Cloud: A new battlefield

The cloud storage market is gradually moving forward, and over the past few years the concept of cloud storage has largely remained in the developer domain, using cloud storage as a cheap, on-demand resource for storing data generated in some projects. Now we've seen the cloud begin to be embraced by traditional IT departments as a new way of storage provisioning and usage. By interacting with some vendors and users, we believe that 2011 will be a key year, and that during the year, some midsize IT organizations and businesses will consider putting cloud storage deployments on the agenda while implementing cloud storage.

The shift from research and development to the product sector is a trend in the early prototypes of the cloud storage market, as reflected in the Taneja group's "cloud storage Emerging market forecasts" report. Based on our research findings, the cloud storage market is now basically in the 4 billion-dollar total, and may grow to $10 billion by 2014. The cloud will have a very sharp impact on the next generation of data storage technologies, including where and how to deploy the storage system.

To understand the weather vane of the cloud storage market, we found that the entire cloud storage market would be useful for subsequent analysis if it were divided into two major parts. These two parts are the primary storage technologies that are located behind the cloud, as well as technologies that allow users, systems, and applications to connect to the cloud. The first wave of competition is focused on the latter part, so let's look at the market first.

Cloud Access Technology

In the area of cloud access technology, we have summed up three broad categories of technologies as follows:

• Common Gateway form: As the number of public and private cloud deployments increases, users need faster and more efficient access to data in the cloud. Access to the cloud can be achieved in a variety of forms, including Common gateway forms, which are responsible for connecting users to the corresponding content and primary storage devices. Cirtas, Nasuni, Twinstrata and other manufacturers have been able to provide such equipment. Although such markets now look small, they are growing fast, with growth reaching 100% by 2014.

• cloud-based Backup: The second way to access the cloud is cloud-based backup, which allows users to transfer the data they need to back up through the wire to the cloud storage space. Vendors with cloud backup services such as CommVault, Riverbed Marvell (whose product is Whiterwater) and Symantec have corresponding cloud backup scenarios. This kind of market will grow rapidly, although it is unlikely to be the same as the cloud Access gateway at twice times a year.

• Data mobility and access: As a result of virtualized environments, a wide range of applications and the continued growth of storage systems, as well as the need to address latency issues in data transmission over long distances, focus on product emerged for data mobility and access that enable users to efficiently move large amounts of data and interact with data in the cloud. Cisco, Juniper and riverbed (their steelhead products) are the main players in this segment. Riverbed, who can actually see it as the leader in this niche, has been focusing on access to distributed data, and now they are doing some innovative technology.

Redesign primary storage for the cloud infrastructure

Storage systems at the back of the cloud will undergo a different transformation from traditional storage, including those in the storage world, as well as the need to adapt their technologies to the cloud to provide a new storage model. We divide the primary storage system in the cloud into two main parts: content and I/O.

Content needs to be stored, accessed, and distributed, and these policies are different in the cloud from traditional primary storage. File system and file access related technologies that meet the requirements of content access in the traditional infrastructure are in the cloud infrastructure and are not able to meet service requirements in terms of scalability and accessibility. A significant portion of the content data in the cloud is given to object technology to support, driving content and archive storage toward highly scalable, multi-tenant, web-based access. In the short term, the market will be driven primarily by service providers, but ultimately in the private cloud within the enterprise. We expect small vendors like datadirect receptacle (whose products are web Object scaler,wos), Nirvanix (Hnode), and Mezeo (Cloud Storage Platform) Also join mainstream vendors such as EMC, HP, and NetApp to serve as a cloud-based content storage platform provider. This growth will be swift and steady, but not as strong as the cloud linking technology markets.

We were led to the biggest cloud storage opportunity, the main storage market behind the cloud, which is now over $2 billion trillion. The market is now being hosted by a series of next-Generation FC technologies, while the unified storage product also occupies a certain share. We believe that the primary storage system in the cloud will undergo a variety of processes, a significant part of which is a wide variety of intelligent block-level storage and access, processing technology. In a private cloud or public cloud, intelligent block-level technology will replace traditional block storage, and these technologies will directly affect the storage vendor's winning and losing relationships. We believe that Dell (EqualLogic), Hp/3par, and NetApp will be successful in the cloud back-end primary storage system. For HP's 3PAR storage System, it's a good insight. 3PAR has been targeting the market for a long time, seeing service providers and business markets as a big contributor to tool storage, and intends to launch a revolution in this area.

Cloud: A new battlefield

Although it is not yet certain to win or lose, we believe that the rapidly growing cloud market will shuffle the list of competitors for the storage vendors. Only vendors that execute the right business model on Scale-out extensions and application storage platforms can be the winners of this war.

(Responsible editor: Lu Guang)