[
https://issues.apache.org/jira/browse/ZOOKEEPER-1802?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13807657#comment-13807657
]
Thawan Kooburat commented on ZOOKEEPER-1802:
--------------------------------------------
As part of fixing duplicate NEWLEADER packet (ZOOKEEPER-1324),
lastProcessedZxid on each server can be different when the quorum start up and
there is no new request (They may point to the last txn from the previous
epoch). As shown in the log here
2013-10-24 10:42:07,301 [myid:] - INFO
[main:FollowerResyncConcurrencyTest@588] - Timeout waiting for zxid to sync:
leader 0x100003ecc clean 0x200000000 restarted 0x100003ecc
I can switch to rely on other method to verify that all server has up-to-date
data instead of checking lastProcessedZxid
> flakey test testResyncByTxnlogThenDiffAfterFollowerCrashes
> ----------------------------------------------------------
>
> Key: ZOOKEEPER-1802
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-1802
> Project: ZooKeeper
> Issue Type: Bug
> Components: tests
> Affects Versions: 3.5.0
> Reporter: Patrick Hunt
> Assignee: Thawan Kooburat
>
> This test fails intermittently on trunk:
> https://builds.apache.org/view/S-Z/view/ZooKeeper/job/ZooKeeper-trunk-jdk7/691/testReport/junit/org.apache.zookeeper.test/FollowerResyncConcurrencyTest/testResyncByTxnlogThenDiffAfterFollowerCrashes/
--
This message was sent by Atlassian JIRA
(v6.1#6144)
