| Bug #17741 | readlock persists after mysqldump | ||
|---|---|---|---|
| Submitted: | 27 Feb 2006 14:38 | Modified: | 17 Nov 2011 16:36 |
| Reporter: | Bart Verwilst | Email Updates: | |
| Status: | No Feedback | Impact on me: | |
| Category: | MySQL Server | Severity: | S1 (Critical) |
| Version: | 5.0.18 | OS: | Linux (Linux) |
| Assigned to: | CPU Architecture: | Any | |
[27 Feb 2006 14:45]
Valeriy Kravchuk
Thank you for a problem report. Do you use MyISAM tables or tables with different storage engines? Can you provide a simple test case?
[28 Feb 2006 9:01]
Bart Verwilst
I was using a mix of MyISAM and InnoDB.. But since last night i switched to InnoDB only. When i'm starting a backup, the processlist shows "FLUSH TABLES WITH READ LOCK" and it stays like that, even when the backup is done. Doing "UNLOCK TABLES;" at the prompt doesn't help, only killing the server and restarting it clears the lock.
[28 Feb 2006 9:02]
Bart Verwilst
For the test sample, i could supply my structure, and an overview of how much data is in each of the tables ( 360Meg total for now ) ?
[6 Apr 2006 16:11]
Valeriy Kravchuk
Sorry for a long delay with this bug report... Please, try to repeat with a newer version of server, 5.0.19. If the problem is still obviously repeatable, please, upload the SHOW CREATE TABLE results for your databases (no need for data yet).
[6 May 2006 23:00]
Bugs System
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".
[13 Jul 2007 6:54]
Oliver Hookins
I think I am experiencing this bug. Platform is RHEL4 with the Redhat Application Stack Beta 1.0 Channel version of MySQL (5.0.18-4.el4s1.1). We have a backup script that runs every 10 minutes. In one thread it issues 'FLUSH TABLES WITH READ LOCK", then performs a mysqldump without using locks, then issues 'UNLOCK TABLES' in the first thread. The server is a replication slave, and we notice the locks not being removed since the slave sql server thread reports "waiting on read lock", and can't process any updates or inserts from the master. The only way to remove the locks seems to be to restart mysql server, the original thread that locked the tables is no longer connected.
[14 Jul 2007 20:07]
Valeriy Kravchuk
All reporters: Please, try to repeat with a newer version, 5.0.45, and inform about the results.
[14 Aug 2007 23:00]
Bugs System
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".
[7 Mar 2008 23:55]
James Day
This was reported in 5.1 in bug #29088, reproduced with 5.1.19 and found not to be reproducible with 5.1.20. I couldn't reproduce with 5.0.54a. So probably fixed. If you can reproduce in 5.0.54a or later please let us know. Requires both master-data and server option log.
[17 Oct 2011 12:02]
Minto Mohan
Got the error in 5.1.44-community-log after the mysqldump was completed, the processlist displayed queries waiting for release of readlock. +-----+--------+-----------------------+---------------+---------+------+---------------------------------+------------------------------------------------------------------------------------------------------+ | Id | User | Host | db | Command | Time | State | Info | +-----+--------+-----------------------+---------------+---------+------+---------------------------------+------------------------------------------------------------------------------------------------------+ | 4 | myuser | C3002:4101 | schema_00001 | Query | 176 | Waiting for release of readlock | update scheduler set PRIORITY=12, VERSION=null, CREATED_BY='SCHEDULER', CREATED_ON='2011-10-14 17:40 | | 5 | myuser | C3002:4104 | schema_00002 | Sleep | 25 | | NULL | | 6 | myuser | C3002:4103 | schema_00002 | Sleep | 25 | | NULL | | 7 | myuser | C3002:4102 | schema_00002 | Sleep | 25 | | NULL | | 11 | myuser | C3002:4130 | schema_00001 | Query | 176 | Waiting for release of readlock | update scheduler set PRIORITY=3, VERSION=null, CREATED_BY='ADMIN', CREATED_ON='2011-10-15 09:48:47', | | 12 | myuser | C3002:4131 | schema_00001 | Query | 176 | Waiting for release of readlock | insert into scheduler_job_history (REMARKS, STATUS_FLG, SCHEDULER_ID, JOB_NM, JOB_START_TIME, JOB_EN | | 13 | myuser | C3002:4133 | schema_00002 | Sleep | 25 | | NULL | | 14 | myuser | C3002:4134 | schema_00001 | Query | 176 | Waiting for release of readlock | insert into scheduler_job_history (REMARKS, STATUS_FLG, SCHEDULER_ID, JOB_NM, JOB_START_TIME, JOB_EN | | 15 | myuser | C3002:4132 | schema_00001 | Query | 176 | Waiting for release of readlock | insert into scheduler_job_history (REMARKS, STATUS_FLG, SCHEDULER_ID, JOB_NM, JOB_START_TIME, JOB_EN | | 16 | myuser | C5141:50728 | schema_00001 | Sleep | 3 | | NULL | +-----+--------+-----------------------+---------------+---------+------+---------------------------------+------------------------------------------------------------------------------------------------------+
[17 Oct 2011 16:36]
Valeriy Kravchuk
Please, check if this problem ever happens with recent versions, 5.0.91+ or 5.1.59+.
[18 Nov 2011 7:00]
Bugs System
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".
Description: I'm running replication. When doing a backup on the master with "mysql --master-data ..." all tables are locked. When the backup is done, all tables stay locked: | 10973 | avas | 192.168.139.145:38708 | avas | Query | 1458 | Waiting for release of readlock | INSERT INTO awl (username,email,ip,count,totscore) VALUES ('amavis','ecolinet@aghfgfghfhg.be','195. | | 10981 | avas | 192.168.139.149:41950 | avas | Query | 1458 | Waiting for release of readlock | INSERT INTO awl (username,email,ip,count,totscore) VALUES ('amavis','eswwhwpuglx@hotmail.com','151.2 | | 10982 | avas | 192.168.139.149:41951 | avas | Query | 1458 | Waiting for release of readlock | INSERT INTO awl (username,email,ip,count,totscore) VALUES ('amavis','yuuaywyrcgwh@thedarkhorses.com' | | 10991 | avas | 192.168.139.145:38741 | avas | Query | 1457 | Waiting for release of readlock | UPDATE awl SET count = count + 1, totscore = totscore + | Only restarting the server clears the locks.. If you need any more info, don't hesitate to ask! How to repeat: mysqldump --master-data on master Suggested fix: unknown