Friday, December 01, 2006

relay_log_space_limit stopped working after replication reset?

I recently reset replication on one of my servers and since then my relay_log_space_limit=8G setting has stopped working since the relay log files continue to grow beyond the 8G limit (currently 15G).

Has anyone else experienced this before? Why would it suddenly stop honoring the relay log space limit?

5 comments:

Mike Kruckenberg said...

Hey Frank,

Does it show correctly when you look at if from within the client? Try something like:

show variables like '%relay%';

Just curious if it's something with the server not picking up options in the my.cnf or if there is a problem within mysqld of it not being respected.

Mike

Frankly Speaking! said...

Hi Mike,

I should have included the output earlier but here it is:


mysql> SHOW variables like '%relay%';
+-----------------------+------------+
| Variable_name | Value |
+-----------------------+------------+
| max_relay_log_size | 0 |
| relay_log_purge | ON |
| relay_log_space_limit | 8589934592 |
+-----------------------+------------+
3 rows in set (0.00 sec)

mysqld is recogizing the limit but not enforcing it as I see 18 log files (18G) already.

To prevent running out of disk space, I have turned off SLAVE IO_THREAD.

Thanks for any insight you can provide.
Frank

Anonymous said...

If you want manage the database you should check the MySQL Turbo Manager product.

Product page:
www.mentattech.com

Kind regards,

Teksure

Anonymous said...

If you want to manage, tune or develop MySQL you should check MySQL Turbo Manager.

Product page:
www.mentattech.com

Regards,

Teksure

Sheeri said...

MySQL should FLUSH the relay logs when you reset the replication information; I'm guessing it didn't.

Are those relay logs all current and in order? My guess is it's keeping around some older files it never purged from when you reset the replication information and that current relay log flushing is working appropriately.