archlinux.org mirror list is a single point of failure

[nroi]

cpcache fetches a JSON document from https://www.archlinux.org/mirrors/status/json/. The fallback strategy when this URL is not available is to look for a locally cached version of the mirror status.

This obviously only works when cpcache has been started before. A nicer approach would be to make a cached version of the JSON document available on the web (e.g. gist.github.com) and use this as a fallback. This would also be cleaner from a DevOps perspective: We could run integration tests with Docker that would succeed even if archlinux.org is down.