flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From tillrohrmann <...@git.apache.org>
Subject [GitHub] flink pull request #5746: [FLINK-8943] [ha] Fail Dispatcher if jobs cannot b...
Date Wed, 28 Mar 2018 12:30:54 GMT
Github user tillrohrmann commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5746#discussion_r177731740
  
    --- Diff: flink-runtime/src/test/java/org/apache/flink/runtime/jobmanager/JobManagerHARecoveryTest.java
---
    @@ -396,15 +398,22 @@ public void testFailingJobRecovery() throws Exception {
     
     			jobManager = system.actorOf(jobManagerProps);
     
    +			final TestProbe testProbe = new TestProbe(system);
    +
    +			testProbe.watch(jobManager);
    +
     			Future<Object> started = Patterns.ask(jobManager, new Identify(42), deadline.timeLeft().toMillis());
     
     			Await.ready(started, deadline.timeLeft());
     
     			// make the job manager the leader --> this triggers the recovery of all jobs
     			myLeaderElectionService.isLeader(leaderSessionID);
     
    -			// check that we have successfully recovered the second job
    -			assertThat(recoveredJobs, containsInAnyOrder(jobId2));
    +			// check that we did not recover any jobs
    +			assertThat(recoveredJobs, Matchers.is(empty()));
    --- End diff --
    
    True, will make it consistent.


---

Mime
View raw message