Are you considering cloud storage for your business? There are many reasons you should. Using innovative cloud technology, IT is solving data storage problems in new ways. Whether it’s for off-site data protection, disaster recovery or just storage capacity expansion, the pay-as-you go model pioneered by a number of cloud storage providers can be very compelling.
Rather than use cloud storage directly by writing to custom APIs, building your own security policies and architecting a performance framework to meet application needs, you may find that on-premise cloud storage software or hardware (i.e. gateways, on-ramps, enablers) make integration simpler. Purchasing a product that handles security, performance, data reduction and plug-and-play integration can significantly accelerate and simplify deployment.
With a handful of gateway products already on the market that can connect your on-premise environment to cloud storage, a natural question may (or should) be “what is the difference between these products?” above and beyond the aforementioned functionality.
To answer this, we’ve put together a list of 6 differentiating features you should consider when choosing a cloud storage gateway:
1) Dynamic caching policies to meet application needs: A monolithic cloud storage cache may not be able to handle the performance needs for all applications. A backup application may benefit from a cache consisting of low-cost storage optimized for large sequential access, while an NTFS file system may benefit more from an SSD-based cache, optimized for smaller, more randomized access. Each application may require more or less cache over time. Having application-specific caching policies that are dynamic means you can meet needs of different applications using a single solution.
2) Option to replicate a local copy to the cloud: Some vendors argue that having a full local copy defeats the purpose of cloud storage – not at all true! Imagine replacing a real-time replicated secondary site requiring hardware, infrastructure and maintenance costs with a pay-as-you-go cloud! Or imagine not having a secondary site to begin with and now finding a 2-site replication solution within easy reach. This is a very compelling business proposition, particularly for transactional applications that require a full local copy for latency reasons.
3) In-cloud snapshots: Snapshots are rapidly becoming a key part of modernized backup and, when using the cloud, it is important to find out whether a gateway solution offers snapshots. If yes, are the snapshots copy-on-write and on-premise, meaning potential bandwidth thrashing between the local site and the cloud? Or are the snapshots in-cloud, redirect-on write, meaning no bandwidth penalty or performance penalty and readily available in case of disaster? If you have the option of the latter, you may have gathered that it is far superior.
4) Block and file-level access: It’s amazing to hear arguments from vendors trying to convince users that file access is better than block access for cloud gateways. The reality is that there are advantages to file access and advantages to block access. Supporting both means supporting that widest variety of operating systems, file systems and applications; and there is no longer any argument. Hint: having native block access (like iSCSI) means you can support both.
5) “Zero-friction” entry point to cloud storage: Deploying cloud storage should not mean continuing to spend additional CapEx/OpEx associated with traditional storage infrastructure and incurring the same 3-yr upgrade cycles. Sure, there are advantages to optimized hardware appliances for accessing cloud storage, but only when needs and budget dictate. A choice of software, hardware and subscription models with upgrade paths between each are the ideal way to start using cloud storage with minimal risk/cost and the ability to grow.
6) In-cloud disaster recovery and Compute-Anywhere capability: Once your data is in the cloud, you can access it anywhere, but how about in the cloud? Why not be able to leverage unlimited pay-as-you-go cloud compute cycles for disaster recovery or test. Beyond disaster recovery, your data or snapshots of data can and should “work” for you in the cloud. You can even leverage Big Data without dedicated processing resources by using cloud compute. Think about a vision of a hybrid data center and how this capability can enhance IT.
In summary, all cloud gateways, on-ramps, or enablers are not equal and it takes looking beyond the similarities in features to understand whether they will meet the needs specific to your environment and grow to meet your future needs. It pays to look under the covers before purchasing…
Perhaps you have found a cloud storage solution that has all of these features. If you haven’t, we suggest you consider a cloud storage solution that does….
Del.icio.us
digg
Twitter
MySpace
FaceBook
reddit
Stumble Upon
