Threads Stuck In 'In Progress' Status Having 0 Records Without Errors (Doc ID 2357194.1)

Last updated on OCTOBER 18, 2024

Applies to:

Goal

A custom batch that was ran in 200 threads, then majority of the threads got completed but a few was stuck in 'In Progress' status having 0 records processed and no errors. Visiting the logs directory, those threads got 0KB on their respective stdout and stderr. Looking at the TPW logs, we saw a lot of the following messages:
 - 2018-01-10 08:01:52,867 [pool-1-thread-3] INFO (support.cluster.MemberLeftThread) Removing member id: 2 from cache because it left cluster
 - 2018-01-10 08:01:52,867 [pool-1-thread-1] INFO (support.cluster.MemberLeftThread) Removing member id: 8 from cache because it left cluster
 - 2018-01-10 08:01:52,867 [pool-1-thread-4] INFO (support.cluster.MemberLeftThread) Removing member id: 16 from cache because it left cluster
 - 2018-01-10 08:01:52,868 [pool-1-thread-1] INFO (support.cluster.ClusteredNode) Removing member 8 from the cluster cache
 - 2018-01-10 08:01:52,868 [pool-1-thread-3] INFO (support.cluster.ClusteredNode) Removing member 2 from the cluster cache
 - 2018-01-10 08:01:52,868 [pool-1-thread-4] INFO (support.cluster.ClusteredNode) Removing member 16 from the cluster cache

After resubmitting the job, it got completed successfully, but how do we prevent this from occurring in the first place?
 

Solution

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.