AWS Cloud Per Second Billing — A Game Changer for DevOps Engineers

When AWS launched it’s cloud services in 2006, the per hour billing model for compute resources was a game changer for the IT world. Since then, AWS has taken over the cloud computing market as a leader, while broadening the range of cloud services offered every year. However, the per hour billing was one of the major issues for many customers, who had applications that used EC2 for only a few minutes, but had to pay for the entire hour.

Recently, AWS announced the support for per second billing of it’s EC2 service (Compute) that could further revolutionize its usage, and help reduce costs for companies that take advantage of auto-scaling architectures. Here are the top 5 per second billing implications for DevOps teams.

Autoscale As You Want Without Cost Worries — Most engineers were vary of auto-scaling at will, due to hourly charges of EC2 even though they might need additional compute capacity for 15 minutes. With the per second billing, you can be worry free and just scale your EC2 (compute) as and when you need.

No More Custom Logic For Hourly Batches — Many customers use EC2 for jobs and batch processing purposes. Due to the hourly pricing, the engineers have written custom logic to benchmark and constantly tune the number of nodes in their batch clusters to ensure they are cost optimised. This one will free up headaches for DevOps engineers using EC2 for batch processing and jobs cluster.

Need To Speed Up EC2 Launch — With per second billing, it’s important to take advantage of pre-baked AMI’s for launching your EC2 instances, instead of complete configuration setup on every instance launch. This will reduce costs anywhere from 5 to 15 minutes. It’s time to create pre-baked AMI’s for faster instance launches and to save costs.

Bench-mark AWS Lambda or EC2 — Be prepared to compare whether to use Lambda or EC2 for new workloads and applications than ever before. Choose the best path to your goal stringently as the serverless Lambda has just gotten a lot more competitive than EC2.

Your Big Data Just Got Easier — With EMR also supporting per second billing, your job has become a lot easier. As a DevOps engineer, you no longer need to constantly benchmark your EMR clusters for the number of nodes, and the job processing optimization for the number of times per hour billing.

The per second billing is not only beneficial for AWS customers to lower costs in many cases, but also in making the DevOps engineer’s life easy.

These implications are just few of the many ways in which DevOps engineers can benefit from the per-second billing for AWS. Did we miss anything? We would love to know if it is a boon or a bane for all the DevOps engineers out there. Leave a comment with your thoughts on this.