Is your feature request related to a problem? Please describe.
As a system administrator, I need to migrate large OpenCloud instances (potentially 1PB+) from one deployment to another (e.g., cloud to on-premises, or between different infrastructures, or cephfs different erasure coded pools) without causing downtime for users. Currently, there's no built-in way to safely perform these migrations. Manual migration approaches require significant downtime, risk data loss, and are complex to execute, especially for large instances that can take weeks to transfer.
Describe the solution you'd like
I would like an admin UI feature (and/or CLI tool) that allows connecting an old OpenCloud instance to a new one and managing a safe migration with zero downtime:
Core workflow:
- Connect the old instance (source) to the new instance (destination) through the admin interface
- Configure and start an initial bulk synchronization of all files, users, and metadata in the background
- After initial sync completes, enable continuous delta synchronization to keep the new instance updated with any changes
- When ready for cutover, trigger a one-click switchover that performs a final sync, sets the old instance to read-only mode, then disables it while activating the new instance
Key capabilities needed:
- Support for multi-tenant instances (sync all tenants or specific ones)
- Progress monitoring showing sync status, data transferred, and estimated completion time
- Resume capability for interrupted syncs (critical for week-long transfers)
- File integrity verification using checksums
- Bandwidth and schedule controls to avoid overloading networks
- Final delta sync during cutover to catch last-minute changes
- Safe rollback capability if issues are detected
This would allow sysadmins to sync data over weeks while the old instance remains active, then perform a quick, safe cutover with minimal risk.
Describe alternatives you've considered
Manual rsync/rclone approach: Copy files manually using external tools, then migrate database separately. This is error-prone, doesn't handle live changes well, requires significant downtime, and risks data inconsistency.
Snapshot and restore: Take a full snapshot, disable the old instance, restore to new instance. This causes extended downtime proportional to data size (unacceptable for large instances).
Third-party migration tools: External tools don't understand OpenCloud's internal structure, metadata, and multi-tenant architecture, leading to incomplete or broken migrations.
Custom scripts: Possible but requires deep OpenCloud knowledge, significant development effort, and ongoing maintenance. Not feasible for most administrators.
None of these alternatives provide the safety, ease of use, and zero-downtime guarantees that a built-in migration tool would offer.
Additional context
This feature would be particularly valuable for:
- Cloud-to-on-premises migrations driven by cost or compliance requirements
- Infrastructure upgrades requiring fresh installations
- Instance consolidation projects
- Disaster recovery scenarios
Similar migration capabilities exist in enterprise database systems (replication with failover) and storage platforms, but are currently absent from OpenCloud. For organizations running large deployments where downtime translates to significant business impact, this feature would make OpenCloud a more viable enterprise solution.
The ability to test the new instance while syncing, and knowing the cutover can be done safely with a final incremental sync taking minutes rather than days, would dramatically reduce migration risk and complexity.
Is your feature request related to a problem? Please describe.
As a system administrator, I need to migrate large OpenCloud instances (potentially 1PB+) from one deployment to another (e.g., cloud to on-premises, or between different infrastructures, or cephfs different erasure coded pools) without causing downtime for users. Currently, there's no built-in way to safely perform these migrations. Manual migration approaches require significant downtime, risk data loss, and are complex to execute, especially for large instances that can take weeks to transfer.
Describe the solution you'd like
I would like an admin UI feature (and/or CLI tool) that allows connecting an old OpenCloud instance to a new one and managing a safe migration with zero downtime:
Core workflow:
Key capabilities needed:
This would allow sysadmins to sync data over weeks while the old instance remains active, then perform a quick, safe cutover with minimal risk.
Describe alternatives you've considered
Manual rsync/rclone approach: Copy files manually using external tools, then migrate database separately. This is error-prone, doesn't handle live changes well, requires significant downtime, and risks data inconsistency.
Snapshot and restore: Take a full snapshot, disable the old instance, restore to new instance. This causes extended downtime proportional to data size (unacceptable for large instances).
Third-party migration tools: External tools don't understand OpenCloud's internal structure, metadata, and multi-tenant architecture, leading to incomplete or broken migrations.
Custom scripts: Possible but requires deep OpenCloud knowledge, significant development effort, and ongoing maintenance. Not feasible for most administrators.
None of these alternatives provide the safety, ease of use, and zero-downtime guarantees that a built-in migration tool would offer.
Additional context
This feature would be particularly valuable for:
Similar migration capabilities exist in enterprise database systems (replication with failover) and storage platforms, but are currently absent from OpenCloud. For organizations running large deployments where downtime translates to significant business impact, this feature would make OpenCloud a more viable enterprise solution.
The ability to test the new instance while syncing, and knowing the cutover can be done safely with a final incremental sync taking minutes rather than days, would dramatically reduce migration risk and complexity.