Like a lot of people out there in IT departments, I am on call for certain processes every night (not one of the more glamorous parts of the job). This week is my week and so far I have been called 3 times. So I am starting to wonder how to improve some of the systems. One of the ideas I have (since re-writing is not an option) is monitoring processes for certain characteristics to help the nightly guys to make a better decision in order to wake me up or not. With my day job (development) I use some different tools to monitor and troubleshoot processes like procexp.exe, taskmgr.exe, eventvwr.msc, and perfmon.exe, but the nightly guys are not developers and need something to tell them to call or not call.
Greg has used WMI to write some scripts that help monitor some services as well as restart them when needed ... this alone has helped minimize the amount of people needed to handle certain instances.
What are some of the characteristics of a process that is "hanging" vs. a process that is just running behind? Any one want to offer some advice? Anyone in the same situation out there??