Cloud storage has moved out of the experimental mode and into some form of production for many organizations. To date most of the use cases are either to backup data to the cloud or to archive data to the cloud. Now though the move is on to provide leverage the cloud for primary data storage. If successful it could change the way many businesses buy storage.First let me say that leveraging the cloud for primary storage works. There are at least seven shipping products. We have several tests of the various platforms underway with more to come in our labs. We have found the platforms for the most part to perform very well and to be very stable. You should have confidence installing these systems and beginning testing their use.

As we discussed in our article "Getting Your Arms Around The Cloud" primary storage solutions that leverage the cloud typically have some form of hybrid or local storage cache to hold the more immediate information. This cache can be in varying sizes depending on what your active data set looks like. Then changes to the data set are replicated in the background to cloud storage. The advantage to this approach is that you get immediate access to your most active storage so user and/or application performance stays the same. In fact you could even use solid state storage for the local cache which would provide better performance than what was there before while still curtailing cost since older data is moved off of the cache as it.

Where primary cloud storage solutions differ is in how the service is delivered and what type of service is provided. There tends to be two types of approaches to delivering the solutions, either as a virtual appliance or as a physical gateway. The virtual appliance has the advantages of being software only and more portable across the environment while just as easy to maintain. The physical appliance has the advantage of being more consistent in its storage performance.

Both of the traditional services (file sharing and block storage) are available in primary cloud storage. The file server use case is ideal for user home directory type of data where modest performance demands are required. The block case is typically offered via an iSCSI connection and tends to focus on specific application data sets that are easily segment-able. Meaning they have an archive or other type of repository that tends to age and is not frequently accessed. The key in either case is to make sure that the local cache is large enough to hold all the data that needs to be accessed on a regular basis.

There is another option; cache-less cloud storage where your access to primary cloud storage is direct, no cache evolved. This will likely require local storage pods be set up by the cloud provider. While this is more similar to the late 90's effort of managed storage services, it differs in that how the systems are architecture as well as the cost and availability of bandwidth. Now these services can be offered at a more attractive price point and better potential for profitability from the provider.

The point is that primary cloud storage is a reality, and it is not limited to customers in the small business market. These vendors are focused on all sections of the market including the high end enterprise. While each level of data center will use the technology differently, there is value that every data center can gain by using primary cloud storage.

In an upcoming entry we will look at some of the early use cases for primary cloud storage.

Track us on Twitter: http://twitter.com/storageswiss

Subscribe to our RSS feed.

George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Find Storage Switzerland's disclosure statement here.