Improve monitoring and observability for kubernetes with oss tools
0 likes206 views
The document discusses the improvement of monitoring and observability for Kubernetes using open-source tools, emphasizing the importance of centralized logging, monitoring, and alerting through technologies like EFK stack, Prometheus, and Sentry. It outlines the challenges and best practices in observability, such as managing telemetry agents and minimizing vendor lock-in, along with recommendations for building portable observability systems. The author provides examples and references to technical resources and demo applications for practical implementation.
Improve monitoring and observability for kubernetes with oss tools
- 2. $whoami { “name” : “Nilesh Gule”, “website” : “https://www.HandsOnArchitect.com", “github” : “https://GitHub.com/NileshGule" “twitter” : “@nileshgule”, “linkedin” : “https://www.linkedin.com/in/nileshgule”, “likes” : “Technical Evangelism, Cricket”, “co-organizer” : “Azure Singapore UG” }
- 3. @nileshgule Pre-requisites Self contained application with all its dependencies Docker ❖ Orchestrates containers ❖ Self healing ❖ Service discovery ❖ Scaling Kubernetes ❖ Scalable apps in dynamic environments (public / private / hybrid clouds) ❖ Exemplified by Containers, service meshes, microservices, immutable infrastructure & declarative APIs ❖ Loosely coupled systems, resilient, observable & manageable ❖ Robust automation Cloud Native Applications
- 4. @nileshgule
- 9. @nileshgule 3 Pillars of Observability Logs Metrics Traces
- 11. @nileshgule ❑ Application specific ❖ Long term log retention for compliance reasons ❖ Workloads scheduled on different nodes during application restarts / updates ❖ Autoscaling workloads ❑ Kubernetes upgrades ❖ Auto healing can reschedule workloads ❖ Underlying nodes added / deleted during cluster scaling ❖ Underlying nodes replaced during cluster upgrades Container based workloads Why centralized logging ❖ Not much control over underlying infra ❖ Relies on cloud prover specific logging and monitoring solution PaaS / Serverless services
- 12. @nileshgule Tech Talks EFK integration Log collector Log storage Log search, visualise, dashboards rabbitmq-producer-service rabbitmq-consumer-deployment
- 13. @nileshgule Demo 1 – Log Aggregation with EFK
- 15. @nileshgule • Application specific • Monitor resource usage • Monitor scaling needs • Monitor anomalies / outliers • Kubernetes platform level • Monitor cluster resources (CPU / RAM) • API health • Autoscaling Container based workloads Why Monitoring & Alerting • Monitor resource usage • Scaling • Bottlenecks PaaS / Serverless services
- 17. @nileshgule Demo 2 – Metrics using Prometheus & Grafana
- 18. @nileshgule Spring Boot Conference App integration https://github.com/NileshGule/spring-boot-conference-app/tree/mssql-server conference-demo-service-monitor conference-demo-service
- 21. @nileshgule Spring Boot Sentry integration conference-demo-service Managed Kubernetes cluster
- 22. @nileshgule Demo 3 – Exception aggregation using Sentry
- 23. @nileshgule End to End Observability
- 24. @nileshgule Observability challenges ➢ Too many telemetry agents ➢ Instrumentation of Apps ➢ Dynamic & small units in Cloud Native Applications ➢ Right retention period for each type of metric and usage ➢ Minimize vendor or feature lock-in ➢ Buy vs Build ➢ Transition from Monitoring to Observability ➢ Single pane of glass for consuming different information ➢ Correlation of signals
- 25. @nileshgule Analogy - Use right tool for right purpose
- 26. @nileshgule Summary ✓ Use best-of-class for given use case ✓ Rely on open standards (e.g. OpenTelemetry) ✓ Build portable observability systems (e.g. hybrid cloud migration) Log Aggregation ✓ EFK stack helps in centralized logging ✓ Kibana is used to visualize logs and build dashboards Monitoring & Alerting ✓ Prometheus provides easy to use metrics for platforms, applications ✓ Grafana provides visualization capabilities to build intuitive dashboards Exception Aggregation ✓ Sentry provides Exception Aggregation capabilities ✓ Excellent telemetry data captured by Sentry to help diagnose problems
- 27. @nileshgule Some Recommendations ♣ Too many agents ♣ Instrumentation, vendor lock-in ♣ Cloud native logs ♣ Cloud native metrics ♣ Cloud native traces ♣ Single pane of glass, correlation ∞ OpenTelemetry collector ∞ OpenTelemetry, OpenMetrics ∞ Fluent Bit / Fluentd, OpenSearch, Loki ∞ Prometheus, Cortex, Thanos ∞ OpenTelemetry, Jaeger, Grafana ∞ Grafana Challenges Tools
- 28. @nileshgule References Log Aggregation ❖ Elastic stack ❖ Kibana ❖ Fluentbit Monitoring & Alerting ❖ Prometheus ❖ Grafana ❖ Kube Prometheus stack ❖ Dynatrace – Monitoring vs Observability ❖ Houssem Dellai – Prometheus & Grafana for monitoring Kubernetes Sentry ❖ Sentry docs
- 29. @nileshgule Source Code & slide deck Tech Talks https://github.com/NileshGule/pd-tech-fest-2019 Observability & Monitoring markdown Conference app https://github.com/NileshGule/spring-boot-conference-app/tree/mssql-server https://speakerdeck.com/nileshgule/ https://www.slideshare.net/nileshgule/
- 30. Nilesh Gule ARCHITECT | MICROSOFT MVP “Code with Passion and Strive for Excellence” nileshgule @nileshgule Nilesh Gule NileshGule www.handsonarchitect.com https://bit.ly/youtube-nileshgule
- 31. Q&A