Applies to Product - FnO Platform
What’s happening?
Customers are experiencing performance issues with batch jobs, including jobs being slow, stuck, or timing out, which can lead to table locks and delayed business processes.
Reason:
The underlying causes identified include: - Dead connection errors causing batch jobs to be stuck in an executing state. - Custom code within batch jobs that may lead to performance issues and slowness. - Blocking statements during the execution of batch jobs, resulting in deadlock scenarios.
Resolution:
- For batch jobs that are stuck, it is recommended to cancel the job using Enhanced abort functionality and monitor for any dead connection errors.
- If a batch job is identified as slow due to custom code, review and modify the code as necessary, then rerun the batch job to check for improvements.
- Monitor batch jobs for blocking statements and deadlocks, and investigate any recent changes in the environment that may impact performance, such as enabling change tracking or recent upgrades.
- If any SQL related Data stack errors any environment failures , please reach out SQL Performance & Availability Team.
- Please ensure customer is following the PBS Best Practices to configure batch jobs - PBS Best Practices