JJB Blog

do you know which user the safe daemon is running with? Maybe it’s not ‘mysql’? Anyway, can you just chmod -R 777 * in the data dir (or even the whole mysql dir), then pull back perms after that?

Posted by josh on 30 January 2005 @ 4pm

hmm, well it dies differently now. i’ve seen this behvior before. i’m not sure what the circumstances are.

john@g5bachir:~$ sudo chmod -R 777 /usr/local/mysql
john@g5bachir:~$ mysqld_safe
Starting mysqld daemon with databases from /usr/local/mysql/data
STOPPING server from pid file /usr/local/mysql/data/g5bachir.local.pid
050130 17:20:22 mysqld ended

Posted by JJB on 30 January 2005 @ 5pm

try this:

sudo mysqld_safe

to start it.

mysqld_safe probably changes it’s effective user id, which may be something only root can do.

Posted by Troy Johnson on 1 February 2005 @ 3pm

If that works, you can probably set the file ownership back to their original values and see if it still works. It will be a bit more secure that way.

Posted by Troy Johnson on 1 February 2005 @ 3pm

Hi troy (the first non-acquaintance commenter on my blog )

That doesn’t work either:

john@g5bachir:~$ sudo mysqld_safe
Password:
Starting mysqld daemon with databases from /usr/local/mysql/data
STOPPING server from pid file /usr/local/mysql/data/g5bachir.local.pid
050202 03:01:14 mysqld ended

john@g5bachir:~$ su -
Password:
g5bachir:~ root# /usr/local/mysql/bin/mysqld_safe
Starting mysqld daemon with databases from /usr/local/mysql/data
STOPPING server from pid file /usr/local/mysql/data/g5bachir.local.pid
050202 03:01:41 mysqld ende

Posted by JJB on 2 February 2005 @ 3am

here’s some more info. from the error log (and no, i don’t have another mysqld running):

# cat /usr/local/mysql/data/127.0.0.1.err
050205 14:18:37 mysqld started
050205 14:18:37 [Warning] Setting lower_case_table_names=2 because file system for /usr/local/mysql/data/ is case insensitive
050205 14:18:37 [ERROR] Can’t start server : Bind on unix socket: Permission denied
050205 14:18:37 [ERROR] Do you already have another mysqld server running on socket: /tmp/mysql.sock ?
050205 14:18:37 [ERROR] Aborting

050205 14:18:37 [Note] /usr/local/mysql/bin/mysqld: Shutdown complete

050205 14:18:37 mysqld ended

Posted by JJB on 5 February 2005 @ 2pm

It’s a known bug.
And it hasn’t been fixed, no matter what this thread claims.
http://bugs.mysql.com/bug.php?id=7535

Posted by Peter Pan on 6 February 2005 @ 9pm

Check out this link for my solution

Posted by JJB on 6 February 2005 @ 10pm

this link

Posted by JJB on 6 February 2005 @ 10pm

Great. Thats the solution. Spread the word!

Posted by Peter Pan on 7 February 2005 @ 8pm

I had the same problem. Seems I already had another MySQL process running somehow! After killing it, I was able to run mysqld_safe successfully.

Posted by Trevor on 7 June 2005 @ 2am

I fixed this problem on my server.

Somehow when I stopped mysqld, the owner of /var/run/mysqld (the location of the pid) was obliterated — the folder was owned by no one. I chown’d it back to mysql and mysqld started up normally. Hope this helps someone.

Posted by Robert on 4 January 2008 @ 11am