I noticed that SPIDS were being killed on a regular basis although the timing was Intermittent. There appears a record in the SQL Server errorlog which reads as follows: Process ID 121 was killed by hostname xxxxx-SPOTL-APP, host process ID 0.
After some investigation I was able to see the process being killed was the "SQL Server Performance Investigator".
Question #1 Why is the application killing its threads/queries? Should the application not manage its connections and queries through the API rather than forcibly killing a SPID?
Question #2 It looks like the input buffer shows that the Foglight application is running queries that may have been written for SQL Server 2005/2008. Has Foglight updated their queries to adjust for DMV's and other changes that have occurred from the 2005 version to 2019 version, as that is the version of SQL Server we are running?
Question #3 What algorithm is used to determine how, when and why a SPID is killed?
Question #4 Is this documented anywhere??
any help is appreciated!!!
Thanks,
Greg Moss