TechPrufer
Get started
Atlassian
Atlassian logoAtlassian · Feature trail

Streaming Server-Side Rendering in Confluence

2 posts · Jun 2020 – Apr 2026

What this trail covers

Confluence has significantly improved page load performance by implementing React 18's streaming Server-Side Rendering (SSR) capabilities. This allows for faster First Contentful Paint (FCP) and Time to Interactive (TTI) by rendering and streaming page components as they become available, rather than waiting for the entire page to render. The system also optimizes client asset loading by predicting and preloading necessary JavaScript bundles during the streaming process, further reducing interaction-ready time. This builds upon previous performance improvements and the adoption of React hydration and Suspense boundaries.

All Atlassian trails
  1. Streaming Server-Side Rendering in ConfluenceThis post details the integration of React 18's streaming SSR capabilities into Confluence to reduce page load times. It explains the implementation of Suspense boundaries, data streaming abstractions using NodeJS streams, and optimizations for client asset loading through a feedback loop for predicting and preloading JavaScript bundles. The post also discusses key performance metrics (FCP, TTVC, TTI, Hydration Success Rate) and the evolution from traditional SPAs and SSR to SSR streaming.
  2. Improving performance through better architectureThis post details the creation and adoption of a new 'react-resource-router' for Jira Cloud, designed to address performance issues and compatibility problems with React Router, particularly in server-side rendered (SSR) single-page applications. The new router couples data fetching with route matching, utilizing a static route configuration that includes a `resources` property for asynchronous data fetching. It leverages 'react-sweet-state' for state management and provides immediate resource execution on the client and imperative fetching on the server. The post also discusses the progressive migration strategy and the router's internal data storage and caching mechanisms.