Is bigger better?
NME asks the region’s top storage vendors if there is really a justification for continuing to add storage in a world of tightening belts.
More pics ›
For many, many years, the way enterprises approached buying storage wasn’t too far adrift from the way you might approach buying toothpaste.
At second glance, that’s not such an asinine comparison. After all, everyone’s got a full tube of toothpaste in their bathroom cabinet. Or rather, they think they do. Most days, the average person is far more concerned about quickly scrubbing his pearly whites and getting to work on time rather than checking on the health of their toothpaste supply. And one day, you’re caught by surprise when you hurriedly squeeze it and nothing escapes the tip. A trip to the shops later, you’re back in business.
And that, in a nutshell, is how companies have been buying storage for years. The mechanics of managing a vast storage array aside, the actual purchase decision has largely been dictated by how much of it is still available. This has led to a certain amount of neglect from IT managers, a negative zone which internal users are all too happy to exploit. Anthony Harrison, senior principal solutions specialist for storage and server management at Symantec explains how the loophole comes to exist.
“The problem is that it’s not the end-users that are actually buying the storage. There’s this big disconnect between IT and their users. You’ve got to have policies in place saying what they are actually allowed to store there. We found that when we did storage assessment exercises, a lot of people [in the organisation are amazed at how much MP3 and AVI files are stored on their expensive Tier 1 storage. If you don’t have any policies in place, people will do it – it’s human nature. No one’s told me I can’t, so I will,” he says.
“Our perspective is making people more accountable for what they actually use. When you look at comparing the amount of raw storage capacity that you have versus what is actually used for application data, generally we see around 30%-35% utilisation. In some cases, we’ve seen it as low as 12%,” he cautions.
EMC’s Walid Yehia, technology solutions manager for Turkey, Egypt, and Libya (TEL and Turkey, emerging Africa and the Middle East suggests that incorrect reporting and overcautious administrators may also have a role to play in these low utilisation rates.
“Users do allocate their storage capacity to the enterprise applications. Therefore if you issue a utilisation report at the storage level, you will find it high. However, when you further scrutinise how these applications are using the underlying storage capacity, you will find that most of the applications have allocated capacity more than the amoun that they actually host. This is due to the safe approach that database administrator take when requesting storage capacity from the infrastructure team,” he adds.
This wasteful attitude to internal storage clashes with reports from the market that demand continues to climb and climb. While there may be significant data capacity going spare, it’s also an empirical fact that the kinds of data being generated today simply necessitate vast amounts of storage.
“The sorts of numbers that we’re seeing at the moment are from people like IDC that are talking about 400% growth over the next four years. Obviously there’s some peaks and troughs within there but the underlying element is that the majority of that growth is in unstructured data, they’re quoting like 62%,” cites Harrison.
“We’re seeing particular requirements for that in areas like CCTV where you have hundreds of cameras recording lots of video. As a filetype, that’s not very compressible, you can’t do deduplication on it so naturally it consumes a lot of storage. In oil and gas, when you’re doing any kind of exploration, that generates huge amounts of data that needs to be stored somewhere,” he adds.
In this kind of emerging scenario, CIOs are caught in a classic catch-22. From one side, the business is demanding increasing storage, some of which is potentially being wasted. On the other, the CFO and the boardroom demand drastic cuts to keep the business lean and viable. The answer, says Harrison, is to take the long view to storage purchases.
“The requirements have tightened up in terms of getting the justification for many projects so people have to justify their investments far more. That underlying demand is still strong. You’ve got to have a three-year ROI model, total cost of ownership – not just for the hardware purchases but what you need for maintenance as well. Do you need one person or six to run something? That’s a big factor in overall TCO,” he confirms.
Steve Bailey from data management specialists CommVault, suggests that companies are rapidly rethinking their purchase and storage models and thinking about new alternatives – which in turn, puts pressure on vendors to conceive and execute models to their customer’s satisfaction.
“Worldwide, a lot of the change has been focused on pay-as-you-go models or utility-based models. The switch is from capital expenditure to operational expenditure. These utility models could be off-site or on-site, it doesn’t really matter. This has been a talking point with thin provisioning or remote storage tools but it’s now been met with a commercial model to suit this sort of trend,” he says.
“Traditional tiered storage models are still a large part of storage vendor strategy offerings but it’s important to remember that having different storage models on the floor doesn’t guarantee a true tiered storage model. It requires control, process and automation and continual analysis. There is definitely a transition and it’s really about giving the users to ability to pay as they consume, rather than paying for it all up front,” continues Bailey.