[
https://issues.apache.org/jira/browse/CASSSIDECAR-457?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jyothsna Konisa updated CASSSIDECAR-457:
----------------------------------------
Fix Version/s: 0.4
0.4.0
Source Control Link:
https://github.com/apache/cassandra-sidecar/commit/9f8832b65f2ab57d4e8afd09bc0f7f27788f3fe3
Resolution: Fixed
Status: Resolved (was: Ready to Commit)
> Fix CDC message loss during Sidecar restarts
> --------------------------------------------
>
> Key: CASSSIDECAR-457
> URL: https://issues.apache.org/jira/browse/CASSSIDECAR-457
> Project: Sidecar for Apache Cassandra
> Issue Type: Task
> Components: CDC
> Reporter: Jyothsna Konisa
> Assignee: Jyothsna Konisa
> Priority: Major
> Fix For: 0.4, 0.4.0
>
>
> The analytics library's SidecarStatePersister calls EventConsumer.flush()
> before writing the CDC offset checkpoint at shutdown. Because flush() was a
> no-op, in-flight Kafka records were abandoned, and the checkpoint was
> advanced past them on restart, silently dropping events. The fix overrides
> flush() in CdcEventConsumer to call the blocking KafkaPublisher.flush(),
> ensuring all records are broker-acknowledged before the checkpoint is written.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
