I have installed an OpenVZ server on Debian Squeeze. The virtual server has 8GB RAM, 586GB hard disk space, and I'm having forking issues with amavis. At times I go into the virtual server using vzctl enter mail, and begin getting fork errors right away. Additionally there seems to be a lock issue with mysql. Here are the results of: netstat -tap | grep mysql root@mail:/# netstat -tap | grep mysql tcp 0 0 *:mysql *:* LISTEN 1393/mysqld tcp 0 0 localhost:34323 localhost:mysql ESTABLISHED 2090/smtp tcp 0 0 localhost:mysql localhost:34322 ESTABLISHED 1393/mysqld tcp 0 0 mail.domain:52579 www.domain.:mysql TIME_WAIT - tcp 0 0 localhost:mysql localhost:34326 ESTABLISHED 1393/mysqld tcp 0 0 localhost:34322 localhost:mysql ESTABLISHED 2089/smtp tcp 0 0 localhost:mysql localhost:34323 ESTABLISHED 1393/mysqld tcp 0 0 localhost:34326 localhost:mysql ESTABLISHED 2094/error Here is the tail of my mail error log: Mar 28 09:23:31 mail postfix/trivial-rewrite[1334]: fatal: proxy:mysql:/etc/postfix/mysql-virtual_domains.cf(0,lock|fold_fix): table lookup problem Mar 28 09:23:33 mail postfix/trivial-rewrite[1413]: fatal: proxy:mysql:/etc/postfix/mysql-virtual_domains.cf(0,lock|fold_fix): table lookup problem I don't understand why I'm still having issues with 8GB of memory allocated: root@mail:/# free total used free shared buffers cached Mem: 8388608 582428 7806180 0 0 0 -/+ buffers/cache: 582428 7806180 Swap: 0 0 0 Can anyone help? Thanks, Darin
Have a look in the syslog and your beancounters and look for resource shortages. Allocating 8GB isn't enough for openvz, you'll probably need to increase kmemsize, and the various tcpsnd|rcvbuf will need upped as well.
Hi Tod, So my hardware node has 24GB I am running 8GB on the mail server node only. You're saying that isn't enough? Wow, I thought that should be more than enough. What is considered enough? Thanks, Darin
