[
https://issues.apache.org/jira/browse/CASSSIDECAR-460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18082423#comment-18082423
]
Andrés Beck-Ruiz commented on CASSSIDECAR-460:
----------------------------------------------
CI:
https://app.circleci.com/pipelines/github/andresbeckruiz/cassandra-sidecar/46/workflows/b2485d2a-1613-41ce-9b2e-5037c28e0352
> Refactor OperationalJob to have data separate from execution logic
> ------------------------------------------------------------------
>
> Key: CASSSIDECAR-460
> URL: https://issues.apache.org/jira/browse/CASSSIDECAR-460
> Project: Sidecar for Apache Cassandra
> Issue Type: Sub-task
> Reporter: Andrés Beck-Ruiz
> Assignee: Andrés Beck-Ruiz
> Priority: Major
> Labels: CEP-53
> Time Spent: 2h
> Remaining Estimate: 0h
>
> The {{OperationalJob}} class combines data fields (jobId, status,
> operationType, node lists) with execution logic. This creates two issues:
> 1. A {{DurableOperationalJobTracker}} cannot persist a job record before
> execution starts because {{OperationalJobManager}} bundles creation and
> execution in the same mapping function.
> 2. Retrieving historical jobs from storage requires extending
> {{OperationalJob}} and stubbing execution methods with
> {{UnsupportedOperationException}}.
> In order to resolve these issues, we should extract a data-only interface
> from {{OperationalJob}}, update the {{OperationalJobManager#trySubmitJob}} to
> separate creation from execution, and update {{OperationalJobTracker#get}}
> handlers to use the data interface for read paths.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
