Archiving data to cloud storage: How to choose the right cloud storage provider

Before archiving data to the cloud, learn the five questions you should ask to ensure that your choice of a public or private cloud storage provider will meet your requirements.

Today's tough economy is forcing many enterprise data storage administrators to re-evaluate their storage processes and identify more cost-effective approaches to data storage. Archiving data to value tiers of storage, such as storage clouds, is now being viewed as a desirable alternative to traditional storage systems. But before storage professionals begin archiving data to the cloud, here are the five questions they need to ask to ensure that their choice of a cloud storage provider will meet their business needs.

1. Are you doing a green field cloud storage implementation or migrating data to the storage cloud?

Most cloud storage offerings (public cloud or private cloud) are based on object-oriented storage, meaning that each file stored to the cloud is first run through a hash algorithm, given a unique identifier and then stored as an object. This is less of an issue when storing data to cloud storage for the first time (green field) because archival data doesn't have to be migrated.

But if you're migrating data from an existing archival storage solution -- such as an EMC Corp. Centera, NetApp Inc. SnapLock or Permabit Technology Corp. Enterprise Archive -- to cloud storage, it isn't just a simple copy-and-paste operation. "There are lots of 'gotchas' when migrating archival data," said Joe Kvidera, founder of Procedo Inc., a data archive migration company. "You can't just read, copy and paste, and expect the applications to understand that the data is now in the storage cloud."

The migration involves moving the object and metadata associated with it, as well as performing application updates. This allows the application to know where to find the data should it need to access it in the future.

To facilitate object-based migrations, products such as the Procedo Archive Migration Manager integrate with a number of archiving applications (CommVault's Simpana and Symantec Corp.'s Enterprise Vault, for example). They can also integrate with content-addressed storage (CAS) appliances so that all metadata on CAS appliances is retained and migrated to the storage cloud and application updates are done automatically.

2. Are your archiving applications compatible with cloud storage APIs?

By default, most archiving applications support standard file system interfaces (CIFS/NFS); many also support APIs for the EMC Centera. However, archiving data to cloud storage requires them to support new APIs like Representational State Transfer (REST) and Simple Object Access Protocol (SOAP).

Of the two APIs, Jon Martin, director of product management at EMC's Cloud Infrastructure Group, recommends that organizations look for storage cloud providers that use REST. "Eighty-five percent of the market is turning to REST while SOAP is fading away," he said.

3. Do you need to store the data in a compliant format?

One of the reasons organizations archive their data to CAS devices is to preserve it in an unalterable format that satisfies internal and external compliance concerns. This is something you won't find in most cloud storage offerings.

Cloud storage offerings were and still are primarily intended to provide low cost, offsite storage options for Web 2.0 providers. But when it comes to storing archival data in a WORM-like format, that feature isn't yet available. However, EMC's Martin notes that write-once functionality is certainly something that can be added to its REST API implementation if there's a demand for it.

4. Can the cloud storage provider dynamically respond to changes in service demands

Getting data into the cloud is only half of the equation. Cloud storage users must also be able to get the data out when it's needed without running up exorbitant access charges.

Organizations that archive financial databases may find cloud storage to be an economical option until quarter or year-end financial records are needed. In this case, the cloud storage provider may have the data stored on storage that's too slow to respond to these new requests. And even if they do respond, the cloud storage user may be hit with higher network charges due to increased network bandwidth associated with these additional demands.

5. Can you move data from a public cloud to a private cloud?

Some cloud storage users are starting to encounter excessive monthly costs associated with storing their archival data with a cloud storage provider. While storing a few hundred GBs may be cost-effective, it may make more sense to move your archival data and cloud storage solution in-house when users start to store tens or hundreds of TBs with a public storage cloud provider

This is feasible if you select a cloud provider that will sell its solution directly to you and provide a mechanism for you to migrate your data. Conversely, you may find that your private cloud storage demands are not as high as you initially anticipated, so you may want the option of moving your data from a private cloud storage provider to a public cloud storage provider.

In either case, look for a cloud storage solution that's available as either a public cloud or a private cloud. This way you aren't stuck with a solution that may someday necessitate a costly migration of your archival data from one cloud storage solution to another.

Dig Deeper on Cloud backup