There are four different replication methods built into VolSync. Choose the method that best fits your use-case:
- Rclone replication
Use Rclone-based replication for multi-way (1:many) scenarios such as distributing data to edge clusters from a central site.
- Restic backup
Create a Restic-based backup of the data in a PersistentVolume.
- Rsync replication (via TLS)
Use Rsync-based replication for 1:1 replication of volumes in scenarios such as disaster recovery, mirroring to a test environment, or sending data to a remote site for processing.
- Rsync replication (via ssh)
This is the original rsync-based mover for 1:1 data replication. New deployments should favor the TLS-based implementation since the mover requires fewer privileges.
- Syncthing replication
Use Syncthing-based replication for multi-way (many:many), live, eventually consistent data replication in scenarios where the data is spread-out and updated in real-time, such as a wiki application, or a private distributed file-store.
The data replication mover Pods run in the user’s source and destination Namespaces. The permissions that are given to these Pods control what data can be replicated. They also affect the security of the cluster. Please see the permission model documentation for more details.
The data replication mover Pods can be configured to set specific resource requirements. This can include CPU and memory resource requirements or resource limits. Please see the resource requirements documentation for more details.
VolSync supports several types of triggers to specify when to schedule the replication.
PVC Annotations for Copy Triggers
VolSync supports source PVC annotations to coordinate triggering when VolSync takes a copy (snapshot or clone) for a replication.
VolSync exposes a number of metrics that permit monitoring the status of replication relationships via Prometheus.
VolSync provides a Volume Populator to allow creation of PVCs that reference a ReplicationDestination as a dataSourceRef.