RSS
Email
Print
ShareCloud Storage For IT
Posted by George Crump on October 6, 2009
Cloud storage is beginning to settle down and move past the hype wave and early adopter stage and into a real market. We are seeing the adoption of cloud storage architectures begin to move beyond just something that an external provider will use and into something that IT organizations want to take advantage of.
It is true that the term 'cloud storage' gets overused and that it seems every new release from a storage company is their "cloud release". Just because you call it cloud storage doesn't mean it is.
Cloud storage is a broad term that typically defines a storage system as highly scalable, can be deployed for internal or external use and typically uses some form of clustered or grid based storage. It may or may not have a host of other features like geographic dispersion, retention capabilities and hardware independence. The importance of these features will vary between users but are not required to be a cloud storage system. The truth is there likely will never be one architecture that we can reference as cloud storage, no more than there is one architecture that can be referenced as the best way to implement a SAN or NAS.
When used internally cloud storage architectures matter, especially when you have specific problems to solve. When IT looks to cloud storage they are looking to solve the problems of meeting performance demands, dealing with massive data growth, overworked IT staffs and escalating storage costs. Each of these challenges will have a different weighting within the organization and that organization is going to select an architecture that best addresses them. The weighting of the importance of the above challenges will vary between organizations. If it is secondary data, then cost and scalability may be a bigger issue than performance. If it is primary data then performance may be the top priority.
IT organizations are starting to see internal or private cloud storage as an ideal target for unstructured or file based storage. Initially this may be as a secondary storage platform using software from ParaScale, Bycast and others but with yesterday's FileStore announcement from Symantec, internal cloud storage may be suitable for performance demanding primary file based storage as well.
IT organizations that choose to use public or external cloud storage of course will be focused on cost reduction. Can the provider store the data cheaper and more reliably than they can internally? Cost is just one aspect of it however. Companies that for example are storing email archive data or medical images in the cloud are going to be very concerned about data authenticity, rapid delivery and data protection. They may also want to know if the provider has an API set that they can integrate their application into like those offered by Amazon, Nirvanix and Iron Mountain.
As cloud storage intrudes its way into the data center, cloud storage infrastructure providers had better be able to communicate what problem their solution solves. IT is less concerned about how "cool" the solution is but is more concerned about which storage challenges it solves.
Add Your Comment:
Sponsored Links
Comment by Joske on October 8, 2009 5:00 AM
Are you aware of the Cloud Storage TWG that has been set up? If not you should look more in depth at what SNIA is doing around Cloud Storage. In any case, nice article!
Reply to this comment
Comment by CloudedThinker on October 9, 2009 11:41 AM
I'm getting a little tired of the "cloud storage" moniker simply because it remains, well, cloudy. That is, it is still not well defined. Is it a wholly new storage technology? As an old timer in the networking arena the cloud was simply an element on my network maps depicting the Internet. It was a cloud because I didn't care what the infrastructure was, switches, routers, servers, OC192, etc., I only cared that I could easily connect and use it. That's not the case when I bring a technology in house. At that point I do care because I have to implement, manage and support it. For my end user community, they would perceive that our corporate network is itself a cloud. They don't care about the underlying infrastructure, just that they can connect and use it. Access to the Internet "cloud" is just simply an extension of the network.
So, as I think about this thing called cloud storage, my interest is in the technology that will deliver to me the benefits cloud storage service providers are offering: low-cost, ease-of-use, scalability, ubiquitous access, reliability, and simplified operations (admin). I want an infrastructure similar to what those providers have and I want to own it because my data assets are not ready to go to a 3rd party and very well may never be. Thus far my research tells me that object based storage in a cluster of standard server hardware is the optimal technology. I'm looking at all the new comers like the ones you mention above, Atmos the new DDN thing and Caringo as well. If anyone else is investigating this area you may want to add these to your list.
Reply to this comment