[ 
https://issues.apache.org/jira/browse/GOBBLIN-1841?focusedWorklogId=867042&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-867042
 ]

ASF GitHub Bot logged work on GOBBLIN-1841:
-------------------------------------------

                Author: ASF GitHub Bot
            Created on: 23/Jun/23 00:57
            Start Date: 23/Jun/23 00:57
    Worklog Time Spent: 10m 
      Work Description: homatthew commented on code in PR #3708:
URL: https://github.com/apache/gobblin/pull/3708#discussion_r1239159140


##########
gobblin-cluster/src/main/java/org/apache/gobblin/cluster/GobblinClusterManager.java:
##########
@@ -504,6 +506,18 @@ private static void printUsage(Options options) {
     formatter.printHelp(GobblinClusterManager.class.getSimpleName(), options);
   }
 
+  public void disableLiveHelixInstances() {
+    HelixManager helixManager = this.multiManager.getJobClusterHelixManager();

Review Comment:
   Upon further investigation, this job cluster manager should be connected 
elsewhere. But I see another issue where you are fetching all live instances 
(include itself!). Which means that when the AM calls this method, it will 
disable itself.
   
   Please take a look at the below code 
https://github.com/apache/gobblin/blob/5af6bca57df909e44b995e5b2d667c70e0399877/gobblin-yarn/src/main/java/org/apache/gobblin/yarn/YarnAutoScalingManager.java#L187C1-L196C6
   
   This code fetches all live participants that are task runners. We should 
only disable taskrunners, so let's do that instead.
   
   And then we can also rename this method to 
`disableTaskRunnersFromPreviousExecutions`. 





Issue Time Tracking
-------------------

    Worklog Id:     (was: 867042)
    Time Spent: 50m  (was: 40m)

> Move disabling of current live instances to the GobblinClusterManager startup
> -----------------------------------------------------------------------------
>
>                 Key: GOBBLIN-1841
>                 URL: https://issues.apache.org/jira/browse/GOBBLIN-1841
>             Project: Apache Gobblin
>          Issue Type: Improvement
>            Reporter: Matthew Ho
>            Priority: Major
>          Time Spent: 50m
>  Remaining Estimate: 0h
>




--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to