[ 
https://issues.apache.org/jira/browse/CURATOR-280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15648428#comment-15648428
 ] 

Vitalii Tymchyshyn commented on CURATOR-280:
--------------------------------------------

Why so? It's very inconvenient. Often zookeeper is used without any really 
persisted data. In this case, it's much easier just to restart Zookeeper with 
data dir wiped in case of any problems and make it go. Adding a "fill data 
step" can be complex.

What is the reasoning for this limitation? Is it documented somewhere? E.g. 
ZKPaths#mkdirs just says "Make sure all the nodes in the path are created.".

> LeaderLatch doesn't work when using a zookeeper chroot
> ------------------------------------------------------
>
>                 Key: CURATOR-280
>                 URL: https://issues.apache.org/jira/browse/CURATOR-280
>             Project: Apache Curator
>          Issue Type: Bug
>          Components: Framework
>    Affects Versions: 2.9.0, 2.9.1
>            Reporter: Vincent Bernat
>
> Hey!
> When using a ZK connection-string with a chroot (for example 
> {{localhost:2181/chroot}}), the leader election by LeaderLatch doesn't work. 
> This may be similar to CURATOR-270. If I query {{.getParticipants}}, I get:
> {code}
>       Actual: org.apache.zookeeper.KeeperException$NoNodeException: 
> KeeperErrorCode = NoNode for /test4
>               
> org.apache.zookeeper.KeeperException.create(KeeperException.java:111)
>               
> org.apache.zookeeper.KeeperException.create(KeeperException.java:51)
>               org.apache.zookeeper.ZooKeeper.getChildren(ZooKeeper.java:1590)
>               
> org.apache.curator.framework.imps.GetChildrenBuilderImpl$3.call(GetChildrenBuilderImpl.java:214)
>               
> org.apache.curator.framework.imps.GetChildrenBuilderImpl$3.call(GetChildrenBuilderImpl.java:203)
>               org.apache.curator.RetryLoop.callWithRetry(RetryLoop.java:107)
>               
> org.apache.curator.framework.imps.GetChildrenBuilderImpl.pathInForeground(GetChildrenBuilderImpl.java:199)
>               
> org.apache.curator.framework.imps.GetChildrenBuilderImpl.forPath(GetChildrenBuilderImpl.java:191)
>               
> org.apache.curator.framework.imps.GetChildrenBuilderImpl.forPath(GetChildrenBuilderImpl.java:38)
>               
> org.apache.curator.framework.recipes.locks.LockInternals.getSortedChildren(LockInternals.java:150)
>               
> org.apache.curator.framework.recipes.locks.LockInternals.getParticipantNodes(LockInternals.java:132)
>               
> org.apache.curator.framework.recipes.leader.LeaderLatch.getParticipants(LeaderLatch.java:430)
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to