Forum Discussion

Infinity05_1794's avatar
Infinity05_1794
Icon for Nimbostratus rankNimbostratus
Apr 20, 2012

Monitor Flicker

We are using the ANON_HTTP_80 monitor for a majority of our servers and recently we had production issues with a set of IBM WebSphere Application servers that are going down / up / down / up. I don't see anything in my JVM logs or WebServer logs that indicate an outage.

 

 

Here is what I see in F5 local traffic log:

 

 

 

Apr 18 01:44:47 local/LTM1 notice mcpd[3130]: 01070638:5: Pool member xx.xx.xx.xxx:80 monitor status down.

 

Apr 18 01:44:51 local/LTM1 notice mcpd[3130]: 01070727:5: Pool member xx.xx.xx.xxx:80 monitor status up.

 

Apr 18 01:46:07 local/LTM1 notice mcpd[3130]: 01070638:5: Pool member xx.xx.xx.xxx:80 monitor status down.

 

Apr 18 01:46:13 local/LTM1 notice mcpd[3130]: 01070727:5: Pool member x.xx.xx.xxx:80 monitor status up.

 

 

 

 

 

 

The Anon_HTTP_80 is just doing a GET /snoop\r\n and expecting a Snoop Servlet in return. Interval is 5 seconds and the timeout is 16 seconds.

 

 

 

 

Anyone have any ideas on how to troubleshoot this further? Thanks in advance!

 

management
monitoring
  • nitass's avatar
    nitass
    Icon for Employee rankEmployee
    Apr 20, 2012
    what bigip version are you running?

     

     

    would it be possible that server was overloaded (so, it was unable to response to monitor on time)?

     

    is bigd memory usage normal? could it be memory leak?

     

     

    sol12531: Troubleshooting health monitors

     

    http://support.f5.com/kb/en-us/solutions/public/12000/500/sol12531.html
  • Infinity05_1794's avatar
    Infinity05_1794
    Icon for Nimbostratus rankNimbostratus
    Apr 20, 2012
    We are using BIG-IP 10.2.2 Build 930.0 Hotfix HF3. Our memory / cpu / traffic is well below max, we don't run a whole lot of traffic through them (yet).

     

     

    Thanks for the link, I'll check it out.