Data Lifecycle
Recover soft-deleted business records, set retention, preview dependent impact, and hard-purge with a protected cleanup ledger.
Data Lifecycle
Business records in EGKits are soft-deleted by default: the row stays in the database with a deleted marker so you can restore or investigate. Data Lifecycle is the Soft-Deleted Records Center — discover what is recoverable, restore when appropriate, set retention, and permanently remove rows only after a structured preview.
Audit Trail rows are different: they are hard history and use their own Audit retention path. Both features share the same cleanup ledger and concurrency locks.
#Who can use it
| Capability | Permission |
|---|---|
| Browse soft-deleted catalog, preview, restore | Permissions.DataLifecycle.View |
| Change retention, execute hard purge, cancel / unlock | Permissions.DataLifecycle.Manage |
Open Data Lifecycle from Administration (/data-lifecycle). Tabs typically include Catalog, Retention, Operations / History, and Settings.
#Soft delete vs hard purge
flowchart TD Live[Live business row] -->|Delete in ERP| Soft[Soft-deleted row] Soft -->|Restore| Live Soft -->|Retention / selected purge| Preview[Preview impact] Preview -->|Confirm DELETE when required| Hard[Hard purge — row gone] Hard --> Ledger[Cleanup operation ledger]
| Concept | Meaning |
|---|---|
| Soft delete | Row remains; filtered out of normal lists; recoverable |
| Restore | Clears the deleted marker (when the module supports restore) |
| Hard purge | Permanent SQL delete after preview; not undoable |
| Cleanup ledger | Immutable history of purge runs (actor, counts, window) |
#Soft-Deleted Records Center
#Catalog
Browse soft-deleted entities by module and type. The catalog is built from the application’s soft-delete model sources (Identity, Modules, Todo, and other registered domains). Use filters and paging — do not attempt to load every deleted row into the browser.
#Preview before purge
Always Preview before Execute. Preview reports:
- Rows selected for removal
- Soft-deleted dependents that will be expanded into the purge set
- Active dependents that block purge of a parent (those parents are skipped until dependents are handled)
- Estimated impact counts for the confirmation dialog
Zero-row previews cannot be executed — refine filters or selection first.
#Selected purge and dependents
When you purge a selection:
- The service expands soft-deleted dependents so orphaned children are not left behind.
- Parents that still have active (non-deleted) dependents are skipped with a clear reason.
- Large scopes may require typing
DELETEbefore execute. - Concurrent purges for the same tenant are locked; wait or cancel from Operations.
#Retention policy
Set how many days soft-deleted rows are kept before the scheduled Soft-Delete Retention sweep can remove them. Align this with tax, contract, and privacy requirements — shorter is not always safer if legal hold applies.
#Operations and history
Successful (and failed) purge runs appear in the cleanup ledger. Ledger entries are protected: they are not removable from the Soft-Deleted Records Center. Use history for auditors who need proof of what was permanently removed and by whom.
#Recommended policy pattern
| Layer | Typical control |
|---|---|
| Soft-delete retention (days) | Business + legal minimum for recoverable mistakes |
| Audit trail retention (days) | Often longer; evidence of who changed what |
| Auto-purge | Enable only after previewing a dry run in a non-prod tenant |
| Manage permission | Break-glass role; fewer holders than View |
#Interaction with Audit Trail
| Action | Soft-deleted business row | Audit Trail |
|---|---|---|
| User deletes a document | Marked deleted | New trail row (Delete/Update) |
| User restores | Marker cleared | New trail for restore/update |
| Hard purge of business row | Row removed | Trails remain until audit retention |
| Hard purge of audit rows | N/A | Batched delete via Audit Retention |
Never treat “empty audit search” as proof that a business row never existed — check Data Lifecycle catalog and module recycle views as well.
#Safety checklist
- Preview first; never execute from memory of an earlier preview.
- Confirm
DELETEwhen the UI requires it (large / year / up-to-date scopes). - Prefer narrow module + date filters over “purge everything older than…”.
- Keep
DataLifecycle.Manageoff standing day-to-day roles. - After a production purge, export or note the ledger operation id for the change record.
#Troubleshooting
| Issue | What to check |
|---|---|
| Catalog empty | Soft-delete sources registered; filters too narrow; permission View |
| Preview shows blockers | Active dependents still reference the parent — soft-delete or reassign them first |
| Execute blocked (zero rows) | Selection resolved to nothing after expansion / skip rules |
| Forbidden on Retention / Execute | Needs DataLifecycle.Manage |
| Purge lock held | Another job is running; wait or cancel with Manage |
#Related guides
- Audit Trail — investigation, export, audit hard purge
- Security Best Practices — least privilege and evidence cadence
- Identity & Access — assigning View / Manage
- System Configuration — tenant operational defaults