Skip to main content

The Machine is Dead, Long Live the Machine! - Service Resilience and Deployment Automation at The BBC

Culture & Organizational Change
Location: 118/119 Level: Intermediate
Average rating: ****.
(4.63, 19 ratings)
Slides:   1-PDF 

The days of fixed physical infrastructure and isolated Ops teams are over. It is no longer excusable for a developer to write their code, throw it over a wall and let someone else deal with the impact.

As hardware becomes software, engineers are empowered to create infrastructure that is reproducible, scalable and specifically dedicated to their service. They can spin up identical environments that give them testing confidence. They can form automated pipelines to continuously push services to their audiences. They have the power and responsibility to shape their application, the machines that run it and the infrastructure that glues their service together. For the BBC Platform team, which I am part of, this is exciting – we no longer have to be gate keepers and we could focus on being enablers.

In this talk, we will speak about the main ingredients that shape a service in the cloud and the path from a developers keyboard to the audience. We will cover:

  • Building software cleanly and in a reproducible way and why this matters
  • The fastest way to bake an AMI and some considerations for what to bake into the machine.
  • The key benefits of being able to declaratively define and provision infrastructure as code
  • Cosmos, the BBC tool that orchestrates and automates the deployment of software
  • Integrating automated testing with a deployment pipeline.
  • How teams change the way they develop and operate in order to fully maintain their services.
  • What we did well, what mistakes we made and what is our way for improvement
  • How all this makes the world a better place

We will fly over our main architecture designs and we will take some deep dives into technologically curious places without drowning in boring details. Hopefully the journey will leave the audience as inspired as we are.

Photo of Yavor Atanasov

Yavor Atanasov

BBC

Yavor is a software engineer who has in-depth experience within the whole spectrum of web development at a very large scale.
He is from Bulgaria, living and working in London UK.
Part of BBC’s central Platform Team, working with a wide range of technologies and languages.
Currently building BBC’s next generation cloud based platform and continuous delivery pipeline using Amazon Web Services.
Pioneering continuous delivery and dev-ops culture within the development teams. Above all, having fun.