I need to find reason for Event ID 2262, Deadlock Detected of our ASP.NET application (64bit). It happens 1-3 times per day on our development server.
I created mini dump with full memory by orphaning the Application Pool as is described here: http://support.microsoft.com/kb/828222
I opened the dump with windbg and tried to find some managed thread: ~*e !clrstack
I received for all threads: Unable to walk the managed stack. The current thread is likely not a managed thread. You can run !threads to get a list of managed threads in the process
The same result I get when I take dump from command line (executing the same command as orphaning is set up for).
Should I take or view dump in different way?
For make windbg working I execute following commands: .load c:\Windows\Microsoft.NET\Framework64\v2.0.50727\mscorwks.dll .load c:\Windows\Microsoft.NET\Framework64\v2.0.50727\mscordacwks.dll .load c:\Windows\Microsoft.NET\Framework64\v2.0.50727\sos.dll .symfix "c:\Program Files\Debugging Tools for Windows (x64)\cache" .cordll -ve -u -l
And I have following dll in the dbg tools folder: mscordacwks_AMD64_X64_2.0.50727.3603.dll (taken from Framework64 directory).
I tried to explore all our logs and I didn't find any correlated data with deadlock failure time stamps. I reviewed and stressed the code committed into svn on the date of fist deadlock and several days before - no deadlock happened during the stress.
I bet on clr stack, thanks for any help, Jan