MariaDB suddly not starting

[zeke]

I just upgraded packages today, and now I cannot start MariaDB. This was working fine on Friday before I upgraded. I have not changed anything in my.cnf since then, and I confirmed that the directory where the databases live is mounted.

Code: Select all

[root@localhost etc]# systemctl status  mysqld.service
mysqld.service - MySQL database server
          Loaded: loaded (/lib/systemd/system/mysqld.service; enabled)
          Active: failed (Result: exit-code) since Mon, 19 Nov 2012 15:45:46 -0600; 1s ago
         Process: 1834 ExecStartPre=/usr/sbin/mysqld-prepare-db-dir (code=exited, status=1/FAILURE)
          CGroup: name=systemd:/system/mysqld.service


Not really sure what to look for next....

[tom_]

in your /etc/my.conf there should be an option to raise the log level.

then you can check the log file in /var/log

[zeke]

I do not see the option to increase the log level?

The most recent entry in the log appears to be from several days ago:

Code: Select all

121108 08:29:21 mysqld_safe Starting mysqld daemon with databases from /home/brett/ginsburg/databases
121108  8:29:23 InnoDB: The InnoDB memory heap is disabled
121108  8:29:23 InnoDB: Mutexes and rw_locks use GCC atomic builtins
121108  8:29:23 InnoDB: Compressed tables use zlib 1.2.6
121108  8:29:23 InnoDB: Using Linux native AIO
121108  8:29:23 InnoDB: Initializing buffer pool, size = 128.0M
121108  8:29:23 InnoDB: Completed initialization of buffer pool
121108  8:29:23 InnoDB: highest supported file format is Barracuda.
121108  8:29:25  InnoDB: Waiting for the background threads to start
121108  8:29:27 Percona XtraDB (http://www.percona.com) 1.1.8-26.0 started; log sequence number 577174709
121108  8:29:27 [Note] Plugin 'PBXT' is disabled.
121108  8:29:28 [Note] Event Scheduler: Loaded 0 events
121108  8:29:28 [Note] /usr/sbin/mysqld: ready for connections.
Version: '5.5.25-MariaDB'  socket: '/var/lib/mysql/mysql.sock'  port: 0  Mageia - MariaDB Community Edition (GPL)
121112 10:03:59 [Note] /usr/sbin/mysqld: Normal shutdown

121112 10:03:59 [Note] Event Scheduler: Purging the queue. 0 events
121112 10:04:00  InnoDB: Starting shutdown...
121112 10:04:05  InnoDB: Shutdown completed; log sequence number 577519070
121112 10:04:05 [Note] /usr/sbin/mysqld: Shutdown complete

121112 10:05:25 mysqld_safe Starting mysqld daemon with databases from /home/brett/ginsburg/databases
121112 10:05:26 InnoDB: The InnoDB memory heap is disabled
121112 10:05:26 InnoDB: Mutexes and rw_locks use GCC atomic builtins
121112 10:05:26 InnoDB: Compressed tables use zlib 1.2.6
121112 10:05:26 InnoDB: Using Linux native AIO
121112 10:05:26 InnoDB: Initializing buffer pool, size = 128.0M
121112 10:05:26 InnoDB: Completed initialization of buffer pool
121112 10:05:26 InnoDB: highest supported file format is Barracuda.
121112 10:05:28  InnoDB: Waiting for the background threads to start
121112 10:05:31 Percona XtraDB (http://www.percona.com) 1.1.8-26.0 started; log sequence number 577519070
121112 10:05:31 [Note] Plugin 'PBXT' is disabled.
121112 10:05:32 [Note] Event Scheduler: Loaded 0 events
121112 10:05:32 [Note] /usr/sbin/mysqld: ready for connections.
Version: '5.5.25-MariaDB'  socket: '/var/lib/mysql/mysql.sock'  port: 0  Mageia - MariaDB Community Edition (GPL)
121116 14:51:13 [Note] /usr/sbin/mysqld: Normal shutdown



If I delete /etc/my.cnf I can start mysql. Also, if I change the data directory to /var/lib/mysql. But, I had my databases in a samba share that is mounted in my home directory. So of course I do not see my existing databases with these two work-arounds.

I do notice that all of the permissions are root:root, and if I try to chown to mysql:mysql, it reports success but no change in ownership occurs.

[oj]

If the updates that broke your db included samba packages, I'd assume that's where the problem is.