[
https://issues.apache.org/jira/browse/CASSANDRA-957?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13093909#comment-13093909
]
Nick Bailey commented on CASSANDRA-957:
---------------------------------------
Just a couple more things
* Looks like some tests are failing with these patches applied. Want to look
into that?
* Can you revert the logging change in gossiper for when a node restarts? You
changed it to, 'logger.info("Node {} has been restarted...", ep);' which isn't
particularly more informative. The old logging statement allowed you to easily
grep for 'UP' in the logs.
* Can we only sleep for BROADCAST_INTERVAL if we are actually replacing a
token? No reason to add extra sleeping to the normal bootstrap case.
* Nitpick: you added some random whitespace throughout those 2 patches at the
ends of lines. Can we remove that when you fix the above.
Another thing, when we document this let's be sure to mention that clients
should not connect to a node that is replacing another node until the process
is complete. They will see inconsistent results if they do since gossiper
always assumes the local node is alive.
> convenience workflow for replacing dead node
> --------------------------------------------
>
> Key: CASSANDRA-957
> URL: https://issues.apache.org/jira/browse/CASSANDRA-957
> Project: Cassandra
> Issue Type: Wish
> Components: Core, Tools
> Affects Versions: 0.8.2
> Reporter: Jonathan Ellis
> Assignee: Vijay
> Fix For: 1.0
>
> Attachments: 0001-Support-Token-Replace.patch,
> 0001-Support-bringing-back-a-node-to-the-cluster-that-exi.patch,
> 0001-Support-token-replace.patch, 0001-support-for-replace-token-v3.patch,
> 0001-support-token-replace-v4.patch,
> 0002-Do-not-include-local-node-when-computing-workMap.patch,
> 0002-Rework-Hints-to-be-on-token.patch,
> 0002-Rework-Hints-to-be-on-token.patch, 0002-hints-on-token-than-ip-v4.patch,
> 0002-upport-for-hints-on-token-v3.patch,
> 0003-Make-HintedHandoff-More-reliable.patch,
> 0003-Make-hints-More-reliable.patch, 0003-making-bootstrap-sleep-longer.patch
>
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> Replacing a dead node with a new one is a common operation, but "nodetool
> removetoken" followed by bootstrap is inefficient (re-replicating data first
> to the remaining nodes, then to the new one) and manually bootstrapping to a
> token "just less than" the old one's, followed by "nodetool removetoken" is
> slightly painful and prone to manual errors.
> First question: how would you expose this in our tool ecosystem? It needs to
> be a startup-time option to the new node, so it can't be nodetool, and
> messing with the config xml definitely takes the "convenience" out. A
> one-off -DreplaceToken=XXY argument?
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
