Resync Cluster
Manually trigger a full cluster-wide rebalance walk.
| Endpoint | Method | Admin Required |
|---|---|---|
/api/v1/admin/cluster/resync | POST | Yes |
Triggers a full rebalance walk across every bucket, on every node, under the current membership view. This is the same walk that runs automatically on membership change (a node joining, draining, or being removed) - use this to manually re-trigger it, most commonly for a node that's been down long enough to fall past the async replication outbox's retry limit and needs a full resync rather than just catching up.
This is always cluster-wide, not scoped to one node - every node independently walks its own local disk and reconciles it against current placement, regardless of which node the request was made against.
Response
Returns 200 OK once the walk has been scheduled. The walk itself runs in the background - check rebalance status for progress.