS_OP_WRITE || other_op->opcode == BS_OP_WRITE_STABLE))
+        {
+            // Mark operations to cancel them
+            PRIV(other_op)->real_version = UINT64_MAX;
+            other_op->retval = retval;
+        }
+    }
+    op->retval = retval;
+    FINISH_OP(op);
+}
+
 // First step of the write algorithm: dequeue operation and submit initial write(s)
 int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
 {
@@ -143,6 +166,12 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
     }
     if (PRIV(op)->real_version != 0)
     {
+        if (PRIV(op)->real_version == UINT64_MAX)
+        {
+            // This is the flag value used to cancel operations
+            FINISH_OP(op);
+            return 2;
+        }
         // Restore original low version number for unblocked operations
 #ifdef BLOCKSTORE_DEBUG
         printf("Restoring %lx:%lx version: v%lu -> v%lu\n", op->oid.inode, op->oid.stripe, op->version, PRIV(op)->real_version);
@@ -152,11 +181,9 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         if (prev_it->first.oid == op->oid && prev_it->first.version >= PRIV(op)->real_version)
         {
             // Original version is still invalid
-            // FIXME Oops. Successive small writes will currently break in an unexpected way. Fix it
-            dirty_db.erase(dirty_it);
-            op->retval = -EEXIST;
-            FINISH_OP(op);
-            return 1;
+            // All subsequent writes to the same object must be canceled too
+            cancel_all_writes(op, dirty_it, -EEXIST);
+            return 2;
         }
         op->version = PRIV(op)->real_version;
         PRIV(op)->real_version = 0;
@@ -174,7 +201,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
     if ((dirty_it->second.state & BS_ST_TYPE_MASK) == BS_ST_BIG_WRITE)
     {
         blockstore_journal_check_t space_check(this);
-        if (!space_check.check_available(op, unsynced_big_writes.size() + 1, sizeof(journal_entry_big_write), JOURNAL_STABILIZE_RESERVATION))
+        if (!space_check.check_available(op, unsynced_big_write_count + 1, sizeof(journal_entry_big_write), JOURNAL_STABILIZE_RESERVATION))
         {
             return 0;
         }
@@ -189,18 +216,18 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
                 PRIV(op)->wait_for = WAIT_FREE;
                 return 0;
             }
-            // FIXME Oops. Successive small writes will currently break in an unexpected way. Fix it
-            dirty_db.erase(dirty_it);
-            op->retval = -ENOSPC;
-            FINISH_OP(op);
-            return 1;
+            cancel_all_writes(op, dirty_it, -ENOSPC);
+            return 2;
         }
         write_iodepth++;
         BS_SUBMIT_GET_SQE(sqe, data);
         dirty_it->second.location = loc << block_order;
         dirty_it->second.state = (dirty_it->second.state & ~BS_ST_WORKFLOW_MASK) | BS_ST_SUBMITTED;
 #ifdef BLOCKSTORE_DEBUG
-        printf("Allocate block %lu\n", loc);
+        printf(
+            "Allocate block %lu for %lx:%lx v%lu\n",
+            loc, op->oid.inode, op->oid.stripe, op->version
+        );
 #endif
         data_alloc->set(loc, true);
         uint64_t stripe_offset = (op->offset % bitmap_granularity);
@@ -226,11 +253,8 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         PRIV(op)->min_flushed_journal_sector = PRIV(op)->max_flushed_journal_sector = 0;
         if (immediate_commit != IMMEDIATE_ALL)
         {
-            // Remember big write as unsynced
-            unsynced_big_writes.push_back((obj_ver_id){
-                .oid = op->oid,
-                .version = op->version,
-            });
+            // Increase the counter, but don't save into unsynced_writes yet (can't sync until the write is finished)
+            unsynced_big_write_count++;
             PRIV(op)->op_state = 3;
         }
         else
@@ -243,7 +267,7 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         // Small (journaled) write
         // First check if the journal has sufficient space
         blockstore_journal_check_t space_check(this);
-        if (unsynced_big_writes.size() && !space_check.check_available(op, unsynced_big_writes.size(), sizeof(journal_entry_big_write), 0)
+        if (unsynced_big_write_count && !space_check.check_available(op, unsynced_big_write_count, sizeof(journal_entry_big_write), 0)
             || !space_check.check_available(op, 1, sizeof(journal_entry_small_write), op->len + JOURNAL_STABILIZE_RESERVATION))
         {
             return 0;
@@ -335,18 +359,10 @@ int blockstore_impl_t::dequeue_write(blockstore_op_t *op)
         {
             journal.next_free = journal_block_size;
         }
-        if (immediate_commit == IMMEDIATE_NONE)
-        {
-            // Remember small write as unsynced
-            unsynced_small_writes.push_back((obj_ver_id){
-                .oid = op->oid,
-                .version = op->version,
-            });
-        }
         if (!PRIV(op)->pending_ops)
         {
             PRIV(op)->op_state = 4;
-            continue_write(op);
+            return continue_write(op);
         }
         else
         {
@@ -360,24 +376,24 @@ int blockstore_impl_t::continue_write(blockstore_op_t *op)
 {
     io_uring_sqe *sqe = NULL;
     journal_entry_big_write *je;
+    int op_state = PRIV(op)->op_state;
+    if (op_state != 2 && op_state != 4)
+    {
+        // In progress
+        return 1;
+    }
     auto dirty_it = dirty_db.find((obj_ver_id){
         .oid = op->oid,
         .version = op->version,
     });
     assert(dirty_it != dirty_db.end());
-    if (PRIV(op)->op_state == 2)
+    if (op_state == 2)
         goto resume_2;
-    else if (PRIV(op)->op_state == 4)
+    else if (op_state == 4)
         goto resume_4;
-    else
-        return 1;
 resume_2:
     // Only for the immediate_commit mode: prepare and submit big_write journal entry
-    sqe = get_sqe();
-    if (!sqe)
-    {
-        return 0;
-    }
+    BS_SUBMIT_GET_SQE_DECL(sqe);
     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)
@@ -407,7 +423,7 @@ resume_2:
 resume_4:
     // Switch object state
 #ifdef BLOCKSTORE_DEBUG
-    printf("Ack write %lx:%lx v%lu = state %x\n", op->oid.inode, op->oid.stripe, op->version, dirty_it->second.state);
+    printf("Ack write %lx:%lx v%lu = state 0x%x\n", op->oid.inode, op->oid.stripe, op->version, dirty_it->second.state);
 #endif
     bool imm = (dirty_it->second.state & BS_ST_TYPE_MASK) == BS_ST_BIG_WRITE
         ? (immediate_commit == IMMEDIATE_ALL)
@@ -421,11 +437,31 @@ resume_4:
         | (imm ? BS_ST_SYNCED : BS_ST_WRITTEN);
     if (imm && ((dirty_it->second.state & BS_ST_TYPE_MASK) == BS_ST_DELETE || (dirty_it->second.state & BS_ST_INSTANT)))
     {
-        // Deletions are treated as immediately stable
+        // Deletions and 'instant' operations are treated as immediately stable
         mark_stable(dirty_it->first);
     }
-    if (immediate_commit == IMMEDIATE_ALL)
+    if (!imm)
     {
+        if ((dirty_it->second.state & BS_ST_TYPE_MASK) == BS_ST_BIG_WRITE)
+        {
+            // Remember big write as unsynced
+            unsynced_big_writes.push_back((obj_ver_id){
+                .oid = op->oid,
+                .version = op->version,
+            });
+        }
+        else
+        {
+            // Remember small write as unsynced
+            unsynced_small_writes.push_back((obj_ver_id){
+                .oid = op->oid,
+                .version = op->version,
+            });
+        }
+    }
+    if (imm && (dirty_it->second.state & BS_ST_TYPE_MASK) == BS_ST_BIG_WRITE)
+    {
+        // Unblock small writes
         dirty_it++;
         while (dirty_it != dirty_db.end() && dirty_it->first.oid == op->oid)
         {
@@ -440,7 +476,7 @@ resume_4:
     op->retval = op->len;
     write_iodepth--;
     FINISH_OP(op);
-    return 1;
+    return 2;
 }
 
 void blockstore_impl_t::handle_write_event(ring_data_t *data, blockstore_op_t *op)
@@ -459,10 +495,7 @@ void blockstore_impl_t::handle_write_event(ring_data_t *data, blockstore_op_t *o
     {
         release_journal_sectors(op);
         PRIV(op)->op_state++;
-        if (!continue_write(op))
-        {
-            submit_queue.push_front(op);
-        }
+        ringloop->wakeup();
     }
 }
 
@@ -500,6 +533,10 @@ void blockstore_impl_t::release_journal_sectors(blockstore_op_t *op)
 
 int blockstore_impl_t::dequeue_del(blockstore_op_t *op)
 {
+    if (PRIV(op)->op_state)
+    {
+        return continue_write(op);
+    }
     auto dirty_it = dirty_db.find((obj_ver_id){
         .oid = op->oid,
         .version = op->version,
@@ -558,18 +595,10 @@ int blockstore_impl_t::dequeue_del(blockstore_op_t *op)
         PRIV(op)->min_flushed_journal_sector = PRIV(op)->max_flushed_journal_sector = 1 + journal.cur_sector;
         PRIV(op)->pending_ops++;
     }
-    else
-    {
-        // Remember delete as unsynced
-        unsynced_small_writes.push_back((obj_ver_id){
-            .oid = op->oid,
-            .version = op->version,
-        });
-    }
     if (!PRIV(op)->pending_ops)
     {
         PRIV(op)->op_state = 4;
-        continue_write(op);
+        return continue_write(op);
     }
     else
     {

src/cluster_client.h

@@ -10,7 +10,8 @@
 #define MAX_BLOCK_SIZE 128*1024*1024
 #define DEFAULT_DISK_ALIGNMENT 4096
 #define DEFAULT_BITMAP_GRANULARITY 4096
-#define DEFAULT_CLIENT_DIRTY_LIMIT 32*1024*1024
+#define DEFAULT_CLIENT_MAX_DIRTY_BYTES 32*1024*1024
+#define DEFAULT_CLIENT_MAX_DIRTY_OPS 1024
 
 struct cluster_op_t;
 
@@ -22,8 +23,7 @@ struct cluster_op_part_t
     pg_num_t pg_num;
     osd_num_t osd_num;
     osd_op_buf_list_t iov;
-    bool sent;
-    bool done;
+    unsigned flags;
     osd_op_t op;
 };
 
@@ -37,70 +37,77 @@ struct cluster_op_t
     osd_op_buf_list_t iov;
     std::function<void(cluster_op_t*)> callback;
 protected:
+    int flags = 0;
+    int state = 0;
     void *buf = NULL;
     cluster_op_t *orig_op = NULL;
-    bool is_internal = false;
     bool needs_reslice = false;
     bool up_wait = false;
-    int sent_count = 0, done_count = 0;
+    int inflight_count = 0, done_count = 0;
     std::vector<cluster_op_part_t> parts;
     friend class cluster_client_t;
 };
 
+struct cluster_buffer_t
+{
+    void *buf;
+    uint64_t len;
+    int state;
+};
+
+// FIXME: Split into public and private interfaces
 class cluster_client_t
 {
     timerfd_manager_t *tfd;
     ring_loop_t *ringloop;
 
     uint64_t bs_block_size = 0;
-    uint64_t bs_disk_alignment = 0;
     uint64_t bs_bitmap_granularity = 0;
     std::map<pool_id_t, uint64_t> pg_counts;
     bool immediate_commit = false;
     // FIXME: Implement inmemory_commit mode. Note that it requires to return overlapping reads from memory.
-    uint64_t client_dirty_limit = 0;
+    uint64_t client_max_dirty_bytes = 0;
+    uint64_t client_max_dirty_ops = 0;
     int log_level;
     int up_wait_retry_interval = 500; // ms
 
-    uint64_t op_id = 1;
-    ring_consumer_t consumer;
-    // operations currently in progress
-    std::set<cluster_op_t*> cur_ops;
     int retry_timeout_id = 0;
-    // unsynced operations are copied in memory to allow replay when cluster isn't in the immediate_commit mode
-    // unsynced_writes are replayed in any order (because only the SYNC operation guarantees ordering)
-    std::vector<cluster_op_t*> unsynced_writes;
-    std::vector<cluster_op_t*> syncing_writes;
-    cluster_op_t* cur_sync = NULL;
-    std::vector<cluster_op_t*> next_writes;
+    uint64_t op_id = 1;
     std::vector<cluster_op_t*> offline_ops;
-    uint64_t queued_bytes = 0;
+    std::deque<cluster_op_t*> op_queue;
+    std::map<object_id, cluster_buffer_t> dirty_buffers;
+    std::set<osd_num_t> dirty_osds;
+    uint64_t dirty_bytes = 0, dirty_ops = 0;
 
     bool pgs_loaded = false;
+    ring_consumer_t consumer;
     std::vector<std::function<void(void)>> on_ready_hooks;
+    int continuing_ops = 0;
 
 public:
     etcd_state_client_t st_cli;
     osd_messenger_t msgr;
+    json11::Json config;
 
     cluster_client_t(ring_loop_t *ringloop, timerfd_manager_t *tfd, json11::Json & config);
     ~cluster_client_t();
     void execute(cluster_op_t *op);
+    bool is_ready();
     void on_ready(std::function<void(void)> fn);
-    void stop();
 
-protected:
+    static void copy_write(cluster_op_t *op, std::map<object_id, cluster_buffer_t> & dirty_buffers);
     void continue_ops(bool up_retry = false);
+protected:
+    bool affects_osd(uint64_t inode, uint64_t offset, uint64_t len, osd_num_t osd);
+    void flush_buffer(const object_id & oid, cluster_buffer_t & wr);
     void on_load_config_hook(json11::Json::object & config);
     void on_load_pgs_hook(bool success);
     void on_change_hook(json11::Json::object & changes);
     void on_change_osd_state_hook(uint64_t peer_osd);
-    void continue_rw(cluster_op_t *op);
+    int continue_rw(cluster_op_t *op);
     void slice_rw(cluster_op_t *op);
-    bool try_send(cluster_op_t *op, cluster_op_part_t *part);
-    void execute_sync(cluster_op_t *op);
-    void continue_sync();
-    void finish_sync();
+    bool try_send(cluster_op_t *op, int i);
+    int continue_sync(cluster_op_t *op);
     void send_sync(cluster_op_t *op, cluster_op_part_t *part);
     void handle_op_part(cluster_op_part_t *part);
 };

src/etcd_state_client.cpp

@@ -4,9 +4,24 @@
 #include "osd_ops.h"
 #include "pg_states.h"
 #include "etcd_state_client.h"
+#ifndef __MOCK__
 #include "http_client.h"
 #include "base64.h"
+#endif
 
+etcd_state_client_t::~etcd_state_client_t()
+{
+    etcd_watches_initialised = -1;
+#ifndef __MOCK__
+    if (etcd_watch_ws)
+    {
+        etcd_watch_ws->close();
+        etcd_watch_ws = NULL;
+    }
+#endif
+}
+
+#ifndef __MOCK__
 json_kv_t etcd_state_client_t::parse_etcd_kv(const json11::Json & kv_json)
 {
     json_kv_t kv;
@@ -46,6 +61,23 @@ void etcd_state_client_t::etcd_call(std::string api, json11::Json payload, int t
     http_request_json(tfd, etcd_address, req, timeout, callback);
 }
 
+void etcd_state_client_t::add_etcd_url(std::string addr)
+{
+    if (addr.length() > 0)
+    {
+        if (strtolower(addr.substr(0, 7)) == "http://")
+            addr = addr.substr(7);
+        else if (strtolower(addr.substr(0, 8)) == "https://")
+        {
+            printf("HTTPS is unsupported for etcd. Either use plain HTTP or setup a local proxy for etcd interaction\n");
+            exit(1);
+        }
+        if (addr.find('/') < 0)
+            addr += "/v3";
+        this->etcd_addresses.push_back(addr);
+    }
+}
+
 void etcd_state_client_t::parse_config(json11::Json & config)
 {
     this->etcd_addresses.clear();
@@ -55,13 +87,7 @@ void etcd_state_client_t::parse_config(json11::Json & config)
         while (1)
         {
             int pos = ea.find(',');
-            std::string addr = pos >= 0 ? ea.substr(0, pos) : ea;
-            if (addr.length() > 0)
-            {
-                if (addr.find('/') < 0)
-                    addr += "/v3";
-                this->etcd_addresses.push_back(addr);
-            }
+            add_etcd_url(pos >= 0 ? ea.substr(0, pos) : ea);
             if (pos >= 0)
                 ea = ea.substr(pos+1);
             else
@@ -72,13 +98,7 @@ void etcd_state_client_t::parse_config(json11::Json & config)
     {
         for (auto & ea: config["etcd_address"].array_items())
         {
-            std::string addr = ea.string_value();
-            if (addr != "")
-            {
-                if (addr.find('/') < 0)
-                    addr += "/v3";
-                this->etcd_addresses.push_back(addr);
-            }
+            add_etcd_url(ea.string_value());
         }
     }
     this->etcd_prefix = config["etcd_prefix"].string_value();
@@ -160,7 +180,7 @@ void etcd_state_client_t::start_etcd_watcher()
                     start_etcd_watcher();
                 });
             }
-            else
+            else if (etcd_watches_initialised > 0)
             {
                 // Connection was live, retry immediately
                 start_etcd_watcher();
@@ -308,6 +328,26 @@ void etcd_state_client_t::load_pgs()
         start_etcd_watcher();
     });
 }
+#else
+void etcd_state_client_t::parse_config(json11::Json & config)
+{
+}
+
+void etcd_state_client_t::load_global_config()
+{
+    json11::Json::object global_config;
+    on_load_config_hook(global_config);
+}
+
+void etcd_state_client_t::load_pgs()
+{
+}
+#endif
+
+void etcd_state_client_t::parse_state(const json_kv_t & kv)
+{
+    parse_state(kv.key, kv.value);
+}
 
 void etcd_state_client_t::parse_state(const std::string & key, const json11::Json & value)
 {
@@ -321,8 +361,10 @@ void etcd_state_client_t::parse_state(const std::string & key, const json11::Jso
         {
             pool_config_t pc;
             // ID
-            pool_id_t pool_id = stoull_full(pool_item.first);
-            if (!pool_id || pool_id >= POOL_ID_MAX)
+            pool_id_t pool_id;
+            char null_byte = 0;
+            sscanf(pool_item.first.c_str(), "%u%c", &pool_id, &null_byte);
+            if (!pool_id || pool_id >= POOL_ID_MAX || null_byte != 0)
             {
                 printf("Pool ID %s is invalid (must be a number less than 0x%x), skipping pool\n", pool_item.first.c_str(), POOL_ID_MAX);
                 continue;
@@ -434,16 +476,19 @@ void etcd_state_client_t::parse_state(const std::string & key, const json11::Jso
         }
         for (auto & pool_item: value["items"].object_items())
         {
-            pool_id_t pool_id = stoull_full(pool_item.first);
-            if (!pool_id || pool_id >= POOL_ID_MAX)
+            pool_id_t pool_id;
+            char null_byte = 0;
+            sscanf(pool_item.first.c_str(), "%u%c", &pool_id, &null_byte);
+            if (!pool_id || pool_id >= POOL_ID_MAX || null_byte != 0)
             {
                 printf("Pool ID %s is invalid in PG configuration (must be a number less than 0x%x), skipping pool\n", pool_item.first.c_str(), POOL_ID_MAX);
                 continue;
             }
             for (auto & pg_item: pool_item.second.object_items())
             {
-                pg_num_t pg_num = stoull_full(pg_item.first);
-                if (!pg_num)
+                pg_num_t pg_num = 0;
+                sscanf(pg_item.first.c_str(), "%u%c", &pg_num, &null_byte);
+                if (!pg_num || null_byte != 0)
                 {
                     printf("Bad key in pool %u PG configuration: %s (must be a number), skipped\n", pool_id, pg_item.first.c_str());
                     continue;

src/etcd_state_client.h

@@ -3,8 +3,8 @@
 
 #pragma once
 
+#include "json11/json11.hpp"
 #include "osd_id.h"
-#include "http_client.h"
 #include "timerfd_manager.h"
 
 #define ETCD_CONFIG_WATCH_ID 1
@@ -52,8 +52,15 @@ struct pool_config_t
     std::map<pg_num_t, pg_config_t> pg_config;
 };
 
+struct websocket_t;
+
 struct etcd_state_client_t
 {
+protected:
+    websocket_t *etcd_watch_ws = NULL;
+    uint64_t bs_block_size = DEFAULT_BLOCK_SIZE;
+    void add_etcd_url(std::string);
+public:
     std::vector<std::string> etcd_addresses;
     std::string etcd_prefix;
     int log_level = 0;
@@ -61,8 +68,6 @@ struct etcd_state_client_t
 
     int etcd_watches_initialised = 0;
     uint64_t etcd_watch_revision = 0;
-    websocket_t *etcd_watch_ws = NULL;
-    uint64_t bs_block_size = 0;
     std::map<pool_id_t, pool_config_t> pool_config;
     std::map<osd_num_t, json11::Json> peer_states;
 
@@ -79,6 +84,8 @@ struct etcd_state_client_t
     void start_etcd_watcher();
     void load_global_config();
     void load_pgs();
+    void parse_state(const json_kv_t & kv);
     void parse_state(const std::string & key, const json11::Json & value);
     void parse_config(json11::Json & config);
+    ~etcd_state_client_t();
 };

src/fio_cluster.cpp

@@ -117,8 +117,15 @@ static struct fio_option options[] = {
 
 static int sec_setup(struct thread_data *td)
 {
+    sec_options *o = (sec_options*)td->eo;
     sec_data *bsd;
 
+    if (!o->etcd_host)
+    {
+        td_verror(td, EINVAL, "etcd address is missing");
+        return 1;
+    }
+
     bsd = new sec_data;
     if (!bsd)
     {

src/http_client.cpp

@@ -22,7 +22,6 @@
 #define READ_BUFFER_SIZE 9000
 
 static int extract_port(std::string & host);
-static std::string strtolower(const std::string & in);
 static std::string trim(const std::string & in);
 static std::string ws_format_frame(int type, uint64_t size);
 static bool ws_parse_frame(std::string & buf, int & type, std::string & res);
@@ -673,7 +672,7 @@ static int extract_port(std::string & host)
     return port;
 }
 
-static std::string strtolower(const std::string & in)
+std::string strtolower(const std::string & in)
 {
     std::string s = in;
     for (int i = 0; i < s.length(); i++)

src/http_client.h

@@ -49,6 +49,8 @@ std::vector<std::string> getifaddr_list(bool include_v6 = false);
 
 uint64_t stoull_full(const std::string & str, int base = 10);
 
+std::string strtolower(const std::string & in);
+
 void http_request(timerfd_manager_t *tfd, const std::string & host, const std::string & request,
     const http_options_t & options, std::function<void(const http_response_t *response)> callback);
 

src/messenger.cpp

@@ -10,30 +10,119 @@
 
 #include "messenger.h"
 
-osd_op_t::~osd_op_t()
+void osd_messenger_t::init()
 {
-    assert(!bs_op);
-    assert(!op_data);
-    if (rmw_buf)
+    keepalive_timer_id = tfd->set_timer(1000, true, [this](int)
     {
-        free(rmw_buf);
-    }
-    if (buf)
-    {
-        // Note: reusing osd_op_t WILL currently lead to memory leaks
-        // So we don't reuse it, but free it every time
-        free(buf);
-    }
+        std::vector<int> to_stop;
+        std::vector<osd_op_t*> to_ping;
+        for (auto cl_it = clients.begin(); cl_it != clients.end(); cl_it++)
+        {
+            auto cl = cl_it->second;
+            if (!cl->osd_num || cl->peer_state != PEER_CONNECTED)
+            {
+                // Do not run keepalive on regular clients
+                continue;
+            }
+            if (cl->ping_time_remaining > 0)
+            {
+                cl->ping_time_remaining--;
+                if (!cl->ping_time_remaining)
+                {
+                    // Ping timed out, stop the client
+                    printf("Ping timed out for OSD %lu (client %d), disconnecting peer\n", cl->osd_num, cl->peer_fd);
+                    to_stop.push_back(cl->peer_fd);
+                }
+            }
+            else if (cl->idle_time_remaining > 0)
+            {
+                cl->idle_time_remaining--;
+                if (!cl->idle_time_remaining)
+                {
+                    // Connection is idle for <osd_idle_time>, send ping
+                    osd_op_t *op = new osd_op_t();
+                    op->op_type = OSD_OP_OUT;
+                    op->peer_fd = cl->peer_fd;
+                    op->req = (osd_any_op_t){
+                        .hdr = {
+                            .magic = SECONDARY_OSD_OP_MAGIC,
+                            .id = this->next_subop_id++,
+                            .opcode = OSD_OP_PING,
+                        },
+                    };
+                    op->callback = [this, cl](osd_op_t *op)
+                    {
+                        int fail_fd = (op->reply.hdr.retval != 0 ? op->peer_fd : -1);
+                        cl->ping_time_remaining = 0;
+                        delete op;
+                        if (fail_fd >= 0)
+                        {
+                            printf("Ping failed for OSD %lu (client %d), disconnecting peer\n", cl->osd_num, cl->peer_fd);
+                            stop_client(fail_fd, true);
+                        }
+                    };
+                    to_ping.push_back(op);
+                    cl->ping_time_remaining = osd_ping_timeout;
+                    cl->idle_time_remaining = osd_idle_timeout;
+                }
+            }
+            else
+            {
+                cl->idle_time_remaining = osd_idle_timeout;
+            }
+        }
+        // Don't stop clients while a 'clients' iterator is still active
+        for (int peer_fd: to_stop)
+        {
+            stop_client(peer_fd, true);
+        }
+        for (auto op: to_ping)
+        {
+            outbox_push(op);
+        }
+    });
 }
 
 osd_messenger_t::~osd_messenger_t()
 {
+    if (keepalive_timer_id >= 0)
+    {
+        tfd->clear_timer(keepalive_timer_id);
+        keepalive_timer_id = -1;
+    }
     while (clients.size() > 0)
     {
         stop_client(clients.begin()->first, true);
     }
 }
 
+void osd_messenger_t::parse_config(const json11::Json & config)
+{
+    this->use_sync_send_recv = config["use_sync_send_recv"].bool_value() ||
+        config["use_sync_send_recv"].uint64_value();
+    this->peer_connect_interval = config["peer_connect_interval"].uint64_value();
+    if (!this->peer_connect_interval)
+    {
+        this->peer_connect_interval = DEFAULT_PEER_CONNECT_INTERVAL;
+    }
+    this->peer_connect_timeout = config["peer_connect_timeout"].uint64_value();
+    if (!this->peer_connect_timeout)
+    {
+        this->peer_connect_timeout = DEFAULT_PEER_CONNECT_TIMEOUT;
+    }
+    this->osd_idle_timeout = config["osd_idle_timeout"].uint64_value();
+    if (!this->osd_idle_timeout)
+    {
+        this->osd_idle_timeout = DEFAULT_OSD_PING_TIMEOUT;
+    }
+    this->osd_ping_timeout = config["osd_ping_timeout"].uint64_value();
+    if (!this->osd_ping_timeout)
+    {
+        this->osd_ping_timeout = DEFAULT_OSD_PING_TIMEOUT;
+    }
+    this->log_level = config["log_level"].uint64_value();
+}
+
 void osd_messenger_t::connect_peer(uint64_t peer_osd, json11::Json peer_state)
 {
     if (wanted_peers.find(peer_osd) == wanted_peers.end())
@@ -49,17 +138,14 @@ void osd_messenger_t::connect_peer(uint64_t peer_osd, json11::Json peer_state)
         wanted_peers[peer_osd].port = (int)peer_state["port"].int64_value();
     }
     wanted_peers[peer_osd].address_changed = true;
-    if (!wanted_peers[peer_osd].connecting &&
-        (time(NULL) - wanted_peers[peer_osd].last_connect_attempt) >= peer_connect_interval)
-    {
-        try_connect_peer(peer_osd);
-    }
+    try_connect_peer(peer_osd);
 }
 
 void osd_messenger_t::try_connect_peer(uint64_t peer_osd)
 {
     auto wp_it = wanted_peers.find(peer_osd);
-    if (wp_it == wanted_peers.end())
+    if (wp_it == wanted_peers.end() || wp_it->second.connecting ||
+        (time(NULL) - wp_it->second.last_connect_attempt) < peer_connect_interval)
     {
         return;
     }
@@ -105,10 +191,22 @@ void osd_messenger_t::try_connect_peer_addr(osd_num_t peer_osd, const char *peer
         on_connect_peer(peer_osd, -errno);
         return;
     }
-    int timeout_id = -1;
+    clients[peer_fd] = new osd_client_t();
+    clients[peer_fd]->peer_addr = addr;
+    clients[peer_fd]->peer_port = peer_port;
+    clients[peer_fd]->peer_fd = peer_fd;
+    clients[peer_fd]->peer_state = PEER_CONNECTING;
+    clients[peer_fd]->connect_timeout_id = -1;
+    clients[peer_fd]->osd_num = peer_osd;
+    clients[peer_fd]->in_buf = malloc_or_die(receive_buffer_size);
+    tfd->set_fd_handler(peer_fd, true, [this](int peer_fd, int epoll_events)
+    {
+        // Either OUT (connected) or HUP
+        handle_connect_epoll(peer_fd);
+    });
     if (peer_connect_timeout > 0)
     {
-        timeout_id = tfd->set_timer(1000*peer_connect_timeout, false, [this, peer_fd](int timer_id)
+        clients[peer_fd]->connect_timeout_id = tfd->set_timer(1000*peer_connect_timeout, false, [this, peer_fd](int timer_id)
         {
             osd_num_t peer_osd = clients.at(peer_fd)->osd_num;
             stop_client(peer_fd, true);
@@ -116,20 +214,6 @@ void osd_messenger_t::try_connect_peer_addr(osd_num_t peer_osd, const char *peer
             return;
         });
     }
-    clients[peer_fd] = new osd_client_t((osd_client_t){
-        .peer_addr = addr,
-        .peer_port = peer_port,
-        .peer_fd = peer_fd,
-        .peer_state = PEER_CONNECTING,
-        .connect_timeout_id = timeout_id,
-        .osd_num = peer_osd,
-        .in_buf = malloc_or_die(receive_buffer_size),
-    });
-    tfd->set_fd_handler(peer_fd, true, [this](int peer_fd, int epoll_events)
-    {
-        // Either OUT (connected) or HUP
-        handle_connect_epoll(peer_fd);
-    });
 }
 
 void osd_messenger_t::handle_connect_epoll(int peer_fd)
@@ -281,123 +365,6 @@ void osd_messenger_t::check_peer_config(osd_client_t *cl)
     outbox_push(op);
 }
 
-void osd_messenger_t::cancel_osd_ops(osd_client_t *cl)
-{
-    for (auto p: cl->sent_ops)
-    {
-        cancel_op(p.second);
-    }
-    cl->sent_ops.clear();
-    cl->outbox.clear();
-}
-
-void osd_messenger_t::cancel_op(osd_op_t *op)
-{
-    if (op->op_type == OSD_OP_OUT)
-    {
-        op->reply.hdr.magic = SECONDARY_OSD_REPLY_MAGIC;
-        op->reply.hdr.id = op->req.hdr.id;
-        op->reply.hdr.opcode = op->req.hdr.opcode;
-        op->reply.hdr.retval = -EPIPE;
-        // Copy lambda to be unaffected by `delete op`
-        std::function<void(osd_op_t*)>(op->callback)(op);
-    }
-    else
-    {
-        // This function is only called in stop_client(), so it's fine to destroy the operation
-        delete op;
-    }
-}
-
-void osd_messenger_t::stop_client(int peer_fd, bool force)
-{
-    assert(peer_fd != 0);
-    auto it = clients.find(peer_fd);
-    if (it == clients.end())
-    {
-        return;
-    }
-    uint64_t repeer_osd = 0;
-    osd_client_t *cl = it->second;
-    if (cl->peer_state == PEER_CONNECTED)
-    {
-        if (cl->osd_num)
-        {
-            // Reload configuration from etcd when the connection is dropped
-            if (log_level > 0)
-                printf("[OSD %lu] Stopping client %d (OSD peer %lu)\n", osd_num, peer_fd, cl->osd_num);
-            repeer_osd = cl->osd_num;
-        }
-        else
-        {
-            if (log_level > 0)
-                printf("[OSD %lu] Stopping client %d (regular client)\n", osd_num, peer_fd);
-        }
-    }
-    else if (!force)
-    {
-        return;
-    }
-    cl->peer_state = PEER_STOPPED;
-    clients.erase(it);
-    tfd->set_fd_handler(peer_fd, false, NULL);
-    if (cl->connect_timeout_id >= 0)
-    {
-        tfd->clear_timer(cl->connect_timeout_id);
-        cl->connect_timeout_id = -1;
-    }
-    if (cl->osd_num)
-    {
-        osd_peer_fds.erase(cl->osd_num);
-    }
-    if (cl->read_op)
-    {
-        if (cl->read_op->callback)
-        {
-            cancel_op(cl->read_op);
-        }
-        else
-        {
-            delete cl->read_op;
-        }
-        cl->read_op = NULL;
-    }
-    for (auto rit = read_ready_clients.begin(); rit != read_ready_clients.end(); rit++)
-    {
-        if (*rit == peer_fd)
-        {
-            read_ready_clients.erase(rit);
-            break;
-        }
-    }
-    for (auto wit = write_ready_clients.begin(); wit != write_ready_clients.end(); wit++)
-    {
-        if (*wit == peer_fd)
-        {
-            write_ready_clients.erase(wit);
-            break;
-        }
-    }
-    free(cl->in_buf);
-    cl->in_buf = NULL;
-    close(peer_fd);
-    if (repeer_osd)
-    {
-        // First repeer PGs as canceling OSD ops may push new operations
-        // and we need correct PG states when we do that
-        repeer_pgs(repeer_osd);
-    }
-    if (cl->osd_num)
-    {
-        // Cancel outbound operations
-        cancel_osd_ops(cl);
-    }
-    if (cl->refs <= 0)
-    {
-        delete cl;
-    }
-}
-
 void osd_messenger_t::accept_connections(int listen_fd)
 {
     // Accept new connections
@@ -413,13 +380,12 @@ void osd_messenger_t::accept_connections(int listen_fd)
         fcntl(peer_fd, F_SETFL, fcntl(peer_fd, F_GETFL, 0) | O_NONBLOCK);
         int one = 1;
         setsockopt(peer_fd, SOL_TCP, TCP_NODELAY, &one, sizeof(one));
-        clients[peer_fd] = new osd_client_t((osd_client_t){
-            .peer_addr = addr,
-            .peer_port = ntohs(addr.sin_port),
-            .peer_fd = peer_fd,
-            .peer_state = PEER_CONNECTED,
-            .in_buf = malloc_or_die(receive_buffer_size),
-        });
+        clients[peer_fd] = new osd_client_t();
+        clients[peer_fd]->peer_addr = addr;
+        clients[peer_fd]->peer_port = ntohs(addr.sin_port);
+        clients[peer_fd]->peer_fd = peer_fd;
+        clients[peer_fd]->peer_state = PEER_CONNECTED;
+        clients[peer_fd]->in_buf = malloc_or_die(receive_buffer_size);
         // Add FD to epoll
         tfd->set_fd_handler(peer_fd, false, [this](int peer_fd, int epoll_events)
         {

src/messenger.h

@@ -14,19 +14,15 @@
 
 #include "malloc_or_die.h"
 #include "json11/json11.hpp"
-#include "osd_ops.h"
+#include "msgr_op.h"
 #include "timerfd_manager.h"
-#include "ringloop.h"
-
-#define OSD_OP_IN 0
-#define OSD_OP_OUT 1
+#include <ringloop.h>
 
 #define CL_READ_HDR 1
 #define CL_READ_DATA 2
 #define CL_READ_REPLY_DATA 3
 #define CL_WRITE_READY 1
 #define CL_WRITE_REPLY 2
-#define OSD_OP_INLINE_BUF_COUNT 16
 
 #define PEER_CONNECTING 1
 #define PEER_CONNECTED 2
@@ -34,160 +30,7 @@
 
 #define DEFAULT_PEER_CONNECT_INTERVAL 5
 #define DEFAULT_PEER_CONNECT_TIMEOUT 5
-
-// Kind of a vector with small-list-optimisation
-struct osd_op_buf_list_t
-{
-    int count = 0, alloc = OSD_OP_INLINE_BUF_COUNT, done = 0;
-    iovec *buf = NULL;
-    iovec inline_buf[OSD_OP_INLINE_BUF_COUNT];
-
-    inline osd_op_buf_list_t()
-    {
-        buf = inline_buf;
-    }
-
-    inline osd_op_buf_list_t(const osd_op_buf_list_t & other)
-    {
-        buf = inline_buf;
-        append(other);
-    }
-
-    inline osd_op_buf_list_t & operator = (const osd_op_buf_list_t & other)
-    {
-        reset();
-        append(other);
-        return *this;
-    }
-
-    inline ~osd_op_buf_list_t()
-    {
-        if (buf && buf != inline_buf)
-        {
-            free(buf);
-        }
-    }
-
-    inline void reset()
-    {
-        count = 0;
-        done = 0;
-    }
-
-    inline iovec* get_iovec()
-    {
-        return buf + done;
-    }
-
-    inline int get_size()
-    {
-        return count - done;
-    }
-
-    inline void append(const osd_op_buf_list_t & other)
-    {
-        if (count+other.count > alloc)
-        {
-            if (buf == inline_buf)
-            {
-                int old = alloc;
-                alloc = (((count+other.count+15)/16)*16);
-                buf = (iovec*)malloc(sizeof(iovec) * alloc);
-                if (!buf)
-                {
-                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
-                    exit(1);
-                }
-                memcpy(buf, inline_buf, sizeof(iovec) * old);
-            }
-            else
-            {
-                alloc = (((count+other.count+15)/16)*16);
-                buf = (iovec*)realloc(buf, sizeof(iovec) * alloc);
-                if (!buf)
-                {
-                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
-                    exit(1);
-                }
-            }
-        }
-        for (int i = 0; i < other.count; i++)
-        {
-            buf[count++] = other.buf[i];
-        }
-    }
-
-    inline void push_back(void *nbuf, size_t len)
-    {
-        if (count >= alloc)
-        {
-            if (buf == inline_buf)
-            {
-                int old = alloc;
-                alloc = ((alloc/16)*16 + 1);
-                buf = (iovec*)malloc(sizeof(iovec) * alloc);
-                if (!buf)
-                {
-                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
-                    exit(1);
-                }
-                memcpy(buf, inline_buf, sizeof(iovec)*old);
-            }
-            else
-            {
-                alloc = alloc < 16 ? 16 : (alloc+16);
-                buf = (iovec*)realloc(buf, sizeof(iovec) * alloc);
-                if (!buf)
-                {
-                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
-                    exit(1);
-                }
-            }
-        }
-        buf[count++] = { .iov_base = nbuf, .iov_len = len };
-    }
-
-    inline void eat(int result)
-    {
-        while (result > 0 && done < count)
-        {
-            iovec & iov = buf[done];
-            if (iov.iov_len <= result)
-            {
-                result -= iov.iov_len;
-                done++;
-            }
-            else
-            {
-                iov.iov_len -= result;
-                iov.iov_base += result;
-                break;
-            }
-        }
-    }
-};
-
-struct blockstore_op_t;
-
-struct osd_primary_op_data_t;
-
-struct osd_op_t
-{
-    timespec tv_begin;
-    uint64_t op_type = OSD_OP_IN;
-    int peer_fd;
-    osd_any_op_t req;
-    osd_any_reply_t reply;
-    blockstore_op_t *bs_op = NULL;
-    void *buf = NULL;
-    void *rmw_buf = NULL;
-    osd_primary_op_data_t* op_data = NULL;
-    std::function<void(osd_op_t*)> callback;
-
-    osd_op_buf_list_t iov;
-
-    ~osd_op_t();
-};
+#define DEFAULT_OSD_PING_TIMEOUT 5
 
 struct osd_client_t
 {
@@ -198,6 +41,8 @@ struct osd_client_t
     int peer_fd;
     int peer_state;
     int connect_timeout_id = -1;
+    int ping_time_remaining = 0;
+    int idle_time_remaining = 0;
     osd_num_t osd_num = 0;
 
     void *in_buf = NULL;
@@ -225,6 +70,12 @@ struct osd_client_t
     int write_state = 0;
     std::vector<iovec> send_list, next_send_list;
     std::vector<osd_op_t*> outbox, next_outbox;
+
+    ~osd_client_t()
+    {
+        free(in_buf);
+        in_buf = NULL;
+    }
 };
 
 struct osd_wanted_peer_t
@@ -249,37 +100,41 @@ struct osd_op_stats_t
 
 struct osd_messenger_t
 {
-    timerfd_manager_t *tfd;
-    ring_loop_t *ringloop;
+protected:
+    int keepalive_timer_id = -1;
 
-    // osd_num_t is only for logging and asserts
-    osd_num_t osd_num;
     // FIXME: make receive_buffer_size configurable
     int receive_buffer_size = 64*1024;
     int peer_connect_interval = DEFAULT_PEER_CONNECT_INTERVAL;
     int peer_connect_timeout = DEFAULT_PEER_CONNECT_TIMEOUT;
+    int osd_idle_timeout = DEFAULT_OSD_PING_TIMEOUT;
+    int osd_ping_timeout = DEFAULT_OSD_PING_TIMEOUT;
     int log_level = 0;
     bool use_sync_send_recv = false;
 
-    std::map<osd_num_t, osd_wanted_peer_t> wanted_peers;
-    std::map<uint64_t, int> osd_peer_fds;
-    uint64_t next_subop_id = 1;
-
-    std::map<int, osd_client_t*> clients;
     std::vector<int> read_ready_clients;
     std::vector<int> write_ready_clients;
     std::vector<std::function<void()>> set_immediate;
 
+public:
+    timerfd_manager_t *tfd;
+    ring_loop_t *ringloop;
+    // osd_num_t is only for logging and asserts
+    osd_num_t osd_num;
+    uint64_t next_subop_id = 1;
+    std::map<int, osd_client_t*> clients;
+    std::map<osd_num_t, osd_wanted_peer_t> wanted_peers;
+    std::map<uint64_t, int> osd_peer_fds;
     // op statistics
     osd_op_stats_t stats;
 
-public:
+    void init();
+    void parse_config(const json11::Json & config);
     void connect_peer(uint64_t osd_num, json11::Json peer_state);
     void stop_client(int peer_fd, bool force = false);
     void outbox_push(osd_op_t *cur_op);
     std::function<void(osd_op_t*)> exec_op;
     std::function<void(osd_num_t)> repeer_pgs;
-    void handle_peer_epoll(int peer_fd, int epoll_events);
     void read_requests();
     void send_replies();
     void accept_connections(int listen_fd);
@@ -288,6 +143,7 @@ public:
 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);
+    void handle_peer_epoll(int peer_fd, int epoll_events);
     void handle_connect_epoll(int peer_fd);
     void on_connect_peer(osd_num_t peer_osd, int peer_fd);
     void check_peer_config(osd_client_t *cl);

src/mock/build.sh

@@ -0,0 +1 @@
+g++ -D__MOCK__ -fsanitize=address -g -Wno-pointer-arith pg_states.cpp osd_ops.cpp test_cluster_client.cpp cluster_client.cpp msgr_op.cpp msgr_stop.cpp mock/messenger.cpp etcd_state_client.cpp timerfd_manager.cpp ../json11/json11.cpp -I mock -I . -I ..; ./a.out

src/mock/messenger.cpp

@@ -0,0 +1,44 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 or GNU GPL-2.0+ (see README.md for details)
+
+#include <unistd.h>
+#include <stdexcept>
+#include <assert.h>
+
+#include "messenger.h"
+
+void osd_messenger_t::init()
+{
+}
+
+osd_messenger_t::~osd_messenger_t()
+{
+    while (clients.size() > 0)
+    {
+        stop_client(clients.begin()->first, true);
+    }
+}
+
+void osd_messenger_t::outbox_push(osd_op_t *cur_op)
+{
+    clients[cur_op->peer_fd]->sent_ops[cur_op->req.hdr.id] = cur_op;
+}
+
+void osd_messenger_t::parse_config(const json11::Json & config)
+{
+}
+
+void osd_messenger_t::connect_peer(uint64_t peer_osd, json11::Json peer_state)
+{
+    wanted_peers[peer_osd] = (osd_wanted_peer_t){
+        .port = 1,
+    };
+}
+
+void osd_messenger_t::read_requests()
+{
+}
+
+void osd_messenger_t::send_replies()
+{
+}

src/mock/ringloop.h

@@ -0,0 +1,25 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 or GNU GPL-2.0+ (see README.md for details)
+
+#pragma once
+
+#include <functional>
+
+struct ring_consumer_t
+{
+    std::function<void(void)> loop;
+};
+
+class ring_loop_t
+{
+public:
+    void register_consumer(ring_consumer_t *consumer)
+    {
+    }
+    void unregister_consumer(ring_consumer_t *consumer)
+    {
+    }
+    void submit()
+    {
+    }
+};

src/msgr_op.cpp

@@ -0,0 +1,22 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 or GNU GPL-2.0+ (see README.md for details)
+
+#include <assert.h>
+
+#include "msgr_op.h"
+
+osd_op_t::~osd_op_t()
+{
+    assert(!bs_op);
+    assert(!op_data);
+    if (rmw_buf)
+    {
+        free(rmw_buf);
+    }
+    if (buf)
+    {
+        // Note: reusing osd_op_t WILL currently lead to memory leaks
+        // So we don't reuse it, but free it every time
+        free(buf);
+    }
+}

src/msgr_op.h

@@ -0,0 +1,171 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 or GNU GPL-2.0+ (see README.md for details)
+
+#pragma once
+
+#include <sys/uio.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdlib.h>
+
+#include "osd_ops.h"
+
+#define OSD_OP_IN 0
+#define OSD_OP_OUT 1
+
+#define OSD_OP_INLINE_BUF_COUNT 16
+
+// Kind of a vector with small-list-optimisation
+struct osd_op_buf_list_t
+{
+    int count = 0, alloc = OSD_OP_INLINE_BUF_COUNT, done = 0;
+    iovec *buf = NULL;
+    iovec inline_buf[OSD_OP_INLINE_BUF_COUNT];
+
+    inline osd_op_buf_list_t()
+    {
+        buf = inline_buf;
+    }
+
+    inline osd_op_buf_list_t(const osd_op_buf_list_t & other)
+    {
+        buf = inline_buf;
+        append(other);
+    }
+
+    inline osd_op_buf_list_t & operator = (const osd_op_buf_list_t & other)
+    {
+        reset();
+        append(other);
+        return *this;
+    }
+
+    inline ~osd_op_buf_list_t()
+    {
+        if (buf && buf != inline_buf)
+        {
+            free(buf);
+        }
+    }
+
+    inline void reset()
+    {
+        count = 0;
+        done = 0;
+    }
+
+    inline iovec* get_iovec()
+    {
+        return buf + done;
+    }
+
+    inline int get_size()
+    {
+        return count - done;
+    }
+
+    inline void append(const osd_op_buf_list_t & other)
+    {
+        if (count+other.count > alloc)
+        {
+            if (buf == inline_buf)
+            {
+                int old = alloc;
+                alloc = (((count+other.count+15)/16)*16);
+                buf = (iovec*)malloc(sizeof(iovec) * alloc);
+                if (!buf)
+                {
+                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    exit(1);
+                }
+                memcpy(buf, inline_buf, sizeof(iovec) * old);
+            }
+            else
+            {
+                alloc = (((count+other.count+15)/16)*16);
+                buf = (iovec*)realloc(buf, sizeof(iovec) * alloc);
+                if (!buf)
+                {
+                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    exit(1);
+                }
+            }
+        }
+        for (int i = 0; i < other.count; i++)
+        {
+            buf[count++] = other.buf[i];
+        }
+    }
+
+    inline void push_back(void *nbuf, size_t len)
+    {
+        if (count >= alloc)
+        {
+            if (buf == inline_buf)
+            {
+                int old = alloc;
+                alloc = ((alloc/16)*16 + 1);
+                buf = (iovec*)malloc(sizeof(iovec) * alloc);
+                if (!buf)
+                {
+                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    exit(1);
+                }
+                memcpy(buf, inline_buf, sizeof(iovec)*old);
+            }
+            else
+            {
+                alloc = alloc < 16 ? 16 : (alloc+16);
+                buf = (iovec*)realloc(buf, sizeof(iovec) * alloc);
+                if (!buf)
+                {
+                    printf("Failed to allocate %lu bytes\n", sizeof(iovec) * alloc);
+                    exit(1);
+                }
+            }
+        }
+        buf[count++] = { .iov_base = nbuf, .iov_len = len };
+    }
+
+    inline void eat(int result)
+    {
+        while (result > 0 && done < count)
+        {
+            iovec & iov = buf[done];
+            if (iov.iov_len <= result)
+            {
+                result -= iov.iov_len;
+                done++;
+            }
+            else
+            {
+                iov.iov_len -= result;
+                iov.iov_base += result;
+                break;
+            }
+        }
+    }
+};
+
+struct blockstore_op_t;
+
+struct osd_primary_op_data_t;
+
+struct osd_op_t
+{
+    timespec tv_begin;
+    uint64_t op_type = OSD_OP_IN;
+    int peer_fd;
+    osd_any_op_t req;
+    osd_any_reply_t reply;
+    blockstore_op_t *bs_op = NULL;
+    void *buf = NULL;
+    void *rmw_buf = NULL;
+    osd_primary_op_data_t* op_data = NULL;
+    std::function<void(osd_op_t*)> callback;
+
+    osd_op_buf_list_t iov;
+
+    ~osd_op_t();
+};

src/msgr_send.cpp

@@ -180,7 +180,7 @@ void osd_messenger_t::handle_send(int result, osd_client_t *cl)
     cl->refs--;
     if (cl->peer_state == PEER_STOPPED)
     {
-        if (!cl->refs)
+        if (cl->refs <= 0)
         {
             delete cl;
         }

src/msgr_stop.cpp

@@ -0,0 +1,137 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 or GNU GPL-2.0+ (see README.md for details)
+
+#include <unistd.h>
+#include <assert.h>
+
+#include "messenger.h"
+
+void osd_messenger_t::cancel_osd_ops(osd_client_t *cl)
+{
+    std::vector<osd_op_t*> cancel_ops;
+    cancel_ops.resize(cl->sent_ops.size());
+    int i = 0;
+    for (auto p: cl->sent_ops)
+    {
+        cancel_ops[i++] = p.second;
+    }
+    cl->sent_ops.clear();
+    cl->outbox.clear();
+    for (auto op: cancel_ops)
+    {
+        cancel_op(op);
+    }
+}
+
+void osd_messenger_t::cancel_op(osd_op_t *op)
+{
+    if (op->op_type == OSD_OP_OUT)
+    {
+        op->reply.hdr.magic = SECONDARY_OSD_REPLY_MAGIC;
+        op->reply.hdr.id = op->req.hdr.id;
+        op->reply.hdr.opcode = op->req.hdr.opcode;
+        op->reply.hdr.retval = -EPIPE;
+        // Copy lambda to be unaffected by `delete op`
+        std::function<void(osd_op_t*)>(op->callback)(op);
+    }
+    else
+    {
+        // This function is only called in stop_client(), so it's fine to destroy the operation
+        delete op;
+    }
+}
+
+void osd_messenger_t::stop_client(int peer_fd, bool force)
+{
+    assert(peer_fd != 0);
+    auto it = clients.find(peer_fd);
+    if (it == clients.end())
+    {
+        return;
+    }
+    osd_client_t *cl = it->second;
+    if (cl->peer_state == PEER_CONNECTING && !force || cl->peer_state == PEER_STOPPED)
+    {
+        return;
+    }
+    if (log_level > 0)
+    {
+        if (cl->osd_num)
+        {
+            printf("[OSD %lu] Stopping client %d (OSD peer %lu)\n", osd_num, peer_fd, cl->osd_num);
+        }
+        else
+        {
+            printf("[OSD %lu] Stopping client %d (regular client)\n", osd_num, peer_fd);
+        }
+    }
+    // First set state to STOPPED so another stop_client() call doesn't try to free it again
+    cl->refs++;
+    cl->peer_state = PEER_STOPPED;
+    if (cl->osd_num)
+    {
+        // ...and forget OSD peer
+        osd_peer_fds.erase(cl->osd_num);
+    }
+#ifndef __MOCK__
+    // Then remove FD from the eventloop so we don't accidentally read something
+    tfd->set_fd_handler(peer_fd, false, NULL);
+    if (cl->connect_timeout_id >= 0)
+    {
+        tfd->clear_timer(cl->connect_timeout_id);
+        cl->connect_timeout_id = -1;
+    }
+    for (auto rit = read_ready_clients.begin(); rit != read_ready_clients.end(); rit++)
+    {
+        if (*rit == peer_fd)
+        {
+            read_ready_clients.erase(rit);
+            break;
+        }
+    }
+    for (auto wit = write_ready_clients.begin(); wit != write_ready_clients.end(); wit++)
+    {
+        if (*wit == peer_fd)
+        {
+            write_ready_clients.erase(wit);
+            break;
+        }
+    }
+#endif
+    if (cl->osd_num)
+    {
+        // Then repeer PGs because cancel_op() callbacks can try to perform
+        // some actions and we need correct PG states to not do something silly
+        repeer_pgs(cl->osd_num);
+    }
+    // Then cancel all operations
+    if (cl->read_op)
+    {
+        if (!cl->read_op->callback)
+        {
+            delete cl->read_op;
+        }
+        cl->read_op = NULL;
+    }
+    if (cl->osd_num)
+    {
+        // Cancel outbound operations
+        cancel_osd_ops(cl);
+    }
+#ifndef __MOCK__
+    // And close the FD only when everything is done
+    // ...because peer_fd number can get reused after close()
+    close(peer_fd);
+#endif
+    // Find the item again because it can be invalidated at this point
+    it = clients.find(peer_fd);
+    if (it != clients.end())
+    {
+        clients.erase(it);
+    }
+    cl->refs--;
+    if (cl->refs <= 0)
+    {
+        delete cl;
+    }
+}

src/osd.cpp

@@ -8,16 +8,20 @@
 #include <arpa/inet.h>
 
 #include "osd.h"
+#include "http_client.h"
 
-osd_t::osd_t(blockstore_config_t & config, blockstore_t *bs, ring_loop_t *ringloop)
+osd_t::osd_t(blockstore_config_t & config, ring_loop_t *ringloop)
 {
+    config["entry_attr_size"] = "0";
+
     this->config = config;
-    this->bs = bs;
     this->ringloop = ringloop;
 
+    // FIXME: Create Blockstore from on-disk superblock config and check it against the OSD cluster config
+    this->bs = new blockstore_t(config, ringloop);
+
     this->bs_block_size = bs->get_block_size();
-    // FIXME: use bitmap granularity instead
-    this->bs_disk_alignment = bs->get_disk_alignment();
+    this->bs_bitmap_granularity = bs->get_bitmap_granularity();
 
     parse_config(config);
 
@@ -37,6 +41,7 @@ osd_t::osd_t(blockstore_config_t & config, blockstore_t *bs, ring_loop_t *ringlo
     c_cli.ringloop = this->ringloop;
     c_cli.exec_op = [this](osd_op_t *op) { exec_op(op); };
     c_cli.repeer_pgs = [this](osd_num_t peer_osd) { repeer_pgs(peer_osd); };
+    c_cli.init();
 
     init_cluster();
 
@@ -48,6 +53,7 @@ osd_t::~osd_t()
 {
     ringloop->unregister_consumer(&consumer);
     delete epmgr;
+    delete bs;
     close(listen_fd);
 }
 
@@ -55,6 +61,7 @@ void osd_t::parse_config(blockstore_config_t & config)
 {
     if (config.find("log_level") == config.end())
         config["log_level"] = "1";
+    log_level = strtoull(config["log_level"].c_str(), NULL, 10);
     // Initial startup configuration
     json11::Json json_config = json11::Json(config);
     st_cli.parse_config(json_config);
@@ -66,6 +73,8 @@ void osd_t::parse_config(blockstore_config_t & config)
         throw std::runtime_error("osd_num is required in the configuration");
     c_cli.osd_num = osd_num;
     run_primary = config["run_primary"] != "false" && config["run_primary"] != "0" && config["run_primary"] != "no";
+    no_rebalance = config["no_rebalance"] == "true" || config["no_rebalance"] == "1" || config["no_rebalance"] == "yes";
+    no_recovery = config["no_recovery"] == "true" || config["no_recovery"] == "1" || config["no_recovery"] == "yes";
     // Cluster configuration
     bind_address = config["bind_address"];
     if (bind_address == "")
@@ -92,6 +101,9 @@ void osd_t::parse_config(blockstore_config_t & config)
     recovery_queue_depth = strtoull(config["recovery_queue_depth"].c_str(), NULL, 10);
     if (recovery_queue_depth < 1 || recovery_queue_depth > MAX_RECOVERY_QUEUE)
         recovery_queue_depth = DEFAULT_RECOVERY_QUEUE;
+    recovery_sync_batch = strtoull(config["recovery_sync_batch"].c_str(), NULL, 10);
+    if (recovery_sync_batch < 1 || recovery_sync_batch > MAX_RECOVERY_QUEUE)
+        recovery_sync_batch = DEFAULT_RECOVERY_BATCH;
     if (config["readonly"] == "true" || config["readonly"] == "1" || config["readonly"] == "yes")
         readonly = true;
     print_stats_interval = strtoull(config["print_stats_interval"].c_str(), NULL, 10);
@@ -100,14 +112,7 @@ void osd_t::parse_config(blockstore_config_t & config)
     slow_log_interval = strtoull(config["slow_log_interval"].c_str(), NULL, 10);
     if (!slow_log_interval)
         slow_log_interval = 10;
-    c_cli.peer_connect_interval = strtoull(config["peer_connect_interval"].c_str(), NULL, 10);
-    if (!c_cli.peer_connect_interval)
-        c_cli.peer_connect_interval = DEFAULT_PEER_CONNECT_INTERVAL;
-    c_cli.peer_connect_timeout = strtoull(config["peer_connect_timeout"].c_str(), NULL, 10);
-    if (!c_cli.peer_connect_timeout)
-        c_cli.peer_connect_timeout = DEFAULT_PEER_CONNECT_TIMEOUT;
-    log_level = strtoull(config["log_level"].c_str(), NULL, 10);
-    c_cli.log_level = log_level;
+    c_cli.parse_config(json_config);
 }
 
 void osd_t::bind_socket()
@@ -171,7 +176,7 @@ bool osd_t::shutdown()
     {
         return false;
     }
-    return bs->is_safe_to_stop();
+    return !bs || bs->is_safe_to_stop();
 }
 
 void osd_t::loop()
@@ -191,6 +196,8 @@ void osd_t::exec_op(osd_op_t *cur_op)
         delete cur_op;
         return;
     }
+    // Clear the reply buffer
+    memset(cur_op->reply.buf, 0, OSD_PACKET_SIZE);
     inflight_ops++;
     if (cur_op->req.hdr.magic != SECONDARY_OSD_OP_MAGIC ||
         cur_op->req.hdr.opcode < OSD_OP_MIN || cur_op->req.hdr.opcode > OSD_OP_MAX ||
@@ -198,19 +205,25 @@ void osd_t::exec_op(osd_op_t *cur_op)
             cur_op->req.hdr.opcode == OSD_OP_SEC_WRITE ||
             cur_op->req.hdr.opcode == OSD_OP_SEC_WRITE_STABLE) &&
             (cur_op->req.sec_rw.len > OSD_RW_MAX ||
-            cur_op->req.sec_rw.len % bs_disk_alignment ||
-            cur_op->req.sec_rw.offset % bs_disk_alignment)) ||
+            cur_op->req.sec_rw.len % bs_bitmap_granularity ||
+            cur_op->req.sec_rw.offset % bs_bitmap_granularity)) ||
         ((cur_op->req.hdr.opcode == OSD_OP_READ ||
             cur_op->req.hdr.opcode == OSD_OP_WRITE ||
             cur_op->req.hdr.opcode == OSD_OP_DELETE) &&
             (cur_op->req.rw.len > OSD_RW_MAX ||
-            cur_op->req.rw.len % bs_disk_alignment ||
-            cur_op->req.rw.offset % bs_disk_alignment)))
+            cur_op->req.rw.len % bs_bitmap_granularity ||
+            cur_op->req.rw.offset % bs_bitmap_granularity)))
     {
         // Bad command
         finish_op(cur_op, -EINVAL);
         return;
     }
+    if (cur_op->req.hdr.opcode == OSD_OP_PING)
+    {
+        // Pong
+        finish_op(cur_op, 0);
+        return;
+    }
     if (readonly &&
         cur_op->req.hdr.opcode != OSD_OP_SEC_READ &&
         cur_op->req.hdr.opcode != OSD_OP_SEC_LIST &&
@@ -261,9 +274,9 @@ void osd_t::reset_stats()
 
 void osd_t::print_stats()
 {
-    for (int i = 0; i <= OSD_OP_MAX; i++)
+    for (int i = OSD_OP_MIN; i <= OSD_OP_MAX; i++)
     {
-        if (c_cli.stats.op_stat_count[i] != prev_stats.op_stat_count[i])
+        if (c_cli.stats.op_stat_count[i] != prev_stats.op_stat_count[i] && i != OSD_OP_PING)
         {
             uint64_t avg = (c_cli.stats.op_stat_sum[i] - prev_stats.op_stat_sum[i])/(c_cli.stats.op_stat_count[i] - prev_stats.op_stat_count[i]);
             uint64_t bw = (c_cli.stats.op_stat_bytes[i] - prev_stats.op_stat_bytes[i]) / print_stats_interval;
@@ -284,7 +297,7 @@ void osd_t::print_stats()
             prev_stats.op_stat_bytes[i] = c_cli.stats.op_stat_bytes[i];
         }
     }
-    for (int i = 0; i <= OSD_OP_MAX; i++)
+    for (int i = OSD_OP_MIN; i <= OSD_OP_MAX; i++)
     {
         if (c_cli.stats.subop_stat_count[i] != prev_stats.subop_stat_count[i])
         {

src/osd.h

@@ -37,6 +37,7 @@
 #define DEFAULT_AUTOSYNC_INTERVAL 5
 #define MAX_RECOVERY_QUEUE 2048
 #define DEFAULT_RECOVERY_QUEUE 4
+#define DEFAULT_RECOVERY_BATCH 16
 
 //#define OSD_STUB
 
@@ -64,6 +65,8 @@ class osd_t
     bool readonly = false;
     osd_num_t osd_num = 1; // OSD numbers start with 1
     bool run_primary = false;
+    bool no_rebalance = false;
+    bool no_recovery = false;
     std::string bind_address;
     int bind_port, listen_backlog;
     // FIXME: Implement client queue depth limit
@@ -74,6 +77,7 @@ class osd_t
     int immediate_commit = IMMEDIATE_NONE;
     int autosync_interval = DEFAULT_AUTOSYNC_INTERVAL; // sync every 5 seconds
     int recovery_queue_depth = DEFAULT_RECOVERY_QUEUE;
+    int recovery_sync_batch = DEFAULT_RECOVERY_BATCH;
     int log_level = 0;
 
     // cluster state
@@ -95,9 +99,11 @@ class osd_t
     std::map<pool_pg_num_t, pg_t> pgs;
     std::set<pool_pg_num_t> dirty_pgs;
     std::set<osd_num_t> dirty_osds;
+    int copies_to_delete_after_sync_count = 0;
     uint64_t misplaced_objects = 0, degraded_objects = 0, incomplete_objects = 0;
     int peering_state = 0;
     std::map<object_id, osd_recovery_op_t> recovery_ops;
+    int recovery_done = 0;
     osd_op_t *autosync_op = NULL;
 
     // Unstable writes
@@ -109,7 +115,7 @@ class osd_t
     bool stopping = false;
     int inflight_ops = 0;
     blockstore_t *bs;
-    uint32_t bs_block_size, bs_disk_alignment;
+    uint32_t bs_block_size, bs_bitmap_granularity;
     ring_loop_t *ringloop;
     timerfd_manager_t *tfd = NULL;
     epoll_manager_t *epmgr = NULL;
@@ -160,6 +166,7 @@ class osd_t
     void submit_list_subop(osd_num_t role_osd, pg_peering_state_t *ps);
     void discard_list_subop(osd_op_t *list_op);
     bool stop_pg(pg_t & pg);
+    void reset_pg(pg_t & pg);
     void finish_stop_pg(pg_t & pg);
 
     // flushing, recovery and backfill
@@ -191,6 +198,7 @@ class osd_t
     void continue_primary_del(osd_op_t *cur_op);
     bool check_write_queue(osd_op_t *cur_op, pg_t & pg);
     void remove_object_from_state(object_id & oid, pg_osd_set_state_t *object_state, pg_t &pg);
+    void free_object_state(pg_t & pg, pg_osd_set_state_t **object_state);
     bool remember_unstable_write(osd_op_t *cur_op, pg_t & pg, pg_osd_set_t & loc_set, int base_state);
     void handle_primary_subop(osd_op_t *subop, osd_op_t *cur_op);
     void handle_primary_bs_subop(osd_op_t *subop);
@@ -198,9 +206,12 @@ class osd_t
     void pg_cancel_write_queue(pg_t & pg, osd_op_t *first_op, object_id oid, int retval);
     void submit_primary_subops(int submit_type, uint64_t op_version, int pg_size, const uint64_t* osd_set, osd_op_t *cur_op);
     void submit_primary_del_subops(osd_op_t *cur_op, uint64_t *cur_set, uint64_t set_size, pg_osd_set_t & loc_set);
-    void submit_primary_sync_subops(osd_op_t *cur_op);
+    void submit_primary_del_batch(osd_op_t *cur_op, obj_ver_osd_t *chunks_to_delete, int chunks_to_delete_count);
+    int submit_primary_sync_subops(osd_op_t *cur_op);
     void submit_primary_stab_subops(osd_op_t *cur_op);
 
+    uint64_t* get_object_osd_set(pg_t &pg, object_id &oid, uint64_t *def, pg_osd_set_state_t **object_state);
+
     inline pg_num_t map_to_pg(object_id oid, uint64_t pg_stripe_size)
     {
         uint64_t pg_count = pg_counts[INODE_POOL(oid.inode)];
@@ -210,7 +221,7 @@ class osd_t
     }
 
 public:
-    osd_t(blockstore_config_t & config, blockstore_t *bs, ring_loop_t *ringloop);
+    osd_t(blockstore_config_t & config, ring_loop_t *ringloop);
     ~osd_t();
     void force_stop(int exitcode);
     bool shutdown();

src/osd_cluster.cpp

@@ -4,6 +4,7 @@
 #include "osd.h"
 #include "base64.h"
 #include "etcd_state_client.h"
+#include "http_client.h"
 #include "osd_rmw.h"
 
 // Startup sequence:
@@ -37,7 +38,7 @@ void osd_t::init_cluster()
                 .pg_cursize = 0,
                 .pg_size = 3,
                 .pg_minsize = 2,
-                .parity_chunks = 1,
+                .pg_data_size = 2,
                 .pool_id = 1,
                 .pg_num = 1,
                 .target_set = { 1, 2, 3 },
@@ -557,7 +558,7 @@ void osd_t::apply_pg_config()
                 }
                 if (currently_taken)
                 {
-                    if (pg_it->second.state & (PG_ACTIVE | PG_INCOMPLETE | PG_PEERING))
+                    if (pg_it->second.state & (PG_ACTIVE | PG_INCOMPLETE | PG_PEERING | PG_REPEERING))
                     {
                         if (pg_it->second.target_set == pg_cfg.target_set)
                         {
@@ -606,7 +607,8 @@ void osd_t::apply_pg_config()
                     .pg_cursize = 0,
                     .pg_size = pool_item.second.pg_size,
                     .pg_minsize = pool_item.second.pg_minsize,
-                    .parity_chunks = pool_item.second.parity_chunks,
+                    .pg_data_size = pg.scheme == POOL_SCHEME_REPLICATED
+                         ? 1 : pool_item.second.pg_size - pool_item.second.parity_chunks,
                     .pool_id = pool_id,
                     .pg_num = pg_num,
                     .reported_epoch = pg_cfg.epoch,
@@ -616,7 +618,7 @@ void osd_t::apply_pg_config()
                 };
                 if (pg.scheme == POOL_SCHEME_JERASURE)
                 {
-                    use_jerasure(pg.pg_size, pg.pg_size-pg.parity_chunks, true);
+                    use_jerasure(pg.pg_size, pg.pg_data_size, true);
                 }
                 this->pg_state_dirty.insert({ .pool_id = pool_id, .pg_num = pg_num });
                 pg.print_state();
@@ -664,7 +666,21 @@ void osd_t::report_pg_states()
         auto & pg = pg_it->second;
         reporting_pgs.push_back({ *it, pg.history_changed });
         std::string state_key_base64 = base64_encode(st_cli.etcd_prefix+"/pg/state/"+std::to_string(pg.pool_id)+"/"+std::to_string(pg.pg_num));
-        if (pg.state == PG_STARTING)
+        bool pg_state_exists = false;
+        if (pg.state != PG_STARTING)
+        {
+            auto pool_it = st_cli.pool_config.find(pg.pool_id);
+            if (pool_it != st_cli.pool_config.end())
+            {
+                auto pg_it = pool_it->second.pg_config.find(pg.pg_num);
+                if (pg_it != pool_it->second.pg_config.end() &&
+                    pg_it->second.cur_state != 0)
+                {
+                    pg_state_exists = true;
+                }
+            }
+        }
+        if (!pg_state_exists)
         {
             // Check that the PG key does not exist
             // Failed check indicates an unsuccessful PG lock attempt in this case
@@ -676,9 +692,7 @@ void osd_t::report_pg_states()
         }
         else
         {
-            // Check that the key is ours
-            // Failed check indicates success for OFFLINE pgs (PG lock is already deleted)
-            // and an unexpected race condition for started pgs (PG lock is held by someone else)
+            // Check that the key is ours if it already exists
             checks.push_back(json11::Json::object {
                 { "target", "LEASE" },
                 { "lease", etcd_lease_id },
@@ -800,17 +814,16 @@ void osd_t::report_pg_states()
             for (auto pp: reporting_pgs)
             {
                 auto pg_it = this->pgs.find(pp.first);
-                if (pg_it != this->pgs.end())
+                if (pg_it != this->pgs.end() &&
+                    pg_it->second.state == PG_OFFLINE &&
+                    pg_state_dirty.find(pp.first) == pg_state_dirty.end())
                 {
-                    if (pg_it->second.state == PG_OFFLINE)
+                    // Forget offline PGs after reporting their state
+                    if (pg_it->second.scheme == POOL_SCHEME_JERASURE)
                     {
-                        // Remove offline PGs after reporting their state
-                        if (pg_it->second.scheme == POOL_SCHEME_JERASURE)
-                        {
-                            use_jerasure(pg_it->second.pg_size, pg_it->second.pg_size-pg_it->second.parity_chunks, false);
-                        }
-                        this->pgs.erase(pg_it);
+                        use_jerasure(pg_it->second.pg_size, pg_it->second.pg_data_size, false);
                     }
+                    this->pgs.erase(pg_it);
                 }
             }
             // Push other PG state updates, if any

src/osd_flush.cpp

@@ -149,10 +149,14 @@ void osd_t::handle_flush_op(bool rollback, pool_id_t pool_id, pg_num_t pg_num, p
         {
             continue_primary_write(op);
         }
-        if (pg.inflight == 0 && (pg.state & PG_STOPPING))
+        if ((pg.state & PG_STOPPING) && pg.inflight == 0 && !pg.flush_batch)
         {
             finish_stop_pg(pg);
         }
+        else if ((pg.state & PG_REPEERING) && pg.inflight == 0 && !pg.flush_batch)
+        {
+            start_pg_peering(pg);
+        }
     }
 }
 
@@ -209,32 +213,39 @@ void osd_t::submit_flush_op(pool_id_t pool_id, pg_num_t pg_num, pg_flush_batch_t
 
 bool osd_t::pick_next_recovery(osd_recovery_op_t &op)
 {
-    for (auto pg_it = pgs.begin(); pg_it != pgs.end(); pg_it++)
+    if (!no_recovery)
     {
-        if ((pg_it->second.state & (PG_ACTIVE | PG_HAS_DEGRADED)) == (PG_ACTIVE | PG_HAS_DEGRADED))
+        for (auto pg_it = pgs.begin(); pg_it != pgs.end(); pg_it++)
         {
-            for (auto obj_it = pg_it->second.degraded_objects.begin(); obj_it != pg_it->second.degraded_objects.end(); obj_it++)
+            if ((pg_it->second.state & (PG_ACTIVE | PG_HAS_DEGRADED)) == (PG_ACTIVE | PG_HAS_DEGRADED))
             {
-                if (recovery_ops.find(obj_it->first) == recovery_ops.end())
+                for (auto obj_it = pg_it->second.degraded_objects.begin(); obj_it != pg_it->second.degraded_objects.end(); obj_it++)
                 {
-                    op.degraded = true;
-                    op.oid = obj_it->first;
-                    return true;
+                    if (recovery_ops.find(obj_it->first) == recovery_ops.end())
+                    {
+                        op.degraded = true;
+                        op.oid = obj_it->first;
+                        return true;
+                    }
                 }
             }
         }
     }
-    for (auto pg_it = pgs.begin(); pg_it != pgs.end(); pg_it++)
+    if (!no_rebalance)
     {
-        if ((pg_it->second.state & (PG_ACTIVE | PG_HAS_MISPLACED)) == (PG_ACTIVE | PG_HAS_MISPLACED))
+        for (auto pg_it = pgs.begin(); pg_it != pgs.end(); pg_it++)
         {
-            for (auto obj_it = pg_it->second.misplaced_objects.begin(); obj_it != pg_it->second.misplaced_objects.end(); obj_it++)
+            // Don't try to "recover" misplaced objects if "recovery" would make them degraded
+            if ((pg_it->second.state & (PG_ACTIVE | PG_DEGRADED | PG_HAS_MISPLACED)) == (PG_ACTIVE | PG_HAS_MISPLACED))
             {
-                if (recovery_ops.find(obj_it->first) == recovery_ops.end())
+                for (auto obj_it = pg_it->second.misplaced_objects.begin(); obj_it != pg_it->second.misplaced_objects.end(); obj_it++)
                 {
-                    op.degraded = false;
-                    op.oid = obj_it->first;
-                    return true;
+                    if (recovery_ops.find(obj_it->first) == recovery_ops.end())
+                    {
+                        op.degraded = false;
+                        op.oid = obj_it->first;
+                        return true;
+                    }
                 }
             }
         }
@@ -264,7 +275,6 @@ void osd_t::submit_recovery_op(osd_recovery_op_t *op)
     }
     op->osd_op->callback = [this, op](osd_op_t *osd_op)
     {
-        // Don't sync the write, it will be synced by our regular sync coroutine
         if (osd_op->reply.hdr.retval < 0)
         {
             // Error recovering object
@@ -286,6 +296,17 @@ void osd_t::submit_recovery_op(osd_recovery_op_t *op)
         op->osd_op = NULL;
         recovery_ops.erase(op->oid);
         delete osd_op;
+        if (immediate_commit != IMMEDIATE_ALL)
+        {
+            recovery_done++;
+            if (recovery_done >= recovery_sync_batch)
+            {
+                // Force sync every <recovery_sync_batch> operations
+                // This is required not to pile up an excessive amount of delete operations
+                autosync();
+                recovery_done = 0;
+            }
+        }
         continue_recovery();
     };
     exec_op(op->osd_op);

src/osd_main.cpp

@@ -41,16 +41,13 @@ int main(int narg, char *args[])
     signal(SIGINT, handle_sigint);
     signal(SIGTERM, handle_sigint);
     ring_loop_t *ringloop = new ring_loop_t(512);
-    // FIXME: Create Blockstore from on-disk superblock config and check it against the OSD cluster config
-    blockstore_t *bs = new blockstore_t(config, ringloop);
-    osd = new osd_t(config, bs, ringloop);
+    osd = new osd_t(config, ringloop);
     while (1)
     {
         ringloop->loop();
         ringloop->wait();
     }
     delete osd;
-    delete bs;
     delete ringloop;
     return 0;
 }

src/osd_ops.cpp

@@ -19,4 +19,5 @@ const char* osd_op_names[] = {
     "primary_write",
     "primary_sync",
     "primary_delete",
+    "ping",
 };

src/osd_ops.h

@@ -27,7 +27,8 @@
 #define OSD_OP_WRITE                12
 #define OSD_OP_SYNC                 13
 #define OSD_OP_DELETE               14
-#define OSD_OP_MAX                  14
+#define OSD_OP_PING                 15
+#define OSD_OP_MAX                  15
 // Alignment & limit for read/write operations
 #ifndef MEM_ALIGNMENT
 #define MEM_ALIGNMENT               512

src/osd_peering.cpp

@@ -77,10 +77,11 @@ void osd_t::repeer_pgs(osd_num_t peer_osd)
     // Re-peer affected PGs
     for (auto & p: pgs)
     {
+        auto & pg = p.second;
         bool repeer = false;
-        if (p.second.state & (PG_PEERING | PG_ACTIVE | PG_INCOMPLETE))
+        if (pg.state & (PG_PEERING | PG_ACTIVE | PG_INCOMPLETE))
         {
-            for (osd_num_t pg_osd: p.second.all_peers)
+            for (osd_num_t pg_osd: pg.all_peers)
             {
                 if (pg_osd == peer_osd)
                 {
@@ -91,22 +92,29 @@ void osd_t::repeer_pgs(osd_num_t peer_osd)
             if (repeer)
             {
                 // Repeer this pg
-                printf("[PG %u/%u] Repeer because of OSD %lu\n", p.second.pool_id, p.second.pg_num, peer_osd);
-                start_pg_peering(p.second);
+                printf("[PG %u/%u] Repeer because of OSD %lu\n", pg.pool_id, pg.pg_num, peer_osd);
+                if (!(pg.state & (PG_ACTIVE | PG_REPEERING)) || pg.inflight == 0 && !pg.flush_batch)
+                {
+                    start_pg_peering(pg);
+                }
+                else
+                {
+                    // Stop accepting new operations, wait for current ones to finish or fail
+                    pg.state = pg.state & ~PG_ACTIVE | PG_REPEERING;
+                    report_pg_state(pg);
+                }
             }
         }
     }
 }
 
-// Repeer on each connect/disconnect peer event
-void osd_t::start_pg_peering(pg_t & pg)
+// Reset PG state (when peering or stopping)
+void osd_t::reset_pg(pg_t & pg)
 {
-    pg.state = PG_PEERING;
-    this->peering_state |= OSD_PEERING_PGS;
-    report_pg_state(pg);
-    // Reset PG state
     pg.cur_peers.clear();
     pg.state_dict.clear();
+    copies_to_delete_after_sync_count -= pg.copies_to_delete_after_sync.size();
+    pg.copies_to_delete_after_sync.clear();
     incomplete_objects -= pg.incomplete_objects.size();
     misplaced_objects -= pg.misplaced_objects.size();
     degraded_objects -= pg.degraded_objects.size();
@@ -135,6 +143,15 @@ void osd_t::start_pg_peering(pg_t & pg)
             it++;
     }
     dirty_pgs.erase({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
+}
+
+// Repeer on each connect/disconnect peer event
+void osd_t::start_pg_peering(pg_t & pg)
+{
+    pg.state = PG_PEERING;
+    this->peering_state |= OSD_PEERING_PGS;
+    reset_pg(pg);
+    report_pg_state(pg);
     // Drop connections of clients who have this PG in dirty_pgs
     if (immediate_commit != IMMEDIATE_ALL)
     {
@@ -175,13 +192,18 @@ 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 = false;
+            bool found = true;
             for (auto history_osd: history_set)
             {
-                if (history_osd != 0 && c_cli.osd_peer_fds.find(history_osd) != c_cli.osd_peer_fds.end())
+                if (history_osd != 0)
                 {
-                    found = true;
-                    break;
+                    found = false;
+                    if (history_osd == this->osd_num ||
+                        c_cli.osd_peer_fds.find(history_osd) != c_cli.osd_peer_fds.end())
+                    {
+                        found = true;
+                        break;
+                    }
                 }
             }
             if (!found)
@@ -322,9 +344,10 @@ void osd_t::submit_sync_and_list_subop(osd_num_t role_osd, pg_peering_state_t *p
             {
                 // FIXME: Mark peer as failed and don't reconnect immediately after dropping the connection
                 printf("Failed to sync OSD %lu: %ld (%s), disconnecting peer\n", role_osd, op->reply.hdr.retval, strerror(-op->reply.hdr.retval));
+                int fail_fd = op->peer_fd;
                 ps->list_ops.erase(role_osd);
-                c_cli.stop_client(op->peer_fd);
                 delete op;
+                c_cli.stop_client(fail_fd);
                 return;
             }
             delete op;
@@ -401,9 +424,10 @@ void osd_t::submit_list_subop(osd_num_t role_osd, pg_peering_state_t *ps)
             if (op->reply.hdr.retval < 0)
             {
                 printf("Failed to get object list from OSD %lu (retval=%ld), disconnecting peer\n", role_osd, op->reply.hdr.retval);
+                int fail_fd = op->peer_fd;
                 ps->list_ops.erase(role_osd);
-                c_cli.stop_client(op->peer_fd);
                 delete op;
+                c_cli.stop_client(fail_fd);
                 return;
             }
             printf(
@@ -454,11 +478,11 @@ bool osd_t::stop_pg(pg_t & pg)
     if (pg.peering_state)
     {
         // Stop peering
-        for (auto it = pg.peering_state->list_ops.begin(); it != pg.peering_state->list_ops.end();)
+        for (auto it = pg.peering_state->list_ops.begin(); it != pg.peering_state->list_ops.end(); it++)
         {
             discard_list_subop(it->second);
         }
-        for (auto it = pg.peering_state->list_results.begin(); it != pg.peering_state->list_results.end();)
+        for (auto it = pg.peering_state->list_results.begin(); it != pg.peering_state->list_results.end(); it++)
         {
             if (it->second.buf)
             {
@@ -468,14 +492,17 @@ bool osd_t::stop_pg(pg_t & pg)
         delete pg.peering_state;
         pg.peering_state = NULL;
     }
-    if (!(pg.state & PG_ACTIVE))
+    if (pg.state & (PG_STOPPING | PG_OFFLINE))
     {
         return false;
     }
-    pg.state = pg.state & ~PG_ACTIVE | PG_STOPPING;
-    if (pg.inflight == 0 && !pg.flush_batch &&
-        // We must either forget all PG's unstable writes or wait for it to become clean
-        dirty_pgs.find({ .pool_id = pg.pool_id, .pg_num = pg.pg_num }) == dirty_pgs.end())
+    if (!(pg.state & (PG_ACTIVE | PG_REPEERING)))
+    {
+        finish_stop_pg(pg);
+        return true;
+    }
+    pg.state = pg.state & ~PG_ACTIVE & ~PG_REPEERING | PG_STOPPING;
+    if (pg.inflight == 0 && !pg.flush_batch)
     {
         finish_stop_pg(pg);
     }
@@ -489,6 +516,7 @@ bool osd_t::stop_pg(pg_t & pg)
 void osd_t::finish_stop_pg(pg_t & pg)
 {
     pg.state = PG_OFFLINE;
+    reset_pg(pg);
     report_pg_state(pg);
 }
 

src/osd_peering_pg.cpp

@@ -108,7 +108,7 @@ void pg_obj_state_check_t::start_object()
 
 void pg_obj_state_check_t::handle_version()
 {
-    if (!target_ver && last_ver != list[list_pos].version && (n_stable > 0 || n_roles >= pg->pg_minsize))
+    if (!target_ver && last_ver != list[list_pos].version && (n_stable > 0 || n_roles >= pg->pg_data_size))
     {
         // Version is either stable or recoverable
         target_ver = last_ver;
@@ -171,7 +171,7 @@ void pg_obj_state_check_t::handle_version()
 
 void pg_obj_state_check_t::finish_object()
 {
-    if (!target_ver && (n_stable > 0 || n_roles >= pg->pg_minsize))
+    if (!target_ver && (n_stable > 0 || n_roles >= pg->pg_data_size))
     {
         // Version is either stable or recoverable
         target_ver = last_ver;
@@ -233,7 +233,7 @@ void pg_obj_state_check_t::finish_object()
     {
         return;
     }
-    if (!replicated && n_roles < pg->pg_minsize)
+    if (!replicated && n_roles < pg->pg_data_size)
     {
         if (log_level > 1)
         {
@@ -430,12 +430,13 @@ void pg_t::calc_object_states(int log_level)
 void pg_t::print_state()
 {
     printf(
-        "[PG %u/%u] is %s%s%s%s%s%s%s%s%s%s%s%s%s (%lu objects)\n", pool_id, pg_num,
+        "[PG %u/%u] is %s%s%s%s%s%s%s%s%s%s%s%s%s%s (%lu objects)\n", pool_id, pg_num,
         (state & PG_STARTING) ? "starting" : "",
         (state & PG_OFFLINE) ? "offline" : "",
         (state & PG_PEERING) ? "peering" : "",
         (state & PG_INCOMPLETE) ? "incomplete" : "",
         (state & PG_ACTIVE) ? "active" : "",
+        (state & PG_REPEERING) ? "repeering" : "",
         (state & PG_STOPPING) ? "stopping" : "",
         (state & PG_DEGRADED) ? " + degraded" : "",
         (state & PG_HAS_INCOMPLETE) ? " + has_incomplete" : "",

src/osd_peering_pg.h

@@ -56,6 +56,13 @@ struct obj_piece_id_t
     uint64_t osd_num;
 };
 
+struct obj_ver_osd_t
+{
+    uint64_t osd_num;
+    object_id oid;
+    uint64_t version;
+};
+
 struct flush_action_t
 {
     bool rollback = false, make_stable = false;
@@ -75,7 +82,7 @@ struct pg_t
 {
     int state = 0;
     uint64_t scheme = 0;
-    uint64_t pg_cursize = 0, pg_size = 0, pg_minsize = 0, parity_chunks = 0;
+    uint64_t pg_cursize = 0, pg_size = 0, pg_minsize = 0, pg_data_size = 0;
     pool_id_t pool_id = 0;
     pg_num_t pg_num = 0;
     uint64_t clean_count = 0, total_count = 0;
@@ -101,6 +108,7 @@ struct pg_t
     std::map<pg_osd_set_t, pg_osd_set_state_t> state_dict;
     btree::btree_map<object_id, pg_osd_set_state_t*> incomplete_objects, misplaced_objects, degraded_objects;
     std::map<obj_piece_id_t, flush_action_t> flush_actions;
+    std::vector<obj_ver_osd_t> copies_to_delete_after_sync;
     btree::btree_map<object_id, uint64_t> ver_override;
     pg_peering_state_t *peering_state = NULL;
     pg_flush_batch_t *flush_batch = NULL;

src/osd_primary.cpp

@@ -18,7 +18,7 @@ bool osd_t::prepare_primary_rw(osd_op_t *cur_op)
     // Our EC scheme stores data in fixed chunks equal to (K*block size)
     // K = (pg_size-parity_chunks) in case of EC/XOR, or 1 for replicated pools
     pool_id_t pool_id = INODE_POOL(cur_op->req.rw.inode);
-    // FIXME: We have to access pool config here, so make sure that it doesn't change while its PGs are active...
+    // Note: We read pool config here, so we must NOT change it when PGs are active
     auto pool_cfg_it = st_cli.pool_config.find(pool_id);
     if (pool_cfg_it == st_cli.pool_config.end())
     {
@@ -44,8 +44,8 @@ bool osd_t::prepare_primary_rw(osd_op_t *cur_op)
         return false;
     }
     if ((cur_op->req.rw.offset + cur_op->req.rw.len) > (oid.stripe + pg_block_size) ||
-        (cur_op->req.rw.offset % bs_disk_alignment) != 0 ||
-        (cur_op->req.rw.len % bs_disk_alignment) != 0)
+        (cur_op->req.rw.offset % bs_bitmap_granularity) != 0 ||
+        (cur_op->req.rw.len % bs_bitmap_granularity) != 0)
     {
         finish_op(cur_op, -EINVAL);
         return false;
@@ -64,7 +64,7 @@ bool osd_t::prepare_primary_rw(osd_op_t *cur_op)
     return true;
 }
 
-static uint64_t* get_object_osd_set(pg_t &pg, object_id &oid, uint64_t *def, pg_osd_set_state_t **object_state)
+uint64_t* osd_t::get_object_osd_set(pg_t &pg, object_id &oid, uint64_t *def, pg_osd_set_state_t **object_state)
 {
     if (!(pg.state & (PG_HAS_INCOMPLETE | PG_HAS_DEGRADED | PG_HAS_MISPLACED)))
     {
@@ -177,527 +177,6 @@ resume_2:
     finish_op(cur_op, cur_op->req.rw.len);
 }
 
-bool osd_t::check_write_queue(osd_op_t *cur_op, pg_t & pg)
-{
-    osd_primary_op_data_t *op_data = cur_op->op_data;
-    // Check if actions are pending for this object
-    auto act_it = pg.flush_actions.lower_bound((obj_piece_id_t){
-        .oid = op_data->oid,
-        .osd_num = 0,
-    });
-    if (act_it != pg.flush_actions.end() &&
-        act_it->first.oid.inode == op_data->oid.inode &&
-        (act_it->first.oid.stripe & ~STRIPE_MASK) == op_data->oid.stripe)
-    {
-        pg.write_queue.emplace(op_data->oid, cur_op);
-        return false;
-    }
-    // Check if there are other write requests to the same object
-    auto vo_it = pg.write_queue.find(op_data->oid);
-    if (vo_it != pg.write_queue.end())
-    {
-        op_data->st = 1;
-        pg.write_queue.emplace(op_data->oid, cur_op);
-        return false;
-    }
-    pg.write_queue.emplace(op_data->oid, cur_op);
-    return true;
-}
-
-void osd_t::continue_primary_write(osd_op_t *cur_op)
-{
-    if (!cur_op->op_data && !prepare_primary_rw(cur_op))
-    {
-        return;
-    }
-    osd_primary_op_data_t *op_data = cur_op->op_data;
-    auto & pg = pgs.at({ .pool_id = INODE_POOL(op_data->oid.inode), .pg_num = op_data->pg_num });
-    if (op_data->st == 1)      goto resume_1;
-    else if (op_data->st == 2) goto resume_2;
-    else if (op_data->st == 3) goto resume_3;
-    else if (op_data->st == 4) goto resume_4;
-    else if (op_data->st == 5) goto resume_5;
-    else if (op_data->st == 6) goto resume_6;
-    else if (op_data->st == 7) goto resume_7;
-    else if (op_data->st == 8) goto resume_8;
-    else if (op_data->st == 9) goto resume_9;
-    else if (op_data->st == 10) goto resume_10;
-    assert(op_data->st == 0);
-    if (!check_write_queue(cur_op, pg))
-    {
-        return;
-    }
-resume_1:
-    // Determine blocks to read and write
-    // Missing chunks are allowed to be overwritten even in incomplete objects
-    // FIXME: Allow to do small writes to the old (degraded/misplaced) OSD set for lower performance impact
-    op_data->prev_set = get_object_osd_set(pg, op_data->oid, pg.cur_set.data(), &op_data->object_state);
-    if (op_data->scheme == POOL_SCHEME_REPLICATED)
-    {
-        // Simplified algorithm
-        op_data->stripes[0].write_start = op_data->stripes[0].req_start;
-        op_data->stripes[0].write_end = op_data->stripes[0].req_end;
-        op_data->stripes[0].write_buf = cur_op->buf;
-        if (pg.cur_set.data() != op_data->prev_set && (op_data->stripes[0].write_start != 0 ||
-            op_data->stripes[0].write_end != bs_block_size))
-        {
-            // Object is degraded/misplaced and will be moved to <write_osd_set>
-            op_data->stripes[0].read_start = 0;
-            op_data->stripes[0].read_end = bs_block_size;
-            cur_op->rmw_buf = op_data->stripes[0].read_buf = memalign_or_die(MEM_ALIGNMENT, bs_block_size);
-        }
-    }
-    else
-    {
-        cur_op->rmw_buf = calc_rmw(cur_op->buf, op_data->stripes, op_data->prev_set,
-            pg.pg_size, op_data->pg_data_size, pg.pg_cursize, pg.cur_set.data(), bs_block_size);
-        if (!cur_op->rmw_buf)
-        {
-            // Refuse partial overwrite of an incomplete object
-            cur_op->reply.hdr.retval = -EINVAL;
-            goto continue_others;
-        }
-    }
-    // Read required blocks
-    submit_primary_subops(SUBMIT_RMW_READ, UINT64_MAX, pg.pg_size, op_data->prev_set, cur_op);
-resume_2:
-    op_data->st = 2;
-    return;
-resume_3:
-    if (op_data->errors > 0)
-    {
-        pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
-        return;
-    }
-    // Save version override for parallel reads
-    pg.ver_override[op_data->oid] = op_data->fact_ver;
-    if (op_data->scheme == POOL_SCHEME_REPLICATED)
-    {
-        // Only (possibly) copy new data from the request into the recovery buffer
-        if (pg.cur_set.data() != op_data->prev_set && (op_data->stripes[0].write_start != 0 ||
-            op_data->stripes[0].write_end != bs_block_size))
-        {
-            memcpy(
-                op_data->stripes[0].read_buf + op_data->stripes[0].req_start,
-                op_data->stripes[0].write_buf,
-                op_data->stripes[0].req_end - op_data->stripes[0].req_start
-            );
-            op_data->stripes[0].write_buf = op_data->stripes[0].read_buf;
-            op_data->stripes[0].write_start = 0;
-            op_data->stripes[0].write_end = bs_block_size;
-        }
-    }
-    else
-    {
-        // Recover missing stripes, calculate parity
-        if (pg.scheme == POOL_SCHEME_XOR)
-        {
-            calc_rmw_parity_xor(op_data->stripes, pg.pg_size, op_data->prev_set, pg.cur_set.data(), bs_block_size);
-        }
-        else if (pg.scheme == POOL_SCHEME_JERASURE)
-        {
-            calc_rmw_parity_jerasure(op_data->stripes, pg.pg_size, op_data->pg_data_size, op_data->prev_set, pg.cur_set.data(), bs_block_size);
-        }
-    }
-    // Send writes
-    if ((op_data->fact_ver >> (64-PG_EPOCH_BITS)) < pg.epoch)
-    {
-        op_data->target_ver = ((uint64_t)pg.epoch << (64-PG_EPOCH_BITS)) | 1;
-    }
-    else
-    {
-        if ((op_data->fact_ver & (1ul<<(64-PG_EPOCH_BITS) - 1)) == (1ul<<(64-PG_EPOCH_BITS) - 1))
-        {
-            assert(pg.epoch != ((1ul << PG_EPOCH_BITS)-1));
-            pg.epoch++;
-        }
-        op_data->target_ver = op_data->fact_ver + 1;
-    }
-    if (pg.epoch > pg.reported_epoch)
-    {
-        // Report newer epoch before writing
-        // FIXME: We may report only one PG state here...
-        this->pg_state_dirty.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
-        pg.history_changed = true;
-        report_pg_states();
-resume_10:
-        if (pg.epoch > pg.reported_epoch)
-        {
-            op_data->st = 10;
-            return;
-        }
-    }
-    submit_primary_subops(SUBMIT_WRITE, op_data->target_ver, pg.pg_size, pg.cur_set.data(), cur_op);
-resume_4:
-    op_data->st = 4;
-    return;
-resume_5:
-    if (op_data->errors > 0)
-    {
-        pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
-        return;
-    }
-resume_6:
-resume_7:
-    if (!remember_unstable_write(cur_op, pg, pg.cur_loc_set, 6))
-    {
-        // FIXME: Check for immediate_commit == IMMEDIATE_SMALL
-        return;
-    }
-    if (op_data->fact_ver == 1)
-    {
-        // Object is created
-        pg.clean_count++;
-        pg.total_count++;
-    }
-    if (op_data->object_state)
-    {
-        {
-            int recovery_type = op_data->object_state->state & (OBJ_DEGRADED|OBJ_INCOMPLETE) ? 0 : 1;
-            recovery_stat_count[0][recovery_type]++;
-            if (!recovery_stat_count[0][recovery_type])
-            {
-                recovery_stat_count[0][recovery_type]++;
-                recovery_stat_bytes[0][recovery_type] = 0;
-            }
-            for (int role = 0; role < (op_data->scheme == POOL_SCHEME_REPLICATED ? 1 : pg.pg_size); role++)
-            {
-                recovery_stat_bytes[0][recovery_type] += op_data->stripes[role].write_end - op_data->stripes[role].write_start;
-            }
-        }
-        if (op_data->object_state->state & OBJ_MISPLACED)
-        {
-            // Remove extra chunks
-            submit_primary_del_subops(cur_op, pg.cur_set.data(), pg.pg_size, op_data->object_state->osd_set);
-            if (op_data->n_subops > 0)
-            {
-resume_8:
-                op_data->st = 8;
-                return;
-resume_9:
-                if (op_data->errors > 0)
-                {
-                    pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
-                    return;
-                }
-            }
-        }
-        // Clear object state
-        remove_object_from_state(op_data->oid, op_data->object_state, pg);
-        pg.clean_count++;
-    }
-    cur_op->reply.hdr.retval = cur_op->req.rw.len;
-continue_others:
-    // Remove version override
-    pg.ver_override.erase(op_data->oid);
-    object_id oid = op_data->oid;
-    finish_op(cur_op, cur_op->reply.hdr.retval);
-    // Continue other write operations to the same object
-    auto next_it = pg.write_queue.find(oid);
-    auto this_it = next_it;
-    if (this_it != pg.write_queue.end() && this_it->second == cur_op)
-    {
-        next_it++;
-        pg.write_queue.erase(this_it);
-        if (next_it != pg.write_queue.end() && next_it->first == oid)
-        {
-            osd_op_t *next_op = next_it->second;
-            continue_primary_write(next_op);
-        }
-    }
-}
-
-bool osd_t::remember_unstable_write(osd_op_t *cur_op, pg_t & pg, pg_osd_set_t & loc_set, int base_state)
-{
-    osd_primary_op_data_t *op_data = cur_op->op_data;
-    if (op_data->st == base_state)
-    {
-        goto resume_6;
-    }
-    else if (op_data->st == base_state+1)
-    {
-        goto resume_7;
-    }
-    // FIXME: Check for immediate_commit == IMMEDIATE_SMALL
-    if (immediate_commit == IMMEDIATE_ALL)
-    {
-        if (op_data->scheme != POOL_SCHEME_REPLICATED)
-        {
-            // Send STABILIZE ops immediately
-            op_data->unstable_write_osds = new std::vector<unstable_osd_num_t>();
-            op_data->unstable_writes = new obj_ver_id[loc_set.size()];
-            {
-                int last_start = 0;
-                for (auto & chunk: loc_set)
-                {
-                    op_data->unstable_writes[last_start] = (obj_ver_id){
-                        .oid = {
-                            .inode = op_data->oid.inode,
-                            .stripe = op_data->oid.stripe | chunk.role,
-                        },
-                        .version = op_data->fact_ver,
-                    };
-                    op_data->unstable_write_osds->push_back((unstable_osd_num_t){
-                        .osd_num = chunk.osd_num,
-                        .start = last_start,
-                        .len = 1,
-                    });
-                    last_start++;
-                }
-            }
-            submit_primary_stab_subops(cur_op);
-resume_6:
-            op_data->st = 6;
-            return false;
-resume_7:
-            // FIXME: Free those in the destructor?
-            delete op_data->unstable_write_osds;
-            delete[] op_data->unstable_writes;
-            op_data->unstable_writes = NULL;
-            op_data->unstable_write_osds = NULL;
-            if (op_data->errors > 0)
-            {
-                pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
-                return false;
-            }
-        }
-    }
-    else
-    {
-        if (op_data->scheme != POOL_SCHEME_REPLICATED)
-        {
-            // Remember version as unstable for EC/XOR
-            for (auto & chunk: loc_set)
-            {
-                this->dirty_osds.insert(chunk.osd_num);
-                this->unstable_writes[(osd_object_id_t){
-                    .osd_num = chunk.osd_num,
-                    .oid = {
-                        .inode = op_data->oid.inode,
-                        .stripe = op_data->oid.stripe | chunk.role,
-                    },
-                }] = op_data->fact_ver;
-            }
-        }
-        else
-        {
-            // Only remember to sync OSDs for replicated pools
-            for (auto & chunk: loc_set)
-            {
-                this->dirty_osds.insert(chunk.osd_num);
-            }
-        }
-        // Remember PG as dirty to drop the connection when PG goes offline
-        // (this is required because of the "lazy sync")
-        auto cl_it = c_cli.clients.find(cur_op->peer_fd);
-        if (cl_it != c_cli.clients.end())
-        {
-            cl_it->second->dirty_pgs.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
-        }
-        dirty_pgs.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
-    }
-    return true;
-}
-
-// Save and clear unstable_writes -> SYNC all -> STABLE all
-void osd_t::continue_primary_sync(osd_op_t *cur_op)
-{
-    if (!cur_op->op_data)
-    {
-        cur_op->op_data = (osd_primary_op_data_t*)calloc_or_die(1, sizeof(osd_primary_op_data_t));
-    }
-    osd_primary_op_data_t *op_data = cur_op->op_data;
-    if (op_data->st == 1)      goto resume_1;
-    else if (op_data->st == 2) goto resume_2;
-    else if (op_data->st == 3) goto resume_3;
-    else if (op_data->st == 4) goto resume_4;
-    else if (op_data->st == 5) goto resume_5;
-    else if (op_data->st == 6) goto resume_6;
-    assert(op_data->st == 0);
-    if (syncs_in_progress.size() > 0)
-    {
-        // Wait for previous syncs, if any
-        // FIXME: We may try to execute the current one in parallel, like in Blockstore, but I'm not sure if it matters at all
-        syncs_in_progress.push_back(cur_op);
-        op_data->st = 1;
-resume_1:
-        return;
-    }
-    else
-    {
-        syncs_in_progress.push_back(cur_op);
-    }
-resume_2:
-    if (dirty_osds.size() == 0)
-    {
-        // Nothing to sync
-        goto finish;
-    }
-    // Save and clear unstable_writes
-    // In theory it is possible to do in on a per-client basis, but this seems to be an unnecessary complication
-    // It would be cool not to copy these here at all, but someone has to deduplicate them by object IDs anyway
-    if (unstable_writes.size() > 0)
-    {
-        op_data->unstable_write_osds = new std::vector<unstable_osd_num_t>();
-        op_data->unstable_writes = new obj_ver_id[this->unstable_writes.size()];
-        osd_num_t last_osd = 0;
-        int last_start = 0, last_end = 0;
-        for (auto it = this->unstable_writes.begin(); it != this->unstable_writes.end(); it++)
-        {
-            if (last_osd != it->first.osd_num)
-            {
-                if (last_osd != 0)
-                {
-                    op_data->unstable_write_osds->push_back((unstable_osd_num_t){
-                        .osd_num = last_osd,
-                        .start = last_start,
-                        .len = last_end - last_start,
-                    });
-                }
-                last_osd = it->first.osd_num;
-                last_start = last_end;
-            }
-            op_data->unstable_writes[last_end] = (obj_ver_id){
-                .oid = it->first.oid,
-                .version = it->second,
-            };
-            last_end++;
-        }
-        if (last_osd != 0)
-        {
-            op_data->unstable_write_osds->push_back((unstable_osd_num_t){
-                .osd_num = last_osd,
-                .start = last_start,
-                .len = last_end - last_start,
-            });
-        }
-        this->unstable_writes.clear();
-    }
-    {
-        void *dirty_buf = malloc_or_die(sizeof(pool_pg_num_t)*dirty_pgs.size() + sizeof(osd_num_t)*dirty_osds.size());
-        op_data->dirty_pgs = (pool_pg_num_t*)dirty_buf;
-        op_data->dirty_osds = (osd_num_t*)(dirty_buf + sizeof(pool_pg_num_t)*dirty_pgs.size());
-        op_data->dirty_pg_count = dirty_pgs.size();
-        op_data->dirty_osd_count = dirty_osds.size();
-        int dpg = 0;
-        for (auto dirty_pg_num: dirty_pgs)
-        {
-            pgs.at(dirty_pg_num).inflight++;
-            op_data->dirty_pgs[dpg++] = dirty_pg_num;
-        }
-        dirty_pgs.clear();
-        dpg = 0;
-        for (auto osd_num: dirty_osds)
-        {
-            op_data->dirty_osds[dpg++] = osd_num;
-        }
-        dirty_osds.clear();
-    }
-    if (immediate_commit != IMMEDIATE_ALL)
-    {
-        // SYNC
-        submit_primary_sync_subops(cur_op);
-resume_3:
-        op_data->st = 3;
-        return;
-resume_4:
-        if (op_data->errors > 0)
-        {
-            goto resume_6;
-        }
-    }
-    if (op_data->unstable_writes)
-    {
-        // Stabilize version sets, if any
-        submit_primary_stab_subops(cur_op);
-resume_5:
-        op_data->st = 5;
-        return;
-    }
-resume_6:
-    if (op_data->errors > 0)
-    {
-        // Return PGs and OSDs back into their dirty sets
-        for (int i = 0; i < op_data->dirty_pg_count; i++)
-        {
-            dirty_pgs.insert(op_data->dirty_pgs[i]);
-        }
-        for (int i = 0; i < op_data->dirty_osd_count; i++)
-        {
-            dirty_osds.insert(op_data->dirty_osds[i]);
-        }
-        if (op_data->unstable_writes)
-        {
-            // Return objects back into the unstable write set
-            for (auto unstable_osd: *(op_data->unstable_write_osds))
-            {
-                for (int i = 0; i < unstable_osd.len; i++)
-                {
-                    // Except those from peered PGs
-                    auto & w = op_data->unstable_writes[i];
-                    pool_pg_num_t wpg = {
-                        .pool_id = INODE_POOL(w.oid.inode),
-                        .pg_num = map_to_pg(w.oid, st_cli.pool_config.at(INODE_POOL(w.oid.inode)).pg_stripe_size),
-                    };
-                    if (pgs.at(wpg).state & PG_ACTIVE)
-                    {
-                        uint64_t & dest = this->unstable_writes[(osd_object_id_t){
-                            .osd_num = unstable_osd.osd_num,
-                            .oid = w.oid,
-                        }];
-                        dest = dest < w.version ? w.version : dest;
-                        dirty_pgs.insert(wpg);
-                    }
-                }
-            }
-        }
-    }
-    for (int i = 0; i < op_data->dirty_pg_count; i++)
-    {
-        auto & pg = pgs.at(op_data->dirty_pgs[i]);
-        pg.inflight--;
-        if ((pg.state & PG_STOPPING) && pg.inflight == 0 && !pg.flush_batch)
-        {
-            finish_stop_pg(pg);
-        }
-    }
-    // FIXME: Free those in the destructor?
-    free(op_data->dirty_pgs);
-    op_data->dirty_pgs = NULL;
-    op_data->dirty_osds = NULL;
-    if (op_data->unstable_writes)
-    {
-        delete op_data->unstable_write_osds;
-        delete[] op_data->unstable_writes;
-        op_data->unstable_writes = NULL;
-        op_data->unstable_write_osds = NULL;
-    }
-    if (op_data->errors > 0)
-    {
-        finish_op(cur_op, op_data->epipe > 0 ? -EPIPE : -EIO);
-    }
-    else
-    {
-finish:
-        if (cur_op->peer_fd)
-        {
-            auto it = c_cli.clients.find(cur_op->peer_fd);
-            if (it != c_cli.clients.end())
-                it->second->dirty_pgs.clear();
-        }
-        finish_op(cur_op, 0);
-    }
-    assert(syncs_in_progress.front() == cur_op);
-    syncs_in_progress.pop_front();
-    if (syncs_in_progress.size() > 0)
-    {
-        cur_op = syncs_in_progress.front();
-        op_data = cur_op->op_data;
-        op_data->st++;
-        goto resume_2;
-    }
-}
-
 // Decrement pg_osd_set_state_t's object_count and change PG state accordingly
 void osd_t::remove_object_from_state(object_id & oid, pg_osd_set_state_t *object_state, pg_t & pg)
 {
@@ -736,10 +215,14 @@ void osd_t::remove_object_from_state(object_id & oid, pg_osd_set_state_t *object
     {
         throw std::runtime_error("BUG: Invalid object state: "+std::to_string(object_state->state));
     }
-    object_state->object_count--;
-    if (!object_state->object_count)
+}
+
+void osd_t::free_object_state(pg_t & pg, pg_osd_set_state_t **object_state)
+{
+    if (*object_state && !(--(*object_state)->object_count))
     {
-        pg.state_dict.erase(object_state->osd_set);
+        pg.state_dict.erase((*object_state)->osd_set);
+        *object_state = NULL;
     }
 }
 
@@ -805,22 +288,21 @@ resume_5:
     else
     {
         remove_object_from_state(op_data->oid, op_data->object_state, pg);
+        free_object_state(pg, &op_data->object_state);
     }
     pg.total_count--;
-    object_id oid = op_data->oid;
-    finish_op(cur_op, cur_op->req.rw.len);
-    // Continue other write operations to the same object
-    auto next_it = pg.write_queue.find(oid);
-    auto this_it = next_it;
-    if (this_it != pg.write_queue.end() && this_it->second == cur_op)
+    osd_op_t *next_op = NULL;
+    auto next_it = pg.write_queue.find(op_data->oid);
+    if (next_it != pg.write_queue.end() && next_it->second == cur_op)
     {
-        next_it++;
-        pg.write_queue.erase(this_it);
-        if (next_it != pg.write_queue.end() &&
-            next_it->first == oid)
-        {
-            osd_op_t *next_op = next_it->second;
-            continue_primary_write(next_op);
-        }
+        pg.write_queue.erase(next_it++);
+        if (next_it != pg.write_queue.end() && next_it->first == op_data->oid)
+            next_op = next_it->second;
+    }
+    finish_op(cur_op, cur_op->req.rw.len);
+    if (next_op)
+    {
+        // Continue next write to the same object
+        continue_primary_write(next_op);
     }
 }

src/osd_primary.h

@@ -38,4 +38,8 @@ struct osd_primary_op_data_t
     osd_num_t *dirty_osds = NULL;
     int dirty_osd_count = 0;
     obj_ver_id *unstable_writes = NULL;
+    obj_ver_osd_t *copies_to_delete = NULL;
+    int copies_to_delete_count = 0;
 };
+
+bool contains_osd(osd_num_t *osd_set, uint64_t size, osd_num_t osd_num);

src/osd_primary_subops.cpp

@@ -47,6 +47,10 @@ void osd_t::finish_op(osd_op_t *cur_op, int retval)
             {
                 finish_stop_pg(pg);
             }
+            else if ((pg.state & PG_REPEERING) && pg.inflight == 0 && !pg.flush_batch)
+            {
+                start_pg_peering(pg);
+            }
         }
         assert(!cur_op->op_data->subops);
         assert(!cur_op->op_data->unstable_write_osds);
@@ -192,14 +196,7 @@ void osd_t::submit_primary_subops(int submit_type, uint64_t op_version, int pg_s
                 }
                 subops[i].callback = [cur_op, this](osd_op_t *subop)
                 {
-                    int fail_fd = subop->req.hdr.opcode == OSD_OP_SEC_WRITE &&
-                        subop->reply.hdr.retval != subop->req.sec_rw.len ? subop->peer_fd : -1;
                     handle_primary_subop(subop, cur_op);
-                    if (fail_fd >= 0)
-                    {
-                        // write operation failed, drop the connection
-                        c_cli.stop_client(fail_fd);
-                    }
                 };
                 c_cli.outbox_push(&subops[i]);
             }
@@ -245,6 +242,7 @@ void osd_t::handle_primary_bs_subop(osd_op_t *subop)
     }
     delete bs_op;
     subop->bs_op = NULL;
+    subop->peer_fd = -1;
     handle_primary_subop(subop, cur_op);
 }
 
@@ -286,6 +284,11 @@ void osd_t::handle_primary_subop(osd_op_t *subop, osd_op_t *cur_op)
             op_data->epipe++;
         }
         op_data->errors++;
+        if (subop->peer_fd >= 0)
+        {
+            // Drop connection on any error
+            c_cli.stop_client(subop->peer_fd);
+        }
     }
     else
     {
@@ -353,7 +356,7 @@ void osd_t::cancel_primary_write(osd_op_t *cur_op)
     }
 }
 
-static bool contains_osd(osd_num_t *osd_set, uint64_t size, osd_num_t osd_num)
+bool contains_osd(osd_num_t *osd_set, uint64_t size, osd_num_t osd_num)
 {
     for (uint64_t i = 0; i < size; i++)
     {
@@ -369,83 +372,81 @@ void osd_t::submit_primary_del_subops(osd_op_t *cur_op, osd_num_t *cur_set, uint
 {
     osd_primary_op_data_t *op_data = cur_op->op_data;
     bool rep = op_data->scheme == POOL_SCHEME_REPLICATED;
-    int extra_chunks = 0;
-    // ordered comparison for EC/XOR, unordered for replicated pools
+    obj_ver_osd_t extra_chunks[loc_set.size()];
+    int chunks_to_del = 0;
     for (auto & chunk: loc_set)
     {
-        if (!cur_set || (rep ? !contains_osd(cur_set, set_size, chunk.osd_num) : chunk.osd_num != cur_set[chunk.role]))
+        // ordered comparison for EC/XOR, unordered for replicated pools
+        if (!cur_set || (rep
+            ? !contains_osd(cur_set, set_size, chunk.osd_num)
+            : (chunk.osd_num != cur_set[chunk.role])))
         {
-            extra_chunks++;
+            extra_chunks[chunks_to_del++] = (obj_ver_osd_t){
+                .osd_num = chunk.osd_num,
+                .oid = {
+                    .inode = op_data->oid.inode,
+                    .stripe = op_data->oid.stripe | (rep ? 0 : chunk.role),
+                },
+                // Same version as write
+                .version = op_data->fact_ver,
+            };
         }
     }
-    op_data->n_subops = extra_chunks;
+    submit_primary_del_batch(cur_op, extra_chunks, chunks_to_del);
+}
+
+void osd_t::submit_primary_del_batch(osd_op_t *cur_op, obj_ver_osd_t *chunks_to_delete, int chunks_to_delete_count)
+{
+    osd_primary_op_data_t *op_data = cur_op->op_data;
+    op_data->n_subops = chunks_to_delete_count;
     op_data->done = op_data->errors = 0;
-    if (!extra_chunks)
+    if (!op_data->n_subops)
     {
         return;
     }
-    osd_op_t *subops = new osd_op_t[extra_chunks];
+    osd_op_t *subops = new osd_op_t[chunks_to_delete_count];
     op_data->subops = subops;
-    int i = 0;
-    for (auto & chunk: loc_set)
+    for (int i = 0; i < chunks_to_delete_count; i++)
     {
-        if (!cur_set || (rep ? !contains_osd(cur_set, set_size, chunk.osd_num) : chunk.osd_num != cur_set[chunk.role]))
+        auto & chunk = chunks_to_delete[i];
+        if (chunk.osd_num == this->osd_num)
         {
-            int stripe_num = op_data->scheme == POOL_SCHEME_REPLICATED ? 0 : chunk.role;
-            if (chunk.osd_num == this->osd_num)
-            {
-                clock_gettime(CLOCK_REALTIME, &subops[i].tv_begin);
-                subops[i].op_type = (uint64_t)cur_op;
-                subops[i].bs_op = new blockstore_op_t({
-                    .opcode = BS_OP_DELETE,
-                    .callback = [subop = &subops[i], this](blockstore_op_t *bs_subop)
-                    {
-                        handle_primary_bs_subop(subop);
-                    },
-                    .oid = {
-                        .inode = op_data->oid.inode,
-                        .stripe = op_data->oid.stripe | stripe_num,
-                    },
-                    // Same version as write
-                    .version = op_data->fact_ver,
-                });
-                bs->enqueue_op(subops[i].bs_op);
-            }
-            else
-            {
-                subops[i].op_type = OSD_OP_OUT;
-                subops[i].peer_fd = c_cli.osd_peer_fds.at(chunk.osd_num);
-                subops[i].req.sec_del = {
-                    .header = {
-                        .magic = SECONDARY_OSD_OP_MAGIC,
-                        .id = c_cli.next_subop_id++,
-                        .opcode = OSD_OP_SEC_DELETE,
-                    },
-                    .oid = {
-                        .inode = op_data->oid.inode,
-                        .stripe = op_data->oid.stripe | stripe_num,
-                    },
-                    // Same version as write
-                    .version = op_data->fact_ver,
-                };
-                subops[i].callback = [cur_op, this](osd_op_t *subop)
+            clock_gettime(CLOCK_REALTIME, &subops[i].tv_begin);
+            subops[i].op_type = (uint64_t)cur_op;
+            subops[i].bs_op = new blockstore_op_t({
+                .opcode = BS_OP_DELETE,
+                .callback = [subop = &subops[i], this](blockstore_op_t *bs_subop)
                 {
-                    int fail_fd = subop->reply.hdr.retval != 0 ? subop->peer_fd : -1;
-                    handle_primary_subop(subop, cur_op);
-                    if (fail_fd >= 0)
-                    {
-                        // delete operation failed, drop the connection
-                        c_cli.stop_client(fail_fd);
-                    }
-                };
-                c_cli.outbox_push(&subops[i]);
-            }
-            i++;
+                    handle_primary_bs_subop(subop);
+                },
+                .oid = chunk.oid,
+                .version = chunk.version,
+            });
+            bs->enqueue_op(subops[i].bs_op);
+        }
+        else
+        {
+            subops[i].op_type = OSD_OP_OUT;
+            subops[i].peer_fd = c_cli.osd_peer_fds.at(chunk.osd_num);
+            subops[i].req = (osd_any_op_t){ .sec_del = {
+                .header = {
+                    .magic = SECONDARY_OSD_OP_MAGIC,
+                    .id = c_cli.next_subop_id++,
+                    .opcode = OSD_OP_SEC_DELETE,
+                },
+                .oid = chunk.oid,
+                .version = chunk.version,
+            } };
+            subops[i].callback = [cur_op, this](osd_op_t *subop)
+            {
+                handle_primary_subop(subop, cur_op);
+            };
+            c_cli.outbox_push(&subops[i]);
         }
     }
 }
 
-void osd_t::submit_primary_sync_subops(osd_op_t *cur_op)
+int osd_t::submit_primary_sync_subops(osd_op_t *cur_op)
 {
     osd_primary_op_data_t *op_data = cur_op->op_data;
     int n_osds = op_data->dirty_osd_count;
@@ -453,6 +454,7 @@ void osd_t::submit_primary_sync_subops(osd_op_t *cur_op)
     op_data->done = op_data->errors = 0;
     op_data->n_subops = n_osds;
     op_data->subops = subops;
+    std::map<uint64_t, int>::iterator peer_it;
     for (int i = 0; i < n_osds; i++)
     {
         osd_num_t sync_osd = op_data->dirty_osds[i];
@@ -469,30 +471,35 @@ void osd_t::submit_primary_sync_subops(osd_op_t *cur_op)
             });
             bs->enqueue_op(subops[i].bs_op);
         }
-        else
+        else if ((peer_it = c_cli.osd_peer_fds.find(sync_osd)) != c_cli.osd_peer_fds.end())
         {
             subops[i].op_type = OSD_OP_OUT;
-            subops[i].peer_fd = c_cli.osd_peer_fds.at(sync_osd);
-            subops[i].req.sec_sync = {
+            subops[i].peer_fd = peer_it->second;
+            subops[i].req = (osd_any_op_t){ .sec_sync = {
                 .header = {
                     .magic = SECONDARY_OSD_OP_MAGIC,
                     .id = c_cli.next_subop_id++,
                     .opcode = OSD_OP_SEC_SYNC,
                 },
-            };
+            } };
             subops[i].callback = [cur_op, this](osd_op_t *subop)
             {
-                int fail_fd = subop->reply.hdr.retval != 0 ? subop->peer_fd : -1;
                 handle_primary_subop(subop, cur_op);
-                if (fail_fd >= 0)
-                {
-                    // sync operation failed, drop the connection
-                    c_cli.stop_client(fail_fd);
-                }
             };
             c_cli.outbox_push(&subops[i]);
         }
+        else
+        {
+            op_data->done++;
+        }
     }
+    if (op_data->done >= op_data->n_subops)
+    {
+        delete[] op_data->subops;
+        op_data->subops = NULL;
+        return 0;
+    }
+    return 1;
 }
 
 void osd_t::submit_primary_stab_subops(osd_op_t *cur_op)
@@ -525,24 +532,18 @@ void osd_t::submit_primary_stab_subops(osd_op_t *cur_op)
         {
             subops[i].op_type = OSD_OP_OUT;
             subops[i].peer_fd = c_cli.osd_peer_fds.at(stab_osd.osd_num);
-            subops[i].req.sec_stab = {
+            subops[i].req = (osd_any_op_t){ .sec_stab = {
                 .header = {
                     .magic = SECONDARY_OSD_OP_MAGIC,
                     .id = c_cli.next_subop_id++,
                     .opcode = OSD_OP_SEC_STABILIZE,
                 },
                 .len = (uint64_t)(stab_osd.len * sizeof(obj_ver_id)),
-            };
+            } };
             subops[i].iov.push_back(op_data->unstable_writes + stab_osd.start, stab_osd.len * sizeof(obj_ver_id));
             subops[i].callback = [cur_op, this](osd_op_t *subop)
             {
-                int fail_fd = subop->reply.hdr.retval != 0 ? subop->peer_fd : -1;
                 handle_primary_subop(subop, cur_op);
-                if (fail_fd >= 0)
-                {
-                    // sync operation failed, drop the connection
-                    c_cli.stop_client(fail_fd);
-                }
             };
             c_cli.outbox_push(&subops[i]);
         }
@@ -560,7 +561,7 @@ void osd_t::pg_cancel_write_queue(pg_t & pg, osd_op_t *first_op, object_id oid,
         return;
     }
     std::vector<osd_op_t*> cancel_ops;
-    while (it != pg.write_queue.end())
+    while (it != pg.write_queue.end() && it->first == oid)
     {
         cancel_ops.push_back(it->second);
         it++;

src/osd_primary_sync.cpp

@@ -0,0 +1,265 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 (see README.md for details)
+
+#include "osd_primary.h"
+
+// Save and clear unstable_writes -> SYNC all -> STABLE all
+void osd_t::continue_primary_sync(osd_op_t *cur_op)
+{
+    if (!cur_op->op_data)
+    {
+        cur_op->op_data = (osd_primary_op_data_t*)calloc_or_die(1, sizeof(osd_primary_op_data_t));
+    }
+    osd_primary_op_data_t *op_data = cur_op->op_data;
+    if (op_data->st == 1)      goto resume_1;
+    else if (op_data->st == 2) goto resume_2;
+    else if (op_data->st == 3) goto resume_3;
+    else if (op_data->st == 4) goto resume_4;
+    else if (op_data->st == 5) goto resume_5;
+    else if (op_data->st == 6) goto resume_6;
+    else if (op_data->st == 7) goto resume_7;
+    else if (op_data->st == 8) goto resume_8;
+    assert(op_data->st == 0);
+    if (syncs_in_progress.size() > 0)
+    {
+        // Wait for previous syncs, if any
+        // FIXME: We may try to execute the current one in parallel, like in Blockstore, but I'm not sure if it matters at all
+        syncs_in_progress.push_back(cur_op);
+        op_data->st = 1;
+resume_1:
+        return;
+    }
+    else
+    {
+        syncs_in_progress.push_back(cur_op);
+    }
+resume_2:
+    if (dirty_osds.size() == 0)
+    {
+        // Nothing to sync
+        goto finish;
+    }
+    // Save and clear unstable_writes
+    // In theory it is possible to do in on a per-client basis, but this seems to be an unnecessary complication
+    // It would be cool not to copy these here at all, but someone has to deduplicate them by object IDs anyway
+    if (unstable_writes.size() > 0)
+    {
+        op_data->unstable_write_osds = new std::vector<unstable_osd_num_t>();
+        op_data->unstable_writes = new obj_ver_id[this->unstable_writes.size()];
+        osd_num_t last_osd = 0;
+        int last_start = 0, last_end = 0;
+        for (auto it = this->unstable_writes.begin(); it != this->unstable_writes.end(); it++)
+        {
+            if (last_osd != it->first.osd_num)
+            {
+                if (last_osd != 0)
+                {
+                    op_data->unstable_write_osds->push_back((unstable_osd_num_t){
+                        .osd_num = last_osd,
+                        .start = last_start,
+                        .len = last_end - last_start,
+                    });
+                }
+                last_osd = it->first.osd_num;
+                last_start = last_end;
+            }
+            op_data->unstable_writes[last_end] = (obj_ver_id){
+                .oid = it->first.oid,
+                .version = it->second,
+            };
+            last_end++;
+        }
+        if (last_osd != 0)
+        {
+            op_data->unstable_write_osds->push_back((unstable_osd_num_t){
+                .osd_num = last_osd,
+                .start = last_start,
+                .len = last_end - last_start,
+            });
+        }
+        this->unstable_writes.clear();
+    }
+    {
+        void *dirty_buf = malloc_or_die(
+            sizeof(pool_pg_num_t)*dirty_pgs.size() +
+            sizeof(osd_num_t)*dirty_osds.size() +
+            sizeof(obj_ver_osd_t)*this->copies_to_delete_after_sync_count
+        );
+        op_data->dirty_pgs = (pool_pg_num_t*)dirty_buf;
+        op_data->dirty_osds = (osd_num_t*)(dirty_buf + sizeof(pool_pg_num_t)*dirty_pgs.size());
+        op_data->dirty_pg_count = dirty_pgs.size();
+        op_data->dirty_osd_count = dirty_osds.size();
+        if (this->copies_to_delete_after_sync_count)
+        {
+            op_data->copies_to_delete_count = 0;
+            op_data->copies_to_delete = (obj_ver_osd_t*)(op_data->dirty_osds + op_data->dirty_osd_count);
+            for (auto dirty_pg_num: dirty_pgs)
+            {
+                auto & pg = pgs.at(dirty_pg_num);
+                assert(pg.copies_to_delete_after_sync.size() <= this->copies_to_delete_after_sync_count);
+                memcpy(
+                    op_data->copies_to_delete + op_data->copies_to_delete_count,
+                    pg.copies_to_delete_after_sync.data(),
+                    sizeof(obj_ver_osd_t)*pg.copies_to_delete_after_sync.size()
+                );
+                op_data->copies_to_delete_count += pg.copies_to_delete_after_sync.size();
+                this->copies_to_delete_after_sync_count -= pg.copies_to_delete_after_sync.size();
+                pg.copies_to_delete_after_sync.clear();
+            }
+            assert(this->copies_to_delete_after_sync_count == 0);
+        }
+        int dpg = 0;
+        for (auto dirty_pg_num: dirty_pgs)
+        {
+            pgs.at(dirty_pg_num).inflight++;
+            op_data->dirty_pgs[dpg++] = dirty_pg_num;
+        }
+        dirty_pgs.clear();
+        dpg = 0;
+        for (auto osd_num: dirty_osds)
+        {
+            op_data->dirty_osds[dpg++] = osd_num;
+        }
+        dirty_osds.clear();
+    }
+    if (immediate_commit != IMMEDIATE_ALL)
+    {
+        // SYNC
+        if (!submit_primary_sync_subops(cur_op))
+        {
+            goto resume_4;
+        }
+resume_3:
+        op_data->st = 3;
+        return;
+resume_4:
+        if (op_data->errors > 0)
+        {
+            goto resume_6;
+        }
+    }
+    if (op_data->unstable_writes)
+    {
+        // Stabilize version sets, if any
+        submit_primary_stab_subops(cur_op);
+resume_5:
+        op_data->st = 5;
+        return;
+    }
+resume_6:
+    if (op_data->errors > 0)
+    {
+        // Return PGs and OSDs back into their dirty sets
+        for (int i = 0; i < op_data->dirty_pg_count; i++)
+        {
+            dirty_pgs.insert(op_data->dirty_pgs[i]);
+        }
+        for (int i = 0; i < op_data->dirty_osd_count; i++)
+        {
+            dirty_osds.insert(op_data->dirty_osds[i]);
+        }
+        if (op_data->unstable_writes)
+        {
+            // Return objects back into the unstable write set
+            for (auto unstable_osd: *(op_data->unstable_write_osds))
+            {
+                for (int i = 0; i < unstable_osd.len; i++)
+                {
+                    // Except those from peered PGs
+                    auto & w = op_data->unstable_writes[i];
+                    pool_pg_num_t wpg = {
+                        .pool_id = INODE_POOL(w.oid.inode),
+                        .pg_num = map_to_pg(w.oid, st_cli.pool_config.at(INODE_POOL(w.oid.inode)).pg_stripe_size),
+                    };
+                    if (pgs.at(wpg).state & PG_ACTIVE)
+                    {
+                        uint64_t & dest = this->unstable_writes[(osd_object_id_t){
+                            .osd_num = unstable_osd.osd_num,
+                            .oid = w.oid,
+                        }];
+                        dest = dest < w.version ? w.version : dest;
+                        dirty_pgs.insert(wpg);
+                    }
+                }
+            }
+        }
+        if (op_data->copies_to_delete)
+        {
+            // Return 'copies to delete' back into respective PGs
+            for (int i = 0; i < op_data->copies_to_delete_count; i++)
+            {
+                auto & w = op_data->copies_to_delete[i];
+                auto & pg = pgs.at((pool_pg_num_t){
+                    .pool_id = INODE_POOL(w.oid.inode),
+                    .pg_num = map_to_pg(w.oid, st_cli.pool_config.at(INODE_POOL(w.oid.inode)).pg_stripe_size),
+                });
+                if (pg.state & PG_ACTIVE)
+                {
+                    pg.copies_to_delete_after_sync.push_back(w);
+                    copies_to_delete_after_sync_count++;
+                }
+            }
+        }
+    }
+    else if (op_data->copies_to_delete)
+    {
+        // Actually delete copies which we wanted to delete
+        submit_primary_del_batch(cur_op, op_data->copies_to_delete, op_data->copies_to_delete_count);
+resume_7:
+        op_data->st = 7;
+        return;
+resume_8:
+        if (op_data->errors > 0)
+        {
+            goto resume_6;
+        }
+    }
+    for (int i = 0; i < op_data->dirty_pg_count; i++)
+    {
+        auto & pg = pgs.at(op_data->dirty_pgs[i]);
+        pg.inflight--;
+        if ((pg.state & PG_STOPPING) && pg.inflight == 0 && !pg.flush_batch)
+        {
+            finish_stop_pg(pg);
+        }
+        else if ((pg.state & PG_REPEERING) && pg.inflight == 0 && !pg.flush_batch)
+        {
+            start_pg_peering(pg);
+        }
+    }
+    // FIXME: Free those in the destructor?
+    free(op_data->dirty_pgs);
+    op_data->dirty_pgs = NULL;
+    op_data->dirty_osds = NULL;
+    if (op_data->unstable_writes)
+    {
+        delete op_data->unstable_write_osds;
+        delete[] op_data->unstable_writes;
+        op_data->unstable_writes = NULL;
+        op_data->unstable_write_osds = NULL;
+    }
+    if (op_data->errors > 0)
+    {
+        finish_op(cur_op, op_data->epipe > 0 ? -EPIPE : -EIO);
+    }
+    else
+    {
+finish:
+        if (cur_op->peer_fd)
+        {
+            auto it = c_cli.clients.find(cur_op->peer_fd);
+            if (it != c_cli.clients.end())
+                it->second->dirty_pgs.clear();
+        }
+        finish_op(cur_op, 0);
+    }
+    assert(syncs_in_progress.front() == cur_op);
+    syncs_in_progress.pop_front();
+    if (syncs_in_progress.size() > 0)
+    {
+        cur_op = syncs_in_progress.front();
+        op_data = cur_op->op_data;
+        op_data->st++;
+        goto resume_2;
+    }
+}

src/osd_primary_write.cpp

@@ -0,0 +1,378 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 (see README.md for details)
+
+#include "osd_primary.h"
+#include "allocator.h"
+
+bool osd_t::check_write_queue(osd_op_t *cur_op, pg_t & pg)
+{
+    osd_primary_op_data_t *op_data = cur_op->op_data;
+    // Check if actions are pending for this object
+    auto act_it = pg.flush_actions.lower_bound((obj_piece_id_t){
+        .oid = op_data->oid,
+        .osd_num = 0,
+    });
+    if (act_it != pg.flush_actions.end() &&
+        act_it->first.oid.inode == op_data->oid.inode &&
+        (act_it->first.oid.stripe & ~STRIPE_MASK) == op_data->oid.stripe)
+    {
+        pg.write_queue.emplace(op_data->oid, cur_op);
+        return false;
+    }
+    // Check if there are other write requests to the same object
+    auto vo_it = pg.write_queue.find(op_data->oid);
+    if (vo_it != pg.write_queue.end())
+    {
+        op_data->st = 1;
+        pg.write_queue.emplace(op_data->oid, cur_op);
+        return false;
+    }
+    pg.write_queue.emplace(op_data->oid, cur_op);
+    return true;
+}
+
+void osd_t::continue_primary_write(osd_op_t *cur_op)
+{
+    if (!cur_op->op_data && !prepare_primary_rw(cur_op))
+    {
+        return;
+    }
+    osd_primary_op_data_t *op_data = cur_op->op_data;
+    auto & pg = pgs.at({ .pool_id = INODE_POOL(op_data->oid.inode), .pg_num = op_data->pg_num });
+    if (op_data->st == 1)      goto resume_1;
+    else if (op_data->st == 2) goto resume_2;
+    else if (op_data->st == 3) goto resume_3;
+    else if (op_data->st == 4) goto resume_4;
+    else if (op_data->st == 5) goto resume_5;
+    else if (op_data->st == 6) goto resume_6;
+    else if (op_data->st == 7) goto resume_7;
+    else if (op_data->st == 8) goto resume_8;
+    else if (op_data->st == 9) goto resume_9;
+    else if (op_data->st == 10) goto resume_10;
+    assert(op_data->st == 0);
+    if (!check_write_queue(cur_op, pg))
+    {
+        return;
+    }
+resume_1:
+    // Determine blocks to read and write
+    // Missing chunks are allowed to be overwritten even in incomplete objects
+    // FIXME: Allow to do small writes to the old (degraded/misplaced) OSD set for lower performance impact
+    op_data->prev_set = get_object_osd_set(pg, op_data->oid, pg.cur_set.data(), &op_data->object_state);
+    if (op_data->scheme == POOL_SCHEME_REPLICATED)
+    {
+        // Simplified algorithm
+        op_data->stripes[0].write_start = op_data->stripes[0].req_start;
+        op_data->stripes[0].write_end = op_data->stripes[0].req_end;
+        op_data->stripes[0].write_buf = cur_op->buf;
+        if (pg.cur_set.data() != op_data->prev_set && (op_data->stripes[0].write_start != 0 ||
+            op_data->stripes[0].write_end != bs_block_size))
+        {
+            // Object is degraded/misplaced and will be moved to <write_osd_set>
+            op_data->stripes[0].read_start = 0;
+            op_data->stripes[0].read_end = bs_block_size;
+            cur_op->rmw_buf = op_data->stripes[0].read_buf = memalign_or_die(MEM_ALIGNMENT, bs_block_size);
+        }
+    }
+    else
+    {
+        cur_op->rmw_buf = calc_rmw(cur_op->buf, op_data->stripes, op_data->prev_set,
+            pg.pg_size, op_data->pg_data_size, pg.pg_cursize, pg.cur_set.data(), bs_block_size);
+        if (!cur_op->rmw_buf)
+        {
+            // Refuse partial overwrite of an incomplete object
+            cur_op->reply.hdr.retval = -EINVAL;
+            goto continue_others;
+        }
+    }
+    // Read required blocks
+    submit_primary_subops(SUBMIT_RMW_READ, UINT64_MAX, pg.pg_size, op_data->prev_set, cur_op);
+resume_2:
+    op_data->st = 2;
+    return;
+resume_3:
+    if (op_data->errors > 0)
+    {
+        pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
+        return;
+    }
+    if (op_data->scheme == POOL_SCHEME_REPLICATED)
+    {
+        // Only (possibly) copy new data from the request into the recovery buffer
+        if (pg.cur_set.data() != op_data->prev_set && (op_data->stripes[0].write_start != 0 ||
+            op_data->stripes[0].write_end != bs_block_size))
+        {
+            memcpy(
+                op_data->stripes[0].read_buf + op_data->stripes[0].req_start,
+                op_data->stripes[0].write_buf,
+                op_data->stripes[0].req_end - op_data->stripes[0].req_start
+            );
+            op_data->stripes[0].write_buf = op_data->stripes[0].read_buf;
+            op_data->stripes[0].write_start = 0;
+            op_data->stripes[0].write_end = bs_block_size;
+        }
+    }
+    else
+    {
+        // For EC/XOR pools, save version override to make it impossible
+        // for parallel reads to read different versions of data and parity
+        pg.ver_override[op_data->oid] = op_data->fact_ver;
+        // Recover missing stripes, calculate parity
+        if (pg.scheme == POOL_SCHEME_XOR)
+        {
+            calc_rmw_parity_xor(op_data->stripes, pg.pg_size, op_data->prev_set, pg.cur_set.data(), bs_block_size);
+        }
+        else if (pg.scheme == POOL_SCHEME_JERASURE)
+        {
+            calc_rmw_parity_jerasure(op_data->stripes, pg.pg_size, op_data->pg_data_size, op_data->prev_set, pg.cur_set.data(), bs_block_size);
+        }
+    }
+    // Send writes
+    if ((op_data->fact_ver >> (64-PG_EPOCH_BITS)) < pg.epoch)
+    {
+        op_data->target_ver = ((uint64_t)pg.epoch << (64-PG_EPOCH_BITS)) | 1;
+    }
+    else
+    {
+        if ((op_data->fact_ver & (1ul<<(64-PG_EPOCH_BITS) - 1)) == (1ul<<(64-PG_EPOCH_BITS) - 1))
+        {
+            assert(pg.epoch != ((1ul << PG_EPOCH_BITS)-1));
+            pg.epoch++;
+        }
+        op_data->target_ver = op_data->fact_ver + 1;
+    }
+    if (pg.epoch > pg.reported_epoch)
+    {
+        // Report newer epoch before writing
+        // FIXME: We may report only one PG state here...
+        this->pg_state_dirty.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
+        pg.history_changed = true;
+        report_pg_states();
+resume_10:
+        if (pg.epoch > pg.reported_epoch)
+        {
+            op_data->st = 10;
+            return;
+        }
+    }
+    submit_primary_subops(SUBMIT_WRITE, op_data->target_ver, pg.pg_size, pg.cur_set.data(), cur_op);
+resume_4:
+    op_data->st = 4;
+    return;
+resume_5:
+    if (op_data->scheme != POOL_SCHEME_REPLICATED)
+    {
+        // Remove version override just after the write, but before stabilizing
+        pg.ver_override.erase(op_data->oid);
+    }
+    if (op_data->errors > 0)
+    {
+        pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
+        return;
+    }
+    if (op_data->object_state)
+    {
+        // We must forget the unclean state of the object before deleting it
+        // so the next reads don't accidentally read a deleted version
+        // And it should be done at the same time as the removal of the version override
+        remove_object_from_state(op_data->oid, op_data->object_state, pg);
+        pg.clean_count++;
+    }
+resume_6:
+resume_7:
+    if (!remember_unstable_write(cur_op, pg, pg.cur_loc_set, 6))
+    {
+        return;
+    }
+    if (op_data->fact_ver == 1)
+    {
+        // Object is created
+        pg.clean_count++;
+        pg.total_count++;
+    }
+    if (op_data->object_state)
+    {
+        {
+            int recovery_type = op_data->object_state->state & (OBJ_DEGRADED|OBJ_INCOMPLETE) ? 0 : 1;
+            recovery_stat_count[0][recovery_type]++;
+            if (!recovery_stat_count[0][recovery_type])
+            {
+                recovery_stat_count[0][recovery_type]++;
+                recovery_stat_bytes[0][recovery_type] = 0;
+            }
+            for (int role = 0; role < (op_data->scheme == POOL_SCHEME_REPLICATED ? 1 : pg.pg_size); role++)
+            {
+                recovery_stat_bytes[0][recovery_type] += op_data->stripes[role].write_end - op_data->stripes[role].write_start;
+            }
+        }
+        // Any kind of a non-clean object can have extra chunks, because we don't record objects
+        // as degraded & misplaced or incomplete & misplaced at the same time. So try to remove extra chunks
+        if (immediate_commit != IMMEDIATE_ALL)
+        {
+            // We can't remove extra chunks yet if fsyncs are explicit, because
+            // new copies may not be committed to stable storage yet
+            // We can only remove extra chunks after a successful SYNC for this PG
+            for (auto & chunk: op_data->object_state->osd_set)
+            {
+                // Check is the same as in submit_primary_del_subops()
+                if (op_data->scheme == POOL_SCHEME_REPLICATED
+                    ? !contains_osd(pg.cur_set.data(), pg.pg_size, chunk.osd_num)
+                    : (chunk.osd_num != pg.cur_set[chunk.role]))
+                {
+                    pg.copies_to_delete_after_sync.push_back((obj_ver_osd_t){
+                        .osd_num = chunk.osd_num,
+                        .oid = {
+                            .inode = op_data->oid.inode,
+                            .stripe = op_data->oid.stripe | (op_data->scheme == POOL_SCHEME_REPLICATED ? 0 : chunk.role),
+                        },
+                        .version = op_data->fact_ver,
+                    });
+                    copies_to_delete_after_sync_count++;
+                }
+            }
+            free_object_state(pg, &op_data->object_state);
+        }
+        else
+        {
+            submit_primary_del_subops(cur_op, pg.cur_set.data(), pg.pg_size, op_data->object_state->osd_set);
+            free_object_state(pg, &op_data->object_state);
+            if (op_data->n_subops > 0)
+            {
+resume_8:
+                op_data->st = 8;
+                return;
+resume_9:
+                if (op_data->errors > 0)
+                {
+                    pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
+                    return;
+                }
+            }
+        }
+    }
+    cur_op->reply.hdr.retval = cur_op->req.rw.len;
+continue_others:
+    osd_op_t *next_op = NULL;
+    auto next_it = pg.write_queue.find(op_data->oid);
+    // Remove the operation from queue before calling finish_op so it doesn't see the completed operation in queue
+    if (next_it != pg.write_queue.end() && next_it->second == cur_op)
+    {
+        pg.write_queue.erase(next_it++);
+        if (next_it != pg.write_queue.end() && next_it->first == op_data->oid)
+            next_op = next_it->second;
+    }
+    // finish_op would invalidate next_it if it cleared pg.write_queue, but it doesn't do that :)
+    finish_op(cur_op, cur_op->req.rw.len);
+    if (next_op)
+    {
+        // Continue next write to the same object
+        continue_primary_write(next_op);
+    }
+}
+
+bool osd_t::remember_unstable_write(osd_op_t *cur_op, pg_t & pg, pg_osd_set_t & loc_set, int base_state)
+{
+    osd_primary_op_data_t *op_data = cur_op->op_data;
+    if (op_data->st == base_state)
+    {
+        goto resume_6;
+    }
+    else if (op_data->st == base_state+1)
+    {
+        goto resume_7;
+    }
+    if (immediate_commit == IMMEDIATE_ALL)
+    {
+immediate:
+        if (op_data->scheme != POOL_SCHEME_REPLICATED)
+        {
+            // Send STABILIZE ops immediately
+            op_data->unstable_write_osds = new std::vector<unstable_osd_num_t>();
+            op_data->unstable_writes = new obj_ver_id[loc_set.size()];
+            {
+                int last_start = 0;
+                for (auto & chunk: loc_set)
+                {
+                    op_data->unstable_writes[last_start] = (obj_ver_id){
+                        .oid = {
+                            .inode = op_data->oid.inode,
+                            .stripe = op_data->oid.stripe | chunk.role,
+                        },
+                        .version = op_data->fact_ver,
+                    };
+                    op_data->unstable_write_osds->push_back((unstable_osd_num_t){
+                        .osd_num = chunk.osd_num,
+                        .start = last_start,
+                        .len = 1,
+                    });
+                    last_start++;
+                }
+            }
+            submit_primary_stab_subops(cur_op);
+resume_6:
+            op_data->st = 6;
+            return false;
+resume_7:
+            // FIXME: Free those in the destructor?
+            delete op_data->unstable_write_osds;
+            delete[] op_data->unstable_writes;
+            op_data->unstable_writes = NULL;
+            op_data->unstable_write_osds = NULL;
+            if (op_data->errors > 0)
+            {
+                pg_cancel_write_queue(pg, cur_op, op_data->oid, op_data->epipe > 0 ? -EPIPE : -EIO);
+                return false;
+            }
+        }
+    }
+    else if (immediate_commit == IMMEDIATE_SMALL)
+    {
+        int stripe_count = (op_data->scheme == POOL_SCHEME_REPLICATED ? 1 : op_data->pg_size);
+        for (int role = 0; role < stripe_count; role++)
+        {
+            if (op_data->stripes[role].write_start == 0 &&
+                op_data->stripes[role].write_end == bs_block_size)
+            {
+                // Big write. Treat write as unsynced
+                goto lazy;
+            }
+        }
+        goto immediate;
+    }
+    else
+    {
+lazy:
+        if (op_data->scheme != POOL_SCHEME_REPLICATED)
+        {
+            // Remember version as unstable for EC/XOR
+            for (auto & chunk: loc_set)
+            {
+                this->dirty_osds.insert(chunk.osd_num);
+                this->unstable_writes[(osd_object_id_t){
+                    .osd_num = chunk.osd_num,
+                    .oid = {
+                        .inode = op_data->oid.inode,
+                        .stripe = op_data->oid.stripe | chunk.role,
+                    },
+                }] = op_data->fact_ver;
+            }
+        }
+        else
+        {
+            // Only remember to sync OSDs for replicated pools
+            for (auto & chunk: loc_set)
+            {
+                this->dirty_osds.insert(chunk.osd_num);
+            }
+        }
+        // Remember PG as dirty to drop the connection when PG goes offline
+        // (this is required because of the "lazy sync")
+        auto cl_it = c_cli.clients.find(cur_op->peer_fd);
+        if (cl_it != c_cli.clients.end())
+        {
+            cl_it->second->dirty_pgs.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
+        }
+        dirty_pgs.insert({ .pool_id = pg.pool_id, .pg_num = pg.pg_num });
+    }
+    return true;
+}

src/pg_states.cpp

@@ -3,13 +3,14 @@
 
 #include "pg_states.h"
 
-const int pg_state_bit_count = 14;
+const int pg_state_bit_count = 15;
 
-const int pg_state_bits[14] = {
+const int pg_state_bits[15] = {
     PG_STARTING,
     PG_PEERING,
     PG_INCOMPLETE,
     PG_ACTIVE,
+    PG_REPEERING,
     PG_STOPPING,
     PG_OFFLINE,
     PG_DEGRADED,
@@ -21,11 +22,12 @@ const int pg_state_bits[14] = {
     PG_LEFT_ON_DEAD,
 };
 
-const char *pg_state_names[14] = {
+const char *pg_state_names[15] = {
     "starting",
     "peering",
     "incomplete",
     "active",
+    "repeering",
     "stopping",
     "offline",
     "degraded",

src/pg_states.h

@@ -10,16 +10,17 @@
 #define PG_PEERING (1<<1)
 #define PG_INCOMPLETE (1<<2)
 #define PG_ACTIVE (1<<3)
-#define PG_STOPPING (1<<4)
-#define PG_OFFLINE (1<<5)
+#define PG_REPEERING (1<<4)
+#define PG_STOPPING (1<<5)
+#define PG_OFFLINE (1<<6)
 // Plus any of these:
-#define PG_DEGRADED (1<<6)
-#define PG_HAS_INCOMPLETE (1<<7)
-#define PG_HAS_DEGRADED (1<<8)
-#define PG_HAS_MISPLACED (1<<9)
-#define PG_HAS_UNCLEAN (1<<10)
-#define PG_HAS_INVALID (1<<11)
-#define PG_LEFT_ON_DEAD (1<<12)
+#define PG_DEGRADED (1<<7)
+#define PG_HAS_INCOMPLETE (1<<8)
+#define PG_HAS_DEGRADED (1<<9)
+#define PG_HAS_MISPLACED (1<<10)
+#define PG_HAS_UNCLEAN (1<<11)
+#define PG_HAS_INVALID (1<<12)
+#define PG_LEFT_ON_DEAD (1<<13)
 
 // Lower bits that represent object role (EC 0/1/2... or always 0 with replication)
 // 12 bits is a safe default that doesn't depend on pg_stripe_size or pg_block_size

src/qemu_proxy.cpp

@@ -47,7 +47,6 @@ public:
 
     ~QemuProxy()
     {
-        cli->stop();
         delete cli;
         delete tfd;
     }

src/test_allocator.cpp

@@ -2,20 +2,47 @@
 // License: VNPL-1.1 (see README.md for details)
 
 #include <stdio.h>
+#include <stdlib.h>
 #include "allocator.h"
 
+void alloc_all(int size)
+{
+    allocator *a = new allocator(size);
+    for (int i = 0; i < size; i++)
+    {
+        uint64_t x = a->find_free();
+        if (x == UINT64_MAX)
+        {
+            printf("ran out of space %d allocated=%d\n", size, i);
+            exit(1);
+        }
+        if (x != i)
+        {
+            printf("incorrect block allocated: expected %d, got %lu\n", i, x);
+        }
+        if (a->get(x))
+        {
+            printf("not free before set at %d\n", i);
+        }
+        a->set(x, true);
+        if (!a->get(x))
+        {
+            printf("free after set at %d\n", i);
+        }
+    }
+    uint64_t x = a->find_free();
+    if (x != UINT64_MAX)
+    {
+        printf("extra free space found: %lx (%d)\n", x, size);
+        exit(1);
+    }
+    delete a;
+}
+
 int main(int narg, char *args[])
 {
-    allocator a(8192);
-    for (int i = 0; i < 8192; i++)
-    {
-        uint64_t x = a.find_free();
-        if (x == UINT64_MAX)
-        {
-            printf("ran out of space %d\n", i);
-            return 1;
-        }
-        a.set(x, true);
-    }
+    alloc_all(8192);
+    alloc_all(8062);
+    alloc_all(4096);
     return 0;
 }

src/test_cluster_client.cpp

@@ -0,0 +1,383 @@
+// Copyright (c) Vitaliy Filippov, 2019+
+// License: VNPL-1.1 (see README.md for details)
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <assert.h>
+#include "cluster_client.h"
+
+void configure_single_pg_pool(cluster_client_t *cli)
+{
+    cli->st_cli.on_load_pgs_hook(true);
+    cli->st_cli.parse_state((json_kv_t){
+        .key = "/config/pools",
+        .value = json11::Json::object {
+            { "1", json11::Json::object {
+                { "name", "hddpool" },
+                { "scheme", "replicated" },
+                { "pg_size", 2 },
+                { "pg_minsize", 1 },
+                { "pg_count", 1 },
+                { "failure_domain", "osd" },
+            } }
+        },
+    });
+    cli->st_cli.parse_state((json_kv_t){
+        .key = "/config/pgs",
+        .value = json11::Json::object {
+            { "items", json11::Json::object {
+                { "1", json11::Json::object {
+                    { "1", json11::Json::object {
+                        { "osd_set", json11::Json::array { 1, 2 } },
+                        { "primary", 1 },
+                    } }
+                } }
+            } }
+        },
+    });
+    cli->st_cli.parse_state((json_kv_t){
+        .key = "/pg/state/1/1",
+        .value = json11::Json::object {
+            { "peers", json11::Json::array { 1, 2 } },
+            { "primary", 1 },
+            { "state", json11::Json::array { "active" } },
+        },
+    });
+    json11::Json::object changes;
+    cli->st_cli.on_change_hook(changes);
+}
+
+int *test_write(cluster_client_t *cli, uint64_t offset, uint64_t len, uint8_t c)
+{
+    printf("Post write %lx+%lx\n", offset, len);
+    int *r = new int;
+    *r = -1;
+    cluster_op_t *op = new cluster_op_t();
+    op->opcode = OSD_OP_WRITE;
+    op->inode = 0x1000000000001;
+    op->offset = offset;
+    op->len = len;
+    op->iov.push_back(malloc_or_die(len), len);
+    memset(op->iov.buf[0].iov_base, c, len);
+    op->callback = [r](cluster_op_t *op)
+    {
+        if (*r == -1)
+            printf("Error: Not allowed to complete yet\n");
+        assert(*r != -1);
+        *r = op->retval == op->len ? 1 : 0;
+        free(op->iov.buf[0].iov_base);
+        printf("Done write %lx+%lx r=%d\n", op->offset, op->len, op->retval);
+        delete op;
+    };
+    cli->execute(op);
+    return r;
+}
+
+int *test_sync(cluster_client_t *cli)
+{
+    printf("Post sync\n");
+    int *r = new int;
+    *r = -1;
+    cluster_op_t *op = new cluster_op_t();
+    op->opcode = OSD_OP_SYNC;
+    op->callback = [r](cluster_op_t *op)
+    {
+        if (*r == -1)
+            printf("Error: Not allowed to complete yet\n");
+        assert(*r != -1);
+        *r = op->retval == 0 ? 1 : 0;
+        printf("Done sync r=%d\n", op->retval);
+        delete op;
+    };
+    cli->execute(op);
+    return r;
+}
+
+void can_complete(int *r)
+{
+    // Allow the operation to proceed so the test verifies
+    // that it doesn't complete earlier than expected
+    *r = -2;
+}
+
+void check_completed(int *r)
+{
+    assert(*r == 1);
+    delete r;
+}
+
+void pretend_connected(cluster_client_t *cli, osd_num_t osd_num)
+{
+    printf("OSD %lu connected\n", osd_num);
+    int peer_fd = cli->msgr.clients.size() ? std::prev(cli->msgr.clients.end())->first+1 : 10;
+    cli->msgr.osd_peer_fds[osd_num] = peer_fd;
+    cli->msgr.clients[peer_fd] = new osd_client_t();
+    cli->msgr.clients[peer_fd]->osd_num = osd_num;
+    cli->msgr.clients[peer_fd]->peer_state = PEER_CONNECTED;
+    cli->msgr.wanted_peers.erase(osd_num);
+    cli->msgr.repeer_pgs(osd_num);
+}
+
+void pretend_disconnected(cluster_client_t *cli, osd_num_t osd_num)
+{
+    printf("OSD %lu disconnected\n", osd_num);
+    cli->msgr.stop_client(cli->msgr.osd_peer_fds.at(osd_num));
+}
+
+void check_disconnected(cluster_client_t *cli, osd_num_t osd_num)
+{
+    if (cli->msgr.osd_peer_fds.find(osd_num) != cli->msgr.osd_peer_fds.end())
+    {
+        printf("OSD %lu not disconnected as it ought to be\n", osd_num);
+        assert(0);
+    }
+}
+
+void check_op_count(cluster_client_t *cli, osd_num_t osd_num, int ops)
+{
+    int peer_fd = cli->msgr.osd_peer_fds.at(osd_num);
+    int real_ops = cli->msgr.clients[peer_fd]->sent_ops.size();
+    if (real_ops != ops)
+    {
+        printf("error: %d ops expected, but %d queued\n", ops, real_ops);
+        assert(0);
+    }
+}
+
+osd_op_t *find_op(cluster_client_t *cli, osd_num_t osd_num, uint64_t opcode, uint64_t offset, uint64_t len)
+{
+    int peer_fd = cli->msgr.osd_peer_fds.at(osd_num);
+    auto op_it = cli->msgr.clients[peer_fd]->sent_ops.begin();
+    while (op_it != cli->msgr.clients[peer_fd]->sent_ops.end())
+    {
+        auto op = op_it->second;
+        if (op->req.hdr.opcode == opcode && (opcode == OSD_OP_SYNC ||
+            op->req.rw.inode == 0x1000000000001 && op->req.rw.offset == offset && op->req.rw.len == len))
+        {
+            return op;
+        }
+        op_it++;
+    }
+    return NULL;
+}
+
+void pretend_op_completed(cluster_client_t *cli, osd_op_t *op, int64_t retval)
+{
+    assert(op);
+    printf("Pretend completed %s %lx+%x\n", op->req.hdr.opcode == OSD_OP_SYNC
+        ? "sync" : (op->req.hdr.opcode == OSD_OP_WRITE ? "write" : "read"), op->req.rw.offset, op->req.rw.len);
+    uint64_t op_id = op->req.hdr.id;
+    int peer_fd = op->peer_fd;
+    cli->msgr.clients[peer_fd]->sent_ops.erase(op_id);
+    op->reply.hdr.magic = SECONDARY_OSD_REPLY_MAGIC;
+    op->reply.hdr.id = op->req.hdr.id;
+    op->reply.hdr.opcode = op->req.hdr.opcode;
+    op->reply.hdr.retval = retval < 0 ? retval : (op->req.hdr.opcode == OSD_OP_SYNC ? 0 : op->req.rw.len);
+    // Copy lambda to be unaffected by `delete op`
+    std::function<void(osd_op_t*)>(op->callback)(op);
+}
+
+void test1()
+{
+    json11::Json config;
+    timerfd_manager_t *tfd = new timerfd_manager_t([](int fd, bool wr, std::function<void(int, int)> callback){});
+    cluster_client_t *cli = new cluster_client_t(NULL, tfd, config);
+
+    int *r1 = test_write(cli, 0, 4096, 0x55);
+    configure_single_pg_pool(cli);
+    pretend_connected(cli, 1);
+    cli->continue_ops(true);
+    can_complete(r1);
+    check_op_count(cli, 1, 1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0, 4096), 0);
+    check_completed(r1);
+    pretend_disconnected(cli, 1);
+    int *r2 = test_sync(cli);
+    pretend_connected(cli, 1);
+    check_op_count(cli, 1, 0);
+    cli->continue_ops(true);
+    check_op_count(cli, 1, 1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0, 4096), 0);
+    check_op_count(cli, 1, 1);
+    can_complete(r2);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_SYNC, 0, 0), 0);
+    check_completed(r2);
+    // Check that the client doesn't repeat operations once more
+    pretend_disconnected(cli, 1);
+    pretend_connected(cli, 1);
+    check_op_count(cli, 1, 0);
+
+    // Case:
+    // Write(1) -> Complete Write(1) -> Overwrite(2) -> Complete Write(2)
+    // -> Overwrite(3) -> Drop OSD connection -> Reestablish OSD connection
+    // -> Complete All Posted Writes -> Sync -> Complete Sync
+    // The resulting state of the block must be (3) over (2) over (1).
+    // I.e. the part overwritten by (3) must remain as in (3) and so on.
+
+    // More interesting case:
+    // Same, but both Write(2) and Write(3) must consist of two parts:
+    // one from an OSD 2 that drops connection and other from OSD 1 that doesn't.
+    // The idea is that if the whole Write(2) is repeated when OSD 2 drops connection
+    // then it may also overwrite a part in OSD 1 which shouldn't be overwritten.
+
+    // Another interesting case:
+    // A new operation added during replay (would also break with the previous implementation)
+
+    r1 = test_write(cli, 0, 0x10000, 0x56);
+    can_complete(r1);
+    check_op_count(cli, 1, 1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0, 0x10000), 0);
+    check_completed(r1);
+
+    r1 = test_write(cli, 0xE000, 0x4000, 0x57);
+    can_complete(r1);
+    check_op_count(cli, 1, 1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0xE000, 0x4000), 0);
+    check_completed(r1);
+
+    r1 = test_write(cli, 0x10000, 0x4000, 0x58);
+
+    pretend_disconnected(cli, 1);
+    pretend_connected(cli, 1);
+    cli->continue_ops(true);
+
+    // Check replay
+    {
+        uint64_t replay_start = UINT64_MAX;
+        uint64_t replay_end = 0;
+        std::vector<osd_op_t*> replay_ops;
+        auto osd_cl = cli->msgr.clients.at(cli->msgr.osd_peer_fds.at(1));
+        for (auto & op_p: osd_cl->sent_ops)
+        {
+            auto op = op_p.second;
+            assert(op->req.hdr.opcode == OSD_OP_WRITE);
+            uint64_t offset = op->req.rw.offset;
+            if (op->req.rw.offset < replay_start)
+                replay_start = op->req.rw.offset;
+            if (op->req.rw.offset+op->req.rw.len > replay_end)
+                replay_end = op->req.rw.offset+op->req.rw.len;
+            for (int buf_idx = 0; buf_idx < op->iov.count; buf_idx++)
+            {
+                for (int i = 0; i < op->iov.buf[buf_idx].iov_len; i++, offset++)
+                {
+                    uint8_t c = offset < 0xE000 ? 0x56 : (offset < 0x10000 ? 0x57 : 0x58);
+                    if (((uint8_t*)op->iov.buf[buf_idx].iov_base)[i] != c)
+                    {
+                        printf("Write replay: mismatch at %lu\n", offset-op->req.rw.offset);
+                        goto fail;
+                    }
+                }
+            }
+        fail:
+            assert(offset == op->req.rw.offset+op->req.rw.len);
+            replay_ops.push_back(op);
+        }
+        if (replay_start != 0 || replay_end != 0x14000)
+        {
+            printf("Write replay: range mismatch: %lx-%lx\n", replay_start, replay_end);
+            assert(0);
+        }
+        for (auto op: replay_ops)
+        {
+            pretend_op_completed(cli, op, 0);
+        }
+    }
+    // Check that the following write finally proceeds
+    check_op_count(cli, 1, 1);
+    can_complete(r1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0x10000, 0x4000), 0);
+    check_completed(r1);
+    check_op_count(cli, 1, 0);
+
+    // Check sync
+    r2 = test_sync(cli);
+    can_complete(r2);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_SYNC, 0, 0), 0);
+    check_completed(r2);
+
+    // Check disconnect during write
+    r1 = test_write(cli, 0, 4096, 0x59);
+    check_op_count(cli, 1, 1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0, 0x1000), -EPIPE);
+    check_disconnected(cli, 1);
+    pretend_connected(cli, 1);
+    check_op_count(cli, 1, 0);
+    cli->continue_ops(true);
+    check_op_count(cli, 1, 1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0, 0x1000), 0);
+    check_op_count(cli, 1, 1);
+    can_complete(r1);
+    pretend_op_completed(cli, find_op(cli, 1, OSD_OP_WRITE, 0, 0x1000), 0);
+    check_completed(r1);
+
+    // Free client
+    delete cli;
+    delete tfd;
+    printf("[ok] write replay test\n");
+}
+
+void test2()
+{
+    std::map<object_id, cluster_buffer_t> unsynced_writes;
+    cluster_op_t *op = new cluster_op_t();
+    op->opcode = OSD_OP_WRITE;
+    op->inode = 1;
+    op->offset = 0;
+    op->len = 4096;
+    op->iov.push_back(malloc_or_die(4096*1024), 4096);
+    // 0-4k = 0x55
+    memset(op->iov.buf[0].iov_base, 0x55, op->iov.buf[0].iov_len);
+    cluster_client_t::copy_write(op, unsynced_writes);
+    // 8k-12k = 0x66
+    op->offset = 8192;
+    memset(op->iov.buf[0].iov_base, 0x66, op->iov.buf[0].iov_len);
+    cluster_client_t::copy_write(op, unsynced_writes);
+    // 4k-1M+4k = 0x77
+    op->len = op->iov.buf[0].iov_len = 1048576;
+    op->offset = 4096;
+    memset(op->iov.buf[0].iov_base, 0x77, op->iov.buf[0].iov_len);
+    cluster_client_t::copy_write(op, unsynced_writes);
+    // check it
+    assert(unsynced_writes.size() == 4);
+    auto uit = unsynced_writes.begin();
+    int i;
+    assert(uit->first.inode == 1);
+    assert(uit->first.stripe == 0);
+    assert(uit->second.len == 4096);
+    for (i = 0; i < uit->second.len && ((uint8_t*)uit->second.buf)[i] == 0x55; i++) {}
+    assert(i == uit->second.len);
+    uit++;
+    assert(uit->first.inode == 1);
+    assert(uit->first.stripe == 4096);
+    assert(uit->second.len == 4096);
+    for (i = 0; i < uit->second.len && ((uint8_t*)uit->second.buf)[i] == 0x77; i++) {}
+    assert(i == uit->second.len);
+    uit++;
+    assert(uit->first.inode == 1);
+    assert(uit->first.stripe == 8192);
+    assert(uit->second.len == 4096);
+    for (i = 0; i < uit->second.len && ((uint8_t*)uit->second.buf)[i] == 0x77; i++) {}
+    assert(i == uit->second.len);
+    uit++;
+    assert(uit->first.inode == 1);
+    assert(uit->first.stripe == 12*1024);
+    assert(uit->second.len == 1016*1024);
+    for (i = 0; i < uit->second.len && ((uint8_t*)uit->second.buf)[i] == 0x77; i++) {}
+    assert(i == uit->second.len);
+    uit++;
+    // free memory
+    free(op->iov.buf[0].iov_base);
+    delete op;
+    for (auto p: unsynced_writes)
+    {
+        free(p.second.buf);
+    }
+    printf("[ok] copy_write test\n");
+}
+
+int main(int narg, char *args[])
+{
+    test1();
+    test2();
+    return 0;
+}

src/timerfd_manager.cpp

@@ -121,7 +121,7 @@ again:
             exp.it_value.tv_sec--;
             exp.it_value.tv_nsec += 1000000000;
         }
-        if (exp.it_value.tv_sec < 0 || !exp.it_value.tv_sec && !exp.it_value.tv_nsec)
+        if (exp.it_value.tv_sec < 0 || exp.it_value.tv_sec == 0 && exp.it_value.tv_nsec <= 0)
         {
             // It already happened
             trigger_nearest();
@@ -159,6 +159,6 @@ void timerfd_manager_t::trigger_nearest()
     {
         timers.erase(timers.begin()+nearest, timers.begin()+nearest+1);
     }
-    cb(nearest_id);
     nearest = -1;
+    cb(nearest_id);
 }

test-build-el7.sh

@@ -1,34 +0,0 @@
-#!/bin/bash
-
-# Cheatsheet for CentOS 7 packaging (not a build script)
-
-set -e
-rm -f /etc/yum.repos.d/CentOS-Media.repo
-yum -y --enablerepo=extras install centos-release-scl epel-release
-yum -y --enablerepo='*' install devtoolset-9-gcc-c++ devtoolset-9-libatomic-devel gperftools-devel
-yumdownloader --source qemu
-yumdownloader --source fio
-yum-builddep -y --enablerepo='*' qemu
-yum -y install rpm-build
-. /opt/rh/devtoolset-9/enable
-rpm --nomd5 -i qemu*.src.rpm
-rpm --nomd5 -i fio*.src.rpm
-cd ~/rpmbuild/SPECS
-rpmbuild -bp fio.spec
-perl -i -pe 's/^make V=1/exit 1; make V=1/' qemu.spec
-rpmbuild -bc qemu.spec
-perl -i -pe 's/^exit 1; make V=1/make V=1/' qemu.spec
-cd ~/rpmbuild/BUILD/qemu*/
-make qapi-types.h
-mkdir -p ~/vitastor/qemu/b/qemu
-cp config-host.h ~/vitastor/qemu/b/qemu 
-cp qapi-types.h ~/vitastor/qemu/b/qemu
-cp -r include ~/vitastor/qemu
-cd ~/vitastor
-sh copy-qemu-includes.sh
-mv qemu qemu-old
-mv qemu-copy qemu
-ln -s ~/rpmbuild/BUILD/fio*/ fio
-sh copy-fio-includes.sh
-rm fio
-mv fio-copy fio

tests/test_change_pg_count.sh

@@ -2,6 +2,14 @@
 
 . `dirname $0`/common.sh
 
+if [ "$EC" != "" ]; then
+    POOLCFG='"scheme":"xor","pg_size":3,"pg_minsize":2,"parity_chunks":1'
+    NOBJ=512
+else
+    POOLCFG='"scheme":"replicated","pg_size":2,"pg_minsize":2'
+    NOBJ=1024
+fi
+
 dd if=/dev/zero of=./testdata/test_osd1.bin bs=1024 count=1 seek=$((1024*1024-1))
 dd if=/dev/zero of=./testdata/test_osd2.bin bs=1024 count=1 seek=$((1024*1024-1))
 dd if=/dev/zero of=./testdata/test_osd3.bin bs=1024 count=1 seek=$((1024*1024-1))
@@ -9,17 +17,17 @@ dd if=/dev/zero of=./testdata/test_osd4.bin bs=1024 count=1 seek=$((1024*1024-1)
 dd if=/dev/zero of=./testdata/test_osd5.bin bs=1024 count=1 seek=$((1024*1024-1))
 dd if=/dev/zero of=./testdata/test_osd6.bin bs=1024 count=1 seek=$((1024*1024-1))
 
-build/src/vitastor-osd --osd_num 1 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd1.bin 2>/dev/null) &>./testdata/osd1.log &
+build/src/vitastor-osd --osd_num 1 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd1.bin 2>/dev/null) 2>&1 >>./testdata/osd1.log &
 OSD1_PID=$!
-build/src/vitastor-osd --osd_num 2 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd2.bin 2>/dev/null) &>./testdata/osd2.log &
+build/src/vitastor-osd --osd_num 2 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd2.bin 2>/dev/null) 2>&1 >>./testdata/osd2.log &
 OSD2_PID=$!
-build/src/vitastor-osd --osd_num 3 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd3.bin 2>/dev/null) &>./testdata/osd3.log &
+build/src/vitastor-osd --osd_num 3 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd3.bin 2>/dev/null) 2>&1 >>./testdata/osd3.log &
 OSD3_PID=$!
-build/src/vitastor-osd --osd_num 4 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd4.bin 2>/dev/null) &>./testdata/osd4.log &
+build/src/vitastor-osd --osd_num 4 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd4.bin 2>/dev/null) 2>&1 >>./testdata/osd4.log &
 OSD4_PID=$!
-build/src/vitastor-osd --osd_num 5 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd5.bin 2>/dev/null) &>./testdata/osd5.log &
+build/src/vitastor-osd --osd_num 5 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd5.bin 2>/dev/null) 2>&1 >>./testdata/osd5.log &
 OSD5_PID=$!
-build/src/vitastor-osd --osd_num 6 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd6.bin 2>/dev/null) &>./testdata/osd6.log &
+build/src/vitastor-osd --osd_num 6 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd6.bin 2>/dev/null) 2>&1 >>./testdata/osd6.log &
 OSD6_PID=$!
 
 cd mon
@@ -28,9 +36,7 @@ cd ..
 node mon/mon-main.js --etcd_url http://$ETCD_URL --etcd_prefix "/vitastor" --verbose 1 &>./testdata/mon.log &
 MON_PID=$!
 
-$ETCDCTL put /vitastor/config/global '{"immediate_commit":"all"}'
-
-$ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"replicated","pg_size":2,"pg_minsize":2,"pg_count":16,"failure_domain":"osd"}}'
+$ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool",'$POOLCFG',"pg_count":16,"failure_domain":"osd"}}'
 
 sleep 2
 
@@ -50,7 +56,11 @@ try_change()
 {
     n=$1
 
-    $ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"replicated","pg_size":2,"pg_minsize":2,"pg_count":'$n',"failure_domain":"osd"}}'
+    for i in {1..6}; do
+        echo --- Change PG count to $n --- >>testdata/osd$i.log
+    done
+
+    $ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool",'$POOLCFG',"pg_count":'$n',"failure_domain":"osd"}}'
 
     for i in {1..10}; do
         ($ETCDCTL get /vitastor/config/pgs --print-value-only | jq -s -e '(.[0].items["1"] | map((.osd_set | select(. > 0)) | length == 2) | length) == '$n) && \
@@ -77,6 +87,12 @@ try_change()
         $ETCDCTL get --prefix /vitastor/pg/state/
         format_error "FAILED: $n PGS NOT UP"
     fi
+
+    # Check that no objects are lost !
+    nobj=`$ETCDCTL get --prefix '/vitastor/pg/stats' --print-value-only | jq -s '[ .[].object_count ] | reduce .[] as $num (0; .+$num)'`
+    if [ "$nobj" -ne $NOBJ ]; then
+        format_error "Data lost after changing PG count to $n: $NOBJ objects expected, but got $nobj"
+    fi
 }
 
 # 16 -> 32

tests/test_interrupted_rebalance.sh

@@ -0,0 +1,109 @@
+#!/bin/bash -ex
+
+. `dirname $0`/common.sh
+
+if [ "$IMMEDIATE_COMMIT" != "" ]; then
+    NO_SAME="--journal_no_same_sector_overwrites true --journal_sector_buffer_count 1024 --disable_data_fsync 1 --immediate_commit all --log_level 1"
+    $ETCDCTL put /vitastor/config/global '{"recovery_queue_depth":1,"osd_out_time":5,"immediate_commit":"all"}'
+else
+    NO_SAME="--journal_sector_buffer_count 1024 --log_level 1"
+    $ETCDCTL put /vitastor/config/global '{"recovery_queue_depth":1,"osd_out_time":5}'
+fi
+
+dd if=/dev/zero of=./testdata/test_osd1.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd2.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd3.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd4.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd5.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd6.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd7.bin bs=1024 count=1 seek=$((1024*1024-1))
+
+build/src/vitastor-osd --osd_num 1 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd1.bin 2>/dev/null) 2>&1 >>./testdata/osd1.log &
+OSD1_PID=$!
+build/src/vitastor-osd --osd_num 2 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd2.bin 2>/dev/null) 2>&1 >>./testdata/osd2.log &
+OSD2_PID=$!
+build/src/vitastor-osd --osd_num 3 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd3.bin 2>/dev/null) 2>&1 >>./testdata/osd3.log &
+OSD3_PID=$!
+build/src/vitastor-osd --osd_num 4 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd4.bin 2>/dev/null) 2>&1 >>./testdata/osd4.log &
+OSD4_PID=$!
+build/src/vitastor-osd --osd_num 5 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd5.bin 2>/dev/null) 2>&1 >>./testdata/osd5.log &
+OSD5_PID=$!
+build/src/vitastor-osd --osd_num 6 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd6.bin 2>/dev/null) 2>&1 >>./testdata/osd6.log &
+OSD6_PID=$!
+build/src/vitastor-osd --osd_num 7 --bind_address 127.0.0.1 $NO_SAME --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd7.bin 2>/dev/null) 2>&1 >>./testdata/osd7.log &
+OSD7_PID=$!
+
+cd mon
+npm install
+cd ..
+node mon/mon-main.js --etcd_url http://$ETCD_URL --etcd_prefix "/vitastor" --verbose 1 &>./testdata/mon.log &
+MON_PID=$!
+
+$ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"replicated","pg_size":2,"pg_minsize":1,"pg_count":32,"failure_domain":"osd"}}'
+
+sleep 2
+
+if ! ($ETCDCTL get /vitastor/config/pgs --print-value-only | jq -s -e '(.[0].items["1"] | map((.osd_set | select(. > 0)) | length == 2) | length) == 32'); then
+    format_error "FAILED: 32 PGS NOT CONFIGURED"
+fi
+
+if ! ($ETCDCTL get --prefix /vitastor/pg/state/ --print-value-only | jq -s -e '([ .[] | select(.state == ["active"]) ] | length) == 32'); then
+    format_error "FAILED: 32 PGS NOT UP"
+fi
+
+IMG_SIZE=960
+
+LD_PRELOAD=libasan.so.5 \
+fio -thread -name=test -ioengine=build/src/libfio_vitastor.so -bs=4M -direct=1 -iodepth=16 -fsync=16 -rw=write \
+    -etcd=$ETCD_URL -pool=1 -inode=2 -size=${IMG_SIZE}M -cluster_log_level=10
+
+try_reweight()
+{
+    osd=$1
+    w=$2
+    $ETCDCTL put /vitastor/config/osd/$osd '{"reweight":'$w'}'
+    sleep 3
+}
+
+try_reweight 1 0
+
+try_reweight 2 0
+
+try_reweight 3 0
+
+try_reweight 4 0
+
+try_reweight 5 0
+
+try_reweight 1 1
+
+try_reweight 2 1
+
+try_reweight 3 1
+
+try_reweight 4 1
+
+try_reweight 5 1
+
+# Wait for the rebalance to finish
+for i in {1..60}; do
+    ($ETCDCTL get --prefix /vitastor/pg/state/ --print-value-only | jq -s -e '([ .[] | select(.state == ["active"]) ] | length) == 32') && \
+        break
+    if [ $i -eq 60 ]; then
+        format_error "Rebalance couldn't finish in 60 seconds"
+    fi
+    sleep 1
+done
+
+# Check that PGs never had degraded objects !
+if grep has_degraded ./testdata/mon.log; then
+    format_error "Some copies of objects were lost during interrupted rebalancings"
+fi
+
+# Check that no objects are lost !
+nobj=`$ETCDCTL get --prefix '/vitastor/pg/stats' --print-value-only | jq -s '[ .[].object_count ] | reduce .[] as $num (0; .+$num)'`
+if [ "$nobj" -ne $((IMG_SIZE*8)) ]; then
+    format_error "Data lost after multiple interrupted rebalancings"
+fi
+
+format_green OK

tests/test_move_reappear.sh

@@ -0,0 +1,79 @@
+#!/bin/bash -ex
+
+. `dirname $0`/common.sh
+
+dd if=/dev/zero of=./testdata/test_osd1.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd2.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd3.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd4.bin bs=1024 count=1 seek=$((1024*1024-1))
+dd if=/dev/zero of=./testdata/test_osd5.bin bs=1024 count=1 seek=$((1024*1024-1))
+
+build/src/vitastor-osd --osd_num 1 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd1.bin 2>/dev/null) &>./testdata/osd1.log &
+OSD1_PID=$!
+build/src/vitastor-osd --osd_num 2 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd2.bin 2>/dev/null) &>./testdata/osd2.log &
+OSD2_PID=$!
+build/src/vitastor-osd --osd_num 3 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd3.bin 2>/dev/null) &>./testdata/osd3.log &
+OSD3_PID=$!
+build/src/vitastor-osd --osd_num 4 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd4.bin 2>/dev/null) &>./testdata/osd4.log &
+OSD4_PID=$!
+build/src/vitastor-osd --osd_num 5 --bind_address 127.0.0.1 --etcd_address $ETCD_URL $(node mon/simple-offsets.js --format options --device ./testdata/test_osd5.bin 2>/dev/null) &>./testdata/osd5.log &
+OSD5_PID=$!
+
+$ETCDCTL put /vitastor/config/pools '{"1":{"name":"testpool","scheme":"replicated","pg_size":2,"pg_minsize":1,"pg_count":1,"failure_domain":"osd"}}'
+
+$ETCDCTL put /vitastor/config/pgs '{"items":{"1":{"1":{"osd_set":[1,0],"primary":1}}}}'
+
+sleep 2
+
+if ! ($ETCDCTL get /vitastor/pg/state/1/1 --print-value-only | jq -s -e '(. | length) != 0 and .[0].state == ["active","degraded"]'); then
+    format_error "Failed to start the PG active+degraded"
+fi
+
+LD_PRELOAD=libasan.so.5 \
+fio -thread -name=test -ioengine=build/src/libfio_vitastor.so -bs=4M -direct=1 -iodepth=1 -fsync=1 -rw=write \
+    -etcd=$ETCD_URL -pool=1 -inode=2 -size=32M -cluster_log_level=10
+
+$ETCDCTL put /vitastor/config/pgs '{"items":{"1":{"1":{"osd_set":[1,0],"primary":0}}}}'
+
+sleep 2
+
+if [ "`$ETCDCTL get /vitastor/pg/state/1/1 --print-value-only`" != "" ]; then
+    format_error "Failed to stop the PG"
+fi
+
+$ETCDCTL put /vitastor/pg/history/1/1 '{"all_peers":[1,2,3]}'
+
+$ETCDCTL put /vitastor/config/pgs '{"items":{"1":{"1":{"osd_set":[4,5],"primary":4}}}}'
+
+sleep 5
+
+if ! ($ETCDCTL get /vitastor/pg/state/1/1 --print-value-only | jq -s -e '(. | length) != 0 and .[0].state == ["active"]'); then
+    format_error "Failed to move degraded objects to the clean OSD set"
+fi
+
+$ETCDCTL put /vitastor/config/pgs '{"items":{"1":{"1":{"osd_set":[4,5],"primary":0}}}}'
+
+$ETCDCTL put /vitastor/pg/history/1/1 '{"all_peers":[1,2,3]}'
+
+sleep 2
+
+if [ "`$ETCDCTL get /vitastor/pg/state/1/1 --print-value-only`" != "" ]; then
+    format_error "Failed to stop the PG after degraded recovery"
+fi
+
+cp testdata/osd4.log testdata/osd4_pre.log
+>testdata/osd4.log
+
+$ETCDCTL put /vitastor/config/pgs '{"items":{"1":{"1":{"osd_set":[4,5],"primary":4}}}}'
+
+sleep 2
+
+if grep -q 'PG 1/1.*is.*has_' testdata/osd4.log; then
+    format_error "PG has degraded or misplaced objects after a full re-peer following a degraded recovery"
+fi
+
+if ! ($ETCDCTL get /vitastor/pg/state/1/1 --print-value-only | jq -s -e '(. | length) != 0 and .[0].state == ["active"]'); then
+    format_error "PG not active+clean after a full re-peer following a degraded recovery"
+fi
+
+format_green OK

tests/test_write.sh

@@ -35,6 +35,18 @@ fi
 #    fio -thread -name=test -ioengine=build/src/libfio_vitastor_sec.so -bs=4k -fsync=128 `$ETCDCTL get /vitastor/osd/state/1 --print-value-only | jq -r '"-host="+.addresses[0]+" -port="+(.port|tostring)'` -rw=write -size=32M
 
 LD_PRELOAD=libasan.so.5 \
-    fio -thread -name=test -ioengine=build/src/libfio_vitastor.so -bs=4M -direct=1 -iodepth=1 -fsync=1 -rw=write -etcd=$ETCD_URL -pool=1 -inode=1 -size=1G -cluster_log_level=10
+    fio -thread -name=test -ioengine=build/src/libfio_vitastor.so -bs=4M -direct=1 -iodepth=1 -fsync=1 -rw=write -etcd=$ETCD_URL -pool=1 -inode=1 -size=128M -cluster_log_level=10
+
+LD_PRELOAD=libasan.so.5 \
+    fio -thread -name=test -ioengine=build/src/libfio_vitastor.so -bs=4k -direct=1 -iodepth=1 -fsync=32 -buffer_pattern=0xdeadface \
+        -rw=randwrite -etcd=$ETCD_URL -pool=1 -inode=1 -size=128M -number_ios=1024
+
+qemu-img convert -S 4096 -p \
+    -f raw "vitastor:etcd_host=127.0.0.1\:$ETCD_PORT/v3:pool=1:inode=1:size=$((128*1024*1024))" \
+    -O raw ./testdata/read.bin
+
+qemu-img convert -S 4096 -p \
+    -f raw ./testdata/read.bin \
+    -O raw "vitastor:etcd_host=127.0.0.1\:$ETCD_PORT/v3:pool=1:inode=1:size=$((128*1024*1024))"
 
 format_green OK