[Tarantool-patches] [PATCH v4 17/16] replication: fix flaky election_qsync.test

[Serge Petrenko via Tarantool-patches <tarantool-patches@dev.tarantool.org>]

Fix the test failing occasionally with the following result mismatch:

[001] replication/election_qsync.test.lua memtx           [ fail ]
[001]
[001] Test failed! Result content mismatch:
[001] --- replication/election_qsync.result    Thu Jul 15 17:15:48 2021
[001] +++ var/rejects/replication/election_qsync.reject    Thu Jul 15 
20:46:51 2021
[001] @@ -145,8 +145,7 @@
[001]   | ...
[001]  box.space.test:select{}
[001]   | ---
[001] - | - - [1]
[001] - |   - [2]
[001] + | - - [2]
[001]   | ...
[001]  box.space.test:drop()
[001]   | ---
[001]

The issue happened because row [1] wasn't delivered to the 'default'
instance from the 'replica' at all. The test does try to wait for [1] to
be written to WAL and replicated, but sometimes it fails to wait until
this event happens:

box.ctl.promote() is issued asynchronously once the instance becomes the
Raft leader. So issuing `box.ctl.wait_rw()` doesn't guarantee that the
replica has already written the PROMOTE (the limbo is initially
unclaimed so replica becomes writeable as soon as it becomes the Raft
leader).

Right after `wait_rw()` we wait for lsn propagation and for 'default'
instance to reach replica's lsn. It may happen that lsn propagation
happens due to PROMOTE being written to WAL, and not row [1].

When this is the case, the 'default' instance doesn't receive row [1] at
all, resulting in the test error shown above.

Fix the issue by waiting for the promotion to happen explicitly.

Part of #5430
---
  test/replication/election_qsync.result   | 8 +++++++-
  test/replication/election_qsync.test.lua | 7 ++++++-
  2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/test/replication/election_qsync.result 
b/test/replication/election_qsync.result
index 2402c8578..c6ec5e352 100644
--- a/test/replication/election_qsync.result
+++ b/test/replication/election_qsync.result
@@ -75,13 +75,19 @@ box.cfg{
   | ---
   | ...

-box.ctl.wait_rw()
+-- Promote is written asynchronously to the instance becoming the 
leader, so
+-- wait for it. As soon as it's written, the instance's definitely a 
leader.
+test_run:wait_cond(function() \
+    return box.info.synchro.queue.owner == 
box.info.id                          \
+end)
   | ---
+ | - true
   | ...
  assert(box.info.election.state == 'leader')
   | ---
   | - true
   | ...
+
  lsn = box.info.lsn
   | ---
   | ...
diff --git a/test/replication/election_qsync.test.lua 
b/test/replication/election_qsync.test.lua
index e1aca8351..f3c7c290b 100644
--- a/test/replication/election_qsync.test.lua
+++ b/test/replication/election_qsync.test.lua
@@ -39,8 +39,13 @@ box.cfg{
      replication_timeout = 
0.1,                                                  \
  }

-box.ctl.wait_rw()
+-- Promote is written asynchronously to the instance becoming the 
leader, so
+-- wait for it. As soon as it's written, the instance's definitely a 
leader.
+test_run:wait_cond(function() \
+    return box.info.synchro.queue.owner == 
box.info.id                          \
+end)
  assert(box.info.election.state == 'leader')
+
  lsn = box.info.lsn
  _ = fiber.create(function() \
      ok, err = pcall(box.space.test.replace, box.space.test, 
{1})                \
-- 
2.30.1 (Apple Git-130)