A process serving application pool 'Classic .NET AppPool' suffered a fatal communication error with the Windows Process Activation Service

This one is a pretty nasty one. I recommend getting the IIS Debug Diagnostics tools. You can probably alleviate the symptom by turning off "Rapid Fail Protection" in the properties of your application pool, but that will just mask the problem. What is going on is that your application is experiencing an exception so often/so repeatedly that the service shuts itself off. You'll need to run the diagnostic tool to capture everything going in and out to find the exception that's occuring.

This can be a very laborious task because the tool will gather gigabytes of junk to sift through. If you're lucky, the turning off of the "Rapid Fail Protection" might allow the exception to proceed through to the browser, but my guess is that the server will just end up hanging.

If you are using NServiceBus make sure whatever account the application pool is running under has permissions to read off of the message queue.

NServiceBus runs a background task on application startup that timesout after 10 seconds, causing the w3wp process to be killed if it can't read from the queue. After n (default of 5) w3wp crashes Rapid-Fail Protection kicks in and stops the application pool.

http://blog.jonathanoliver.com/2010/11/nservicebus-w3wp-exe-processes-crashing/

Its most likely caused by an infinite loop.

(I just had the same problem.) So you will just have to look for places in the code that use loops and see if removing one of them stops the error. Once you have found the culprit, then you can try to figure out whats causing it to loop indefinitely.