Abstract
Many organizations have had success dabbling with with Linux Containers. Once you take a small project and have success, the epiphany happens – and you ask yourself: 1. What else can we containerize? 2. Can we put everything in containers? 3. How do we get traditional applications into containers? This talk will highlight technical and architectural considerations when moving existing applications to containers. Ranging from systemd, and storage to backups, and debugging applications in production, there are a lot of things to think about when migrating existing applications to containers and running them in production.
The audience is anyone interested in migrating existing applications to containers and running them in production. This is architects, systems administrators, and perhaps some ops focused developers (release engineers, etc).
This presentation will give operational confidence to people on the fence or just dabbling in containers. This will in turn lead to more shared experience and a bigger community of people running containers for real workloads. Currently, there is an extreme lack of operational confidence. Everyone is inventing the wheels right now. For example, people are just figuring out that they need HA with their registry servers. Second, there is a huge lack of knowledge around how to containerize optimally – one big container, many smaller containers (more difficult). Migrating an existing application is almost a Turing complete problem, we need to start collecting and discussing the best practices more to encourage wider adoption.