10 Best Practices for Implementing Cloud Storage

by Chris Preimesberger

Know your options: Private versus public clouds—which strategy is right for your organization? The simplest way to describe these two options is: Public cloud means storing your information and running your application on the Internet; private cloud means having this capability in-house, which means much greater flexibility, control and security.

Have a comprehensive virtualization strategy in place as the foundation of the data center. Virtualization of computing power and storage is the key to making cloud computing work. The first thing IT people think of is virtualizing servers. But IT data centers need to virtualize storage as well because if it isn’t, it slows down what they can do on the application side.

Know when to use a public cloud versus a private cloud to deliver the right level of service to the end user in the most cost-effective way. Testing and R&D may be better served from a private cloud due to speed of deployment and security. Less critical business applications could be hosted on a public service.

Ensure the storage foundation for the private cloud is virtualization-based and interoperable with multiple server and desktop virtualization platforms.

Select a cloud storage environment that offers tools such as automated storage tiering and thin provisioning that allow for the automation of common, complex tasks. This all starts with virtualized storage; you can’t do advanced features—like thin provisioning and storage tiering—without it. You need to be able to automate the movement of the workload to the right storage resource at the right time.

Be sure the company’s data storage solution will scale on demand to support rapid, unplanned data growth. Choose a storage architecture that allows nondisruptive scaling in any direction. Ask yourself: What does it take to add capacity to a SAN? If it takes downtime, a lot of staff attention and maintenance, then maybe you’re not getting the best value from the system.

Determine if there are hardware and/or software upgrade costs associated with a system’s “limitless” scalability. Be careful: Some architectures require significant relicensing and hidden costs when a system’s capacity gets to a limit. Example: Will you take a big hit in software licensing when you have to add that 101st drive or storage array?

Confirm that the storage environment will remain online and data will be available during system maintenance or hardware reconfigurations. All storage systems need updates occasionally, including firmware. When that needs to be updated, be sure you know what the impact will be on the system.

Choose data and disaster recovery products that allow snapshots and replication to be automated. But watch out for hidden costs such as bandwidth requirements, limits on the number of snapshots that can be taken, or how they are replicated (i.e., just changes or entire volumes replicated each time).

Set up management tools and reporting capabilities. Don’t overlook SRM tools, such as storage chargeback to analyze use and profitability accurately. Companies such as Aptare, CA, BMC, Symantec and others have respected products in this area.