Consider the following scenario:
You create an XEvent session with the sqlos.wait_info event in Microsoft SQL Server 2014, 2016 or 2017.
The XEvent session includes one or more of the following global actions: sql_text, tsql_frame, tsql_stack, query_hash, or query_hash_signed.
The server is under high load serving many concurrent parallel queries.
In this scenario, SQL Server may experience delays in parallel thread cleanup which fires the wait_info event and triggers an unhandled access violation. Therefore, SQL Server may stop unexpectedly.
Note: The sqlos.wait_info event is also present in the default system health session. Therefore, it's also susceptible to cause the same issue.
Microsoft has confirmed that this is a problem in the Microsoft products that are listed in the "Applies to" section.
This issue is fixed in the following cumulative updates for SQL Server:
About cumulative updates for SQL Server:
Each new cumulative update for SQL Server contains all the hotfixes and all the security fixes that were included with the previous cumulative update. Check out the latest cumulative updates for SQL Server:
Learn about the terminology that Microsoft uses to describe software updates.