[
https://issues.apache.org/jira/browse/ZOOKEEPER-231?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12650743#action_12650743
]
Patrick Hunt commented on ZOOKEEPER-231:
----------------------------------------
@Krishna: the paths I specified are zk subtree path specs.
I should have been more clear: /app/app1 corresponds to the /app/app1 znode
subtree w/in ZK namespace. To take my example further you might have:
/app/app1/cluster1/members/...
/app/app1/cluster1/leader
/app/app1/cluster2/members/...
/app/app1/cluster2/leader
where members is storing the hosts & ports of the members of cluster1, for
app1, etc... and leader is storing the leader node for the cluster (this is but
one simple example).
The ZK ACLs feature could be used to lock down a particular acl id to
particular subtree(s).
http://hadoop.apache.org/zookeeper/docs/r3.0.0/zookeeperProgrammers.html#sc_ZooKeeperAccessControl
So in this case you might have acl ids for app1, app2 and app3, each only able
to access it's particular corresponding subtree of znodes, where each of those
trees has some associated quota.
Does that help?
> Quotas in zookeeper
> -------------------
>
> Key: ZOOKEEPER-231
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-231
> Project: Zookeeper
> Issue Type: New Feature
> Affects Versions: 3.0.0
> Reporter: Mahadev konar
> Assignee: Mahadev konar
> Fix For: 3.1.0
>
>
> creating quota's in zookeeper so that a runaway application does not bring
> down the zookeeper cluster.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
