By Yuval Shavit, Features Writer
It's no secret that companies create and consume more information than ever, but chances are good that most of your customers don't have a well-structured plan for accommodating their data growth. Planning how much storage capacity your client needs is an important balancing act: not enough and applications can crash, but too much and you're wasting your client's money. In this article, we'll look at the benefits of storage capacity planning and your role in the process.
The goal of storage capacity planning is to predict how much storage your client will need in a given timeframe so that you can buy just enough disk space for it. Although it's important to keep a buffer of spare storage in case an application spikes in terms of its data usage, buying too much is not economical. Overbuying mean a larger initial cost, which ties up resources that could otherwise go to other projects. Also, since prices on storage devices are constantly falling, buying 100 GB today will cost more than buying 50 GB today and another 50 GB in six months.
For consultants or resellers, one of the benefits of storage capacity planning is that customers better understand how much storage they need and are therefore more willing to pay for it. If a customer's storage system isn't scalable -- for instance, if their servers use direct-attached storage (DAS) -- making them aware of storage capacity planning may be a good way to sell them on a higher-end system when it comes time to replace old hardware.
Although plenty of vendors have storage resource management (SRM) tools that automatically monitor storage, most companies treat storage capacity as an afterthought, said Kurtis Lindemann, director of technical services at RoundTower Technologies, a storage consultancy in Cincinnati. Rather than putting together a plan to centrally monitor and plan for storage growth, companies sometimes try to save money by giving managers monetary incentives to use up as much of their storage as possible, he said. One of the benefits of storage capacity planning is that it helps managers prove how much storage they need, making it easier for them to get funding.
Running too lean on storage can bring about performance problems, and applications that run out of disk space usually crash, often causing data to be lost forever, said Henry Baltazar, storage analyst at The 451 Group in San Francisco. As a disk fills up, it becomes harder for operating systems and applications to move data around within it, and data fragmentation becomes more of a problem. Applications are likely to crash after disk utilization gets past about 90%, Baltazar said, especially if an unexpected spike in data eats up the remaining space. Even if applications don't crash, your client will likely have to waste time constantly monitoring and adding to the available storage.
On the other hand, buying too much disk space is wasteful, and customers may not be aware of how much they need to buy. Whether you use SRM tools specifically designed to track storage capacity or more manual methods, it's important to take your client's growth rate, and not just current utilization, into consideration, Baltazar said. Large companies that grow their data relatively slowly and steadily may be able to keep disks as much as 80% to 90% full, while smaller companies or companies that are constantly creating data may want to be closer to 20% to 30%.
You should also find out how often the company wants to buy extra storage capacity; a company that wants relatively few upgrades needs to buy more storage up front. This is an important factor, especially if your client is using DAS or a low-end storage area network (SAN) that can't be easily expanded. You can take advantage of the benefits of storage capacity planning to help your client more accurately predict how much storage it will need over that longer period of time.
In the next installment of our Hot Spot Tutorial on storage capacity planning, we'll look at some of the specific tools and methods you can use to help your client plan its next storage upgrade.
This was first published in April 2008