Skip to content
This repository

This is the currently known best practice for Picky:

  • A web accelerator (e.g. Varnish) which
  • splits the full (with results) and the live queries: full go to the app servers, live directly to the Picky server.
  • caches the results of both types of requests.
  • A number of app servers (with or without load balancer, proxy, nginx, etc.).
  • A number of Picky servers (with or without load balancer, proxy, nginx, etc.).
  • An indexing server which indexes periodically.
  • A file server that holds the shared indexes for all Picky servers.

Something went wrong with that request. Please try again.