lasdf1234 opened a new issue, #13760: URL: https://github.com/apache/gravitino/issues/13760
### Describe the proposal Implement the persistence and in-job execution foundation for the in-process Table Maintenance Service (TMS) as a `plugins/tms` module (same SPI / mapper packaging pattern as `idp-basic`, not in `core`). This epic delivers: - Relational DDL and DAOs for `scheduled_tasks` (db-scheduler), `table_maintenance_job`, and `table_snapshot_metrics` - Three db-scheduler pools (expand / cron / commit), REST Feature lifecycle, CAS claim / heartbeat, and occupancy reconcile - In-Spark Job gates: sample metrics at snapshot `S0`, snapshot-drift early exit, and recommender maintain/skip - Execution of four maintenance types (compaction, manifest-rewrite, snapshot-expiry, orphan-cleanup) inside Spark Jobs - Commit-chain runs with Iceberg OCC (`validateFromSnapshot`) and validation metrics Design: [apache/gravitino#13386](https://github.com/apache/gravitino/pull/13386). Related: [apache/gravitino#13488](https://github.com/apache/gravitino/issues/13488). **PR size:** ~800–1100 lines per subtask (including tests), max 1100. ### Task list - [ ] `plugins/tms` skeleton, DDL, and DAOs - [ ] Scheduler pools, Feature lifecycle, CAS, and occupancy - [ ] In-job S0 sample, snapshot drift, and recommend gate - [ ] Four maintenance types in Spark Job execution - [ ] Commit-chain, Iceberg OCC, and validation metrics -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
