[
https://issues.apache.org/jira/browse/CASSANDRA-564?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12785697#action_12785697
]
Jaakko Laine commented on CASSANDRA-564:
----------------------------------------
Not directly related to this, but close anyway: Currently Gossiper will evict a
node from membership if it has not heard anything from it for 3 days. At this
stage the node is removed from gossiper's own data structures
(unreachableEndpoints), but not from token metadata. Currently token metadata
is never cleaned up unless the node is decommissioned, should we do that at
some point, too?
Another thing is: when a node enters the ring (that is, broadcasts
STATE_NORMAL), there is no check in token metadata or storage service to
prevent another IP from taking a token. Token metadata is simply updated with
the new endpoint/token information. From this it would seem that another node
can take the same token.
>From this it would seem that this improvement request is to decommission a
>dead node from the ring? Replacing refers to taking the same token, but that
>would seem to be possible already.
> Provide recoverability when a node dies and it is impossible to get the same
> IP.
> --------------------------------------------------------------------------------
>
> Key: CASSANDRA-564
> URL: https://issues.apache.org/jira/browse/CASSANDRA-564
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Anthony Molinaro
> Priority: Minor
> Fix For: 0.5
>
>
> From the descriptions on the mailing list, when a node dies permanently from
> hardware or other failure and you need to replace it, it must have the same
> IP. For people running in cloud environments, this is often times
> impossible. So it would be very useful if there was a way to replace a node
> with a new node without requiring the same IP.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
