Finding the optimal amount of resources needed to get the job done, and obtaining them at the most efficient cost —a tradeoff between SLA, availability and cost? These eternal concerns of capacity planning experts are more relevant than ever, despite the dematerialization of many of a company's “resources” as computing power migrates to the cloud. The main attributes of the cloud environment: flexibility and scalability, rhyme with unpredictability. Capacity planning is confronted with a range of new challenges as the scope shifts from hard iron to IaaS.
Here, we look at five ways in which capacity planning is evolving within a IaaS model.
1) Mental shift: stable to dynamic
In the past companies were concerned about having just the right amount of hardware and network to deal with projected demand for the next few years down the line. The job consisted in investing the right amount of money into infrastructure that requires maintenance and rapidly depreciates in value. Today’s IaaS cloud infrastructure environment means we are in a situation where supply is basically endless, expensive, exponentially complex and non-conform to any traditional procurement process. The capacity planning mindset has gone from stable: negotiating necessary resources for a set timeframe and budget, to dynamic: a “pay-as-you-go model where consumption is in real-time and with a constantly evolving (and quite frankly often unclear or at least impossible to easily forecast) price configuration.
2) Budget blues
The days when capacity planning was centered around requesting quotes from several hardware providers and datacenter hosts and then negotiating discounts seem so quaint, compared with today’s mission of trying to keep tabs on, let alone control IaaS procurement. While cloud infrastructure boasts flexibility and scalability, the flip side is it’s been near impossible to stick to a budget. Cloud infrastructure lets you adjust IT resources to meet a constantly shifting baseline. In turn, the corresponding spend is constantly shifting. Although this budget-blocking platform resolves cloud overspend.
3) (Im)possible cost comparison
Nowadays, nearly all cloud-provider pricing is public, but the complexity of the servers and services going up and down, scaling and being served from several locations worldwide make it extremely difficult to make a comparative quote. In addition, network exchange costs—the most difficult to evaluate—which were fixed price in the past, are now calculated on the cross geographical locations of servers and storage. It’s complex for the IT team to understand what they’re buying and near-impossible for c-levels who aren’t technical experts (but are all too aware of budget constraints) to judge which provider best suits their needs. To top this off, there is no longer a friendly salesman ready to find you the best deal and throw a cooling fan into the deal. It’s a sellers market, so no-one’s investing in sales teams or promotions and there certainly aren’t a lot of tailored offers out there. Cloud providers reps have no or little freedom for cutting a good deal, and when they do it, it's almost impossible to compare the new offers. We’ve decided to do something about this, and simplify cloud cost comparisons, with this easy-to-use tool.
4) Price no longer rules
It’s not just the fact that it’s hard to compare prices that makes the budget side of capacity planning so much more complicated in a cloud environment. Like with any purchasing decision it’s the intangibles, not just the price, that play a big part. Optimizing IaaS spend is not the same as optimizing investment in hardware. On-premise iron depreciates in value over-time and needs constant maintenance, with IaaS this is no longer your problem. But (sorry) there is a problem: uncontrolled overprovisioning, without real-time cost monitoring and a controlled workflow of your procurement approval up to half of companies cloud spend 30% goes to waste.
5) It’s a process not an end-game
Being on the cloud is about flexibility, not saving money. This change in values tips the balance for capacity planning.
Depending on your industry and company, there may eventually come a point where it’s no longer economical to operate in the cloud because the monthly costs exceed what you would pay to run it on-premises. As was the case for DropBox last year. Cost reduction should not be a motivation, IaaS is a great (and necessary) tool to adapt to and remain competitive in the increasingly dynamic global market landscape. The challenge for capacity planning, alongside a labyrinth of complexity, is to accept that companies must pay a premium for something as intangible as flexibility. How do you calculate the ROI on that?
How is your company dealing with complex, expensive and limitless cloud infrastructure? Are there any other challenges for capacity planning in the cloud? Leave your comment or question below and one of our experts will respond.