Percona Tests Show High Scalability and Sustained Performance of Clustrix Database Appliances Compared to MySQL Systems

Largest Independent MySQL Services Vendor Confirms Performance and Scale Benefits of the Clustrix Parallel Distributed Database Architecture


Clustrix announced the results of a Percona head-to-head evaluation of Clustrix and MySQL systems, showing the high scalability and sustained performance of Clustrix database appliances compared to MySQL. The tests showed once more what Clustrix users have already been experiencing in web environments: with the Clustrix systems, they do not have to give up transactions or ACID properties, relational semantics, or any of the rich SQL features to meet the performance and scale demands of rapidly growing high-traffic Internet applications. Full tests and findings can be downloaded at

Percona ran a Percona-written, TPCC benchmark against Percona’s MySQL Server on an Intel- SSD based machine, MySQL on a Fusion-io-based machine, and different sizes of Clustrix database appliances based on a parallel distributed database architecture. On top of the high performance and scalability delivered by the Clustrix system, it also provides built-in fault tolerance and no single point of failure. The MySQL systems tested either have a single Fusion-io card or SSD drives in a RAID 0 configuration. If any component – such as drive, motherboard, or memory – fails, the entire database will fail.

The first conclusion that Percona noted was that the Clustrix database really was indeed scalable. As a website gets more and more users, concurrency goes up and aggregate performance throughput requirements increase while users expect the same or better response time.

According to Percona, the experts in MySQL performance, “For highly concurrent workloads, Clustrix provides great performance, and if you need better throughput, it scales just by adding more nodes.”

“The most interesting conclusion in my opinion is the fact that transactional, relational, full-featured SQL can scale and Clustrix has proved it in this testing and at multiple production customer deployments around the world,” said Aaron Passey, Clustrix CTO, who recently wrote a blog post describing the tests and results. “The test showed a complicated workload with multi-statement transactions, aggregates, foreign key constraints, and a large percentage of writes, which are what the NoSQL movement has proclaimed as too difficult to scale. Clustrix has solved this problem; once more, we deliver the unlimited scale promised by NoSQL without giving up the features of SQL, at unmatched performance.”