Talks
Speakers
Events
Topics
Sign in
Home
Talks
Speakers
Events
Topics
Leaderboard
Use
Analytics
Sign in
Suggest modification to this talk
Title
Description
Developing large and high load application without monitoring is a hard and dangerous task, just like piloting an aircraft without gauges. Why it is so important to keep an eye on how application's “flight“ goes, what things you should care more, and how graphs may help to resolve occurring performance problems quickly. On an example of Sidekiq, Prometheus, and Grafana, we will learn how to take metrics from the Ruby application, what can we see from their visualization, and I will tell a couple of sad tales about how it helps in everyday life (with happy-end!) RubyKaigi 2019 https://rubykaigi.org/2019/presentations/Envek.html#apr19
Date
Summarized using AI?
If this talk's summary was generated by AI, please check this box. A "Summarized using AI" badge will be displayed in the summary tab to indicate that the summary was generated using AI.
Show "Summarized using AI" badge on summary page
Summary
Markdown supported
In the presentation titled "Yabeda: Monitoring monogatari" by Andrey Novikov at RubyKaigi 2019, the importance of monitoring in software development is highlighted, drawing parallels with aerospace engineering. Novikov emphasizes that developing large-scale applications without adequate monitoring is risky, akin to piloting an aircraft without instruments. Key points discussed in the talk include: - **Monitoring Needs**: Understanding the necessity of monitoring to detect performance issues and ensure seamless operation in applications, particularly those using Sidekiq for managing background jobs. - **Performance Indicators**: Highlighting critical metrics to track, such as queue sizes, job dynamics, and application performance to identify bottlenecks or failures early. - **Analogy with Aeronautics**: Comparing software applications to modes of transport, stating that as applications scale, they require more extensive metrics and monitoring akin to an aircraft's cockpit filled with gauges. - **Types of Metrics**: Discussing different metric types like counters and gauges, explaining their roles in monitoring job execution times, queue sizes, and system performance. - **Practical Implementation**: Demonstrating how to integrate monitoring using Prometheus and Grafana, enabling detailed tracking of Sidekiq jobs through customizable metrics. - **Case Study**: Sharing a real-life experience where monitoring helped troubleshoot a bottleneck with the Sidekiq queues, leading to a clearer understanding of the root cause involving database locks due to a surge in jobs from new feature usage. - **Yabeda Framework**: Introducing Yabeda, an open-source monitoring framework that simplifies the setup of custom metrics for Ruby applications, encouraging the use of open source to enhance monitoring capabilities. The conclusion underscores the need for thorough monitoring in software applications to mitigate unforeseen issues and expedite debugging processes. Effective monitoring allows developers to confidently assess their application's performance and verify the impact of optimizations. Novikov invites the audience to explore Yabeda and its functionalities for better monitoring practices, reinforcing the idea that proactive monitoring is essential for robust application performance.
Suggest modifications
Cancel
