TheDocumentation Index
Fetch the complete documentation index at: https://mintlify.com/getployz/ployz/llms.txt
Use this file to discover all available pages before exploring further.
migrate command moves a workload — container and persistent volumes — from one machine to another within the cluster. Migration uses ZFS incremental send to transfer volume state efficiently, with minimal downtime. It is a single explicit operation with a clear result: the workload is running on the target machine or the command fails cleanly.
Migration transfers both the container configuration and any persistent ZFS volumes attached to the service. The source volumes are snapshotted, sent to the target incrementally, and the workload is restarted on the target. Always preview before applying in production.
Identifying a service
Theservice-ref positional argument identifies the workload to migrate. It takes the form namespace/service-name, for example production/web or staging/worker.
Subcommands
preview — preview a migration
preview — preview a migration
Compute the migration plan without executing it. Returns what would be transferred, which volumes are involved, and the target machine. Use this before any production migration.The preview response includes the target machine, the service’s current placement, and the list of volumes that will be transferred.
The service to migrate, in
namespace/service format.ID or address of the target machine to migrate the workload to.
apply — execute a migration
apply — execute a migration
render-manifest — render the migration manifest
render-manifest — render the migration manifest
Render the manifest that describes the migration operation, without applying it. Useful for auditing, storing in version control, or applying manually.Outputs the manifest to stdout.
The service to migrate, in
namespace/service format.ID or address of the target machine.
How migration works
- The daemon snapshots all ZFS volumes attached to the service on the source machine.
- The snapshot is sent to the target machine using ZFS incremental send over the cluster’s WireGuard overlay network.
- Once the transfer is complete, the workload is stopped on the source and started on the target from the received volumes.
- The cluster membership and routing records are updated to reflect the new placement.
Recommended workflow
Drain the source machine (optional)
If migrating because of planned maintenance, drain the source first to stop new workload placements.
Apply the migration
Execute the migration. The daemon returns when the workload is running on the target.