[
https://issues.apache.org/jira/browse/CASSANDRA-10130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16052165#comment-16052165
]
Paulo Motta commented on CASSANDRA-10130:
-----------------------------------------
bq. I think we should preserve the original behaviour; it is not perfect, but
CASSANDRA-13606 should provide a proper fix.
I really think it makes the code cleaner and less prone to errors to update
{{SecondaryIndexManager}} state ({{needsFullRebuild}}, {{queryableIndexes}} and
{{inProgressBuilds}}) exclusively in the {{mark*}} methods rather than in many
places throughout the code.
The SASI test failures was due to the fact that the
{{invalidateAllIndexesBlocking}} name was misleading, since this was actually
dropping all indexes, so I renamed it to {{dropAllIndexes}} and extracted the
invalidate part into another method which can be used by the SASI tests.
Here is the [commit with these
changes|https://github.com/pauloricardomg/cassandra/commit/ad8beb0bb4bf40d0e5cc4aee3c1bf02ace3f6929].
In the same spirit, I also removed the marking of index as failed before
fetching the initialization task [on this
commit|https://github.com/pauloricardomg/cassandra/commit/33f56de12f3573f42e0426cd915fb049405c883d].
Regarding the exception being passed as the last parameter to the logger,
that's not a bug since sl4j [will properly handle that
case|https://www.slf4j.org/faq.html#paramException], so I added [this
commit|https://github.com/pauloricardomg/cassandra/commit/b5771f010191b06198880b5c2bb1dfc0c46b3b1c]
reinstating this behavior.
[~sbtourist] let me know what do you think before [~adelapena] updates the
branch with these suggestions.
Thanks!
> Node failure during 2i update after streaming can have incomplete 2i when
> restarted
> -----------------------------------------------------------------------------------
>
> Key: CASSANDRA-10130
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10130
> Project: Cassandra
> Issue Type: Bug
> Components: Coordination
> Reporter: Yuki Morishita
> Assignee: Andrés de la Peña
> Priority: Minor
>
> Since MV/2i update happens after SSTables are received, node failure during
> MV/2i update can leave received SSTables live when restarted while MV/2i are
> partially up to date.
> We can add some kind of tracking mechanism to automatically rebuild at the
> startup, or at least warn user when the node restarts.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
