[jira] [Comment Edited] (CASSANDRA-9317) Populate TokenMetadata earlier during startup

["Marcus Eriksson (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CASSANDRA-9317?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14532552#comment-14532552 ] 

Marcus Eriksson edited comment on CASSANDRA-9317 at 5/7/15 12:39 PM:
---------------------------------------------------------------------

http://cassci.datastax.com/view/Dev/view/krummas/job/krummas-marcuse-tmdearly-testall/
http://cassci.datastax.com/view/Dev/view/krummas/job/krummas-marcuse-tmdearly-dtest/

look kind of similar to trunk, but the important ones (bootstrap) seem to fail in both


was (Author: krummas):
http://cassci.datastax.com/view/Dev/view/krummas/job/krummas-marcuse-tmdearly-testall/
http://cassci.datastax.com/view/Dev/view/krummas/job/krummas-marcuse-tmdearly-dtest/

look kind of similar, but the important ones (bootstrap) seem to fail in both

> Populate TokenMetadata earlier during startup
> ---------------------------------------------
>
>                 Key: CASSANDRA-9317
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9317
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Marcus Eriksson
>            Assignee: Marcus Eriksson
>             Fix For: 3.x
>
>         Attachments: 0001-populate-tmd-early.patch
>
>
> For CASSANDRA-6696 we need to know the local tokens earlier when starting up, this is due to the fact that we need to know how to split the local ranges when we flush for example (we can flush during commitlog replay etc).
> Breaking out the TMD population into its own method makes it possible to use for the offline tools post 6696 as well - we will need to know tokens when doing scrub etc



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