[Tarantool-patches] [PATCH 1/1] test: fix flaky qsync_basic.test.lua

[Tarantool-patches] [PATCH 1/1] test: fix flaky qsync_basic.test.lua

[Vladislav Shpilevoy]

In one of the test cases 2 fibers were started making a
transaction. In the first fiber the transaction was rolled back,
and the second fiber was expected to do the same.

It did rollback too, but not always immediately after the first
one. Because the first fiber needed not just do rollback right
away, but write a ROLLBACK entry into WAL before applying the
rollback to all next transactions. This led to a yield, during
which it was possible to observe the second fiber not dead yet.

The patch makes the test explicitly wait for the fibers death.

Closes #5162
---
Branch: http://github.com/tarantool/tarantool/tree/gerold103/gh-5162-flaky-qsync_basic
Issue: https://github.com/tarantool/tarantool/issues/5162

 test/replication/qsync_basic.result   | 8 ++++----
 test/replication/qsync_basic.test.lua | 4 ++--
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/test/replication/qsync_basic.result b/test/replication/qsync_basic.result
index 6b55a0e5e..b355a5c06 100644
--- a/test/replication/qsync_basic.result
+++ b/test/replication/qsync_basic.result
@@ -382,13 +382,13 @@ end
  | ---
  | - error: A rollback for a synchronous transaction is received
  | ...
-f1:status()
+test_run:wait_cond(function() return f1:status() == 'dead' end)
  | ---
- | - dead
+ | - true
  | ...
-f2:status()
+test_run:wait_cond(function() return f2:status() == 'dead' end)
  | ---
- | - dead
+ | - true
  | ...
 box.space.sync:select{9}
  | ---
diff --git a/test/replication/qsync_basic.test.lua b/test/replication/qsync_basic.test.lua
index dcd1d6c76..205fb0664 100644
--- a/test/replication/qsync_basic.test.lua
+++ b/test/replication/qsync_basic.test.lua
@@ -153,8 +153,8 @@ do
     f2 = fiber.create(box.space.locallocal.replace, box.space.locallocal, {9})  \
     box.space.test:replace{9}                                                   \
 end
-f1:status()
-f2:status()
+test_run:wait_cond(function() return f1:status() == 'dead' end)
+test_run:wait_cond(function() return f2:status() == 'dead' end)
 box.space.sync:select{9}
 box.space.locallocal:select{9}
 box.space.test:select{9}
-- 
2.21.1 (Apple Git-122.3)

Re: [Tarantool-patches] [PATCH 1/1] test: fix flaky qsync_basic.test.lua

[Cyrill Gorcunov]

On Sat, Jul 11, 2020 at 07:01:52PM +0200, Vladislav Shpilevoy wrote:
> In one of the test cases 2 fibers were started making a
> transaction. In the first fiber the transaction was rolled back,
> and the second fiber was expected to do the same.
> 
> It did rollback too, but not always immediately after the first
> one. Because the first fiber needed not just do rollback right
> away, but write a ROLLBACK entry into WAL before applying the
> rollback to all next transactions. This led to a yield, during
> which it was possible to observe the second fiber not dead yet.
> 
> The patch makes the test explicitly wait for the fibers death.
> 
> Closes #5162
> ---
> Branch: http://github.com/tarantool/tarantool/tree/gerold103/gh-5162-flaky-qsync_basic
> Issue: https://github.com/tarantool/tarantool/issues/5162

Reviewed-by: Cyrill Gorcunov <gorcunov@gmail.com>

Re: [Tarantool-patches] [PATCH 1/1] test: fix flaky qsync_basic.test.lua

[Kirill Yukhin]

Hello,

On 11 июл 19:01, Vladislav Shpilevoy wrote:
> In one of the test cases 2 fibers were started making a
> transaction. In the first fiber the transaction was rolled back,
> and the second fiber was expected to do the same.
> 
> It did rollback too, but not always immediately after the first
> one. Because the first fiber needed not just do rollback right
> away, but write a ROLLBACK entry into WAL before applying the
> rollback to all next transactions. This led to a yield, during
> which it was possible to observe the second fiber not dead yet.
> 
> The patch makes the test explicitly wait for the fibers death.
> 
> Closes #5162
> ---
> Branch: http://github.com/tarantool/tarantool/tree/gerold103/gh-5162-flaky-qsync_basic
> Issue: https://github.com/tarantool/tarantool/issues/5162

LGTM.
I've checked your patch into master.

--
Regards, Kirill Yukhin