I have a Centos server which is running httpd, mysql, memcached and postfix. Normally, my server load is OK
Code:
Usage: /usr/sbin/httpd [-D name] [-d directory] [-f file]
top - 07:43:55 up 9:38, 1 user, load average: 1.04, 1.08, 1.11
Tasks: 236 total, 1 running, 234 sleeping, 0 stopped, 1 zombie
Cpu(s): 4.2%us, 1.0%sy, 0.0%ni, 93.6%id, 1.1%wa, 0.0%hi, 0.1%si, 0.0%st
Mem: 8015636k total, 5743920k used, 2271716k free, 254304k buffers
Swap: 8392696k total, 0k used, 8392696k free, 4224184k cached
But sometimes it crashes ( httpd is not working, can't connect ssh to server ) and I need to ask the datacenter to reboot it to be able to ssh.
After reconnecting to the server, I check with the top command and see that the server load average is still high ( more than 100 ), CPU ( wait ) is more than 70% , and there are too many httpd process in D state.
Now I'm stuck on this problem. Can anyone offer some advice?