11:40 - 12:40 (UTC+02)
Talk (60 min)
Changing your system based on metrics from within your infrastructure is nothing new. Scaling up and down based on how much strain your services are under has been around since forever. But what if you wanted to change your system based on events from external services?
Imagine something like a PagerDuty alert causing changes in your Kubernetes roles to allow the on-call engineers access to a production cluster that's having issues, or a bastion host to be created in your environment with the keys necessary to allow someone to gain access, or deploying new infrastructure for a customer because your payment system has announced a new successful card transaction.
In this talk, Piers will go through how you can make use of techniques that are well established when building complex event-driven software systems when creating your infrastructure.