Oplægget er optaget til SAOS hos Lund&Bendsen A/S den 04/11-2021.
Abstract:
Designing organisations with a high degree of team independence, very short iteration cycles and an inherent ability to make fast decisions is the holy grail of modern agile teams. Driving that autonomy and ownership while at the same time achieving consistently high service level objectives that your enterprise customers demand can be a daunting challenge.
This talk covers usage of technologies such as LinkerD, Kubernetes and Prometheus to set service error budgets for highly independent teams through the use of service level objectives and stop-the-line policies for microservices.
We will cover the implementation details of this system including challenges with the technology and ensuring a scalable foundation for a monitoring system that can query success rates on billions of http requests over hundreds of microservices for extended periods of time. We will also discuss how this was introduced organizationally, custom software components we had to write to make this possible and what value this transformation provided to our customers.
Speakers profile:
Jesper’s background is in software engineering, with interests spanning: typesafe languages, distributed systems, cloud infrastructure, dev-ops automation, software architecture, project management, leadership, organisational theory and human processes. Working as a consultant for 10+ years in telecom and finance. Jesper has spent the last 8 years in Tradeshift helping the company grow by leading engineering teams, building software systems and software development processes.