Cloud storage is a broad term that covers a lot of different storage types. Generally it should be reserved for storage that resides outside of an organization’s data center. Typically your data goes to one of the larger public cloud providers or smaller managed service providers. There is nothing magic about the storage designs cloud providers use. Most are hyper-scale storage architectures designed to minimize costs and maximize efficiency. And like a “regular” data center, they use a variety of storage designs ranging from high-performance architectures to capacity architectures.
It is important for organizations to understand the use cases for cloud storage so IT professionals can best select the type that best fits the organization’s needs. Once they understand the types of cloud storage available, comprehending how to best move data to and from cloud storage becomes the next critical component of the cloud toolkit.
Cloud Storage for Hosted Applications
When an application is hosted in the cloud, it needs to store data. That data is typically stored in the same cloud data center as the application. Cloud providers typically offer at least two classes of storage: a performance class for this type of application data and a lower performance, lower cost, high capacity tier for storing archive or backup data.
The challenge is most providers provide only block or object access to both of these classes of storage. Of course, some applications, especially those that were migrated to the cloud instead of being born there, may require file (NFS or SMB) access to cloud storage resources. IT professionals should look for a solution that can provide a file interface to cloud storage, which would enable more applications to easily migrate to the cloud. They should also look for solutions that can automatically move data between the provider’s high performance tier and their capacity tier, to minimize costs.
For some organizations, a major inhibitor to cloud migration is the sensitivity of the data the application creates and stores. But there still may be good reason to move the application to the cloud. Organizations should look for solutions that can cache active data to the cloud from on-prem storage. This technique makes sure only a small amount of data is in the cloud at any given point in time, lowering cloud storage costs and better protecting organizational assets.
Cloud Storage for Archive Data
The provider’s capacity tier is also an ideal location for data centers to migrate older data that is on premises seldom accessed by users. To enable this, the organization needs to look for a solution that can interface between on premises storage and the cloud archive tier. But, it should be more than a simple gateway. It should enable the movement of data automatically and seamlessly between tiers. This solution requires the caching technique described above and a global file system capability that enables transparent data movement.
Private Cloud Storage
Finally, some organizations will never put certain data sets in the cloud, either because of data footprint, data security concerns or both. These organizations can still take advantage of key cloud storage attributes like data reliability and cloud economics. In this use case, organizations can select object storage systems, most of which are scale out designs that leverage commodity hardware. Again the challenge is moving data to this new storage system and again solutions exist that enable this private cloud storage tier to be interfaced with like a normal NFS/SMB share.
There are very valid reasons for organizations to have more than one of these use cases active as part of their overall storage strategy. The challenge is that until recently each of these use cases required a different solution, creating a management nightmare for the IT personnel tasked to administering it. Now though, companies like Avere Systems provide a series of solutions built from a common foundation that enable easy management for each of these use cases.