Forum Discussion

jamied66_40325's avatar
jamied66_40325
Icon for Nimbostratus rankNimbostratus
Jul 02, 2009

simple health check... confusing log files

I have (what I hope) is a simple fix.

 

 

I inherited a set of installed F5 BigIP LTM's that were already "configured". For reasons passing understanding, the previous regime had all server pools using the same health check; a custom Health Monitor built off the default http_receive monitor, but looked for a file named "cssindex.html".

 

 

This file, of course, did not exist on any of our servers, so the web logs were constantly populated with 404 entries.

 

 

To test out my intelligence, I set up a new health monitor for one server pool, and had it look for a file called wx.php (this file does exist on each server in the pool I'm testing).

 

 

The new health check seems to be working, HOWEVER my servers access files are still getting filled up with 404 requests for "cssindex.html".

 

 

Is there somewhere else a health monitor can be specified for a pool, other than Local Traffic > Pools > Health Monitors via the GUI?

 

 

Thanks for any thoughts.

 

 

management
monitoring

5 Replies

Sort By
  • James_Quinby_46's avatar
    James_Quinby_46
    Historic F5 Account
    Jul 02, 2009
    Are your LTMs in an HA pair? Could the standby unit still be running the old health check?
  • jamied66_40325's avatar
    jamied66_40325
    Icon for Nimbostratus rankNimbostratus
    Jul 02, 2009
    They are an active/passive failover pair. I checked the standby unit just in case and it is in fact using the same health-check.

     

     

    This one has me stumped. I can't find another reason as to why they'd be looking for this non-existent file.
  • James_Quinby_46's avatar
    James_Quinby_46
    Historic F5 Account
    Jul 02, 2009
    Check Local Traffic > Nodes to see if any node-specific health checks have been enabled.
  • hooleylist's avatar
    hooleylist
    Icon for Cirrostratus rankCirrostratus
    Jul 06, 2009
    A simple way to check is to search the bigip.conf file for the monitor name.

     

     

    You could also have a port specific HTTP default node monitor configured: Local Traffic > Nodes > Default Monitor. Or less likely, you could have pool member specific monitors configured (Local Traffic >> Pools >> pool name >> Members >> member >> Health Monitors).

     

     

    Aaron
  • jamied66_40325's avatar
    jamied66_40325
    Icon for Nimbostratus rankNimbostratus
    Jul 06, 2009
    Thanks for the input hoolio!

     

     

    Someone had gone in and set a pool member specific monitor on 2 of the nodes for reasons passing understanding. Have probably looked at that screen before, just not hit the Advanced view. The logs cleared up almost immediately.