-
Notifications
You must be signed in to change notification settings - Fork 2.4k
Open
Labels
in: corerelated-to: transaction-managementstatus: waiting-for-triageIssues that we did not analyse yetIssues that we did not analyse yettype: bug
Description
Marrache Mickael opened BATCH-2452 and commented
I have multiple threads that concurrently launch jobs using a shared synchronous job launcher. For some reason, I get deadlock while issuing the INSERT statements for creating the job instances. The deadlock occurs even with a single INSERT per transaction because of the unique constraint defined on the JOB_INSTANCE table (on job name and job key). Ideally, I would like to avoid such deadlocks but I didn't find a solution. Inserting job instances in a well defined order is not possible. Therefore, I expect Spring Batch to retry the transaction instead of directly throwing the deadlock exception. However, it doesn't seem to be the case.
No further details from BATCH-2452
Metadata
Metadata
Assignees
Labels
in: corerelated-to: transaction-managementstatus: waiting-for-triageIssues that we did not analyse yetIssues that we did not analyse yettype: bug