Three ways to add capacity to an existing environment

Sooner or later, you'll have to increase storage capacity. Will you buy more or stretch your existing systems? Find out your options so you'll be prepared when the time comes.

In an age where exponential data growth has become the norm, administrators are all too familiar with their storage estate running out of capacity. After making the umpteenth request for users to remove unwanted data from their home drives, there comes a time when a storage purchase is inevitable. The question then becomes "What's the best way to increase my capacity?" This may differ for some organisations, and will likely depend on where in the lifecycle the existing infrastructure resides.

There are three main ways to add capacity to an existing environment:

  1. Reconfigure the current plant so that more storage is available to hosts
  2. Increase the capacity of an existing array
  3. Purchase a new storage array

The cheapest option in terms of capital outlay is to reconfigure the current plant to create more available capacity. Unfortunately, this can also be the most complex and time consuming of the three options, which equates to the highest risk and may also incur more cost than intended. An increasing trend in recent years has been to implement storage tiering, which attempts to align the resilience and performance of storage with the applications that use it. Mission-critical data will reside on the highest performing and most resilient hardware, while temporary and unimportant data may reside on JBOD or "cabbage" storage. An accepted philosophy is that the value of data usually decreases over time, meaning that storage administrators should be constantly moving aged data from tier 1 storage to lower tiers of storage throughout its lifecycle. This constant movement of data is the Holy Grail of information lifecycle management (ILM), which is very rarely (if ever) achieved on a company-wide basis. What can be done is a periodic check of the storage plant to verify that all hosts on the high-performance disk still require that resilience. If it's no longer required, it can be moved to lower performing disk, which will free tier 1 storage for new applications. Lower-tier storage may need to be purchased to fill this need, but it will be considerably cheaper than tier 1 capacity.

The second option is to increase the capacity of existing arrays. Current arrays allow for further capacity to be added online, so interruptions to service won't be an issue. The most important thing to consider is the current utilisation of the array to be expanded. For modular arrays, the maximum performance tends to be constant no matter how many drives are present. As you add drives and shelves, the unit performance per GB reduces, which may cause an array to become a bottleneck for high capacities. It's therefore imperative to assess the current utilisation of a modular array before you consider adding capacity. If current utilisation is high, expand a different array or purchase a new one. For enterprise arrays, the maximum performance generally increases as you add disk frames, so the current utilisation isn't as much of a consideration.

The final option is to purchase a new storage array. The amount of extra capacity and the tier of storage required will affect this purchasing decision, as will the depreciation of current arrays and any vendor deals that are on offer. While this option should be the easiest to implement, it will probably turn out to be the most complicated in terms of available choices.

Unfortunately, there's no standard route to follow when additional storage is needed, and the decisions will depend on what category of storage is required, what's currently deployed and the utilisation of the current estate. Only when these questions have been answered can an educated decision be made, lowering the risk of it being the wrong one.

About the author: Steve Pinder is a principal consultant at GlassHouse Technologies (UK) Ltd. Steve has more than 11 years experience in backup and storage technologies, and has been involved in many deployments for companies of varying size, with responsibilities ranging throughout the sales and deployment lifecycle. Prior to working for GlassHouse, Pinder was an IT contractor concentrating on backup and network management roles. He has a BSc Hons in Computer and Communication Systems.

Read more on Storage management and strategy