[ https://issues.apache.org/jira/browse/YARN-3349?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14364175#comment-14364175 ]
zhihai xu commented on YARN-3349: --------------------------------- [~ozawa], many thanks for the review and commit. > Treat all exceptions as failure in > TestFSRMStateStore#testFSRMStateStoreClientRetry > ----------------------------------------------------------------------------------- > > Key: YARN-3349 > URL: https://issues.apache.org/jira/browse/YARN-3349 > Project: Hadoop YARN > Issue Type: Improvement > Components: test > Affects Versions: 2.6.0 > Reporter: zhihai xu > Assignee: zhihai xu > Priority: Minor > Fix For: 2.7.0 > > Attachments: YARN-3349.000.patch > > > treat all exceptions as failure in testFSRMStateStoreClientRetry. > Currently the exception "could only be replicated to 0 nodes instead of > minReplication (=1)" is not treated as failure in > testFSRMStateStoreClientRetry. > {code} > // TODO 0 datanode exception will not be retried by dfs client, > fix > // that separately. > if (!e.getMessage().contains("could only be replicated" + > " to 0 nodes instead of minReplication (=1)")) { > assertionFailedInThread.set(true); > } > {code} > With YARN-2820(Retry in FileSystemRMStateStore), we needn't treat this > exception specially. We can remove the check and treat all exceptions as > failure in testFSRMStateStoreClientRetry. -- This message was sent by Atlassian JIRA (v6.3.4#6332)