Sometimes an exception causes application pool
to shutdown. I start it manually but the question is how can I automate this behavior in IIS 7.0
(Windows server 2008).
If an application pool dies, the next request for a resource served by that pool will automatically restart it. If, however, you have rapid fail protection enabled on the app pool, and the pool dies more times than the number specified by the maximum failures property within the interval specified by the failure interval property, then you will receive a 503 Service Unavailable message. At this point, you will have to manually restart the app pool.
To work around this, either disable rapid fail protection for the app pool, or try increasing the number of faults within the time period, and then determine the root cause of the exceptions which are terminating the app pool.