traveler_94007
Nimbostratus
Jul 09, 2012CPU intensive in LTM v11
Hi all,
We are having an issue with two loadbalancer in Active/standby mode. A couple of times (randomly )we have experienced in one of them cpu loads that arrive to the 97% of the system, when trying to access via https to the LTM (making impossible to access to it)
Platform: BIG-IP 3600
Version: BIG-IP 11.1.0 Build 1943.0 Final
When having the issue we checked that 3 processes consumes most ot the available cpu: fpdd, ksoftirqd/1 and ksoftirqd/cpu with 39,7%, 31,9 and 29,8 aprox.
From the httpd logs we can only see the attemps to log:
Jul 9 17:01:43 hostname_lb err httpd[27996]: [error] [client x.x.x.x] proxy: error processing end, referer: https://1.1.1.1/tmui/login.jsp
Jul 9 17:01:43 hostname_lb err httpd[5187]: [error] [client x.x.x.x] proxy: error processing end, referer: https://1.1.1.1/tmui/login.jsp
Jul 9 17:01:43 hostname_lb err httpd[5184]: [error] [client x.x.x.x] proxy: error processing end, referer: https://1.1.1.1/tmui/login.jsp
And the ltm logs it is seems the device is unable to monitor any check because the high cpu:
Jul 9 17:01:49 hostname_lb notice mcpd[4719]: 01070638:5: Pool /Common/pool_CHK_s_EE20LTP_WDSL_Internal member /Common/node_2.2.2.158:8103 monitor status unchecked. [ w as node down for 0hr:0min:6sec ] Jul 9 17:01:49 hostname_lb notice mcpd[4719]: 01070638:5: Pool /Common/pool_CHK_s_EE20LTP_WDSL_Internal member /Common/node_2.2.2.161:8103 monitor status unchecked. [ w as node down for 0hr:0min:6sec ]
Any help/advice for the troubleshooting would be appreciated
Thanks.
Regards