“a Peek Thru the Clouds” – Cloud Storage in 2009

2008 is the year that “cloud” captured the imagination of the technology early-adopters. And the interest is not just in Silicon Valley or the U.S.  ParaScale, for instance, has had hundreds of software download requests over the past few months from around the globe. But vendors should be under no illusion at this point that mainstream IT shops understand all the hype.

However, “cloud-ware” is more than the next shiny thing on the IT roadmap. Cloud applications/compute/storage is driving the next generational shift in the evolution of IT.  Over the next decade, IT deployments are going to include a blend of mainframe, client-server, and 3-tier-apps with cloud configurations becoming an increasing and very significant part of the mix. Even with the bottom falling out of the market, cloud technologies have continued to generate interest as it is seen as a way to optimize IT costs.

In 2009, users will be presented with a lot of choice, especially with cloud computing and cloud storage. When users thought of cloud-ware in 2008 they were thinking about cloud applications like SalesForce and Google Apps, cloud computing services like Amazon EC2, or cloud storage like Amazon S3. These are all public cloud services in that the offering is delivered across a public network. The private cloud option, where infrastructure services (i.e. IaaS – Infrastructure as a Service) like cloud computing or cloud storage are made available as a service over a private network, became strongly established in the vernacular in the latter half of 2008. But it will be 2009 before customers have enough vendor choice and adoption starts taking root.



In the cloud storage space, 2008 saw announcements from vendors like EMC (ATMOS hardware + software offering) and ParaScale (software on commodity hardware offering) about packaged cloud storage technologies. With the excitement around cloud storage, there was also activity from vendors making forays from adjacent storage segments, as they announced “cloud-like” storage. Overall, these various packaged options are going to enable both businesses and service providers with their deployments, as they don’t have to undertake custom engineering projects as was the case in 2008.

In 2009 we will see two driving trends:
1. The arrival of private cloud storage options with increasing choices for businesses looking to deploy their own storage clouds inside the firewall
2. A rapid proliferation of public cloud storage service options, with many service providers coming to market with varied services, competing with a market-leading Amazon S3.

Let‘s peek through these clouds.

Many early-adopter enterprises, especially in the U.S. and Europe, will start trialing private storage clouds in 2009. They are going to discover that cloud storage is a very effective way to constrain burgeoning tier 2 storage needs especially in archival applications. By starting small, these IT shops will be able to take measured steps into the cloud. Enterprises will also discover that private clouds will allow data-access via familiar enterprise protocols like NFS, FTP, CIFS allowing drop-in to existing enterprise IT, versus public clouds that typically work with newer web-friendly protocols like WebDAV and REST. On a different note, many rich content businesses (delivering applications like image hosting, video streaming), where storage is a very material cost of doing business, will also be looking to see if cloud storage can make the underlying economics of their businesses more robust. These kinds of businesses typically look to monetize eyeballs (in a weak advertising market), or are trying to get to affordable subscription models, and anything that contains their cost-to-serve in an uncertain economy can only be good.

Looking at the second key trend, 2009 will be the year that public cloud services really breakout. Today the storage service buyer is faced with a few limited choices (Amazon S3, Rackspace, etc). Amazon’s S3 is a very credible solution, but it is a 1-size-fits-all-cloud. With archiving application and streaming application, you have one cloud. Many customers like what S3 offers, but it is a case of “you can have any color so long as it is black”.

In 2009, the many service providers of all sizes, observing the success of Amazon S3 and looking to compete, will be able to leverage packaged cloud storage platforms to enter the market. They can avoid undertaking an open-ended engineering project with unknown scalability challenges, as was the situation in 2008. Instead service providers and hosting companies will be able to focus on that which they do best - service creation and service delivery. Expect to see the following dynamics in the service provider-driven public cloud storage space:

* Real customer choice! Providers creating different kinds of storage clouds by selecting different kinds of underlying commodity hardware. Deployments of archival clouds (cheap CPU, less memory, many SATA disks) versus. streaming media clouds (powerful multi-core CPU, more RAM, SAS drives).
* Smaller providers getting into the cloud storage service business, with focused CAPEX outlays, and satisfying the needs of local SMBs and regional businesses. Customers could even drive down to the local hosting company and get TBs for a fast emergency recovery.
* Service providers building many small and mid-sized (50-200 node) clouds, vs one uber-cloud, thus ensuring better stability, more isolation, and flexible DR-options across multiple locations.
* Hosting providers getting into the business by offering managed cloud storage services for enterprises by bringing up mini-clouds for enterprises in their own racks. Now hosting providers have a way to participate in the wave of interest in private cloud storage.

Recessions in many cases assist in the adoption of cost-saving technologies. This seems to be the case with the various cloud technologies. The brakes are being applied for non-urgent infrastructure projects, yielding more time for IT staff to focus on ways to be frugal in must-spend areas like storage. On current trajectory, 2009 is definitely going to be the break-out year for private and public cloud storage in the early-adopter segment, with 2010 starting the penetration into the mainstream market.