High CPU Usage Issue

uk01

• December 15, 2018 19:22

Hi there, For the last 2 days we've spent many hours on a high cpu issue. Looking at the top command and also processes in WHM, nothing obvious like usually you see. We've tried csf port flood settings, etc etc Tonight even moved 17 accounts off the server and migrated them just so we have less to work with. The server continuously runs at 75-90% cpu (checking in vcenter) - it usually uses around 20-30% Very odd, it's almost as if there is something running at root level all the time but the only thing which shows up in top is cdp (r1soft) for a few minutes which is expected. We've even installed a trial of cloudlinux, but even that doesn't show what's using the cpu... Any ideas? :-( My brain hurts and I'm used to this server admin.

• 

  GOT

  • December 16, 2018 13:01

  So, CPU load in top on the VM is very low, but the CPU usage reported in VCenter is high? You check other VM's running in there? Are most of them reporting high?

  0
• 

  uk01

  • December 16, 2018 14:10

  In vcemter that vm shows as high, but the others show as normal usage. Network packets for that vm also show higher The load in WHM shows as 4-6 Yet top doesn"t show anything obvious using the cpu. Very odd Early this morning it suddenly stopped, load dropped straight down, several hours ok, then back up.

  0
• 

  GOT

  • December 16, 2018 14:26

  You check netstat to see what is connecting?

  0
• 

  uk01

  • December 16, 2018 14:38

  I did at some stage, just checked again with a different command netstat -at Then noticed load has dropped again 10 mins ago, boom, down from 80% to 10%. I"ll check Netstat again when it shoots up

  0
• 

  HostSane

  • December 17, 2018 05:08

  Go through the below article and it will help you out

  0
• 

  uk01

  • December 17, 2018 07:49

  Thanks, that"s one of the pages I"ve already referenced in the 24 hours before posting here.

  0
• 

  cPanelMichael

  • December 17, 2018 20:25

  Hello @uk01, Do you notice any output to /var/log/messages or /var/log/dmesg that could indicate an issue with the hardware? Thank you.

  0
• 

  uk01

  • December 17, 2018 20:32

  I didn"t check, but will do next time load goes up. The other vm"s are ok though on the same host. Weirdly the load has been stable for about 24hrs or so, no idea what caused it but it seemed to disappear overnight. If it comes back I"ll check those folders

  0
• 

  uk01

  • December 17, 2018 20:33

  Ps thank appreciate all help

  0

Please sign in to leave a comment.