Channel uniquely placed to make outages history
Load Dynamix's Gavin Tweedie on how the channel can ensure outages like the one that brought down Tinder last year never happen again
In the channel, you know that data storage systems are your customers' lifeline.
You don't need me to tell you that if your customer can't access its data its whole business can grind to a halt. Take a minute to imagine what would happen if they lost all their billing data for example. Or if they couldn't process sales orders - you would not want to be at the receiving end of *that* phone call. It's not just frustrating: it can spell disaster.
This is where the channel comes into its own. As a reseller, you're the go-to person for both the customer and the vendor. You're there right at the beginning, when a customer works out what it needs from its storage, and you're there throughout the lifetime of that storage kit. You're crucial in making sure everything runs smoothly - or, if the worst does happen, that it's sorted out quickly without some of the huge outages that have been in public eye recently.
I'm thinking of the AWS failure that brought down Netflix for example. Upsetting for users who couldn't download the next episode of Orange is the New Black - and for Tinder users who had to hold off swiping for their next hook-ups - but more than that, potentially hugely damaging for those brands.
In fact, an outage can impact both a company's reputation and revenue. Take Facebook: in 2014 its systems went down for no more than 30 minutes. Yet the cost to the social network was estimated to be as much as $600,000. More recently, when Apple's App Store went down for almost 12 hours (yes - taking out everything from iTunes to the Mac App Store), the outage was valued to have cost the company £17 million in lost revenue. The numbers are high and the staggering costs are due to multiple applications often depending on the same IT infrastructure, which can become a single point of failure.
But let's return to the channel- it might seem like, as a vendor, I'm passing the buck, telling you that you have an important role to play and then washing my hands of it all. Of course the vendor has a huge responsibility to make sure there aren't enormous, or even just highly aggravating, slowdowns. But your position is unique. You speak to the customer before vendors do. You have the opportunity to find out what they really need, rather than what they like the look of.
And we've all been there - just about everyone's bought something because it's all the rage, rather than because it does what we need it to. But when your customer's investing hundreds of thousands, even millions, of pounds in a storage system, it needs to work. And it needs to work whatever's thrown at it: whether it's millions of people logging onto a site as happened with the NHS Choices website not long ago, or a glitch in the system that hasn't been picked up in time.
I'm sure you have seen examples when vendors have proposed products and configurations to customers which fail to meet business performance goals and expectations. In part this is due to a lack of understanding of the workloads and their I/O profiles. If the vendors don't have relevant workload data to work with, how can they be expected to provide solutions and configurations which are fit for purpose?
Properly sizing the storage infrastructure configuration has one of the greatest impacts on the cost and TCO of the solution. Building more precision into the workload analysis and modelling exercise ensures an optimised configuration leading to a better return on investment with fewer nasty surprises.
It starts with a good understanding of your customers' application workloads and the capacity, performance and availability needed to keep everything running smoothly. Without that, there's no way you can even begin to design a fit-for-purpose storage infrastructure.
The storage infrastructure needs to get smarter: we're seeing unstoppable levels of data growth (did you know that 90 per cent of the world's data has been created in the last two years alone?) and generally the number of users continues to increase. These users are impatient and they're fickle: if it takes too long to access information or if a website doesn't load quickly, they will go somewhere else. And your customer will lose revenue. And you might lose the customer.
So in this new demanding age of storage, you need to help your customer look beyond capacity planning and forecasting. There's more to an infrastructure than planning for redundant storage.
You'll have to dig deeper into your customers' application workload requirements, probe into their performance needs and get a real understanding of how these are likely to change over the lifetime of the kit you recommend.
You won't be relying on guesswork then as you'll know the right vendor and propose the right kit. You'll know how different storage systems measure up to your customers' unique workload needs. You won't need to overprovision, because when it comes to performance sizing you'll know how much is enough.
By doing that you'll be offering your customer a streamlined, faster system tailored to its unique workloads. And cutting down on overprovisioning means the system will be much more cost effective for your customers.
The big AWS and Facebook outages made the news. But think about how many don't. Getting the right kit in the first place could make sure these outages don't happen to your customers. You're in the perfect position to do something about it.
Gavin Tweedie is director of EMEA operations at Load DynamiX