Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Duplicate
-
10.0(EOL)
Description
So far it's been a one-time failure, but given the platform and problems associated with it, I find it suspicious, so I'll file it.
The priority is to be raised if the problem occurs again.
Or, maybe it's a duplicate of what Kristian has already analyzed, in this case please close it as such.
10.0 revno 4478
http://buildbot.askmonty.org/buildbot/builders/bintar-rhel6-p8-debug/builds/216/steps/test/logs/stdio
worker[4] mysql-test-run: WARNING: Timeout after mysql-test-run waited 180 seconds for the process [mysqld.1 - pid: 350, winpid: 350] to create a pid file.
|
binlog.binlog_checkpoint 'row,xtradb' w4 [ fail ]
|
Test ended at 2014-11-11 23:57:50
|
|
|
CURRENT_TEST: binlog.binlog_checkpoint
|
|
|
|
|
Failed to start mysqld.1
|
mysqltest failed but provided no output
|
Here is the corresponding part of the error log:
http://buildbot.askmonty.org/buildbot/builders/bintar-rhel6-p8-debug/builds/216/steps/test/logs/mysqld.1.err.4
CURRENT_TEST: binlog.binlog_checkpoint
|
141111 23:54:50 [Note] InnoDB: Using mutexes to ref count buffer pool pages
|
2014-11-11 23:54:50 fffa3b16fe0 InnoDB: !!!!!!!! UNIV_DEBUG switched on !!!!!!!!!
|
2014-11-11 23:54:50 fffa3b16fe0 InnoDB: !!!!!!!! UNIV_SYNC_DEBUG switched on !!!!!!!!!
|
141111 23:54:50 [Note] InnoDB: The InnoDB memory heap is disabled
|
141111 23:54:50 [Note] InnoDB: Mutexes and rw_locks use GCC atomic builtins
|
141111 23:54:50 [Note] InnoDB: GCC builtin __atomic_thread_fence() is used for memory barrier
|
141111 23:54:50 [Note] InnoDB: Compressed tables use zlib 1.2.6
|
141111 23:54:50 [Note] InnoDB: Using Linux native AIO
|
141111 23:54:50 [Note] InnoDB: Not using CPU crc32 instructions
|
141111 23:54:50 [Note] InnoDB: Initializing buffer pool, size = 8.0M
|
141111 23:54:50 [Note] InnoDB: Completed initialization of buffer pool
|
141111 23:54:50 [Note] InnoDB: Highest supported file format is Barracuda.
|
CURRENT_TEST: binlog.binlog_checksum
|
141111 23:57:50
|
So, it looks like during startup InnoDB just stopped.
The next line after 'Highest format' is supposed to be
InnoDB: 128 rollback segment(s) are active.
|
and in previous and next tests on that machine it only takes 1 second for InnoDB to proceed to this step, but not this time. This time it waited for 3 minutes, after which the test was interrupted.
Same happened on the latest 10.0 build revno 4493, bintar-trusty-p8-debug builder, power8-vlp04, innodb.innodb-autoinc-44030 test:
test log: http://buildbot.askmonty.org/buildbot/builders/bintar-trusty-p8-debug/builds/187/steps/test/logs/stdio
server log: http://buildbot.askmonty.org/buildbot/builders/bintar-trusty-p8-debug/builds/187/steps/test/logs/mysqld.1.err.1
CURRENT_TEST: innodb.innodb-autoinc-44030141114 4:52:43 [Note] InnoDB: Using mutexes to ref count buffer pool pages2014-11-14 04:52:43 3fff7aa878b0 InnoDB: !!!!!!!! UNIV_DEBUG switched on !!!!!!!!!2014-11-14 04:52:43 3fff7aa878b0 InnoDB: !!!!!!!! UNIV_SYNC_DEBUG switched on !!!!!!!!!141114 4:52:43 [Note] InnoDB: The InnoDB memory heap is disabled141114 4:52:43 [Note] InnoDB: Mutexes and rw_locks use GCC atomic builtins141114 4:52:43 [Note] InnoDB: GCC builtin __atomic_thread_fence() is used for memory barrier141114 4:52:43 [Note] InnoDB: Compressed tables use zlib 1.2.8141114 4:52:43 [Note] InnoDB: Using Linux native AIO141114 4:52:43 [Note] InnoDB: Not using CPU crc32 instructions141114 4:52:43 [Note] InnoDB: Initializing buffer pool, size = 8.0M141114 4:52:43 [Note] InnoDB: Completed initialization of buffer pool141114 4:52:43 [Note] InnoDB: Highest supported file format is Barracuda.CURRENT_TEST: parts.partition_debug_sync_innodb