Hello I have two problems which might or might not be related. After my VPS ran out of memory last night I had to ask my provider to restart it for me. When it came back up I then had SSH and FTP access but I noticed that I was getting mail errors: "Sending of password did not succeed. Mail server responded: chdir Maildir failed" After going through the forums here I saw that one of the recommendations was to check the ISPConfig (I'm running v2) to make sure that the maildir option was checked. It was then that I noticed mysql was not running as I couldn't access the ISPConfig webpage. I checked the syslog and saw: Code: Oct 6 08:33:11 first mysqld_safe[3154]: started Oct 6 08:33:12 first mysqld[3157]: 101006 8:33:12 InnoDB: Started; log sequence number 0 43705 Oct 6 08:33:12 first mysqld[3157]: 101006 8:33:12 [ERROR] Can't start server : Bind on unix socket: Permission denied Oct 6 08:33:12 first mysqld[3157]: 101006 8:33:12 [ERROR] Do you already have another mysqld server running on socket: /var/run/mysqld/mysqld.sock ? Oct 6 08:33:12 first mysqld[3157]: 101006 8:33:12 [ERROR] Aborting Oct 6 08:33:12 first mysqld[3157]: Oct 6 08:33:12 first mysqld[3157]: 101006 8:33:12 InnoDB: Starting shutdown... Oct 6 08:33:14 first mysqld[3157]: 101006 8:33:14 InnoDB: Shutdown completed; log sequence number 0 43705 Oct 6 08:33:14 first mysqld[3157]: 101006 8:33:14 [Note] /usr/sbin/mysqld: Shutdown complete Oct 6 08:33:14 first mysqld[3157]: Oct 6 08:33:14 first mysqld_safe[3185]: ended Oct 6 08:33:26 first /etc/init.d/mysql[3304]: 0 processes alive and '/usr/bin/mysqladmin --defaults-file=/etc/mysql/debian.cnf ping' resulted in Oct 6 08:33:26 first /etc/init.d/mysql[3304]: #007/usr/bin/mysqladmin: connect to server at 'localhost' failed Oct 6 08:33:26 first /etc/init.d/mysql[3304]: error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)' Oct 6 08:33:26 first /etc/init.d/mysql[3304]: Check that mysqld is running and that the socket: '/var/run/mysqld/mysqld.sock' exists! Oct 6 08:33:26 first /etc/init.d/mysql[3304]: Again, looking through the forums here I did the following: ran updatedb ran locate mysqld.sock - but it came up blank shutdown mysql and renamed my.cnf but mysql would not then start. So I backed up my data files and removed mysql and then reinstalled but it must have kept the config files as it didn't ask me to set a password during the install and I had the same error when I tried to start it. So repeated the process with apt-get --purge remove but still had the same issue (again didn't ask me to set a password as it does in a clean install). I shutdown the server and did an offline recovery through HyperVM to retrieve the mysqld.sock file but now the server is not responding and I have had to send a support request to ask for another hard reset. So now I am stumped - does anyone have any ideas? If I can get mysql running then I will be able to check my ISPConfig setup (although I know that I had the maildir option checked as it's been running fine for months). By the way its Debian Lenny and I used the Perfect Server guide to setup apache, courier, proftpd etc. regards and many thanks if you've got this far, Andrew
Solved After many dead-ends, it turned out to be a file permissions issue where, among other things, mysql was not the owner of /var/run/mysqld/. Hope this helps someone else. regards Andrew
