Laura Nolan
Troubleshooting is one of the hardest parts of software operations. Observability tools show us system state, but rarely how it got there. Problems may be intermittent, difficult to reproduce, or involve systems that are slow rather than broken. We work under uncertainty and time pressure, often without a complete understanding of the system.
So how do expert troubleshooters actually do this work?
This talk combines research on “cognition in the wild” across complex domains with my own research into troubleshooting practices in software operations to explore what effective troubleshooting looks like.
Laura Nolan is a software engineer and SRE who has worked at Google, Slack, and Reddit, focusing on reliability, distributed systems, and incident management. Laura recently completed an MSc in Human Factors and Systems Safety at Lund University, where she investigated how experts troubleshoot complex software systems. She is now at Trace Cognitive Engineering, applying her experience in software operations and research in human factors to help organisations build and run their complex systems more effectively. Laura is a long-time SREcon volunteer and a member of the USENIX Board of Directors.
