[jira] [Commented] (CASSANDRA-5051) Allow automatic cleanup after gc_grace

["Jonathan Ellis (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CASSANDRA-5051?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13649471#comment-13649471 ] 

Jonathan Ellis commented on CASSANDRA-5051:
-------------------------------------------

I wanted to see if I could decouple things a bit and https://github.com/jbellis/cassandra/commits/5051-v7 is the result (also rebased to current trunk).  I added IRangeProvider to allow tests to inject what they want to test directly.  I 

I also switched off of using bootstrapMode in favor of Mode, but either way I don't think we're quite in the clear -- when we tell node A to move or relocate data to B, it relies on gossip to tell node B about that (the Pending Ranges).  But if gossip is extra slow for whatever reason -- even just a large cluster where everything else is working as planned -- it might take longer than the hardcoded ring delay for B to learn about his new responsibility, and compaction would delete the data.

I think we might need to add a "tell the recipient about the new ranges he's going to have" step to streaming. /cc [~yukim] [~brandon.williams]
                
> Allow automatic cleanup after gc_grace
> --------------------------------------
>
>                 Key: CASSANDRA-5051
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-5051
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Core
>            Reporter: Brandon Williams
>            Assignee: Vijay
>              Labels: vnodes
>             Fix For: 2.0
>
>         Attachments: 0001-5051-v4.patch, 0001-5051-v6.patch, 0001-5051-with-test-fixes.patch, 0001-CASSANDRA-5051.patch, 0002-5051-remove-upgradesstable.patch, 0002-5051-remove-upgradesstable-v4.patch, 0004-5051-additional-test-v4.patch, 5051-v2.txt
>
>
> When using vnodes, after adding a new node you have to run cleanup on all the machines, because you don't know which are affected and chances are it was most if not all of them.  As an alternative to this intensive process, we could allow cleanup during compaction if the data is older than gc_grace (or perhaps some other time period since people tend to use gc_grace hacks to get rid of tombstones.)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira