Talks
Speakers
Events
Topics
Search
Sign in
Search
Home
Talks
Speakers
Events
Topics
Leaderboard
Use
Analytics
Sign in
search talks for
⏎
Suggest modification to this talk
Title
Description
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 his presentation at the wroc_love.rb 2023 conference, Cristian Planas discusses the challenges and techniques associated with scaling Ruby on Rails applications, especially in high-demand environments. Emphasizing the theme that 'Scaling is hard,' Planas reflects on his extensive experience in the tech community and his current role at Zendesk, where he deals with performance issues. He begins by highlighting the misconception that Rails does not scale, citing successful examples from companies like Shopify and GitLab. Throughout the presentation, Planas covers several key points regarding performance optimization: - **Monitoring and Error Budgets**: Planas stresses the importance of identifying and monitoring performance issues using error budgets, which help define acceptable latency for key endpoints. - **Database Optimization**: He discusses the need for proper database indexing, the implications of query complexity, and the importance of sharding to improve performance in large datasets. - **Caching Strategies**: Planas explains how caching can significantly enhance application performance, advocating for techniques like write-through caching and asynchronous processing to reduce response times. - **Data Management**: He shares his approach to data regression, where unnecessary data from closed tickets is moved to cold storage, thus reducing load on primary databases. - **Product Limits**: Highlighting the necessity of setting limits on database queries and user operations to prevent abuse and system overload, he suggests implementing timeout limits. - **Handling Large Payloads**: Planas warns against feature creep in API responses, advocating for the delivery of only essential data while managing optional or aggregate data separately. As he wraps up the presentation, Planas emphasizes the mindset required for engineers tackling performance issues, invoking the story of Alexander the Great and the Gordian Knot to illustrate the need for thinking outside conventional constraints. He advocates for a community spirit within the Ruby ecosystem that balances developer happiness with performance optimization. The talk concludes with an invitation for questions, showcasing an interactive engagement with the audience to further explore these complex topics in performance engineering.
Suggest modifications
Cancel