Thanks to one of the members of these forums, I was able to get HTTP probes up and running on our local director.
Recently, the local director failed a number of the servers via the probe.
The servers are setup to request a header from a ColdFusion page on the server with expected result of 200. LD-CRIT and LD-ERR messages are logged into the syslog of a local Solaris box.
Looking at the failure message and subsequent testing message for one of them shows the following:
July 24 13:34:55 LD-ERR Real machine 'websrv05:80:0:tcp': edited from In service to Failed (External) -- Probe.
July 24 13:37:59 LD-ERR Real machine 'websrv05:80:0:tcp': edited from Failed (External) to Testing - HTTP probes.
I'm kind of confused about the ending of those messages. Does that mean that just a plain probe failed the server but the HTTP probe brought it back? Or am I misunderstanding how the probes work? I'd be happy to post more information about the probe configuration if required.
Also, is there any way to have the HTTP probe tell us whether it was a timeout trying to pull the header of the requested page vs. a non-matching header? Or is it simply, return 200 otherwise list as a fail.
Best regards,
Robert Kropiewnicki
