I’m curious if that works with unfederated servers or servers that simple just get shutdown. Ie xyz government decides to raid the servers, (is there redundancy in the data?)
I guess the main challenge would be proving to the new instance that the old offline instance authorized the transfer, maybe something like a keypair could be generated with each account and a signed proof attached to the user profile that gets federated around as other servers receive user profile objects, then provide an account backup function that lets you save the keys as a file so the importing server can verify the key and federate the change of ownership of content to other instances somehow.
Currently, no. Right now you tell your new instance to expect a transfer from your old. Then you tell your old your new instance and if they match, the transfer begins. In your example, you wouldn’t be able to do half the steps needed so it would fail. And since each server is unique, it would be up to them whether or not there were any backups or not.
Thanks! Are the systems standalone or can they be distributed or mirrored? Seems like a potental single point of failure if the instance is literally running on someone’s personal server.
I’m curious if that works with unfederated servers or servers that simple just get shutdown. Ie xyz government decides to raid the servers, (is there redundancy in the data?)
I guess the main challenge would be proving to the new instance that the old offline instance authorized the transfer, maybe something like a keypair could be generated with each account and a signed proof attached to the user profile that gets federated around as other servers receive user profile objects, then provide an account backup function that lets you save the keys as a file so the importing server can verify the key and federate the change of ownership of content to other instances somehow.
Currently, no. Right now you tell your new instance to expect a transfer from your old. Then you tell your old your new instance and if they match, the transfer begins. In your example, you wouldn’t be able to do half the steps needed so it would fail. And since each server is unique, it would be up to them whether or not there were any backups or not.
Thanks! Are the systems standalone or can they be distributed or mirrored? Seems like a potental single point of failure if the instance is literally running on someone’s personal server.