Performance Monitoring
Talks about performance
- "How my app isn't janky anymore"
- "10 awesome performance hacks"
Talks about performance
- "Use GraphQL and Relay / Apollo to load only what you need"
- "Prefetch your components and their data before they load to get better performance"
Too little talks about
- "Our #perf optimization had an impact on our customers and this is the way we measured"
- "This premature optimization had a bad impact on our customers with older phones and this is how we noticed"
Why do we measure?
- Ship the fastest product possible
- Real world performance data
- User type
- Devices
- Usage Path
- Gain confidence that optimizations
actually improve the performance
What we want to see?
What we want to see?
Text
Text
root span
span
service
OpenZipkin
- Server
- Client
- Integrations
docker run -p 9411:9411 openzipkin/zipkin
Zipkin.js
OpenTracing
The Standard
- standard and implementation combined
- switch out your implementation, but keep the same source code
Zipkin-Javascript-Opentracing
jaeger by uber
TL;DR
- Zipkin: awesome tool to get performance insights
- Opentracing: standards are nice
- Integrating Tracing in your App is not that hard
Why I think this is useful
- Backing (or disagreeing) with performance tips based on data is nice
- Adding Tracing on a few key parts helps to understand a lot of behaviour
- Better collaboration with backend-oriented team members
or
My try to get everyone into this without needing to learn that much about all these systems
Goals
- Enable to give this whole world a try
- Abstract common use-cases in the frontend
- Network requests
- User Interactions
- Navigation
- Allow for different backend systems
react-tracing
react-tracing
- Tracing Network Requests: ✅
- User Interaction: ❌
- Navigation: ❌
Thank you for your time ❤️
Performance Monitoring
By Daniel Schmidt
Performance Monitoring
My slides for the React Native Meetup HH #3
- 839