I’m not very fond of proprietary hardware. But, as I wrote in March when I saw it for the first time, Pure Storage FlashBlade seems great and demonstrates that Pure Storage has all the attributes to become a primary storage vendor and compete on equal terms against the usual suspects.

No longer a startup

Pure Storage is demonstrating to the world that this is not a startup any longer and is maturing very quickly. The product line up is growing as are the number of use cases Pure can now cover with its products. And it is still doing this with the speed and flexibility that is common to startups. A real threat to established vendors, IMO.

Pure Storage is demonstrating to the world that this is not a startup any longer and is maturing very quickly.

The recipe is quite simple: nice guys, good marketing, great products and an amazing leadership which is shaping a great mid/long term strategy. Pure has all the characteristics to do well in the storage space and I’m sure they won’t be stopping with the FlashBlade (or, at least, the FlashBlade as we know it…).

A great new product

ICYMI, Pure gave one of the best presentations of the entire SFD10 event.

FlashBlade is simply impressive.

The level of detail in the presentation and the prompt answers to all the questions from delegates highly impressed me. They look very confident about what they did and I wouldn’t do them justice by saying the contrary. This product sounds amazingly good (on paper at least). FlashBlade is simply impressive. It’s still at the beginning of its life (and this is also why it is not GA yet), but you can’t say that the architecture is not designed to face all the challenges they are going to meet in the high performance scale-out file and object field, with HPC-like and analytics workloads in the front row and other kinds of high demanding use cases following immediately after.

Closing the circle

FlashBlade is a cool product and I hope to see some integration between FlashArray and FB in the future (like, for example, snapshot offloading from FA to FB or FB as a hub for multiple FA replicas… somewhat similar to what NetApp is doing with its Data Fabric). But this is just more whishful thinking at the moment, Pure will probably concentrate its efforts on improving the number of supported protocols (only NFS and S3 at launch), scalability of the system (now only one chassis with plans to have a full rack configuration soon) as well as additional features.

One thing I’d like to note here is that just a few weeks later Pure Accelerate (the event where FlashBlade was presented), EMC introduced an AFA version of Isilon (project Nitro) at EMCWorld which addresses the same identical use cases (and with a very similar form factor!). It’s a great validation for Pure, isn’t it?