[
https://issues.apache.org/jira/browse/HBASE-3095?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12928318#action_12928318
]
HBase Review Board commented on HBASE-3095:
-------------------------------------------
Message from: "Jean-Daniel Cryans" <[email protected]>
bq. On 2010-11-04 12:30:24, stack wrote:
bq. >
/trunk/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperWatcher.java,
line 326
bq. > <http://review.cloudera.org/r/1167/diff/1/?file=16666#file16666line326>
bq. >
bq. > Why a new exception? Why not pass the one that got us here? (Or,
maybe there isn't one? Or you can't get at it)
Yeah there's no exception leading us there, it's the ZK event. Will add more
documentation tho.
- Jean-Daniel
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://review.cloudera.org/r/1167/#review1804
-----------------------------------------------------------
> Client needs to reconnect if it expires its zk session
> ------------------------------------------------------
>
> Key: HBASE-3095
> URL: https://issues.apache.org/jira/browse/HBASE-3095
> Project: HBase
> Issue Type: Bug
> Reporter: stack
> Assignee: Jean-Daniel Cryans
> Fix For: 0.90.0
>
>
> Clients use an HConnection down in their guts to connect to the hbase
> cluster. Master-is-running and root-region-location are up in zk. Setup of
> a new HConnection sets up a connection to ZooKeeper. If the session with ZK
> expires for whatever reason -- in tests they would expire because zk ensemble
> was restarted across tests or we might expire because of a long GC, well,
> it'll be frustrating to users if we do not just try and resetup the zk
> connection.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
