The method
Use this prompt in a chat-based LLM like ChatGPT. First, describe your website's architecture, tech stack, and performance metrics. Then, paste the prompt. Analyze the LLM's output and iteratively provide more details for refined diagnostics.
The prompts
Prompt 1
I need help diagnosing performance issues on my e-commerce website. The site is built with React on the frontend and Node.js with Express on the backend. We use a PostgreSQL database hosted on AWS RDS. We've recently seen a significant increase in page load times, especially during peak hours. Our key performance indicators (KPIs) are: First Contentful Paint (FCP), Largest Contentful Paint (LCP), and Time to Interactive (TTI). FCP is averaging around 3 seconds, LCP around 5 seconds, and TTI around 7 seconds during peak hours. Provide a detailed analysis of potential bottlenecks in the frontend, backend, database, and network. Also, suggest tools and techniques for further investigation and optimization in each area. Focus on the most impactful areas for optimization, considering the current architecture. Prioritize suggestions that require minimal code changes for initial investigation.
Prompt 2
I'm working on a single-page application built with Vue.js and a REST API backend using Python Flask. The application is hosted on a Linux server with Nginx as a reverse proxy. We've noticed slow API response times, particularly for endpoints that retrieve large datasets. Provide a checklist of common performance bottlenecks in such setups, focusing on the following areas: Vue.js component rendering and optimization, API endpoint efficiency (including database queries), Nginx configuration, and potential issues with the underlying Linux server. For each area, list specific tools and techniques to diagnose and resolve the problems. Consider factors like caching strategies, database indexing, and server resource utilization.