From 861803eec5962036bc711a3aa7dc08f8b535c196 Mon Sep 17 00:00:00 2001
From: Konstantin Osipov <kostja@tarantool.org>
Date: Tue, 22 May 2018 23:08:23 +0300
Subject: [PATCH] replication: add logging to replication connect/sync
---
src/box/box.cc | 4 ++--
src/box/replication.cc | 23 ++++++++++++++++++++---
2 files changed, 22 insertions(+), 5 deletions(-)
diff --git a/src/box/box.cc b/src/box/box.cc
index 4cf7dd13d0..e3eb2738f3 100644
--- a/src/box/box.cc
+++ b/src/box/box.cc
@@ -1925,13 +1925,13 @@ box_cfg_xc(void)
/* Follow replica */
replicaset_follow();
- say_info("ready to accept requests");
-
fiber_gc();
is_box_configured = true;
if (!is_bootstrap_leader)
replicaset_sync();
+
+ say_info("ready to accept requests");
}
void
diff --git a/src/box/replication.cc b/src/box/replication.cc
index 0b770c9131..89e9eb4a5c 100644
--- a/src/box/replication.cc
+++ b/src/box/replication.cc
@@ -520,6 +520,7 @@ replicaset_connect(struct applier **appliers, int count,
replicaset_update(appliers, count);
return;
}
+ say_verbose("connecting to %d replicas", count);
/*
* Simultaneously connect to remote peers to receive their UUIDs
@@ -564,9 +565,13 @@ replicaset_connect(struct applier **appliers, int count,
timeout -= ev_monotonic_now(loop()) - wait_start;
}
if (state.connected < count) {
+ say_crit("failed to connect to %d out of %d replicas",
+ count - state.connected, count);
/* Timeout or connection failure. */
if (connect_all)
goto error;
+ } else {
+ say_verbose("connected to %d replicas", state.connected);
}
for (int i = 0; i < count; i++) {
@@ -631,29 +636,41 @@ replicaset_sync(void)
{
int quorum = replicaset_quorum();
+ if (quorum > 0)
+ say_verbose("synchronizing with %d replicas", quorum);
+
if (replicaset.applier.connected < quorum) {
/*
* Not enough replicas connected to form a quorum.
* Do not stall configuration, leave the instance
* in 'orphan' state.
*/
+ say_crit("entering orphan mode");
return;
}
/*
- * Wait until a quorum is formed. Abort waiting if
- * a quorum cannot be formed because of errors.
+ * Wait until all connected replicas synchronize up to
+ * replication_sync_lag
*/
while (replicaset.applier.synced < quorum &&
replicaset.applier.connected >= quorum)
fiber_cond_wait(&replicaset.applier.cond);
+
+ if (quorum > 0) {
+ say_crit("replica set sync complete, quorum of %d "
+ "replicas formed", quorum);
+ }
}
void
replicaset_check_quorum(void)
{
- if (replicaset.applier.synced >= replicaset_quorum())
+ if (replicaset.applier.synced >= replicaset_quorum()) {
+ if (replicaset_quorum() > 0)
+ say_crit("leaving orphan mode");
box_clear_orphan();
+ }
}
void
--
GitLab