Skip to main content

Hi folks,

 

For some unresolved reason, our ArcGIS Server 10.7.1 crashes intermittently. Like randomly after a week or a few weeks or month or two, it just crashes. When I check the Server logs for what appears around the crash timestamp, there are a number of SEVERE errors that follow the pattern of "Error performing query operation Wait time of the request to the service <SERVICE> has expired." It's a specific set of services, enterprise feature classes as their source, usually the same culprits appear in these error messages so I thought it might be associated with some web apps that may be experiencing strangely heavy use at sometimes outside of main business hours. I've already had Esri Tech Support investigate this but they couldn't figure out what's causing the issue either.

But I recently changed the source of one of these services to point to a shapefile instead. At the next Server crash, this layer did not have its name show up among the error messages. We have other feature classes served up through Server and they don't show up in the log error messages, so I asked my boss, what commonality could there be among these services?

 

That these erroring services are drawing from feature classes that are regularly updated by FME.

 

So something, maybe in the way we've constructed our Workbench models or I don't know what, could be causing these intermittent ArcGIS Server crashes?

 

Looking for ideas on what I can check for to either fix this problem or eliminate the FME factor, etc. Thank you!

Hi @kirs​, sorry to hear you're running into this issue.

I wouldn't expect FME updating a feature service to cause a crash on ArcGIS Server. It would be more likely that FME could cause an error when accessing the service. Are the logs you noted the only ones that appear related to the crash in ArcGIS Server logs? I would have expected the cause of the crash to be logged.


Hi @danminneyatsaf​ , thanks for getting back to me:

Yes, I wouldn't expect it either, but I just wanted to try presenting my situation here to see if someone has a possible explanation. The crashing has happened before during business hours, at night, early hours of the morning before the workday.. it doesn't seem to happen only at a specific time so I don't think it would be FME accessing the services (our FME processes run at night).

These are often the only suspicious Server logs I see when I check after a crash has happened with no other error-message patterns.

If you have any ideas on what else I can check for, that'd be great!


Reply