Expand an AIStor Deployment on Linux

The following procedure adds a server pool to an existing AIStor deployment running on baremetal infrastructure. Each pool expands the total available storage capacity of the cluster while maintaining the overall availability of the cluster.

All commands provided below use example values. Replace these values with those appropriate for your deployment.

Deploy AIStor on new hardware

Follow the installation guide to install AIStor onto the new hardware.

Do not start the new deployment - stop at the step Review the environment file
Update the deployment topology

AIStor uses the MINIO_VOLUMES environment variable for determining the cluster topology.

When adding a new pool, append the new set of hostnames and volumes to the existing set. For example, consider the following MINIO_VOLUMES describing a single-pool consisting of 16 nodes and 16 drives:
```
MINIO_VOLUMES="https://minio-{1...16}.example.net:9000/mnt/drive-{1...16}/minio"
```
To expand that topology with another 16x16 deployment, append the new set of hosts to the existing value:
```
MINIO_VOLUMES="https://minio-{1...16}.example.net:9000/mnt/drive-{1...16}/minio https://minio-{17...32}.example.net:9000/mnt/drive-{1...16}/minio""
```
You must include the protocol http[s]:// for each new pool. All pools should have matching protocols.

Make this change to all hosts in the AIStor deployment, including the existing pools. The environment files should match across all nodes. Use shasum or a similar utility to validate that all contents match.
Initialize the new deployment topology

You must simultaneously
1. Restart the existing AIStor deployment and
2. Start the new AIStor server nodes
For the existing nodes, you can use the mc admin service restart command to restart all nodes at once. For the new nodes, use systemctl start minio on all nodes in parallel using your preferred shell tooling.
Monitor the deployment status

Use the mc admin info command and journalctl -u minio system utility to monitor status and logging across the cluster. While the nodes come online and synchronize, the system logs may showcase increased churn through warnings and errors.
Next steps

After validating cluster health and expanded storage, you can update any load balancers, reverse proxies, or other network management tools with the new hostnames of the added server pool. This ensures that incoming connections balance appropriately across the expanded topology.