Anatomy of an Application Pool Crash

re: Anatomy of an Application Pool Crash

Vlad — Sat, 24 Jun 2006 09:07:35 GMT

Wait a min - aren't you describign behaviour of a Crash rule above?

"when you select Hang, DebugDiag attaches a debugger onto the w3wp.exe process and waits for an unhandled exception to happen"

I thought when you used a CRASH rule, DD attached a debugger; HANG rules only attach and dump processes when the hang condition fails? Or did I misunderstand it?

If thats the case the guy is asking a different question to the one you answered which is why the process isn't being dumped and resuming with a HANG rule, I think

re: Anatomy of an Application Pool Crash

David.Wang — Sat, 24 Jun 2006 12:19:20 GMT

Vlad - I wouldn't get "hung" up with the names of the rules. The names are simply how we describe actions to users.

Behind the scenes, a debugger has to attach to the process(s) no matter what and wait for specified conditions to trigger and then take action.

When the condition triggers, whether it is an unhandled exception causing a crash, or a hang test, or memory leak, the debugger has to halt the process to get an accurate guage and take action.

It is this interaction with the debugger that the user observes, and that interaction stays consistent by-design.

//David

re: Anatomy of an Application Pool Crash

Vlad — Tue, 27 Jun 2006 15:44:46 GMT

Thanks for correcting me!

So Hang and Crash rules both attach the debug host to the process as soon as the rule is activated, and then it's just the timing of when the dbghost dumps the process that varies.

My mistaken belief was that the Hang rule only triggered an attach at the point at which the Hang test failed.

Tks!

re: Anatomy of an Application Pool Crash

Michael — Sat, 09 Jun 2007 20:08:45 GMT

I have been dealing with an application pool that has been causing the iis to crash.. back on 6/5/07 at 9:22 a virus entered the network aimed specifically at the iis6.0, i run a medical imaging center network and because of this the iis went down and i have 3 companys trying to figure out what is going on.. i got the virus/grayware/malware out but the problem is still there.. i have been working for 4 days straight with the main application vendor, and another company trying to get this back up and it has all of us looking stupid.. we have it down to a process that pings the app pool and the app pool doesn't respond and iis crashes i need some help it gives memory read errors when it crashes. we have removed the iis, and reinstalled the iis and the problem is still there.. do you have any idea's you could throw my way?? and so far with the ssl cert in it still ccrashed one app pool not its crashing the whole iis. any help would be greatly appreciated.