[ https://issues.apache.org/jira/browse/MAPREDUCE-4010?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13230414#comment-13230414 ]
Robert Joseph Evans commented on MAPREDUCE-4010: ------------------------------------------------ OK I have debugged the root cause of the failure. When we read back in a value. It sets both the original value and it also sets the deprecated value. This way the original 273 map entries can become much bigger 364 entries. Which looking at your original patch is probably what you already knew was happening, and why you were ignoring the deprecated values when doing the comparison. Which means that not only the size comparison should fail, but so should the map comparison. Looking at the test there is a 3 and a half year old copy and past error where we are comparing conf1 to conf1!!! iterator2 is being set to conf1.iterator not conf2.iterator. Alejandro, Thanks for putting up with me digging down into a root cause that you probably already understood. I am fine with your first patch, so long as you fix the assertEquals bug I mentioned above and explain in a comment why we are ignoring the deprecated values. > TestWritableJobConf fails on trunk > ---------------------------------- > > Key: MAPREDUCE-4010 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-4010 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2 > Affects Versions: 0.24.0 > Reporter: Jason Lowe > Assignee: Alejandro Abdelnur > Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-4010.patch, MAPREDUCE-4010.patch > > > TestWritableJobConf is currently failing two tests on trunk: > * testEmptyConfiguration > * testNonEmptyConfiguration > Appears to have been caused by HADOOP-8167. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira