MySQL Bugs: #15934: im_daemon_life_cycle fails sporadically

Bug #15934	im_daemon_life_cycle fails sporadically
Submitted:	22 Dec 2005 11:55	Modified:	9 Nov 2006 21:31
Reporter:	Joerg Bruehe	Email Updates:
Status:	Closed	Impact on me:	None
Category:	Instance Manager	Severity:	S3 (Non-critical)
Version:	5.0.18	OS:	Linux (Linux (RHEL4, x86_64))
Assigned to:	Alexander Nozdrin	CPU Architecture:	Any

Description:
This is a known effect, has just not been filed before (it seems).

This is different from bug#15603, which was a port conflict (solved).

Symptom:
im_daemon_life_cycle           [ fail ]

Errors are (from PATH/mysqltest-time) :
mysqltest: Failed in mysql_real_connect(): Can't connect to local MySQL server through socket 'PATH/im.sock' (2)
(the last lines may be the most important ones)

Ending Tests

Same effect in tests "im_life_cycle", "im_options_set", "im_options_unset", and "im_utils".

Occurs: all tests on "cane"

How to repeat:
Test suite failure.

This bug is expected to be fixed by a patch for BUG#14106.
Please, re-verify it again.

No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".

Sorry for the bad news: Still occurs with 5.0.24

im_daemon_life_cycle           [ fail ]

Errors are (from /PATH/mysqltest-time) :
mysqltest: Could not open connection 'default': 2002 Can't connect to local MySQL server through socket '/PATH/im.sock' (2)
(the last lines may be the most important ones)

Killing Possible Leftover Processes
===---===---===---===---===---===--- OCCURRED:
Do-rpm-rh-x86-64-5.0-pro-gpl.rhel4.log   normal
Do-rpm-rh-x86-64-5.0-pro.rhel4.log   normal
Do-rpm-rh-x86-64-5.0-standard.rhel4.log   normal

Same symptom in this build / these runs also in tests 
"im_life_cycle", "im_options_set", "im_options_unset", and "im_utils"

A patch for this bug:
http://lists.mysql.com/commits/10998

See BUG#21331 for more info.

Merged into 5.0 tree, currently tagged 5.0.25;
merged into 5.1 tree, currently tagged 5.1.12-beta.

Test suite change. No changelog entry needed.

The problem has not been completely solved.

The problem is in test case (im_daemon_life_cycle.imtest) -- it
should wait for Instance Manager to start accepting connections
after restart. Otherwise race condition occurs.

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/13019

ChangeSet@1.2274, 2006-10-03 18:42:59+04:00, anozdrin@alik. +4 -0
  Patch for BUG#15934: im_daemon_life_cycle fails sporadically.
  
  The problem was a race condition in a test case.
  
  The fix eliminates the race condition by explicit
  wait on UNIX socket to start accepting connections.
  
  The patch affects only test suite (i.e. does not touch
  server codebase).

Approved by email.

Small improvement has been pushed into 5.0-rt tree.

Pushed into 5.0.29 and 5.1.13.
No documentation entry is needed as this is a test suite bug.