[
https://issues.apache.org/jira/browse/CASSANDRA-213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12738463#action_12738463
]
Sandeep Tata commented on CASSANDRA-213:
----------------------------------------
>>Also, the right approach for nodeprobe is to tell the new node "bootstrap
>>yourself" and have it tell the actual bootstrap commands to the other nodes.
Yeah, this is probably cleaner, but this also means
1) You have to bootstrap each node separately instead of supplying a list of
nodes to an existing node.
2) I'm not sure how long it would take for a new node to get a stable version
of the token-to-endpoint map. Each gossip message might send a random subset of
the endpoint map to the new node (although it currently sends everything, I
think).
I think it is safer for an existing, stable node to send out the bootstrap
commands.
> complete bootstrap code
> -----------------------
>
> Key: CASSANDRA-213
> URL: https://issues.apache.org/jira/browse/CASSANDRA-213
> Project: Cassandra
> Issue Type: New Feature
> Components: Core
> Reporter: Jonathan Ellis
> Assignee: Sandeep Tata
> Priority: Critical
> Fix For: 0.4
>
> Attachments: 213-v1.patch, 213-v2.patch, 213-v3-addendum.patch,
> 213-v3.patch
>
>
> The bootstrap code sends files from
> DatabaseDescriptor.getBootstrapFileLocation to the node that is joining the
> ring, but nothing actually puts files into that directory.
> There are probably other bugs / omissions we will find as we flesh this out.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
