Maximum number of peers?

General
1

Hi all,

I want to share one folder containing around 400 files and weighing around 5Go with as many people as possible, through a snowball effect.

Is there a maximum number of peers I can share to? For example with Resilio it’s 100.

Thanks,

2

It’s mainly limited by your system’s resources. As every connection involves tracking the state of the remote device’s files, partly in memory. But there is no hard limit programmatically burned into Syncthing.

On a modern mid-range desktop PC, I would expect a folder as you indicated to be easily manageable with 100 devices. Maybe more, watch your system performance monitors.

3

Syncthing is not a “and now make this data public/accessible to anyone” type of distribution mechanism.

You are probably better off with something like torrents.

2 Likes
4

Hi,

thanks for your answers.

#Limit. OK so 100 would be a technical number to work with. But then, this 100 limit is for each individual peer. So we can still have thousands of peers sharing the folder, it’s just that each of them will be connected to “only” 100 other peers.

#Torrent. This is good for static files. But I need to synchronise an entire folder because the content will be updated almost daily with new files, as information comes in.

#Public. Yes this is what I need. Something like “here the sync key, join the peer network”. And as each peer is able to share the sync key with its own contacts, we pass the bus test, it’s truly decentralised.

Thanks,

5

This is not how syncthing works. Any one of those peers will be able to delete all the files, and the deletion will propagate everywhere.

There is no “sync key”. To add a new peer, some device will have to add a remote device first, and then share the folder. This doesn’t mean that this device will start getting data from everyone, it will keep getting data from that one device. If that device is offline lets say, the peer will never get any updates.

6

I would user a GitLab or GitHub repo and add Collaborators.

This also works with huge repos (greater than your 5 GB) and a high update rate (hourly) as COVID-19 Dataset by Our World in Data shows for example.