Data Mirroring Service¶
A local machine with a copy of data often downloaded. The usual data is software distributions (Debian/Ubuntu deb packages, Fedora rpm, etc), but mirroring can apply to other kinds of data too.
- Downloading data from a local machine is faster then from a distant one.
- It requires less external traffic, so it costs less in bandwidth.
- The data is still available when the external network link has problems.
steller server is available for this service. Everything else needs to be decided and done.
Basic mirroring service is not very complex, it only requires setting up a few services, typically HTTP, FTP and rsync servers (to distribute the mirrored data) and cron jobs to update the local data from the original source. But once the service is running, it is VERY important to maintain it on the long term in order to have numerous user gradually use it.
nil — Nicolas Limare) am working on it. A few details:
- I am usually in Japan, ie different time zone.
- I don't read/write/understand spanish. French, English, Japanese OK.
- It's a side (hobby?) project. My main work is on IPOL (http://www.ipol.im/), with a server at FING.