In my case there was a dstat process which was scheduled by cron went viral and due to this the system start throwing the error as
Jun 28 09:49:01 machine
kernel: VFS: file-max limit 10000 reached
Jun 28 09:49:01 machine kernel: VFS: file-max limit 10000 reached
Jun 28 09:50:01 machine
kernel: VFS: file-max limit 10000 reached
Jun 28 09:50:01 machine
kernel: VFS: file-max limit 10000 reached
By login as root and running the below command you can get the no if times it run with PID.
now it will be very time consuming if you kill each pid one by one.
better approach
# ps -ef | grep dstat | grep -v grep | awk '{print $2}'
then pass the command to the kill -9 and it will kill all the PID related to this process.
# kill -9 `ps -ef | grep dstat | grep -v grep | awk '{print $2}'`
To kill a process
ReplyDelete$ kill 11900
or
$ kill -9 11900
(-9 is signal 9 SIGKILL)
http://namhuy.net/453/ubuntu-how-to-kill-a-process.html