Peeking into Rails apps using OpenTelemetry

While Rails makes getting the application out quickly without losing the developer productivity, it needs Observability once it hits production. Because we don’t want the business to suffer in case the application performance suffers. Learn how OpenTelemetry offers the best open standards to observe a Rails app by allowing you to ask complex, improvement-focused questions before your customers ask you about them.

Transcript

1. Peeking into Rails apps using OpenTelemetry 💎🔭 1 Prathamesh Sonpatki @last9io Ruby on Rails Global Summit’23
2. 2 Instrumentation? 🤔 🤨
3. 3 Instrumentation? 🤔 🤨 - How do you know your application is running as expected?
4. 4 Instrumentation? 🤔 🤨 - How do you know your application is running as expected? - Service Level Agreements(SLA)
5. 5 Instrumentation? 🤔 🤨 - How do you know your application is running as expected? - Service Level Agreements(SLA) - Good night’s sleep 😴 💤
6. 💡“Hope is not a strategy!” 6 https://sre.google/sre-book/introduction/
7. 💡A Reliability mandate starts with Instrumentation You can only improve what you measure. 7
8. 🌈 Landscape of the Instrumentation 8
9. 🌈 Landscape of the Instrumentation 9 - Your application is not standalone
10. 🌈 Landscape of the Instrumentation 10 - Your application is not standalone - It’s actually a 🍔
11. 🌈 Landscape of the Instrumentation 11 - Your application is not standalone - It’s actually a 🍔 - The Bun(Cloud/VM) - Patty(application) - Along with Mayo sauce(RDS/DB) - And Ketchup(Third party services)
12. 🌈 The Landscape of Instrumentation 12 - Your application is not standalone - It’s actually a 🍔 - The Bun(Cloud/VM) - Patty(application) - Along with Mayo sauce(RDS/DB) - And Ketchup(Third party services) “Full stack observability” FTW!
13. 💡Modern applications are like living organisms that grow and shrink in all possible directions. And also communicate with their peers! 13
14. Agenda 📄 14 - Need of instrumentation/observability - Basic Terminology - What is OpenTelemetry - Otel Architecture - Otel + Rails - Way forward
15. Basic Terminology 󰣖 15
16. Observability 🪩 16 The measurement and attribution of performance in a complex software environment is called Observability. @realmeson10
17. Telemetry 📊 17 Collection of measurements or data at remote points and their automatic transmission to receiving equipment (telecommunication) for monitoring.
18. Logs ⛈ 18 - Why is something wrong?? - Structured vs unstructured - Easier to adopt but consistency is hard Started GET "/posts" for 192.168.0.102 at 2023-01-25 11:53:32 +0530 Processing by PostsController#index as HTML Post Load (0.3ms) SELECT "posts".* FROM "posts" Rendered posts/index.html.erb within layouts/application (3.0ms) Completed 200 OK in 16ms (Views: 14.0ms | ActiveRecord: 0.3ms)
19. Metrics 📊 19 - What is wrong? - Overview via aggregates - Consistency and adoption - Visualization using external tools https://guides.rubyonrails.org/active_support_instrumentation.html
20. Traces 󰙪 20 - Entire journey of a workflow. - Data that tracks an application request as it flows through the various parts of an application. - Provides context for what exactly happened “Tracing is just structured logging at DEBUG level. Would you run production logs on debug level? NO! ” - @nishantmodak
21. Span 🔧 21 - Basic building block of a trace - Span is equivalent to an event happening. - A tree of spans make a trace Image credit https://docs.splunk.com/Observability/apm/apm-spans-traces/traces-spans.html
22. Observability back-ends 📇 22 - A tool/service that gets the telemetry data and presents way of visualization, aggregation, alerting, notifications.
23. Standardization Challenges 📜 23 - Different vendors - Different formats - Agents - Nomenclature https://last9.io/blog/observability-is-dead-long-live-observability/
24. OpenCensus 💻 OpenTracing 🪞 24
25. OpenTelemetry 🔭 25 https://www.cncf.io/blog/2021/08/26/opentelemetry-becomes-a-cncf-incubating-project/
26. What OpenTelemetry is not ⛔ 26 OpenTelemetry is not an observability back-end.
27. Architecture 🏛 27
28. Architecture Demo󰙥 28
29. Architecture 🏛 29
30. Architecture 🏛 30
31. Architecture 🏛 31 https://lightstep.com/blog/opentelemetry-collector-design-and- architecture
32. OTLP 🧬 - Describes the encoding, transport, and delivery mechanism of telemetry data between telemetry sources, intermediate nodes such as collectors and backends. - gRPC and HTTP implementations https://opentelemetry.io/docs/reference/specification/protocol/ 32
33. Automatic and Manual instrumentations ⚒ - Automatic instrumentation by just including relevant gems. - Out of the box telemetry data. - Common bases are covered. - Don’t have application specific data. - Getting started with auto-instrumentation is easiest. 33
34. Automatic and Manual instrumentations ⚒ - Automatic instrumentation by just including relevant gems. - Out of the box telemetry data. - Common bases are covered. - Don’t have application specific data. - Getting started with auto-instrumentation is easiest. 34 - Deeper insights into application. - Custom metadata. - More control. - Needs application code change.
35. Way forward 🛣 35
36. Way forward 🛣 36 - Ruby Otel SDK Complete support for traces - WIP for logs and metrics - https://github.com/open-telemetry/opentelemetry-ruby - https://opentelemetry.io/docs/instrumentation/ruby/
37. Who are already onboard on the Otel Train 🚝 37 - GitHub - Shopify - Heroku - Puppet - Dropbox - You??
38. Recap 💹 38 - Need for standardized instrumentation - What is the essence of OpenTelemetry? - Otel Architecture - Rails app using Otel - Automatic and Manual instrumentation - Way forward
39. Thanks 🤝 39 Prathamesh Sonpatki Last9.io 󰜼 prathamesh.tech 🐧 twitter.com/_cha1tanya 🐘hachyderm.io/@Prathamesh “Last9 of Reliability” Discord