[ https://issues.apache.org/jira/browse/HIVE-20547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16615033#comment-16615033 ]
Hive QA commented on HIVE-20547: -------------------------------- Here are the results of testing the latest attachment: https://issues.apache.org/jira/secure/attachment/12939601/HIVE-20547.01.patch {color:red}ERROR:{color} -1 due to build exiting with an error Test results: https://builds.apache.org/job/PreCommit-HIVE-Build/13784/testReport Console output: https://builds.apache.org/job/PreCommit-HIVE-Build/13784/console Test logs: http://104.198.109.242/logs/PreCommit-HIVE-Build-13784/ Messages: {noformat} Executing org.apache.hive.ptest.execution.TestCheckPhase Tests exited with: Exception: Patch URL https://issues.apache.org/jira/secure/attachment/12939601/HIVE-20547.01.patch was found in seen patch url's cache and a test was probably run already on it. Aborting... {noformat} This message is automatically generated. ATTACHMENT ID: 12939601 - PreCommit-HIVE-Build > HS2: support Tez sessions started by someone else (part 1) > ---------------------------------------------------------- > > Key: HIVE-20547 > URL: https://issues.apache.org/jira/browse/HIVE-20547 > Project: Hive > Issue Type: Bug > Reporter: Sergey Shelukhin > Assignee: Sergey Shelukhin > Priority: Major > Attachments: HIVE-20547.01.patch, HIVE-20547.patch > > > The registry/configs/some code is based on a private patch by [~prasanth_j]. > The patch refactors tez pool session to use composition instead of > implementation inheritance from TezSessionState, to allow for two > implementations of TezSessionState. > For now it's blocked on getClient API in Tez that will be available after > 0.9.3 release; however I commented out that path to check that refactoring > passes tests. > When 0.9.3 becomes available, we can uncomment and commit. > In part 2, we may add some tests, and also consider other changes that are > required for external sessions (e.g. KillQuery, where we cannot assume YARN > is present). > We may also consider a WM change that allows for proportional session > distribution when the number of external sessions and the number of > admin-specified sessions doesn't match, or at least some validation to see > that the external sessions are available when applying a RP. -- This message was sent by Atlassian JIRA (v7.6.3#76005)