Redis Cluster

Redis Cluster is a distributed implementation of Redis, the open-source in-memory data structure store. It can be used for data storage, caching, message brokering, and similar workloads while providing high performance and high availability.

Redis Cluster Overview

The Redis Cluster package creates a topology with at least three Primary servers. Every Primary node is paired with a Secondary node to distribute read load and support automatic recovery when a Primary becomes unavailable.

The cluster can be scaled manually or automatically. Every scaling operation adds or removes a pair of nodes: one Primary and one Secondary.

Redis Cluster Installation

Automatic installation and configuration are available through the Redis Auto-Clustering option in the topology wizard.

Alternatively, import the Redis Cluster package manifest or install the package from Marketplace. The installation window provides the following parameters:

Default Horizontal Auto-Scaling Conditions

• Add 2 nodes when CPU or Memory load remains above 70% of total cluster capacity for more than 5 minutes, up to 12 total nodes.
• Remove 2 nodes when CPU or Memory load remains below 40% of total cluster capacity for more than 15 minutes, while keeping at least 6 total nodes.

After configuring the required settings, click Install.

When deployment finishes, the platform displays a success notification with Redis Cluster administration-panel credentials. The same access information is also sent by email.

Cluster Access and Verification

Once the cluster is deployed, open the administration panel to review cluster details and make the required configuration changes.

You can also connect to any Redis cluster node through SSH, such as Web SSH, and use redis-cli to verify the cluster state.

redis-cli
auth {passw0rd}
cluster nodes

The cluster nodes output can be requested from any Primary or Secondary node. It shows the cluster node count, node IDs, addresses, roles, and the hash slots assigned to each node.

Expected Result

A Redis Cluster is deployed with at least three Primary-Secondary pairs. The cluster can be scaled in node pairs, optionally uses automatic horizontal scaling, and can be inspected through the administration panel or redis-cli.

Important Notes

• The minimum documented topology is 6 nodes: 3 Primary and 3 Secondary.
• The maximum documented topology is 12 nodes: 6 Primary and 6 Secondary.
• Scaling always occurs in Primary-Secondary pairs.
• Auto-scaling performs automatic resharding and rebalancing during node changes.
• Enabling external IP addresses assigns a public IP to every cluster node and switches cluster communication to public IPs only.
• Administration-panel credentials are shown after deployment and sent by email.