[
https://issues.apache.org/jira/browse/CASSANDRA-20489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17939223#comment-17939223
]
Sam Tunnicliffe commented on CASSANDRA-20489:
---------------------------------------------
+1 LGTM
> Avoid failing queries when epoch changes and replica goes up/down
> -----------------------------------------------------------------
>
> Key: CASSANDRA-20489
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20489
> Project: Apache Cassandra
> Issue Type: Bug
> Components: Transactional Cluster Metadata
> Reporter: Marcus Eriksson
> Assignee: Marcus Eriksson
> Priority: Normal
> Fix For: 5.x
>
> Attachments: ci_summary.html
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> To make sure that we got responses from the correct replicas we recompute the
> replica plan for queries if the epoch changes during the query to make sure
> we got responses from a quorum of the nodes in the recomputed replica plan.
> Problem is that this recomputation is filtering away down nodes, so we can
> fail queries if the coordinator marks a replica down during the query and the
> epoch changes, even though the placements have not changed.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
