Service Instance Memory Limits

Hello,

I did not find anything in the community on service instance memory limits and hence posting this as an FYI in case others run into it.

Had an issue where the users could not download a Pyramid discover to Excel (about 200K rows) in production.

The discover itself renders in a couple of seconds and when printing to Excel, would run for about 30 seconds and then create an empty excel file (any output format) with a message "User Query cannot be completed". Also on the Spool Jobs window, Successful Queries will show as 0.

Same discover in UAT works fine (same volume of data)

After adjusting cell limits, etc, the issue still persisted

I started looking through the logs and buried in there was a heap size error which led to me to think there was some memory parameter that was causing this issue

This led to me the looking at Service Instance memory limits and noticed that UAT and Dev had 5 times what prod was set to, I am assuming they came with the default install and not sure why it was set to such a low number, as I did not even know about these limits until this research.

Runtime, Task and Web memory limits were set really low in Prod compared to UAT

I edited these to take the default recommendation which was 5 times (3832 vs 15772 for Runtime, 1670 vs 7886 for Task and 500 vs 1893 for Web). Keep in mind this will restart the service for each update and needs to be done off hours (in the docs)

That fixed the issue with the rendering of the report to Excel

Thanks

Mad