Urgent request as servers are having mysql recurring problem

[aushost]

In a previous post I outline that the mysql.sock file is mysteriously deleting itself and leaving an error message about TCP/ip conflict with port 3306. As most databases run on port 3306 how can this be changed. I now have this problem with 3 servers and one server does not have a database on that server.
DA is 1.2.2 Could it be a control panel fault? the OS is Red Hat 9.
I am at my wits end as thishas been going on for a week now.

[toml]

Sounds like perhaps DA does not believe MySQL is running and attempting to restart it. That could wipe out the mysql.sock and give the address already in use error.

Some things that might make DA believe are permissions on the /proc filesystem (just read a post about this) or if MySQL was not running (core dump).

I would scan all your log files to see if you can find something that gives you any hints. Look in /var/log/directadmin and even the /var/lib/mysql/www.xxxx.xxx.err file, those should tell you the most about what is going on.

[aushost]

In the /var/lib/mysql directory the server.name.err error file reads:

mysqld started

Cant start server: Bind on TCP/IP port: address already in use

Do you already have another mysqld server running on port 3306

Aborting

/usr/bin/mysqld: Shutdown complete

mysqld ended.

Does this help......

[aushost]

Can I have an answer from Directadmin support on this problem I have three servers down for the last 36 hours.

[skruf]

Hey,

If you haven't already, send an email to "support@directadmin.com".

Realize that they are in the final stages of releasing the next version of DA so they are focusing on that and probably not checking the forums as much as they usually do...

David

[toml]

Have looked at the log files in /var/log/directadmin? Like the error.log or errortasq.log, they may have something in there as to why it thought they were down.

[aushost]

Looked at the error files all mention is mysql tried to start but aborted no mention of cause.

[CrazyMouse]

question. What does the your Service Monitor say?

and are they really down? or you can use mysql? and the server is trying to start it up again and again and again?

[aushost]

On manual reboot mysql starts up then stops after 10-20 minutes gives the same error message as above, then apache and named goes down. 10-20 minutes later the same process is executed.

[skruf]

Hey,

Does this show anything?

ps ax | grep mysql

And, since you have three servers with problems, I'd have to ask what was the last thing you did to each of those servers?

Or, did this just suddenly start?

David

[aushost]

rebooted them...the situation started over a week ago with one of the servers with 24 hours the thgree developed the same problem. Over the last 36 hours its got worse.

[toml]

If you are having all these services going down, perhaps you have something bigger happening. Are you running a process monitor that kills process if their load gets to high? I can see that killing mysql and apache, but named shouldn't.

As skruf said, is this something that just recently started happening? If they are new boxes, what are the exact steps you took to install all the software. What software did you install besides RH9/DA.

[aushost]

No other software installed besides RH9 and DA....I've sent an email off to support see if they know of any reasons plus I will grep mysql as suggested to see what happens...Thanks to all for suggestions and help it is appreciated.

[toml]

Any new info on why this was happening to your servers? Just in case someone else runs into a similar problem, it is good to have a solution posted in the forum.

[aushost]

John At DA Support went into the box and this was his conclusions:
"It looks like when mysql was being stopped/started through DA, it was loading up too slowly (more than 10 seconds) so the taskq thought it wasn't working and tried to restart it. The best bet is to control it from the command line."

Many thanks to one and all for your guidance and help.