How are my microservices doing?

A talk about the basics of observability. Why it is needed? Why it is important? What are distributed tracing, logging and metrics? We talk about Prometheus, Jaeger, OpenTelemetry, Grafana, Kibana, ElasticSearh, Docker, OpenShift and more tools in the observability context.

Transcript

1. +Israel Blancas @iblancasa How are my microservices doing?
2. Israel Blancas @iblancasa isra.rocks Quality Engineer at Red Hat Google Developer Group organizer
3. Before starting
4. Before starting
5. • Microservices Who knows about...
6. • Microservices • Containers (Docker) Who knows about...
7. • Microservices • Containers (Docker) • Container orchestration (Kubernetes) Who knows about...
8. Observability?
9. Observability? Application
10. Observability? Application
11. Observability? Application
12. What do we usually do? Check the stacktrace and read the code until find what is wrong
13. Observability? Application Metrics Logs Traces
14. Why is “observability” something interesting? A quick introduction
15. Scaling up the infra Introduction
16. Hypervisor OS OS Scaling up the infra OS Introduction
17. Scaling up the infra Introduction
18. Hypervisor OS OS Scaling up the infra Hypervisor OS OS Introduction
19. Scaling up the infra Cattle (containers) • Has a number • One is much like any other • Run as a group • If it gets ill, you sacrifice it Pet (VM) • Has a name • Is unique or rare • Personal attention • If it gets ill, you make it better Introduction
20. Application Microservices Service 1 Service 2 Service 3 Service 4 Introduction
21. Service meshs Introduction Service 1 Service 2 Service 3 Sidecar proxy Sidecar proxy Sidecar proxy Control plane
22. To know more! Scalable Microservices with Kubernetes https://www.udacity.com/course/scalable-microservices-with-kubernetes--ud615 Phippy and friends https://www.cncf.io/phippy/ Introduction
23. Why do we need this? Introduction
24. Why do we need this? Introduction It all about the money!
25. Why do we need this? Introduction It all about the money!
26. It all about the money Disaster can not be avoided • Untested code paths • Corner cases • Outages • Human error • Network issues: latency, bandwidth, reliability, homogeneity… Introduction
27. Ok so… observability
28. Application Metrics Logs Traces Observability
29. Observability Observability and monitoring Alerting Overview (dashboards) Profiling, debugging, tracing, logging, dependency analyses... Monitoring Observability
30. Failures appear • What? • When? • Why? Distributed systems as context Observability
31. Observability Some technology examples Prometheus, Thanos Jaeger Elastic, Fluentd, Kibana Log Management Metrics and alerts Tracing Kiali, Grafana, OpenShift UI Management Dashboards
32. Logging
33. Solutions to manage and analyze big amounts of logs from different systems Logging
34. Metrics
35. D Metrics
36. Distributed tracing
37. Distributed tracing Tells the story of a request across services Which services were touched, when, in which order, what parameters were used… D Tracing
38. Spans Units of work created via instrumentation Each one includes information about when it started, how long it took… and can contain references to other spans D Tracing
39. Spans D Tracing
40. Spans D Tracing
41. Spans D Tracing @GET @Produces(MediaType.TEXT_PLAIN) public String submit() { Account account = accountsService .getAccount(); Order order = new Order(UUID.randomUUID().toString(), account); inventoryService .processOrder(order); return String.format("Order submitted: %s" , order); }
42. Spans D Tracing @GET @Produces(MediaType.TEXT_PLAIN) public String submit() { try (Scope scope = tracer.buildSpan("submitOrder").startActive(true)) { Account account = accountsService .getAccount(); Order order = new Order(UUID.randomUUID().toString(), account); inventoryService .processOrder(order); return String.format("Order submitted: %s" , order); } }
43. Jaeger D Tracing
44. dev.to/iblancasa
45. Thank you! Questions? +Israel Blancas @iblancasa