Amazon EC2 now powered by high performance storage (benchmark)

A while ago I had a discussion with someone about the future of server infrastructure. Among other things, we were wondering whether companies will continue to run on dedicated servers or if eventually everyone just ends up in a Cloud environment. During the discussion I raised a point that in principle Cloud is a great idea that will keep attracting more and more people, but it is missing one important piece that stops many from using it – a high performance storage. Apparently, this has just changed.

Yesterday I received an e-mail announcing a new EC2 instance type – hi1.4xlarge. It features 16 logical CPUs (35 ECUs), 60GB of RAM, and… two 1TB SSD-based disk volumes! These are great specs that should work for nearly any database. Even assuming someone has a MySQL database larger than 2TB, not all tables will require fast storage, while more disk space can be easily added by attaching regular EBS volumes.

I decided to give this new instance a spin and try to see whether the storage performance is really that good. I ran Sysbench fileio benchmark with 1, 2, 4, 8 and 16 I/O threads and 16KB block size in both random read and random write modes. Here are the results that I got from a RAID0 array created over both available volumes:

That is not bad at all. The improvement over a standard EBS-based array in both throughput and response time is at least by an order of magnitude. This can actually make EC2 usable even for large and busy MySQL databases, which up until now often wasn’t a viable option. The price does not seem that high either especially after I tried building a similar hardware configuration for a dedicated server at one of the popular hosting providers.

[MySQL Health Check]
About Maciej Dobrzanski

A MySQL consultant with the primary focus on systems, databases and application stacks performance and scalability. Expert on open source technologies such as Linux, BSD, Apache, nginx, MySQL, and many more. @linkedin


  1. Hi Maciej,

    I wouldn’t mind you sharing those price comparisons. So far I’ve always ended up with 2 to 3 times more expensive prices to have high-availability systems on EC2 rater than dedicated. Also, providing an additional comparison with EC2 *without* the SSDs would definitely be more helpful.


    • Well, $3,10/h that Amazon charges for the hi1 instance gives apx. $2300 per month. I created a similar configuration at SoftLayer (8 cores + HT, 48GB RAM, 4x400GB SSD) and the website said it would cost around $2900 per month. I’d say EC2 would not be more expensive than that even assuming some extra fees. Perhaps there are some cheaper options with other managed hosting providers, but I didn’t investigate.

      You can find plenty of EBS benchmarks on the internet. No chance for 10k+ IOPS (or even much less than that) or response times at 0,5ms or less. As far as I can remember a read latency on a regular EBS disk is usually around 5-6ms in the best case, which is about 10 times slower.

  2. Jason Dunkerley says:

    Thanks for the great info and performance benchmarks. This new option really does look to be a game changer. It should be noted that the price for these instances can be lowered substantially if you are willing to commit to a 1year term and use the “Heavy Utilization” reserved instances. It works out to a little over $1k per month. That’s a hell of a deal.


  1. […] benchmark for the amazons new High I/O Quadruple Extra Large Instance. You can check that out here. var dd_offset_from_content = 40; var dd_top_offset_from_content = […]

Speak Your Mind