Help with strange email conf behavoir

[matute]

Hi I've installed almost 1.5 years ago an ISP 3.0.3 in a multiple server config with Debian.
I've got server 1 (S1) with ispconfig
server 2 (S2) mysql
server 3 (S3) mail
I did this installation taking as a guide the tutorial named:
Multiserver Setup With Dedicated Web, Email, DNS & MySQL Database Servers On Debian Squeeze With ISPConfig 3

I also bought the user guide. Everything was working fine 'till yesterday. I entered a new mail forward and when I tried to mail that address I found the mails were rejected by S3 with a message stating "Recipient address rejected: User unknown in virtual mailbox table (state 13)."
After looking at everything for many hours I found that the account information was being saved in the mysql in S1 rather than the one in S3 but postfix looks for this info in S3. Now I don't know which is the proper way in which this should work, should it save the info in S3? should it synchronize the info from S1 to S3? I'm not sure about that. Sorry for my poor english and let me know if you need me to clarify anything.

Also I entered a new email account and I see it saved in table mail_user in S1 but I don't see it in table mail_user in S3 and therefore it doesn't work...

Hope someone out there may help me with this issue because it's kind of driving me nuts.
Regards

[till]

If you installed a ispconfig multiservers etup, then the info should get synced automatically between the servers. If the sync dont work, then you can debug it with the instructions here:

http://www.howtoforge.com/forums/showthread.php?t=58408

The most likely reason for your problem is:

1) You changed the mysql configuration of the master so mysql is not listening on all interfaces anymore.
2) A firewall was started on the master server which blocks the mysql port.
3) A hostname or IP address of one of the server ahs been changed or the /etc/hosts file does not contains the information for all servers of this cluster.

[matute]

Till, thanks for the info! I'll try it and let you (and the rest of the readers) the result. I think neither of the 3 things you mention changed, but I wasn't sure how the synchro worked, so now I will read the info in your link and debug it.
Thanks a lot!
Regards.

[matute]

Finally got it working, everything was related to a block in the mysql (the solution is to perform a mysqladmin -flush-hosts -p.
What made me sick for a while was that the connection failing was from the slave to the master, and I was performing the mysqladmin on the slave, but the debug info on the slave only stated the error without indicating where it was trying to connect. And I wasn't able to figure it until I modified server.php and added an echo with the DBHost, DBUser and DBPass. After I performed it on the master server everything started working fine.
Regards, happy new year.