Honeycomb has finally (begrudgingly) joined the Kubernetes brigade, and we have now migrated our services to k8s. Why? That’s a GREAT question. Honeycomb CTO Charity Majors, Engineering Manager Ian Smith, and SRE Fred Hebert host a casual fireside chat to answer this. They discuss the technical inflection points that helped us decide it was time to migrate, as well as the non-technical reasons (spoiler alert: it’s not because it’s the trendy thing to do right now.) Some key takeaways:
- Migrations are hard—why did we do this one?
- General migration strategies
- How are we monitoring our own Kubernetes stack
- Architects looking for real understanding of their applications
- Developers who want visibility into the impact of their application’s behavior
- Anyone who’s ever felt that monitoring was insufficient