Chapter 3
Right-sizing
Only buy what you need, as you need it
AWS empowers you to flexibly upsize or downsize your AWS resources – and therefore, your AWS bill – as you need. This is a huge benefit over the past, where businesses attempt to predict their future needs, make enormous capex investments in IT infrastructure to cater for this growth, then pay for resources they don’t yet need.
You have a newborn baby and set out to buy a jumper which is the right size. You try to anticipate your baby’s growth spurts from age 0 to 3 years and – after hours researching baby blogs and a few more grey hairs – you realise that baby is going to change jumper sizes at least 7 to 10 times in this period. The size they need at any given point will be based on their individual growth and weight attributes – factors you can’t possibly predict when they’re newborn.
So what do you do? Go and buy multiple jumpers of different sizes and hope that you’ve precisely predicted your baby’s growth and weight increases?
Do you want to make that heavy capex investment upfront, based entirely on your best-guesses? Of course not. You’ll right size according to how your baby grows – only buying the clothes that fit at the time they’re needed.
The ways of the past
If you’re used to buying your IT infrastructure up front, you’re probably familiar with the planning that goes into the purchase. When IT departments receive the go-ahead to scope out your requirements, they commonly ask themselves the following questions:
- What types of workloads (business systems/applications) will need to run on these server(s)?
- How much horsepower (CPU, memory and storage) is needed to support each workload?
- How much will our needs in the future change as we grow and how much extra horsepower do we need to cater for this change?
- What are the hardware expansion capabilities that are available to enable future growth beyond our expectations?
By asking these questions, they’re doing their best to choose the right size IT infrastructure for your business’ needs. But they’re also fearful that they’ll get it wrong.
Overestimation means money wasted on hardware you didn’t need. Underestimation means buying more hardware down the line with money for which there is no budget. Either way, the result is an inefficient use of your resources.
Yet, this is exactly what IT teams have been attempting to do for the last 20 years. They try and predict the growth of their company in different ways, then commit heavy CapEx upfront to buy the IT equipment they think they need to cover them.
Today, there’s no risk of getting it wrong
With AWS, you can flexibly increase or decrease the capacity of your resources at the press of a button (literally). This is of huge benefit to your business at two stages throughout the lifecycle of your infrastructure:
- At launch. The risk of miscalculating your businesses immediate needs is eliminated. If your AWS resources aren’t the right fit, you can adjust in seconds.
- Throughout useful life. The risk of incorrectly anticipating your business’ needs in the future is eliminated. As your business and it’s IT infrastructure requirements change, so can your AWS resources.
Naturally, this flexibility leads us to our next section which aims to help you understand the elasticity of AWS to help make even more cost savings.