[
https://issues.apache.org/jira/browse/HBASE-16144?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrew Purtell updated HBASE-16144:
-----------------------------------
Fix Version/s: (was: 0.98.21)
I had to revert this from 0.98 because it lead to repeatable failures of
TestZooKeeper#testLogSplittingAfterMasterRecoveryDueToZKExpiry
Thanks [~ted_yu] for finding the problem
> Replication queue's lock will live forever if RS acquiring the lock has died
> prematurely
> ----------------------------------------------------------------------------------------
>
> Key: HBASE-16144
> URL: https://issues.apache.org/jira/browse/HBASE-16144
> Project: HBase
> Issue Type: Bug
> Affects Versions: 2.0.0, 1.3.0, 1.4.0, 1.2.2, 0.98.20, 1.1.6
> Reporter: Phil Yang
> Assignee: Phil Yang
> Fix For: 2.0.0, 1.3.0, 1.4.0, 1.1.6, 1.2.3
>
> Attachments: HBASE-16144-0.98.v1.patch,
> HBASE-16144-branch-1-v1.patch, HBASE-16144-branch-1-v2.patch,
> HBASE-16144-branch-1.1-v1.patch, HBASE-16144-branch-1.1-v2.patch,
> HBASE-16144-v1.patch, HBASE-16144-v2.patch, HBASE-16144-v3.patch,
> HBASE-16144-v4.patch, HBASE-16144-v5.patch, HBASE-16144-v6.patch,
> HBASE-16144-v6.patch
>
>
> In default, we will use multi operation when we claimQueues from ZK. But if
> we set hbase.zookeeper.useMulti=false, we will add a lock first, then copy
> nodes, finally clean old queue and the lock.
> However, if the RS acquiring the lock crash before claimQueues done, the lock
> will always be there and other RS can never claim the queue.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
