[ https://issues.apache.org/jira/browse/CONNECTORS-347?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13258120#comment-13258120 ]
Karl Wright commented on CONNECTORS-347: ---------------------------------------- My operating theory is now that the hopcount is for "seconddir" is unavailable even though "seconddir" must have been processed in order to add either it or its child to the tables. Furthermore, the removal of both of these does not clean up properly because the deletion that takes place does not notice the intrinsiclink rows corresponding to the documents for the same reason. In other words, what we may have is the result of either a transaction isolation bug in MySQL or improper usage of transaction isolation in the MySQL ManifoldCF driver and/or the hopcount code. > MySQL HopCount test sometimes fails > ----------------------------------- > > Key: CONNECTORS-347 > URL: https://issues.apache.org/jira/browse/CONNECTORS-347 > Project: ManifoldCF > Issue Type: Bug > Components: Framework core > Affects Versions: ManifoldCF 0.5, ManifoldCF 0.6 > Reporter: Karl Wright > Assignee: Karl Wright > Fix For: ManifoldCF 0.6 > > > The HopcountMySQLIT test sometimes fails complaining that it is expecting to > see 6 documents but only sees 4. > {code} > [junit] org.apache.manifoldcf.core.interfaces.ManifoldCFException: Wrong > number of documents processed - expected 6, saw 4 > [junit] at > org.apache.manifoldcf.filesystem_tests.HopcountMySQLIT.hopcountCheck(HopcountMySQLIT.java:152) > {code} > This does not happen on every run. -- 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