All 7 Clustering questions
More in Clustering
What is the difference between a master node and a replica node within a Redis Cluster? How does Redis Cluster automatically handle failover when a master node becomes unavailable? What are the key limitations of Redis Cluster that developers should be aware of compared to a single instance Redis setup? How do you monitor the overall health and status of a Redis Cluster to catch problems before they affect your application? What is a split brain scenario in the context of Redis Cluster, and how does the cluster protect against it?
Advertisement
Back to results for “Performance”
Redis › Clustering

How would you design a Redis Cluster deployment to properly balance high availability, performance, and cost for a large scale production application?

Advanced

Answer

You would choose a number of master nodes sufficient to spread your data and traffic appropriately, add at least one replica per master for failover protection, place replicas in different physical locations or availability zones than their masters to survive localized failures, and continuously monitor slot distribution and node health to catch imbalances or failures early.
-- Spread masters and their replicas across
-- different availability zones for resilience
redis-cli --cluster create ... --cluster-replicas 1
Real-world example A large financial services company deploys their Redis Cluster with masters and replicas spread across three different availability zones, ensuring the cluster survives an entire zone outage without losing access to any of their data.

Common follow-ups: What is the tradeoff between more replicas for safety versus infrastructure cost?;How do you test that a cluster deployment actually survives a simulated zone failure?

Redis Backup & Disaster Recovery;Redis Monitoring & Observability
All 7 Clustering questions