[
https://issues.apache.org/jira/browse/CASSANDRA-16286?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17269765#comment-17269765
]
Caleb Rackliffe commented on CASSANDRA-16286:
---------------------------------------------
[~cscotta] I know this has been a potential problem for years (and certainly
before 4.0), but I'd like to try to throw a patch together first to see if it's
as simple as I think it is. If I don't get around to it by the end of the
month, I won't argue against pushing it to an early 4.0.x release.
> Make TokenMetadata's ring version increments atomic
> ---------------------------------------------------
>
> Key: CASSANDRA-16286
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16286
> Project: Cassandra
> Issue Type: Bug
> Components: Cluster/Gossip
> Reporter: Caleb Rackliffe
> Assignee: Caleb Rackliffe
> Priority: Normal
> Fix For: 3.0.x, 3.11.x, 4.0-rc
>
>
> The update semantics of the ring version in {{TokenMetadata}} are not clear.
> The instance variable itself is {{volatile}}, but it is still incremented by
> a non-atomic check-and-set, and not all codepaths do that while holding the
> {{TokenMetadata}} write lock. We could make this more intelligible by forcing
> the external callers to use both the write when invalidating the ring and
> read lock when reading the current ring version. Most of the readers of the
> ring version (ex. compaction) don't need it to be fast, but it shouldn't be a
> problem even if they do. If we do this, we should be able to avoid a
> situation where concurrent invalidations don't produce two distinct version
> increments.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
