Don't Let the Public Cloud Snatch Away All Your Budget
“Never put all your eggs in one basket.”
This adage perfectly sums up the advice any veteran cloud manager would pass on to a novice. And it is required more than ever today. Why? Because Gartner reports that global spending on public cloud services hit an insane mark of $214.3 billion this year.
As things stand, spending all your resources on public cloud is the worst decision that you can take. Controlling this is an even harder feat in case you have a multi-cloud strategy in place.
If you have been a victim of wasted resources and higher than appetite costs, here’s what you have to do to keep your cloud environment under check:
1. Analyze Your Monthly Cloud Bills:
If your organizational culture reflects the callous attitude of a rich and spoiled brat, you will be in trouble the moment you get out of bed. The faster your hand reaches for your pocket, the deeper will its hole be. You need to be stingy with your spends, analyze your cloud bills, and narrow down any unexpected services and regions that are in use.
2. Drill Down on Spikes and Anomalies:
But before that, do you have a blueprint of your cloud strategy in place? If not, you should not even be reading this. The next step is to map the frequency and intensity of your cloud resources on a wall (figuratively) and look for instances where they hit the ceiling or the floor. These are the points of inflection that can do with some smart optimization.
3. Compute the Costs of Your Network Traffic:
You don’t bring a needle to a sword fight. And, of course, vice versa. Cloud resources are billed on the hourly rate of machines. And as logic dictates, smaller machines are cheaper than their larger counterparts. While using smaller machines is an obvious initial reflex, you can even deploy a scale-out architecture that makes use of multiple and smaller resources.
4. Consider Cost Differences by Location:
Have you ever been to two different McDonald’s and paid different prices for the same happy meal? The same is possible with Cloud services. Prices are naturally dependent on input variables labor costs and market fluctuations. Whatever the case, don’t jump the gun without a fair share of due diligence.
5. Design Workloads for Scalability:
Before making any decision, time-travel into the future. What would be your cloud needs 2, 5, or even 10 years from now? The efficiency with which you answer these questions is directly correlated to cloud costs down the road. With cloud orchestration, containers, and microservices in the picture today, tipping the decision in favor of one factor may enrage another.
Which is the most nagging issue related to public cloud costs that you are facing? Share your thoughts with us in the comments section below.