| Bug #59123 | rpl_stm_binlog_max_cache_size fails sporadically with found warnings | ||
|---|---|---|---|
| Submitted: | 23 Dec 2010 5:48 | Modified: | 28 Jan 2011 10:45 |
| Reporter: | Zhenxing He | Email Updates: | |
| Status: | Closed | Impact on me: | |
| Category: | MySQL Server: Replication | Severity: | S3 (Non-critical) |
| Version: | 5.6 | OS: | Any |
| Assigned to: | Zhenxing He | CPU Architecture: | Any |
| Tags: | pb2, test | ||
[23 Dec 2010 7:02]
Bugs System
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/127546 3469 He Zhenxing 2010-12-23 BUG#59123 rpl_stm_binlog_max_cache_size fails sporadically with found warnings rpl_stm_binlog_max_cache_size.test fails sporadically on PB2 daily-trunk-bugfixing with binlog-checksum enabled, the expected incident error message was not properly suppressed, and the incident message would have extra strange characters appended. The cause is that the incident error message restored from log event is not properly ended with '\0'. This does not cause problem normally because the error message is at the end of the event, and we added a '\0' at the end of an event. but when binlog-checksum is enabled, the checksum value will be at the end after the message, and so extra strange characters could be shown and cause the suppression rule fail to suppress the message. Fixed the problem by allocate memory for the string and add '\0' at the end of the message.
[24 Dec 2010 11:01]
Alfranio Tavares Correia Junior
See BUG#58967 too.
[27 Dec 2010 8:51]
Bugs System
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/127591 3469 He Zhenxing 2010-12-27 BUG#59123 rpl_stm_binlog_max_cache_size fails sporadically with found warnings rpl_stm_binlog_max_cache_size.test fails sporadically on PB2 daily-trunk-bugfixing with binlog-checksum enabled, the expected incident error message was not properly suppressed, and the incident message would have extra strange characters appended. The cause is that the incident error message restored from log event is not properly ended with '\0'. This does not cause problem normally because the error message is at the end of the event, and we added a '\0' at the end of an event. but when binlog-checksum is enabled, the checksum value will be at the end after the message, and so extra strange characters could be shown and cause the suppression rule fail to suppress the message. Fixed the problem by allocate memory for the string and add '\0' at the end of the message.
[27 Dec 2010 8:56]
Bugs System
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/127592 3469 He Zhenxing 2010-12-27 BUG#59123 rpl_stm_binlog_max_cache_size fails sporadically with found warnings rpl_stm_binlog_max_cache_size.test fails sporadically on PB2 daily-trunk-bugfixing with binlog-checksum enabled, the expected incident error message was not properly suppressed, and the incident message would have extra strange characters appended. The cause is that the incident error message restored from log event is not properly ended with '\0'. This does not cause problem normally because the error message is at the end of the event, and we add a '\0' at the end of an event. but when binlog-checksum is enabled, the checksum value will be after the message, and so extra strange characters could be shown and cause the suppression rule to fail. Fixed the problem by allocate memory for the string and added '\0' at the end of the message.
[28 Dec 2010 6:48]
Bugs System
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/127614 3469 He Zhenxing 2010-12-28 BUG#59123 rpl_stm_binlog_max_cache_size fails sporadically with found warnings rpl_stm_binlog_max_cache_size.test fails sporadically on PB2 daily-trunk-bugfixing with binlog-checksum enabled, the expected incident error message was not properly suppressed, and the incident message would have extra strange characters appended. The cause is that the incident error message restored from log event is not properly ended with '\0'. This does not cause problem normally because the error message is at the end of the event, and we add a '\0' at the end of an event. but when binlog-checksum is enabled, the checksum value will be after the message, and so extra strange characters could be shown and cause the suppression rule to fail. Fixed the problem by allocate memory for the string and added '\0' at the end of the message.
[29 Dec 2010 4:06]
Zhenxing He
BUG#58967 was marked as dup of this one
[11 Jan 2011 5:15]
Bugs System
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/128380 3481 He Zhenxing 2011-01-11 BUG#59123 rpl_stm_binlog_max_cache_size fails sporadically with found warnings rpl_stm_binlog_max_cache_size.test fails sporadically on PB2 daily-trunk-bugfixing with binlog-checksum enabled, the expected incident error message was not properly suppressed, and the incident message would have extra strange characters appended. The cause is that the incident error message restored from log event is not properly ended with '\0'. This does not cause problem normally because the error message is at the end of the event, and we add a '\0' at the end of an event. but when binlog-checksum is enabled, the checksum value will be after the message, and so extra strange characters could be shown and cause the suppression rule to fail. Fixed the problem by allocate memory for the string and added '\0' at the end of the message.
[11 Jan 2011 6:33]
Bugs System
Pushed into mysql-trunk 5.6.2 (revid:zhenxing.he@sun.com-20110111051323-w2xnzvcjn46x6h6u) (version source revid:zhenxing.he@sun.com-20110111051323-w2xnzvcjn46x6h6u) (merge vers: 5.6.2) (pib:24)
[28 Jan 2011 10:45]
Jon Stephens
Does not appear in a release version. Closed.
Description: rpl_stm_binlog_max_cache_size.test fails sporadically on PB2 daily-trunk-bugfixing with binlog-checksum: pl.rpl_stm_binlog_max_cache_size 'stmt' [ fail ] Found warnings/errors in server log file! Test ended at 2010-12-23 03:52:09 line 101223 5:52:06 [ERROR] Slave SQL: The incident LOST_EVENTS occured on the master. Message: error writing to the binary logĂ ^ Found warnings in /export/home/pb2/test/sb_1-2705689-1293055389.61/mysql-5.6.1-m5-linux-x86_64-test/mysql-test/var-rpl_binlog_checksum/log/mysqld.2.err How to repeat: Check: http://pb2.norway.sun.com/?template=mysql_show_test_failure&search=yes&push_id=1854242&tes... Suggested fix: It seems that the test would expect to get the incident event error message, and it the error is currently not suppressed in the test. So probably the fix is to add suppression of the error message, but not quite understand why it only fails sporadically on the daily run and with binlog-checksum enabled, so may need more investigation to make sure there is no other problems.