[This article was written by Matt Kiernan.]
Systems have fundamentally changed with the introduction and adoption of microservices like Docker, resulting in a shift in how we think about log management and analysis. Logging into a server and grepping logs is no longer a reality when dealing with thousands of container instances. While support of logging has improved on Docker over the past year, there is yet to be a widely agreed upon “standard” for Docker logging.
Logentries Chief Scientist Trevor Parsons recently co-hosted a webinar with Peter Elger of nearForm and Bright Fulton of Swipely where we explored three different approaches to logging on Docker, including:
- Logging from within the container using a Daemon/collector
- Logging outside the container, writing logs out and running a collector on the host
- Using a dedicated logging container responsible for collecting and forwarding logs
Over the course of the webinar, Trevor, Peter and Bright engaged in a lively discussion, exploring the pros and cons of each option and debating which one truly represents “best practice.”They also shared insights into common events that are valuable to log, monitor and analyze,, useful Docker APIs, and services for analyzing Docker log data for a deeper understanding of your application and environments.
Missed the webinar? No problem – you can view a video recording of the webinar below and find a list of helpful Docker resources here.