[
https://issues.apache.org/jira/browse/CASSANDRA-18510?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17732275#comment-17732275
]
David Capwell commented on CASSANDRA-18510:
-------------------------------------------
this might be less of an issue in C* as we handle this in messaging layer, we
won't call process until the epoch is ready
> CEP 15: (Accord) push waiting for epoch into CommandStore rather than require
> every command to think about the edge case when the epoch isn't ready
> ---------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-18510
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18510
> Project: Cassandra
> Issue Type: Improvement
> Components: Accord
> Reporter: David Capwell
> Priority: Normal
> Fix For: 5.x
>
>
> We have a common pattern of scheduling work on a CommandStore, but this work
> has to double check the epoch exists, if not it has to wait on the fetch and
> reschedule again… this pattern could likely be pushed to CommandStore if the
> PreLoadContext includes the epochs required (just like we do for commands).
> For things outside of CommandStores, they would have to maintain their
> current logic, the scope of this work is just tasks that run on CommandStores.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
