Lifecycle Management
Apply retention, tiering, and deletion policies to partitions and rows across your lakehouse tables.
Lifecycle management governs how long data stays in your lakehouse, where it lives as it ages, and when it is removed. Crunch treats the lifecycle of a table as a policy you declare once, then enforces it continuously across every partition.
Partition lifecycle policies
A lifecycle policy is defined per table and evaluated on every scheduling cycle. Typical policies combine:
- Retention — keep partitions for a fixed period, then delete them.
- Tiering — move cold partitions to an infrequent-access or archive storage class once they pass an age threshold.
Compliance deletion
Crunch treats compliance as a first-class workflow rather than an ad hoc script:
- Partition-level deletion — delete entire partitions on a defined schedule or on demand, which covers most GDPR "right to be forgotten" requests where personally identifiable information (PII) is partition-isolated.
- Row-level PII deletion — for tables where PII is not partition-isolated, Crunch issues row-level deletes using the native delete mechanism of the table format (general availability coming soon).
- Audit trail — every deletion is recorded in a full audit log that you can export for compliance reporting.
Deletion policies are destructive by design. Validate the scope of a policy with data owners, and confirm that the retention window you configure meets your organization's legal obligations.
Interaction with table formats
| Format | Lifecycle mechanism |
|---|---|
| Apache Iceberg | Partition deletion writes tombstone files; snapshot expiration reclaims the storage; committed through the Iceberg commit protocol |
| Delta Lake | Partition deletion writes tombstone files; VACUUM reclaims the storage; committed through the Delta commit protocol |
| Hive / Parquet | Direct partition removal with catalog metadata updates |
Configure a lifecycle policy
Lifecycle policies are created in the Granica Console or through the Granica API. Each policy specifies the table scope, the age thresholds, the action to take at each threshold, and the SLA window for the resulting jobs.
See also
Table Compaction
Consolidate small files into optimally sized ones to reduce metadata overhead and speed up queries.
Garbage Collection
Reclaim storage by expiring snapshots, removing orphan files, and vacuuming stale data safely.