Synapse Serverless - Deadlock
I think this is something I need to raise a support request over but before I do, thought I would see if it's been experienced elsewhere and how it was resolved (more a setting change than a query change as I'm not looking for "row_number" style resolutions)
In our Synapse serverless environment, if we run a "select *" query, it executes fine, however, if we do a "select top 10 *" it fails with
Transaction (Process ID 153) was deadlocked on generic waitable object resources with another process and has been chosen as the deadlock victim. Rerun the transaction
It seems somehow the "top" is causing a lock that then causes the select to deadlock (or vice versa) but why? and is there something we can set to avoid?