The Storage Future is Cloudy, and it is about time.
One of the things I have talked about quite a bit in the last couple of months is the disjoint between the needs of enterprise IT and the offerings of a wide swath of the cloud marketplace. Some times it seems like many cloud vendors are telling customers “here’s what we choose to offer you, deal with it”. Problem is, oftentimes what they’re offering is not what the enterprise needs.
There are of course some great examples of how to do cloud for the enterprise, Rackspace (among others) has done a smashing job of offering users a server with added services to install a database or web server on those servers. There are still some security concerns that the enterprise needs to address, but at least it’s a solid start toward giving IT the option of using the cloud in the same manner that they use VMs. Microsoft has done a good job of setting up databases that follow a similar approach. If you use MS database products, you know almost all you need to know to run an Azure database.
The storage market has been a bit behind the rest of the cloud movement, with several offerings that aren’t terribly useful to the enterprise, and a few that are, more or less. That is changing, and I personally am thrilled. The most recent delve into cloud storage that is actually useful to the enterprise without rewriting their entire systems or buying a cloud storage gateway is Hitachi Data Systems (HDS). HDS announced their low-risk cloud storage services on the 29th of June, and the press largely yawned. I’m not terribly certain why they didn’t get as excited as I am, other than the fact that they are currently suffering information overload where cloud is concerned.
With the style of offering that HDS has set up, you can use your HDS gear as in “internal” cloud, and their services as “external” cloud, all managed from the same place. And most importantly, all presenting as the storage you’re used to dealing with. The trend of many companies to offer storage as an API is self-defeating, as it seriously limits usefulness to the enterprise and has spawned the entire (mondo-cool, in context) cloud storage gateway market. The HDS solution allows you to hook up your disk like it was disk, not write extra code in every application to utilize disk space “in the cloud” and use the same methods you always have “in the DC”. To do the same with most other offerings requires the purchase of a cloud storage gateway.
So you can have your cloud and internal too. The future of storage is indeed looking cloudy these days, and I’m glad for it. Let the enterprise use cloud, and instead of telling them “everyone is doing it”, give them a way to use it that makes sense in the real world.
The key here is enabling. Now that we’re past early offerings, the winner’s circle will be filled with those that can make cloud storage accessible to IT. After that, the winners’ circle will slowly be filled with those who make it accessible to IT processes and secured against everything else. And it’s coming, so IT actually can take advantage of what is an astounding concept for storage. If you’re not an HDS customer (and don’t want to be), Cloud Storage Gateways can give you similar functionality, or you can hang out for a little bit. Looking like local storage is what must happen to cloud storage for it to be accessible, so more is on its way.
And contrary to some of the hype out there, most of the big storage vendors have the technology to make an environment like – or even better than – the one Hitachi Data Systems has put together, so have no doubt that they will, since the market seems to be going there. There are some bright people working for these companies, so likely they’re almost there now. You might say that EMC has “been there, done that”, but not in a unified manner, in my estimation.
I for one am looking forward to how this all pans out. Now that cloud storage is doing for customers instead of to them, it should be an interesting ride. Until then, cloud storage vendors take note: Storage is not primarily accessed, data files created or copied, or backups performed through an API.
Related Articles and Blogs:
Published Jul 01, 2010
Version 1.0
No CommentsBe the first to comment