Forum Discussion

Eduardo_de_Oliv's avatar
Eduardo_de_Oliv
Icon for Nimbostratus rankNimbostratus
Feb 11, 2015

High consumption of CPU after upgraded to 11.6.0

Hello, could anyone help me?

 

We have a couple of Bigip running with LTM, Webaccelerator and ASM, that is running in Failover. Yesterday we upgrade the version from 11.5.1 Build 5.0.147 to 11.6.0 Build 3.0.412 but now I'm getting high CPU consumption and in less than 24 hours after upgrading my active Bigip did fail over.

 

Follow bellow the messagens logs when the problem starts:

 

Wed Feb 11 15:12:53 BRST 2015sod[6881]Offline Wed Feb 11 15:12:53 BRST 2015sod[6881]Offline for traffic group /Common/traffic-group-1. Wed Feb 11 15:12:53 BRST 2015sod[6881]HA daemon_heartbeat tmm3 fails action is go offline down links and restart. Wed Feb 11 15:12:53 BRST 2015sod[6881]HA daemon_heartbeat tmm2 fails action is go offline down links and restart. Wed Feb 11 15:12:53 BRST 2015sod[6881]HA daemon_heartbeat tmm1 fails action is go offline down links and restart. Wed Feb 11 15:12:53 BRST 2015sod[6881]HA daemon_heartbeat tmm fails action is go offline down links and restart. Wed Feb 11 15:12:48 BRST 2015mcpd[8248]Attempting to connect to CMI peer 192.168.254.2 port 6699 Wed Feb 11 15:12:48 BRST 2015mcpd[8248]CMI reconnect timer: enabled Wed Feb 11 15:12:48 BRST 2015mcpd[8248]Closed connection to device /Common/Bigip-HA02.viajanetmia.local. Wed Feb 11 15:12:48 BRST 2015mcpd[8248]Connection to CMI peer 192.168.254.2 has been removed Wed Feb 11 15:12:42 BRST 2015lacpd[8847]Failover event detected. (Switchboard failsafe disabled while offline) Wed Feb 11 15:12:41 BRST 2015sod[6881]Sod requests links down. Wed Feb 11 15:12:41 BRST 2015sod[6881]HA reports tmm NOT ready.

 

Our top proccess are:

 

tmm, learning_Manage and mysqld

 

Anyone have any ideia about what happens?

 

application acceleration manager
application delivery
ASM Advanced WAF
availability
LTM
security

20 Replies

Sort By
Show More
  • Eduardo_de_Oliv's avatar
    Eduardo_de_Oliv
    Icon for Nimbostratus rankNimbostratus
    Mar 08, 2015

    so.. after uprade we still have a high cpu consume... is the same.. no changes.. We didn't have more crashs but i dont know if stopped because of we turned off a few functions on ASM or if because to the upgrade.. i really think is because to I stopped to uses..

     

  • nsk1's avatar
    nsk1
    Icon for Nimbostratus rankNimbostratus
    Apr 13, 2015

    do you work with a ssl persistence profile on a vip?

     

  • Grayson_149410's avatar
    Grayson_149410
    Icon for Nimbostratus rankNimbostratus
    Jul 20, 2015

    Did you ever get a fix for this? We are having a similar issue where CPU on Core 7 spikes to 90% and then causes a switchboard failure and cause a failover.

     

    • Eduardo_de_Oliv's avatar
      Eduardo_de_Oliv
      Icon for Nimbostratus rankNimbostratus
      Jul 20, 2015
      Yes but I dont what solved I have two Bigip in HA so first I changed from active-passive to active-active and after I delete all previous polices created and made a new polices. and now I expiriencing around 20% CPU on the both Bigip's. The last version released 11.6.0 HF5 has something about CPU monitoring so I think it still problem once i working at previus release. 11.6.0 HF4.. but never had problem anymore.
  • OM's avatar
    OM
    Icon for Nimbostratus rankNimbostratus
    Dec 09, 2015

    starting from 11.6, if ASM is enabled, the logs are not sent to the local file in order to prevent performance issues. So make sure that the logs are not enabled for ASM.

     

  • Eduardo_de_Oliv's avatar
    Eduardo_de_Oliv
    Icon for Nimbostratus rankNimbostratus
    Dec 10, 2015

    Omar,

     

    that problem was solved on HF5, I never had this problem anymore. Anyway, turn off log is not a solution. I using HF5 with logs ON and everything working so good!