Releasing nodes from the storage cluster
To release a node means to remove it from the storage cluster. Once you initiate the release, the storage cluster will start replicating data chunks that were stored on the released node, and then distributing them among other storage nodes in the cluster. Depending on the amount of data to replicate, the process may take as much as several hours. If necessary, you can also release a node forcibly, that is, without replication.
Releasing nodes forcibly may result in data loss.
- If the node runs one of the three required metadata services, the metadata role must be added to another node. You need to ensure that the cluster has at least three metadata services running at any time.
- If the node has any access points, the same access points are configured on other nodes in the cluster as well.
- If the node is in an iSCSI target group, it must be removed from the target group first.
- If the node has an S3 or backup gateway, the node IP addresses must be removed from the DNS records of S3 and Backup Gateway access points. Next, the node should be released from the S3 and Backup Gateway clusters.
- If the node is in the compute cluster, it must be removed from the compute cluster.
- The cluster has enough storage space to accommodate the data from the released node.
To release a node from the storage cluster
- On the Infrastructure > Nodes screen, click the line with the node to release.
- On the right pane, click Release node.
- (Optional, highly not recommended) To release the node without data migration, select Release without data migration.
- Click Release.
The released node will be displayed as Unassigned on the Infrastructure > Nodes screen.
Use the following command:
vinfra node release [--force] <node>
- Release node without data migration
- Node ID or hostname
For example, to release the node
node005 from the storage cluster with migration of data to maintain the set redundancy mode, run:
# vinfra node release node005