From @jantypas on Sun May 29 2016 01:49:15 GMT+0000 (UTC)
A question for the IPFS experts. I am looking at remote IPFS nodes. In some locations, we may have local “IPFS clusters” on Wifi, but the long-haul links are spotty. Consider the following usage case:
A local school district might have multiple IPFS nodes which communicate across local Wifi. Everything works just fine.
The school districts will connect twice a day on long-haul IP links. I am assuming during that connection, the IPFS nodes will synchronize
Can I do this?
Copied from original issue: https://github.com/ipfs/faq/issues/127
From @mattseh on Sat Jun 18 2016 16:01:14 GMT+0000 (UTC)
What do you mean by synchronize? As long as they know of each other and the links work, they can communicate and exchange data if someone requests it (or automatic DHT propagation).
From @jantypas on Sun Jun 19 2016 01:04:40 GMT+0000 (UTC)
I am thinking of the following scenario. This is a very rural school
in Africa. They may only have mobile connections to the net, say,
twice a day. Off-net, they’d be living off a cached-version of IPFS
until the twice-a-day synchronization.
------ Original Message ------
From: “mattseh” firstname.lastname@example.org
To: “ipfs/faq” email@example.com
Cc: “John Antypas” firstname.lastname@example.org; “Author”
Sent: 6/18/2016 9:01:15 AM
Subject: Re: [ipfs/faq] IPFS across temporary links (#127)
What do you mean by synchronize? As long as they know of each other and
the links work, they can communicate and exchange data if someone
requests it (or automatic DHT propagation).
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.
From @Kubuxu on Sun Jun 19 2016 05:37:19 GMT+0000 (UTC)
@jantypas it could work, you would just have to know sites you would like to synchronize.
Then as connection comes in, you would requires the m from IPFS network.
Also for now I would wait as we are still working on making IPFS less transfer hungry.
From @jcalfee on Mon Jun 20 2016 13:22:16 GMT+0000 (UTC)
Wouldn’t you take all the hashes you wanted to sync and grab that file then PIN them in somewhere in the network that needed to download them?