Hello,
I'm running DNN CE 6.02.00 with ASP v.4 on a Windows Server 2003 server with MS SQL Server 2008 R2...
These past two or three weeks we've noticed some serious issues with the application pool's worker process (w3wp.exe) for a particular site running up the CPU and eventually taking the site down (and other sites with it as it is a shared environment). These SocketTimeoutExceptions come in spurts, intermittently, and without warning. A dedicated application pool has been set up for each application, however, this one site / appPool has been isolated as the culprit from using Task Manager to get the PID then using inetmgr to tie the PID to an application pool.
From my limited experience troubleshooting DNN sites, this appears to be the grandaddy of all problems. I've used Windbg, Process Explorer, and DebugDiag to try and make sense of the issue. This doesn't really help, though, because I have no idea how to read the dumps that are tossed out and no real understanding of how application stacks or threads work.
I've opened a ticket directly with Microsoft .NET support to see if they might be able to shed some light on the issue. I've also downloaded ANTS Memory Profiler (http://www.red-gate.com/SupportCenter...) onto the server b/c I've read it's output is easier to understand and make use of.
From what I can see, it says that memory fragmentation is affecting the size of the largest object that can be allocated. It has also suggested that my application is suffering from fragmentation of the large object heap.
Does anybody have some light to shed on this? I'd greatly appreciate it!