Problem with Master-Master MySQL replication of ISPConfig database

[voltron81]

Hi to everybody.
I've two mailservers (Debian 5) based on ISPConfig 3.0.1.6.
This systems are twins, and I've realized them using GlusterFS to replicate /var/vmail and a master-master replication of the ISPConfig database to replicate the ISPConfig configuration.
Few days ago I had a crash on one server, and now that it is up and running again, the replication is not working anymore.
To setup that replication, I've followed this how-to: http://www.howtoforge.com/mysql-5-ma...ation-fedora-8
Running in both server the command SHOW SLAVE STATUS I've this situation:
SRV1:

Slave_IO_Running      Slave_SQL_Running
              YES                    NO

SRV2:

Slave_IO_Running      Slave_SQL_Running
               NO                     YES

Having a look into the log files, I've this files in /var/log/mysql/
SRV1:

mysql-bin.000663  mysql-bin.000667  mysql-bin.000671  mysql-bin.000675
mysql-bin.000664  mysql-bin.000668  mysql-bin.000672  mysql-bin.index
mysql-bin.000665  mysql-bin.000669  mysql-bin.000673
mysql-bin.000666  mysql-bin.000670  mysql-bin.000674

SRV2:

mysql-bin.000495  mysql-bin.000499  mysql-bin.000503  mysql-bin.000507
mysql-bin.000496  mysql-bin.000500  mysql-bin.000504  mysql-bin.index
mysql-bin.000497  mysql-bin.000501  mysql-bin.000505
mysql-bin.000498  mysql-bin.000502  mysql-bin.000506

As far as I can see, it's look like they have lose the syncronization...

Any suggestion?
Thanks
Michele

[falko]

You will have to set up replication from scratch again.

[voltron81]

Hi Falko,
thanks for your reply.
So you mean that I've to cancel the db of SRV2, make a dump of the database from SRV1 and follow the how-to again?

For the future then: there is a way to avoid this kind of problem?

Thanks
Michele

[Toucan]

Originally Posted by voltron81

Hi Falko,
thanks for your reply.
So you mean that I've to cancel the db of SRV2, make a dump of the database from SRV1 and follow the how-to again?

For the future then: there is a way to avoid this kind of problem?

Thanks
Michele

You've used mysql database replication to mirror the db yes?

If you're sure that the database still has the correct data in it you can give the slave the new positions and start it back up again.

I find this tutorial a good guide:

http://www.howtoforge.com/how-to-set...on-ubuntu-9.10