rendertron VS puppeteer-cluster

Rendertron is deprecated

Please note that this project is deprecated. Dynamic rendering is not a recommended approach and there are better approaches to rendering on the web.

Rendertron will not be actively maintained at this point.

https://github.com/GoogleChrome/rendertron

GoogleChrome/Rendertron proxy https://github.com/GoogleChrome/rendertron.

Are you worried that rendertron isn't being maintained? (https://github.com/GoogleChrome/rendertron/issues/865 and https://github.com/GoogleChrome/rendertron/issues/801) Have you thought about using anything else or doing it yourself with Puppeteer or Playwright?

Rendertron, quite old now.

Googlebot: SSR a non-responsive mobile version only for the Googlebot User Agent, to get SEO. But do Client-Side Rendering (CSR) for users, to get full responsivity. Could be done with manual user-agent sniffing, or with a GoogleChrome/Rendertron proxy. Though Google may not like being fed something different than users (different initial markup, even though resulting markup after React’s first CSR should be the same). But it might be problematic long-term to SSR only for the Googlebot, since it’s effectively serving the Googlebot something else than what the users get.. Google might crack down on it, since it could be abused.

This might be another one: https://github.com/GoogleChrome/rendertron (not sure if it does everything that prerender does or not).

Still prefer CRA with Dynamic Rendering with rendertron over Next.js and Gatsby...

A demo Rendertron service is available at https://render-tron.appspot.com/. It is not designed to be used as a production endpoint. You can use it, but there are no uptime guarantees.

If urlsToScrape happens to be large enough, the memory and processor load from spawning dozens or hundreds of pages can bring a system down quickly. Consider puppeteer-cluster.

You could set up a service with something like playwright, puppeteer puppeteer-cluster, browserless, to access the service internally and serve screenshots of it to the outside user. You'd have to set up probably some kind of web service with the appropriate routes.

To make Unlighthouse fast, we combined this lighthouse binary with the package puppeteer-cluster, which allows for multi-threaded lighthouse scans.

https://github.com/thomasdondorf/puppeteer-cluster is doing queue-like mechanism but it is not using redis.

I've used this library: Puppeteer cluster