I have a problem in client`s live environment. Analyzing web server log files I found that some request occasionally get HTTP 404 (GET,POST).
Heavy use during peak times doesn`t seem to affect the number of faults occuring (at least significantly)
XXX.XXX.XXX.XXX- - [25/Sep/2013:17:01:21 +0300] "GET /application/main;jsessionid=LMfvSCpLmyJ6JDyGND67DbVhDyHVxm5jpyLBlBYT2vSCVMSM0Zxs!-296039734?authToken=1380093223928FCAEIOHPZVAXFIIVZRUTVETEUUAWAULKYHFTRCCTVQRPSGJNABL&accoId=1000008357&compCode=mi.dia.07&araTransactionId=override&resetOverlay=true&remoteAppScope=m.f0.rootWidget.topC.f0.menuContainer.f0&lang=et&reset=&vMode=FULL&rUrl=https://client.com:443/application/main;ehealth-jsessionid=LMfvSCpLmyJ6JDyGND67DbVhDyHVxm5jpyLBlBYT2vSCVMSM0Zxs!-296039734&assistStaffId=&assistStaffProfId=&mdl=CLINICIAN_PORTAL HTTP/1.1" 404 1612
Weblogic 12c - clustered (4 nodes)
When trying the same URL through browser, I get HTTP 200 (and users on client side mostly as well, rare cases of 404) Configured debug logging for apache weblogic plugin, everything seems to be ok on webserver side.
Also, since the 404 requests reach weblogic nodes` access log (fault appears on all 4 nodes, so it`s not a case where application is in a bad state on a node) it seems to be a problem regarding weblogic internal request routing/processing
I was hoping for some ideas as how to approach the problem.
* Improve logging? - what specifically and to what level? Restarting nodes is also a bit complicated.
* Some aspects of configuration to check out?
Happily willing to give more information about configuration as well if anyone has some ideas.
Even rough directions as to how to approach the problem would be of great help.