Performance Evaluation: MongoDB over NetApp E-Series

Vladimir Starostenkov


NetApp, a provider of high-performing data storage systems, has been working on adjusting its offerings to the requirements of NoSQL databases (such as MongoDB). As a result, the company now offers two MongoDB-certified flash storage solutions. Altoros joined the effort to evaluate these products.

This blog post reveals some of the performance results for the MongoDB integrated architecture deployed to NetApp E-Series.


Overview of the tested scenarios

The examined system design employs highly effective and scalable MongoDB cluster empowered with the additional storage reliability provided by E-Series. The results shed light on both performance and recovery after disk failure. A sharded MongoDB cluster over NetApp E-Series E5600 was tested under the following scenarios:

  • All-SSD drives (equivalent to an EF560)
  • HDD drives employing SSD drives for read acceleration (SSD Read Cache)
  • E5600 configured with Dynamic Disk Pools (DDP) for enhanced recovery times when a drive fails (performance on failure)


Testing SSD cache

The new SSD Read Cache feature was the most promising to the NoSQL team at Altoros, since it serves as a candidate to improve HDD-backed deployment performance closely to the one observed for the all-SSD configuration. The diagram below demonstrates some of the results:



Disk failure recovery

The following diagram illustrates fast recovery after a disk failure, as well as increased overall storage manageability.




So, as you can see, when implemented for high-load MongoDB deployments, NetApp E-Series outranks direct-access storages against a number of criteria. One of the main advantages is the opportunity to use a regular HDD storage armed with an SSD-based cache. In addition to low latencies and high throughput, the NetApp E-Series integrated architecture for MongoDB provides rapid rebuild after disk failures and other operational advantages.

You can find all the result of our tests in this 25-page technical report.

About the author: Vladimir Starostenkov is a Senior R&D Engineer at Altoros. He is focused on implementing complex software architectures, including data-intensive systems and Hadoop-driven apps. Having background in computer science, Vladimir is passionate about artificial intelligence and machine learning algorithms. His NoSQL and Hadoop studies were published in NetworkWorld,, and other industry media.

For more NoSQL research, subscribe to our blog or follow @altoros.

Get new posts right in your inbox!

No Comments

Benchmarks and Research

Subscribe to new posts

Get new posts right in your inbox!