[ https://issues.apache.org/jira/browse/FLINK-2689?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14803091#comment-14803091 ]
ASF GitHub Bot commented on FLINK-2689: --------------------------------------- Github user srowen commented on the pull request: https://github.com/apache/flink/pull/1136#issuecomment-141122197 @fhueske not sure where else to mention this, but for some reason, PRs like this are causing Apache JIRA to comment on Spark JIRAs (see https://issues.apache.org/jira/browse/SPARK-2689 for instance). Is it maybe because the merge script or other setup from Spark is also used in Flink? Maybe something wasn't changed. Not a huge deal but worth tracking down > Reusing null object for joins with SolutionSet > ---------------------------------------------- > > Key: FLINK-2689 > URL: https://issues.apache.org/jira/browse/FLINK-2689 > Project: Flink > Issue Type: Bug > Components: Local Runtime > Affects Versions: 0.9, 0.10 > Reporter: Fabian Hueske > Assignee: Fabian Hueske > Fix For: 0.9, 0.10 > > > Joins and CoGroups with a solution set have outer join semantics because a > certain key might not have been inserted into the solution set yet. When > probing a non-existing key, the CompactingHashTable will return null. > In object reuse mode, this null value is used as reuse object when the next > key is probed. -- This message was sent by Atlassian JIRA (v6.3.4#6332)