Skip to content
Switch branches/tags

The toolset has been split off, and is now here available.

Current status and any announcements as well as a maintaince log can be found here

Under the domain are package mirrors provided via the IPFS-Network. If you choose this as your mirror, Pacman will download the files from a local http-proxy and the accessed files will be shared with the IPFS-Network.

If you run on multiple machines in your local network IPFS will exchange the files locally rather than downloading the same file multiple times from the internet.

The data is hold and shared by a collaborative cluster, where everyone can participate.

Cluster members

Member ID Server Location Internet-Provider AS Provider
12D3KooWDM4BGmkaxhLtEFbQJekdBHtWHo3ELUL4HE9f4DdNbGZx Nuremberg, Germany netcup AS197540 @RubenKelevra
123... Guthrie, Oklahoma Cox Communications Inc AS22773 @teknomunk
12D3KooWJnEkuYkoWgfGBjQXE6bvn6Y53GmjwuC2vk7d2Xxk9xM5 Vilnius, Lithuania UAB Interneto vizija AS20080814 @RubenKelevra
12D3KooWKSDCCB1Aw5jV4UoVxcnoCA1rAbjAKKys5KbZqJCsNaFM Mumbai, India Oracle Corporation AS31898 @RubenKelevra
123... ~ Tokyo, Japan AS7506 anonymous

Import Server uptime (since 2020-07-02)

Service Status

Join the Cluster

If you want to join and contribute bandwidth and disk space, feel free to do so. The repo-size is about 70 GB. Since the data is rotating quite quickly it's recommended to enable the Garbage Collector with --enable-gc for your IPFS-Daemon.

The default storage size for IPFS needs to be altered in the config file.

If you're running your cluster follower on a computer with a static ip or a static domain name: Feel free to add it to the list of peer_addresses in the JSON config files, found in collab-cluster-config. Then send a pull request.

Details how to join the cluster are available on the collab cluster website.

tl;dr: You need a locally running IPFS node. Your IPFS StorageMax setting may needs to be adjusted. You need ipfs-cluster-follow, then run:

ipfs-cluster-follow run --init

On low power machines use the following command:

ipfs-cluster-follow run --init

Use the pkg cache with Pacman

Install ipfs on each of your systems, you need set it up and and start it as a service.

Then add the following to your /etc/pacman.d/mirrorlist as first entry:

Server =$repo

It is recommended to turn pubsub on, with the gossipsub router for low latency download starts.

This need the flags --enable-pubsub-experiment --enable-namesys-pubsub on the ipfs-daemon and the Pubsub-Router in the config need to be set to gossipsub.

Since directory lookups may be slower over IPNS, you may need to set the pacman option --disable-download-timeout for fetching dbs or packages.


If you have any questions, feel free to ask in the IPFS-chat on matrix, after consulting the FAQ

Get an ISO/bootstrap image from IPFS

ISO and bootstrap files are also stored on the cluster:

ArchLinux ISO/bootstrap images

IPFS-URL Content
/ipns/ cluster setup domain which holds the config file
/ipns/ regular mirror (without '/os/x86_64/' subfolders in path)
/ipns/ current ArchLinux ISO+bootstrap images