[jira] [Created] (SAMZA-1187) TestZkProcessorLatch tests share state causing transient failures in CI builds

["Navina Ramesh (JIRA)" <ji...@apache.org>]

Navina Ramesh created SAMZA-1187:
------------------------------------

             Summary: TestZkProcessorLatch tests share state causing transient failures in CI builds
                 Key: SAMZA-1187
                 URL: https://issues.apache.org/jira/browse/SAMZA-1187
             Project: Samza
          Issue Type: Bug
    Affects Versions: 0.13.0
            Reporter: Navina Ramesh
            Assignee: Navina Ramesh
             Fix For: 0.13.0


TestZkProcessorLatch brings up an EmbeddedZk and uses it to test the latch behavior. However, the unit tests share the latchId, and this can cause failures when tests are not run in order. 

Example failures are listed below:

Gradle test > org.apache.samza.zk.TestZkProcessorLatch.testSingleCountdown > captured output:
    Started single countdown
    targetPath /group1/latch_l1/0000000002
    targetPath /group1/latch_l1/0000000002
    targetPath /group1/latch_l1/0000000002
    countDown created /group1/latch_l1/0000000000

testSingleCountdown FAILED
    java.lang.AssertionError: Failed to get.
        at org.junit.Assert.fail(Assert.java:91)
        at org.apache.samza.zk.TestZkProcessorLatch.testSingleCountdown(TestZkProcessorLatch.java:290)

Gradle test > org.apache.samza.zk.TestZkProcessorLatch.testLatchExpires > captured output:
    countDown created /group1/latch_l1/0000000001
    countDown created /group1/latch_l1/0000000002

testLatchExpires FAILED
    org.I0Itec.zkclient.exception.ZkException: org.apache.zookeeper.KeeperException$NotEmptyException: KeeperErrorCode = Directory not empty for /group1/latch_l1
        at org.I0Itec.zkclient.exception.ZkException.create(ZkException.java:68)
        at org.I0Itec.zkclient.ZkClient.retryUntilConnected(ZkClient.java:1000)
        at org.I0Itec.zkclient.ZkClient.delete(ZkClient.java:1047)
        at org.I0Itec.zkclient.ZkClient.delete(ZkClient.java:1042)
        at org.I0Itec.zkclient.ZkClient.deleteRecursive(ZkClient.java:791)
        at org.I0Itec.zkclient.ZkClient.deleteRecursive(ZkClient.java:786)
        at org.apache.samza.zk.ZkUtils.deleteRoot(ZkUtils.java:268)
        at org.apache.samza.zk.ZkCoordinationUtils.reset(ZkCoordinationUtils.java:45)
        at org.apache.samza.zk.TestZkProcessorLatch.testSetup(TestZkProcessorLatch.java:69)

        Caused by:
        org.apache.zookeeper.KeeperException$NotEmptyException: KeeperErrorCode = Directory not empty for /group1/latch_l1
            at org.apache.zookeeper.KeeperException.create(KeeperException.java:125)
            at org.apache.zookeeper.KeeperException.create(KeeperException.java:51)
            at org.apache.zookeeper.ZooKeeper.delete(ZooKeeper.java:873)
            at org.I0Itec.zkclient.ZkConnection.delete(ZkConnection.java:109)
            at org.I0Itec.zkclient.ZkClient$11.call(ZkClient.java:1051)
            at org.I0Itec.zkclient.ZkClient.retryUntilConnected(ZkClient.java:990)



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)