This week I wanted to pick up the discussion that I started a couple of weeks ago on the major trends of 2009. In that blog, I examined why deduplication was THE success story of 2009. But an equally compelling trend in 2009 was the emergence and growing acceptance of the use of cloud storage as a storage target among organizations of all sizes.
Cloud storage (public or private) is resonating with organizations for three very simple reasons:
First, it is cheap. Just a little over a month ago, Google cut its price for the cost of hosted storage for its Picasa and Gmail applications to $5 per year 20 GBs of storage. Now granted, this is only for the storage of data in Google applications but let’s put that in perspective.
Just a few years ago I started regularly archiving emails in my outlook.pst and at that time the size of the file had only grown to a little over 1 GB. I store that file on an external 120 GB hard disk drive (HDD) (I know that’s dinky by today’s standards but it is not even one quarter full) and that HDD cost me about $120. Now assuming that an external HDD lasts 5 years, it costs me on average $24/year for that HDD not counting power.
Now assume I start storing that same PST file along with every other file on my external HDD with Google’s storage cloud. Since my total used capacity is less than 38 GB, I would only need to spend $10 a year or $50 over the same 5 year time period since I would only buy as much storage capacity as I needed and then I would only buy it when I needed it which would amount to a savings of $70 over that 5 year period. Now granted, that may not sound like much but when consumers can start to justify the use of cloud storage, small, midsize and even enterprise organizations are sure not to be far behind.
Second, data is better protected than when it is stored locally. By better protected, I do not mean that Google or any other storage cloud provider necessarily backs up of all of your data nor do they give you unlimited recoverability options. However because of how they stripe data across multiple HDDs or multiple servers, in the event an HDD or server fails it is unlikely you will lose data. Conversely, if my externally HDD fails, I am forced to do a complete recovery which will minimally take me hours to complete.
Third, network throughput is increasing. I suppose I could also say network connections are getting cheaper but I am not sure that is universally true other than to say organizations can get more network bandwidth for the same price than before. Using this additional bandwidth, organizations can realistically look to start storing archived data, backup data and infrequently accessed data with cloud storage providers and retrieve it in time frames that meet their application or user requirements.
I realize this is a bit of simplistic way to look at why cloud storage is gaining momentum. However many IT managers and administrators that I speak with still lament that the individuals to whom they report, approve the budgets and write the checks still get a glazed look in their eyes when they try to explain to them what “virtualization” means or how cloud storage works. But their bosses do understand the benefits of lowering costs, the ramifications of losing data and getting more for the same price or less. So when the benefits of cloud storage are explained in that context, their interest level rises significantly.
So why wasn’t cloud storage a bigger winner in 2009? There are a couple of logical explanations for this as well.
The first reason is information security. In a conversation I had with Symantec’s vice president of engineering, Ashvin Kamaraju a little over a month ago, he says, “One of the things that Symantec sees as precluding people from moving into cloud environments is, ‘How do I ensure the security of the data?’ Customer A should only have access to Customer A’s data and no one else’s data and the same holds true for Customer B.”
Right now customers who are storing their data in storage clouds are for the most part storing non-sensitive data in formats that is ideally encrypted. While I am sure there are exceptions to this rule, it only makes sense that organizations would first store data with a cloud storage provider that poses the least amount of exposure to the organization. In the event this data is compromised or even lost, their risk is minimal.
The other reason is a lack of standards in the cloud storage space. While this is less of an issue among private storage providers who primarily rely upon CIFS and NFS interfaces for data acces, public storage cloud providers are each developing proprietary forms of the REST API for their cloud storage offerings. While each provider and those application providers that write to these different REST API interfaces tell me that the difference from one API to the other is not significant, I think most customers will feel a bit more comfortable when there is a defined and commonly agreed upon standard for the REST API.
In that vein, work has already commenced in the area of establishing standards for a public storage cloud. In October 2009, SNIA announced its cloud storage initiative and published its first public draft of a new cloud storage management standard, the Cloud Data Management Interface (CDMI). Then while doing some research on the SNIA website last night, I discovered that this morning (Dec 18) at 9 am PST, SNIA is hosting a webcast on its cloud programs for 2010 which provides further insight that the impetus behind cloud storage is alive and well.
Cloud storage was definitely not the “BIG” winner in 2009 but it is now front and center on the radar screen of every large organization while more small and midsize organizations are getting the message loud and clear. It is also my opinion that the trend toward cloud storage will accelerate in 2010 and it continues to be my belief that by 2011 and 2012 cloud storage will be the BIG winner as it matures and its adoption accelerates.
Next week I am going to take a look at why solid state drives (SSDs) are following hot on the heels of cloud storage and why I see these two technologies potentially becoming joined at the hip and becoming such a disruptive force in the storage industry.
Have a good weekend and for those of you who are taking the next two weeks off, have a Merry Christmas and Happy New Year!