[
https://issues.apache.org/jira/browse/CASSSIDECAR-341?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrés Beck-Ruiz updated CASSSIDECAR-341:
-----------------------------------------
Epic Link: CASSSIDECAR-370
Description:
The Sidecar's job management framework
(https://issues.apache.org/jira/browse/CASSSIDECAR-150) currently [stores job
state
in-memory|https://github.com/apache/cassandra-sidecar/blob/0ad3e6fe545abc9254a88bd8de15f24af27d68c7/server/src/main/java/org/apache/cassandra/sidecar/job/OperationalJobTracker.java#L49].
This approach makes some assumptions around the duration of the job and is not
resilient, as all job information is lost if the Sidecar restarts.
To support long-running operations like `repair`
(https://issues.apache.org/jira/browse/CASSSIDECAR-268) that may take days to
complete, or cluster-wide operations such as a rolling restart, we must make
the job state durable.
This ticket will track all the subtasks needed to enable the current
operational job framework to store operational job state in persistent storage.
was:
The Sidecar's job management framework
(https://issues.apache.org/jira/browse/CASSSIDECAR-150) currently stores job
state in-memory. This approach makes some assumptions around the duration of
the job and is not resilient, as all job information is lost if the Sidecar
restarts.
To support long-running operations like `repair`
(https://issues.apache.org/jira/browse/CASSSIDECAR-268) that may take days to
complete, we must make the job state durable. This change will introduce a
persistent storage layer, ensuring that the status and progress of asynchronous
jobs survive restarts.
> Implement Durable State for Long-Running Sidecar Jobs
> -----------------------------------------------------
>
> Key: CASSSIDECAR-341
> URL: https://issues.apache.org/jira/browse/CASSSIDECAR-341
> Project: Sidecar for Apache Cassandra
> Issue Type: Improvement
> Reporter: Arjun Ashok
> Priority: Major
> Labels: RollingRestart
>
> The Sidecar's job management framework
> (https://issues.apache.org/jira/browse/CASSSIDECAR-150) currently [stores job
> state
> in-memory|https://github.com/apache/cassandra-sidecar/blob/0ad3e6fe545abc9254a88bd8de15f24af27d68c7/server/src/main/java/org/apache/cassandra/sidecar/job/OperationalJobTracker.java#L49].
> This approach makes some assumptions around the duration of the job and is
> not resilient, as all job information is lost if the Sidecar restarts.
> To support long-running operations like `repair`
> (https://issues.apache.org/jira/browse/CASSSIDECAR-268) that may take days to
> complete, or cluster-wide operations such as a rolling restart, we must make
> the job state durable.
> This ticket will track all the subtasks needed to enable the current
> operational job framework to store operational job state in persistent
> storage.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
