Riding the Serverless Wave
At Mechanical Rock we are big fans of making things faster and easier.
And that means we just don’t like running servers. With the advent of the Cloud computing and DevOps, servers – along with so much other infrastructure – have become just a form of plumbing.
Why waste discretionary effort on deploying, managing and configuring servers when you can focus on delivering value at the edge of IT - with innovative solutions for users or customers?
But sometimes when you’re living on the bleeding edge it can be hard to tell how far ahead of the pack you are, and to mix my metaphors whether you’re going too far out on a limb. I had a Venture Capitalist tell me once that technology is a lot like surfing, you’ve got to catch the wave and being early can get you dumped.
We hear concerns from some clients who feel that serverless might be too far ahead for them. They worry that they won’t have the skills and capability within the organisation to tap into the value of serverless solutions.
On face value, we think this is unfounded.
Serverless, in all its forms, uses the same underlying programming languages that enterprises have been using for decades to build software. The messaging paradigms and integration patterns are well understood and widely used. Sure the deployment and testing methods might be a little different but that difference saves you time, effort and operating cost – which you can reinvest in new products and features.
We’ve seen a lot of clients in Perth achieve massive benefits from adopting serverless technologies, either as a core strategy or as an edge case for particular compute workloads.
But don’t take our word for it.
This empirical report from DataDog on serverless adoption shows just how widespread serverless has become: https://www.datadoghq.com/state-of-serverless/
Some of the highlights:
- >50% of AWS users have adopted Lambda
- 80% of container users have adopted Lambda
- Half of all Lambda functions run for less than 800ms
Note the data is sourced only from AWS users at this point, but AWS is the cloud leader and where they go, others will follow.
Whether you’re running serverless as AWS Lambdas or GCP Cloud Functions or you’re running Docker containers with Kubernetes in EKS or GKS, there’s value to be extracted.
And if you’re not running serverless yet, then perhaps it’s time to experiment with a sample workload before that wave leaves you too far behind.