[ZOOKEEPER-2381] ReconfigTest.testPortChange has been failing intermittently

• Type: Test
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Component/s: server, tests
• Assignee: Unassigned
• Reporter: Akihiro Suda
• Created: 04/Mar/16 08:35
• Updated: 04/Mar/16 09:09
• Resolved: 04/Mar/16 09:09

Description

ReconfigTest.testPortChange has been failing intermittently:

• Feb 12, 2016: https://builds.apache.org/job/PreCommit-ZOOKEEPER-Build/3045/
• Feb 16, 2015: http://permalink.gmane.org/gmane.comp.java.zookeeper.devel/25521

I can also locally reproduce with 40d0804c (Mar 3, 2016).

The error message looks like as if it is a linearizability violation for sync+read operation, but I'm still not sure.

stack trace:

junit.framework.AssertionFailedError: expected:<test[1]> but was:<test[0]>
	at org.apache.zookeeper.test.ReconfigTest.testNormalOperation(ReconfigTest.java:150)
	at org.apache.zookeeper.test.ReconfigTest.testPortChange(ReconfigTest.java:598)
	at org.apache.zookeeper.JUnit4ZKTestRunner$LoggedInvokeMethod.evaluate(JUnit4ZKTestRunner.java:79)

test code: https://github.com/apache/zookeeper/blob/2cc945442e218d0757983ac42e2a5d86a94ccb30/src/java/test/org/apache/zookeeper/test/ReconfigTest.java#L150

for (int j = 0; j < 30; j++) {
     try {
         ..
         String data = "test" + j;
         writer.setData("/test", data.getBytes(), -1);
         reader.sync("/", null, null);
         byte[] res = reader.getData("/test", null, new Stat());
         Assert.assertEquals(data, new String(res));
         break;
     } catch (KeeperException.ConnectionLossException e) {
         if (j < 29) {
             Thread.sleep(1000);
         } else {
             // test fails if we still can't connect to the quorum after
             // 30 seconds.
             Assert.fail("client could not connect to reestablished quorum: giving up after 30+ seconds.");
         }
     }
}

Issue Links

• duplicates: ZOOKEEPER-2137 Make testPortChange() less flaky