MySql problems

[mishmash]

Following a Red Hat Linux server restart (which may well have enacted some updates) I am unable to start/restart mysql:

/etc/init.d/mysql restart
MySQL PID file could not be found! [FAILED]
Starting MySQL................................... [FAILED]

Simply typing "mysql" gives me

mysql
ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/var/lib/mysql/mysql.sock' (111)

I've looked through various forums etc. and can't find an answer, I've edited /etc/my.cnf
to point to both mysql.sock files I can find on the machine, neither helps.

Any ideas?

 

[Jeewhizz]

What does the mysql error file say? It'll be somewhere like /var/lib/mysql/{hostname}.err

Where {hostname} is the hostname of the machine.

 

[tifosi]

looks like the mysql process that the pid - process ID file holds has died. In the past I've emptied - not removed that file. On rare ocassions I've had to create it as well.

Start and stop mysql to see if it's ok. It may have left tables in a crashed (unclosed correctly) state which need repairing - either through mysqlcheck or myisamchk (with mysql off).

If you use phpMyAdmin then that will have repair options.

I prefer myisamchk (if the db used myisam - default) but you need root/mysql permissions to run that.

S

 

[mishmash]

Thank you for the suggestions - the below is the results from trying them:

What does the mysql error file say?.

Number of processes running now: 0
060902 12:20:31 mysqld restarted
060902 12:20:32 InnoDB: Database was not shut down normally!
InnoDB: Starting crash recovery.
InnoDB: Reading tablespace information from the .ibd files...
InnoDB: Restoring possible half-written data pages from the doublewrite
InnoDB: buffer...
060902 12:20:32 InnoDB: Starting log scan based on checkpoint at
InnoDB: log sequence number 0 1237811.
InnoDB: Doing recovery: scanned up to log sequence number 0 1237811
060902 12:20:32 InnoDB: Flushing modified pages from the buffer pool...
060902 12:20:32 InnoDB: Started; log sequence number 0 1237811
/usr/sbin/mysqld: ready for connections.
Version: '4.1.21-standard' socket: '/var/lib/mysql/mysql.sock' port: 3306 My$

A restart mysql attempt results in

/usr/bin/mysqladmin: connect to server at 'localhost' failed error: 'Can't connect to local MySQL server through socket '/var/lib/mysql/mysql.sock' (111)' Check that mysqld is running and that the socket: '/var/lib/mysql/mysql.sock' exists! mysql has failed, please contact the sysadmin (result was "mysql has failed").

It could be a firewall problem?

I've run the following commands
service iptables stop
iptables --flush

 

[Craig]

Is this a cPanel machine? Do you have a "basedir=" entry in your /etc/my.cnf?

 

[mishmash]

Is this a cPanel machine? Do you have a "basedir=" entry in your /etc/my.cnf?

Yes, this is WHM/cPanel on Red Hat Linux

I do not have a basedir entry in /etc/my.cnf
There is no basedir either in /usr/local/cpanel/whostmgr/my.cnf

 

[Craig]

Ok.. Do you have a symlink in /tmp/ still for mysql.sock?

 

[mishmash]

Ok.. Do you have a symlink in /tmp/ still for mysql.sock?

In /tmp/

I have
mysql.sock@


In response to "locate mysql.sock" (recently having run updatedb) I have:
/var/lib/mysql/mysql.sock
/usr/local/lib/mysql.sock

 

[mishmash]

The result of running the command "ln -s /var/lib/mysql/mysql.sock /tmp/mysql.sock"
Was ln: `/tmp/mysql.sock': File exists

A restart attempt of mysql resulted in the same error as before.

This error was the same following the running of the command: "ln -s -f /var/lib/mysql/mysql.sock /tmp/mysql.sock

 

[yesterday]

are you starting mysql as the right user?

yesterday

 

[mishmash]

are you starting mysql as the right user?

yesterday

I have tried to start it via the WHM control pannel

And as root trying to start mysqld_safe

 

[mishmash]

I tried running mysqlup (updating mysql without success)

su mysql
bash-3.00$ /usr/sbin/mysqld
080627 15:35:22 [ERROR] Can't start server : Bind on unix socket: Permission denied
080627 15:35:22 [ERROR] Do you already have another mysqld server running on socket: /usr/local/lib/mysql.sock ?

No other mysqld :

killall -9 mysqld
mysqld: no process killed

I don't know about the "permission denied".

 

[mishmash]

I don't know about the "permission denied".

Permissions on the mysql.sock file look fine

srwxrwxrwx 1 mysql mysql 0 Jun 26 21:02 mysql.sock=

 

[Jeewhizz]

as silly as it sounds, in the past, a reboot has fixed this... Have you tried that?

However, there are a few issues with cpanel / mysql etc. Try this page: Fun Things in Life Blog Archive WHM/CPanel MySQL 5 Problem

 

[mishmash]

I have tried a number of restarts

I have ensured there are no basedir entries in mysql.cnf

I have also performed a forced upgrade of MySQL and restarted.

 

[yesterday]

I don't think you should be able to start mysql as root very easily - I think you should perhaps be the mysql user or another.

forget that I cant find any evidence online to support my theory!

yesterday

 

[yesterday]

this link:
Cannot start mysqld and mysql - Usenet Forums
suggests (last post) that you delete the mysql.sock file and let mysql_safe rebuild it.

Might be worth a try (probably rename or move rather than delete - just in case!)

yesterday

 

[Jeewhizz]

Alternatively, try looking on the cpanel forums, as you won't have been theonly person with that problem

 

[mishmash]

Craig's team at UK Dedicated Servers, Managed Dedicated Servers, UK Managed Web hosting, cPanel Hosting & Business hosting with Passionate Support solved the problem by commenting out the line:

socket = /usr/local/lib/mysql.sock

In mysql.cnf