Fabian Stäber from Instana, gives a really informative talk about how to tackle monitoring with microservices.
Lots of things to consider:
- Automation of monitoring to avoid bottlenecks with deploying microservices.
- The underlying data
More than health checks which are more common with monoliths.
Define what does it mean for my application to be healthy. Discussing how you take this time series data and apply it to the additional dimensions that microservices bring.
- Aggregation horizontal metrics
- Vertical checks – what is going on in the stack Java, JVM, Docker, K8 node, hardware itself.
- Ensure new and legacy services also provide tracing
- Tie together traces to error logs.
Fabian helpfully points out these three Site Reliability Engineering books from Google that are available for free.
Even finance companies are considering it
Only a young project
1 app = 1 container
What’s the diff
Uses less space, only store what is necessary
Use cases :
Test prod architecture
Can combine with Jenkins CI CD
Can use base images
And then keep the docker configuration file clean
Is command line tool, but can use other tools to make simpler
Fig is a helpful tool for creating an infrastructure
Possible mention to http://deis.io/overview/ and Heroku
Almost no down time, if upgrading a component
Look at mounting volumes for running local dev code
Can use another app to orchestrate the dockers apps