We've had same message this week.

Click image for larger version. 

Name:	2020-08-13_15-39-49.png 
Views:	39 
Size:	36.8 KB 
ID:	13896

In this instance it could be traced back to our global timers running / firing when the database was locked. Inside this block of code it also calls a stored proc. that does some housekeeping operations - delete & update - which also has its own begin transaction and either rollback or commit depending on the outcome. The tasks performed in the stored proc. could take several minutes depending on the volume of transactions. The current code has failed at one site only and the fix was to suspend the global timers before calling the stored proc.