Logging and Monitoring k8s applications with DataDog

A quick overview of logging and monitoring for Kubernetes applications using the DataDog daemonset deployment to assure all pods across the cluster are covered.

Transcript

1. 22 . 03 . 2019 Logging and Monitoring k8s applications Use case with DataDog Felipe Schmitt for PixelsCamp
2. We’re a company that drive innovation and startup culture in big corporations
3. Fun facts
4. Per host, there are eight containers simultaneously, on average. 8 containers
5. Container Lifespan, on organizations using orchestration such as Kubernetes 12 hours
6. Half of Docker environments are orchestrated instead of standalone technology 50% orchestrated
7. You have a K8s cluster. Now what?
8. For a production-grade Kubernetes cluster, monitoring, alerting and log aggregation are essential.
9. Due to increase of increase complexity and ephemeral nature of containers . Infrastructure monitoring . Logging aggregation . Applicational monitoring . Alerting
10. Which enable to easily aggregate logs, analyze them and monitor both containers and hosts very easily
11. 12 Factor methodology for designing microservices and cloud-native apps . logs to stdout . kubectl logs . hundred of pods in Prod . no log history (ephemeral)
12. Logging all the information to stdout in JSON format to enrich with further data
13. Triggering alerts based on log monitoring
14. Across all k8s pods throughout the entire cluster by deploying as DaemonSet
16. Demo
17. Daemonset
19. Logging analysis
20. Alert Monitoring
21. Infrastructure Monitoring
22. Thank you! 22 . 03 . 2019 f.schmitt@kigroup.de /fsschmitt @schmittfelipe