Bug #39000 rpl_ndb_extraCol fails on powermacg5
Submitted: 25 Aug 2008 9:52 Modified: 29 Mar 2009 19:25
Reporter: Sven Sandberg Email Updates:
Status: Can't repeat Impact on me:
Category:Tests: Cluster Severity:S7 (Test Cases)
Version:5.1-rpl OS:Any
Assigned to: Serge Kozlov CPU Architecture:Any
Tags: pushbuild, rpl_ndb_extraCol, sporadic, test failure, timeout

[25 Aug 2008 9:52] Sven Sandberg
Pushbuild failure:

rpl_ndb.rpl_ndb_extraCol                 [ fail ]  timeout after 15 minutes

Test case timeout after 15 minute(s)

 == /Users/henry/pb/bzr_mysql-5.1-rpl/32/mysql-5.1.28-pb32/mysql-test/var-ps_row-6/tmp/analyze-timeout-mysqld.1.1.err ==
mysqltest: Could not open './include/analyze-timeout.test' for reading, errno: 2

 == /Users/henry/pb/bzr_mysql-5.1-rpl/32/mysql-5.1.28-pb32/mysql-test/var-ps_row-6/tmp/analyze-timeout-mysqld.1.slave.err ==
mysqltest: Could not open './include/analyze-timeout.test' for reading, errno: 2

 == /Users/henry/pb/bzr_mysql-5.1-rpl/32/mysql-5.1.28-pb32/mysql-test/var-ps_row-6/tmp/analyze-timeout-mysqld.2.1.err ==
mysqltest: Could not open './include/analyze-timeout.test' for reading, errno: 2

 - saving '/Users/henry/pb/bzr_mysql-5.1-rpl/32/mysql-5.1.28-pb32/mysql-test/var-ps_row-6/log/rpl_ndb.rpl_ndb_extraCol/' to '/Users/henry/pb/bzr_mysql-5.1-rpl/32/mysql-5.1.28-pb32/mysql-test/var-ps_row-6/log/rpl_ndb.rpl_ndb_extraCol/'

Retrying test, attempt(2/3)...

How to repeat:
https://intranet.mysql.com/secure/pushbuild/showpush.pl?dir=bzr_mysql-5.1-rpl&order=32 powermacg5/n_mix and ps_row
[13 Oct 2008 11:15] Serge Kozlov
The test case has 10 included primitives of "wait_for_slave_sql_to_stop.inc". The primitive uses 5 minutes as default timeout so correct minimal timeout for whole case shouldn't be less than 10x5 minutes = 50 min. 
Suggested ways to fix:
1. Increase default timeout for a test case in PB.
2. Make possibility to define an own timeout for each test case.
3. Decrease timeout value for  primitive "wait_for_slave_sql_to_stop.inc". But 15min/10 = 1,5 min. But may be not enough to sync clusters ...
[20 Oct 2008 11:56] Serge Kozlov
It is no information about real reasons of timeouts or log files there and is no way to debug failures like this one.
[20 Oct 2008 13:23] Sven Sandberg
This may be possible to debug after BUG#40150 has been fixed.

See also BUG#40148.
[19 Feb 2009 13:12] Lars Thalmann
BUG#40150 is closed as won't fix, so we need to decide on a new 
strategy for this bug.  Re-opening bug.