ReplicatedPG: clean out completed trimmed objects as we go

Also, explicitely maintain a max number of concurrently trimming
objects.

Fixes: 9113
Backport: dumpling, firefly, giant
Signed-off-by: Samuel Just <sam.just@inktank.com>
Conflicts:
	src/common/config_opts.h
	src/osd/ReplicatedPG.cc

diff --git a/src/common/config_opts.h b/src/common/config_opts.h
index 1fd91c327d83cdded25ea11375f2d38ada52ec09..d688f5b3d29b8b3a24ff62ecafa737890ddbca11 100644 (file)
--- a/src/common/config_opts.h
+++ b/src/common/config_opts.h
@@ -427,6 +427,10 @@ OPTION(osd_heartbeat_addr, OPT_ADDR, entity_addr_t())
 OPTION(osd_heartbeat_interval, OPT_INT, 6)       // (seconds) how often we ping peers
 OPTION(osd_heartbeat_grace, OPT_INT, 20)         // (seconds) how long before we decide a peer has failed
 OPTION(osd_heartbeat_min_peers, OPT_INT, 10)     // minimum number of peers
+
+// max number of parallel snap trims/pg
+OPTION(osd_pg_max_concurrent_snap_trims, OPT_U64, 2)
+
 OPTION(osd_mon_heartbeat_interval, OPT_INT, 30)  // (seconds) how often to ping monitor if no peers
 OPTION(osd_mon_report_interval_max, OPT_INT, 120)
 OPTION(osd_mon_report_interval_min, OPT_INT, 5)  // pg stats, failures, up_thru, boot.

diff --git a/src/osd/ReplicatedPG.cc b/src/osd/ReplicatedPG.cc
index 443e98d3d80dfcc8e3888948c788d05a28570735..5bd2e0c7fc6bcaf79f1bc3fb18fc1677c866f3bd 100644 (file)
--- a/src/osd/ReplicatedPG.cc
+++ b/src/osd/ReplicatedPG.cc
@@ -7898,33 +7898,48 @@ boost::statechart::result ReplicatedPG::TrimmingObjects::react(const SnapTrim&)
 
   dout(10) << "TrimmingObjects: trimming snap " << snap_to_trim << dendl;
 
-  // Get next
-  int r = pg->snap_mapper.get_next_object_to_trim(snap_to_trim, &pos);
-  if (r != 0 && r != -ENOENT) {
-    derr << __func__ << ": get_next returned " << cpp_strerror(r) << dendl;
-    assert(0);
-  } else if (r == -ENOENT) {
-    // Done!
-    dout(10) << "TrimmingObjects: got ENOENT" << dendl;
-    post_event(SnapTrim());
-    return transit< WaitingOnReplicas >();
+  for (set<RepGather *>::iterator i = repops.begin();
+       i != repops.end(); 
+       ) {
+    if ((*i)->done) {
+      (*i)->put();
+      repops.erase(i++);
+    } else {
+      ++i;
+    }
   }
 
-  dout(10) << "TrimmingObjects react trimming " << pos << dendl;
-  RepGather *repop = pg->trim_object(pos);
-  assert(repop);
+  while (repops.size() < g_conf->osd_pg_max_concurrent_snap_trims) {
+    // Get next
+    hobject_t old_pos = pos;
+    int r = pg->snap_mapper.get_next_object_to_trim(snap_to_trim, &pos);
+    if (r != 0 && r != -ENOENT) {
+      derr << __func__ << ": get_next returned " << cpp_strerror(r) << dendl;
+      assert(0);
+    } else if (r == -ENOENT) {
+      // Done!
+      dout(10) << "TrimmingObjects: got ENOENT" << dendl;
+      post_event(SnapTrim());
+      return transit< WaitingOnReplicas >();
+    }
 
-  repop->queue_snap_trimmer = true;
-  eversion_t old_last_update = pg->pg_log.get_head();
-  bool old_exists = repop->obc->obs.exists;
-  uint64_t old_size = repop->obc->obs.oi.size;
-  eversion_t old_version = repop->obc->obs.oi.version;
+    dout(10) << "TrimmingObjects react trimming " << pos << dendl;
+    RepGather *repop = pg->trim_object(pos);
+    assert(repop);
+
+    repop->queue_snap_trimmer = true;
 
-  pg->append_log(repop->ctx->log, eversion_t(), repop->ctx->local_t);
-  pg->issue_repop(repop, repop->ctx->mtime, old_last_update, old_exists, old_size, old_version);
-  pg->eval_repop(repop);
+    eversion_t old_last_update = pg->pg_log.get_head();
+    bool old_exists = repop->obc->obs.exists;
+    uint64_t old_size = repop->obc->obs.oi.size;
+    eversion_t old_version = repop->obc->obs.oi.version;
 
-  repops.insert(repop);
+    pg->append_log(repop->ctx->log, eversion_t(), repop->ctx->local_t);
+    pg->issue_repop(repop, repop->ctx->mtime, old_last_update, old_exists, old_size, old_version);
+    pg->eval_repop(repop);
+
+    repops.insert(repop);
+  }
   return discard_event();
 }
 /* WaitingOnReplicasObjects */
@@ -7957,7 +7972,7 @@ boost::statechart::result ReplicatedPG::WaitingOnReplicas::react(const SnapTrim&
   for (set<RepGather *>::iterator i = repops.begin();
        i != repops.end();
        repops.erase(i++)) {
-    if (!(*i)->applied || !(*i)->waitfor_ack.empty()) {
+    if (!(*i)->done) {
       return discard_event();
     } else {
       (*i)->put();