[
https://issues.apache.org/jira/browse/CASSANDRA-5094?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14705119#comment-14705119
]
David Loegering commented on CASSANDRA-5094:
--------------------------------------------
This is a major issue for organizations that are concerned about security.
Is there a way to lock these actions down at least so someone messing around in
the system can't do so much damage so quickly/ easily?
> Should not reset local schema when there is no other nodes to resync
> --------------------------------------------------------------------
>
> Key: CASSANDRA-5094
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5094
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 1.1.7
> Reporter: Manoj Mainali
> Attachments: trunk-CASSANDRA-5094.txt
>
>
> The nodetool resetlocalschema allows to drop the schema from a node. But, it
> does not check whether there are other nodes present or not to resync the
> schema from. That means, if there is only one instance, whether it is a
> single node or a cluster but other nodes were down, and we reset the local
> schema the instance will not be able to resync from other nodes. In that
> case, the node will not be able to serve requests. I haven't confirmed the
> behavior, i.e. whether the instance will get the schema or not, when other
> nodes in the cluster come back. However, in a single node cluster it will
> never be able to resync itself.
> It is better to check whether there are other nodes or not to resync from,
> and do not clear the schemas if there are none.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
