Dana Asbury is a Senior Engineer at Leafly. He struggles with imposter syndrome and has a hard time writing about himself positively. He has enjoyed working with Kubernetes over the last 4 years and this will be the first time he has given a talk about his experiences with it.
Building from the Ashes: Lessons Learned From Layoffs
Have you ever wished that you could have a do-over on building your system? Our team got to do this in an unlikely way. We were tasked to build an e-commerce marketplace running on Kubernetes with Istio in 16 weeks. Months after launching we were all unexpectedly laid off. In the time in between we learned countless lessons around building and operating in a Kubernetes environment. We were then fortunate enough to be hired together to - you guessed it - build another marketplace running on Kubernetes.
This talk will detail our journey from supporting infrastructure that was always on fire to applying our lessons learned that allowed us to build a maintainable system that allows us to iterate quickly. We will cover:
- Learnings from production outages
- Benefits of GitOps and infrastructure as code
- Leveraging service mesh to improve observability and deployments
- Improving local development and developer experience
Dying With Dignity: How to Exit Your Applications Gracefully
Have you ever wondered why your error rates spike on deployments? Tired of ignoring those 120 second timeouts in your logs? In this talk we will get to the bottom of these common errors and more.
After this talk you will understand:
- What makes a good readiness probe
- The difference between SIGINT vs SIGKILL vs SIGTERM
- The different situations when Kubernetes will kill your pods
- How to handle these different termination cases in your application