[ https://issues.jboss.org/browse/TEIID-4780?page=com.atlassian.jira.plugin... ] Ramesh Reddy commented on TEIID-4780: ------------------------------------- The way I fixed may be not optimal, as code reads like {code} if (elapsed > ttl*1.5) { resetPendingJob(vdb.getVDB(), table); } {code} so that, if the situation is being carried out in clustered environment, then "LOADING" state can be result from another node actively loading the matview load. I did not wanted to kill the on-going job. Instead, I opted to wait upto 1.5 times the configured TTL, if the job is not finished then I consider it as orphaned job. 1.5 is just a arbitrary number. In upstream there is much more comprehensive solution that spans all nodes, and does not wait, however that requires a breaking change, thus this compromise. So, see if you wait enough time, this recovers? -- This message was sent by Atlassian JIRA (v7.2.3#72005)