Monitoring Microservices at Scale on OpenShift

Post on 07-Jan-2017

161 views 2 download

transcript

Martin Etmajer, Technology Lead

Monitoring Microservices at Scaleon OpenShift with Dynatrace

martin.etmajer@dynatrace.com@metmajer

Martin EtmajerTechnology Lead

Digital Performance Management for Biz, Devs and Ops

ModernizeOperations

Optimize Customer Experience

AccelerateInnovation

Dynatrace ♥ OpenShift

Why Microservices?

Why Microservices?

Velocity: Independent Development and Deployment

Growth: Independent Scaling

Innovation: Agile teams

3 Microservices Learnings

#1: Microservices are complex

Lower inner vs. higher outer complexity

Environmental Complexity

Inter-service Complexity

#2: Microservices don‘t fail independently

Design for failure

Your application has failed :-(

What’s affected?What’s the impact?What’s the root cause?

Confidential, Dynatrace LLC

What’s the impact?

What’s affected?

What’s the root cause?

#3: The network isn‘t reliable

Tightly coupled. Really Distribute?

Confidential, Dynatrace LLC

How to Dynatrace?

Dynatrace OneAgent

Option 1a: Dynatrace OneAgent for Full-Stack Monitoring

Option 1a: Dynatrace OneAgent for Full-Stack Monitoring

Option 1b: Dynatrace OneAgent for Full-Stack Monitoring

Option 1b: Dynatrace OneAgent for Full-Stack Monitoring

Confidential, Dynatrace LLC

dynatrace/oneagent Red Hat Container Certified

Option 2: Dynatrace OneAgent for PaaS Monitoring

Option 2: Dynatrace OneAgent for PaaS Monitoring

oc|s2i|Dockerfile

Dynatrace with OpenShift *

Outlook?

Dynatrace InsideContinuous

Delivery

Don’t just optimize for speed…

…instead, release fast and with certainty.

Identify bad code before it gets checked in.

Collect performance metrics from automated tests.

Feed back metrics into Jenkins and auto-stop bad builds.

3 Key Takeaways: AAA

Auto discoveryAuto baseliningAuto problem analysis

martin.etmajer@dynatrace.com@metmajer

Martin EtmajerTechnology Lead

Questions?