Purge All Data¶
Irreversible destructive operation
Purge All Data permanently deletes every transfer record, data file, and transfer history entry across all destinations and all Data Movers. There is no undo. Use only when performing a complete system reset.
The Purge All Data action is available under Administration → Purge All Data (/do/admin/purge) in the Monitoring UI.
Confirmation Flow¶
The action is intentionally protected by a two-step confirmation to prevent accidental execution.
Step 1 — Acknowledge the warning¶
Read the warning panel, then tick:
I understand that this will permanently delete all data from the database and from all data-mover disks, and that this action cannot be reversed.
Click Proceed to final confirmation to advance to step 2. The button is disabled until the checkbox is ticked.
Step 2 — Final confirmation¶
Three controls are presented:
| Control | Required | Description |
|---|---|---|
| Confirmation phrase | Yes | Type PURGE ALL DATA exactly (case-sensitive) to enable the submit button. |
| Critical Password | Only if set | If a Critical Password has been configured, it must be supplied here. If no password has been set yet, this field is not shown — configure one first. |
| Trigger cleanup immediately | Optional | Wakes the purge and file-expiry schedulers immediately rather than waiting for their next scheduled cycle. Recommended in most cases. |
| Also hard-delete all database records immediately | Optional | See Immediate database hard-delete below. |
Click Confirm — Delete All Data to execute.
What Gets Deleted¶
| Data | Mechanism |
|---|---|
All DATA_TRANSFER rows | Marked as deleted; DAT_EXPIRY_TIME set in the past |
All DATA_FILE rows | Marked as deleted; picked up by the file-expiry scheduler |
| Physical files on Data Movers | Removed during the next (or immediate) mover purge scan |
| Transfer history, statistics, publications | Removed during the immediate hard-delete if that option is selected |
Transfers are marked for deletion first; the ExpiredDataFileScheduler and PurgeScheduler then handle physical removal of files from each Data Mover's disk.
Immediate Database Hard-Delete¶
When Also hard-delete all database records immediately is ticked, OpenECPDS performs a second pass that hard-deletes records in foreign-key-safe order:
UPLOAD_HISTORYandINCOMING_HISTORY— foreign-key references toDATA_FILEare nulled out first.TRANSFER_STATISTICSTRANSFER_HISTORYPUBLICATION(rows referencing data files)DATA_TRANSFERMETADATA_VALUEDATA_FILE
After the database records are removed, OpenECPDS immediately triggers a full disk scan on every connected Data Mover via the purgeAll() RMI call. Because the mover scan works by comparing files on disk against entries in the database, any file no longer referenced in the database is deleted at this point — without waiting for the next scheduled scan cycle.
This runs in the background
With large datasets (millions of transfers) the delete loop can take anywhere from a few minutes to several hours. The operation runs in a background thread so the browser does not time out. You can safely leave the page; progress is visible in the MasterServer log.
After the Purge¶
Once the purge completes:
- All destination queues are empty.
- No files remain on any Data Mover's disk (subject to mover scan completion).
- The database contains no transfer, history, or data-file records.
- Destinations, hosts, and all other configuration remain intact.
The system is ready to accept new transfers immediately.
Prerequisites¶
- You must have administrator access to the Monitoring UI.
- A Critical Password should be configured. If none is set, a warning banner is shown on the Administration page and the form will not ask for it — but it is strongly recommended to set one before using this feature in a production environment.