[
https://issues.apache.org/jira/browse/CASSANDRA-14793?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17119674#comment-17119674
]
Benedict Elliott Smith commented on CASSANDRA-14793:
----------------------------------------------------
It would be possible to replicate the data across multiple disks, but only to
compact one of these disks, and to replicate the result of compaction to the
others? This would create redundancy for the relatively cheaply-maintained
system tables?
Note that either way the Paxos table might need to be dealt with differently to
other system tables, since it is potentially high traffic, and constraining it
to one disk might be problematic. I don't recall if we ever got round to
introducing the range-aware compaction Marcus put together, but the Paxos table
seems more suited to this approach.
> Improve system table handling when losing a disk when using JBOD
> ----------------------------------------------------------------
>
> Key: CASSANDRA-14793
> URL: https://issues.apache.org/jira/browse/CASSANDRA-14793
> Project: Cassandra
> Issue Type: Bug
> Components: Legacy/Core
> Reporter: Marcus Eriksson
> Assignee: Benjamin Lerer
> Priority: Normal
> Fix For: 4.0
>
>
> We should improve the way we handle disk failures when losing a disk in a
> JBOD setup
> One way could be to pin the system tables to a special data directory.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
