Bug #33123 | wait_timeout ignored in /etc/my.conf | ||
---|---|---|---|
Submitted: | 10 Dec 2007 20:46 | Modified: | 1 Aug 2009 7:17 |
Reporter: | Thomas Okken | Email Updates: | |
Status: | Not a Bug | Impact on me: | |
Category: | MySQL Server: General | Severity: | S3 (Non-critical) |
Version: | 5.0.45 | OS: | Linux (x86_64-icc-glibc23) |
Assigned to: | CPU Architecture: | Any | |
Tags: | wait_timeout |
[10 Dec 2007 20:46]
Thomas Okken
[10 Dec 2007 23:53]
Sveta Smirnova
Thank you for the report. Please provide output of show global variables like '%timeout%';
[12 Dec 2007 15:20]
Thomas Okken
[mysql@ss-1 ~]$ mysql -u root -p [...] mysql> show global variables like '%timeout%'; +----------------------------+--------+ | Variable_name | Value | +----------------------------+--------+ | connect_timeout | 5 | | delayed_insert_timeout | 300 | | innodb_lock_wait_timeout | 50 | | innodb_rollback_on_timeout | OFF | | interactive_timeout | 28800 | | net_read_timeout | 30 | | net_write_timeout | 60 | | slave_net_timeout | 3600 | | table_lock_wait_timeout | 50 | | wait_timeout | 100000 | +----------------------------+--------+ 10 rows in set (0.00 sec) mysql> show variables like '%timeout%'; +----------------------------+-------+ | Variable_name | Value | +----------------------------+-------+ | connect_timeout | 5 | | delayed_insert_timeout | 300 | | innodb_lock_wait_timeout | 50 | | innodb_rollback_on_timeout | OFF | | interactive_timeout | 28800 | | net_read_timeout | 30 | | net_write_timeout | 60 | | slave_net_timeout | 3600 | | table_lock_wait_timeout | 50 | | wait_timeout | 28800 | +----------------------------+-------+ 10 rows in set (0.00 sec) mysql>
[29 Dec 2007 19:29]
Valeriy Kravchuk
Please, send the entire /etc/my.cnf, including [mysql] and other sections.
[31 Dec 2007 14:59]
Thomas Okken
/etc/my.cnf
Attachment: my.cnf (application/octet-stream, text), 81 bytes.
[31 Jan 2008 23:25]
Sveta Smirnova
Thank you for the feedback. Do you start mysqld with help of some script?
[1 Feb 2008 18:11]
Thomas Okken
My MySQL start script
Attachment: mysql (application/octet-stream, text), 10.89 KiB.
[1 Feb 2008 18:13]
Thomas Okken
I copied the mysql-5.0.45-linux-x86_64-icc-glibc23/support-files/mysql.server script to /etc/rc.d/init.d/mysql, and created the appropriate links to have it started in runlevels 2, 3, and 5. I attached a copy of that script for reference. Note that I used the script straight from the MySQL package, with no modifications.
[25 Apr 2008 17:45]
Sandeep Dubey
Is there any update on this issue? I am also hitting the similar issue. I set global wait_timeout, close the session and reconnect to MySQL. I get the new global wait_timeout but session wait_timeout is reset to default. The variable set as global should be visible to new connection as session variable? Is this correct assumption? I see otherwise in MySQL 5.0.45. Thanks Sandeep Dubey
[13 May 2008 13:40]
Valeriy Kravchuk
Please, specify how you check wait_timeout value for a new session? Using mysql client? Please, read the manual also, http://dev.mysql.com/doc/refman/5.0/en/server-system-variables.html#option_mysqld_wait_tim...: "On thread startup, the session wait_timeout value is initialized from the global wait_timeout value or from the global interactive_timeout value, depending on the type of client (as defined by the CLIENT_INTERACTIVE connect option to mysql_real_connect()). See also interactive_timeout."
[13 Jun 2008 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".
[19 Mar 2009 19:55]
Thomas Okken
Valeriy, I checked the wait_timeout value using the mysql client, as described in my message dated 12 Dec 2007 16:20. I added the line "wait_timeout=100000" in /etc/my.cnf; I provided my version of /etc/my.cnf with my message dated 31 Dec 2007 15:59; I provided my version of the mysqld start script, which is simply an unmodified copy of BASEDIR/support-scripts/mysql.server. I have read the manual page you mentioned, and many other manual pages and mailing list articles as well, but I have found nothing to indicate why setting wait_timeout in /etc/my.cnf would fail to take effect in new sessions. In fact, the manual indicates that setting of the global wait_timeout (which works, see the second set of results in my message dated 12 Dec 2007 16:20) would be used as the default for new sessions, but in fact, new sessions get wait_timeout set to 28800, and I am at a loss as to where that value comes from, and why it doesn't use the one that I provide.
[20 Mar 2009 6:50]
Sveta Smirnova
Thank you for the feedback. Please provide output of SELECT @@global.wait_timeout, @@session.wait_timeout;
[20 Apr 2009 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".
[20 Jul 2009 12:25]
Gabriel Rossetti
I have th same problem with version 5.0.51a-3ubuntu5.4 on linux. Here is the output of your last inquiry : mysql> SELECT @@global.wait_timeout, @@session.wait_timeout; +-----------------------+------------------------+ | @@global.wait_timeout | @@session.wait_timeout | +-----------------------+------------------------+ | 30 | 28800 | +-----------------------+------------------------+ 1 row in set (0.00 sec) mysql> As you see, I set it to 30, but the session timeout is set back to the default value. I restarted the server of course.
[20 Jul 2009 12:31]
Gabriel Rossetti
ok, io get it, since I am using an interactive client, it uses the interactive_timeout instead.
[22 Jul 2009 6:30]
Sveta Smirnova
Gabriel, thank you for the feedback. So in your case this is not a bug. Setting to "Need feedback" while waiting answer from Thomas.
[31 Jul 2009 23:17]
Thomas Okken
When I run Sveta's query in the mysql CLI, I get this: mysql> select @@global.wait_timeout, @@session.wait_timeout; +-----------------------+------------------------+ | @@global.wait_timeout | @@session.wait_timeout | +-----------------------+------------------------+ | 100000 | 28800 | +-----------------------+------------------------+ 1 row in set (0.00 sec) mysql> When I run it over a JDBC connection (using the same user and database), both values are 100000. So, it looks like I'm OK, too, after all; I didn't realize that the timeouts were different in the CLI.
[1 Aug 2009 7:17]
Sveta Smirnova
Thomas, thank you for the feedback. Set to "Not a Bug", because last comment..