Serverless at Capital One
So, what exactly does Serverless mean again?
Firstly, ‘serverless’ doesn’t mean we stop using servers. It just means we don’t have to think about them anymore.
It’s a way of running code without thinking about the infrastructure it’s running on. Think of it as a layer of abstraction on top of what Containers provide, where instead of managing shared infrastructure, you now only need to manage the code you write.
Serverless architecture pushes you to write small units of code because it doesn’t come with the additional overhead of needing to deploy it to new or shared infrastructure. This also makes the testing process nice and simple, because the scope of the code’s functionality is really narrow.
Here at Capital One UK we’re still in the early stages of adopting Serverless, using it mainly for Cron style jobs and reacting to events produced by Amazon Web Services (everything from automatic data encryption, to cost management & compliance exercises like GitHub: capitalone/cloud-custodian).
My first experience with Serverless technology was when Amazon released Lambda, their managed AWS service, at Re:Invent back in 2014. I decided to test it out, so wrote a quick Java API, as well as using it to host a simple Nodejs web app serving content out of S3.
Getting used to working with Serverless technologies has been really straightforward. The learning curve is gentle, and once you’ve cracked the different ways of triggering functions and how to access event data from within its context, the actual processing and sending a response is relatively easy.
I’m not sure every use case that comes along will fit in with what Serverless aims to provide in its current state. But I’m still a believer in it as a technology, and have no doubt Capital One will be using it more and more as we continue to deploy applications to the Cloud.
Of course, within a regulated environment there are always going to be extra hurdles. So it’s encouraging to see Serverless providers working with other technology companies like ourselves, and adding the features needed for new cases to be explored.
One of the biggest benefits of Serverless is removing the requirement for software engineers to understand how to create and manage infrastructure where their code is deployed; or requiring a separate operations team altogether.
It’s had a big impact on time too. Our software developers can now spend more time thinking up great ideas, and less time trying to deploy them into production.
I’m really excited about ServerlessConf London 2016, and can’t wait to see how other organisations and individuals are finding uses for Serverless. Being primarily focused at working within AWS, I’d also like to see some examples of developers using Microsoft Azure and Google Cloud Functions to see how they are competing in this space.
If you’re attending ServerlessConf, come over and say hi at the Capital One stand.
See you there!
The opinions expressed in this article are the author's own and do not reflect the view of Capital One.
ServerlessConf 2016 will be taking over London’s digital playground from 26 – 28 October 2016. The event brings together some of the brightest minds in the industry, and serves as a platform for developers to share experiences and ideas on building applications.
Three of this year’s speakers are from Capital One.