Item: Couchbase Server
Rating: 10
Author: Michael Steuer

Use Cases and Deployment Scope

We use Couchbase to support our online and mobile social casino games backend. In addition to a faster access to relational data, we use Couchbase for low latency large multiplayer gaming engines, where ephemeral data gets stored and distributed amongst many clients at once as they play and compete against each other in multiple games and competitions.

Pros and Cons

Cluster building.
Fast data access.
Integration with other components in the stack.

Process for replacing or upgrading nodes.
Warm up time.
Web interface for ad-hoc data access.

Return on Investment

Ability to run massively multiplayer games.
Flexible data model development.
Reliability and up time.

Scalability

We have run games with tens of thousands of concurrents without problems. We never had to scale past 6 nodes, but realize that with competing for data engines we would have to probably have a much heavier infrastructure to achieve similar results.

Performance

It's predominantly the high data volumes and operations that we were most astonished by. Having hundreds of client operations per second on a single record is quite remarkable. Now multiply that by billions of records.

Alternatives Considered

We have used basic Memcache, Mongo, Redis, and AWS based data services.

Likelihood to Recommend

Well suited in scenarios where flexible scaling and redundancy are absolute requirements. Well suited for geographically spread access models through cross data center replication.

Less suited in scenarios where lots of pre-existing relational data access models already exist in the application stack and need to be replaced in order for Couchbase to be implemented.

Speed, scalability and reliability - the holy trinity that only Couchbase can provide.

Overall Satisfaction with Couchbase