Hi,

Since upgrading to 11.1.7 we have had a few instances where the query service reaches capacity and has to be restarted.  From the XQE logs we were able to identify the source is a scheduled data set refresh.

In the XQE logs I see an error each time the data set refreshes that I do not understand: <![CDATA[Detected a leaked MultiRequestContext (19152_64aaf941-e3ba-11ea-ba5a-fa9e000f7c5b:9ylj8vyCj2vd2M88wjMyvMslvC9h29hG9vC29ywC) for SessionContext with id=19152_64aaf941-e3ba-11ea-ba5a-fa9e000f7c5b]]>

There are no errors or warnings when validating the source data module and data set/report.  The SQL is not overly complex.

Based on feedback from support I have twice increased the amount of memory allocated to the JVM heap and nursery.  However I still see the error message.  My concern is the issue will just continue as it reaches the new limits.

I wonder if anyone else has experienced this and resolved?

Thank you.

------------------------------
Jeremy Aude
------------------------------

#CognosAnalyticswithWatson

FYI - 

APAR logged against the issue: PH28873 REPEATED ERRORS XQE-GEN-0028 YOUR REQUEST WAS REJECTED BECAUSE THE QUERY SERVICE IS BUSY. The source of the problem of the leaked MultiRequestContext has been identified. The code has been fixed and will now go through the process of getting it ready for delivery.

The only solution for me is to continue to restart the query service regularly to prevent an overload.

Thanks to @HENK CAZEMIER for the assistance with this!​


------------------------------