Is there anything built in to Lemmy that would allow a takeover if a big instance went under?

No, and that’s why I treat every account I open on a single admin account as a throwaway that could disappear tomorrow.

I respect single admins, they do a great job, but I also know that life can happen.

I doubt it. You really are relying on the word of another person in that situation. I feel like a best case scenario is multiple admins with backups.

If it’s only been a few weeks though I wouldn’t be too concerned. Chances are the admin isn’t making their living off of Lemmy so they could be preoccupied with work, family, a vacation, or some other thing.

You’d probably be able to tell if a Lemmy instance is coming back based off of how long it’s been around and if there’s some kind of stickied post about what they want the instance to be.

Probably like me and our instance, it runs as an extra on a company server, there’s no risk of it going down, it’s negligible for costs.

It’s just been one of the most popular months for vacation, gone on vacation, will update Lemmy instance to latest version when get around to it.

If your only lemmy account resides on that instance, then I would encourage you to download LASIM (Lemmy Account Settings Instance Migrator) (source: https://github.com/CMahaff/lasim ) and run that to export all your subscriptions. I’d also encourage you to set up one or two more lemmy backup accounts on other instances and import your user subscriptions and settings to those other instances. Doing this now, while your current instance is still running, would save you a lot of pain and suffering later if that instance you are using is just shut down or deleted.

Since lemmy instances are owned fully by their admins, there’s no concept of allowing a hostile takeover of someone else’s instance. Just the thought of this is a little absurd and would already have been exploited for wrongdoing if it were possible.

Mastodon has created a covenant that requires any participating instance to have at least two admins with access to the servers for these type of situations. It also requires instances to give three months notice in case the instance shuts down. Unfortunately it does not have any way to enforce it and there has been cases of admins that broke their pledge, but it does help in bringing some of the admins together.

Lemmy instances are no different to any other website, in this regard. To ‘take over’ an instance would be to take over hosting of a website - which would mean either re-pointing the DNS somewhere else (and getting a copy of the database), or to take over the hosting of it (e.g. if it is hosted with a cloud provider, or you are physically taking possession of the hardware).

Taking over the DNS in nearly any gTLD or ccTLD (short of some kind of compromise at least) generally requires one of the following: 1) a process initiated by the registrant, or 2) proving that you are the registrant, or 3) proving you owned the trademark in the domain before it was registered or 4) waiting until the domain expires, any grace period is up, and then being first to register the domain.

If the admin is completely gone, and they are the individual owner of the instance, you could wait for (4) and try to drop-catch the domain. But domains are generally registered for a minimum of 1 year, and often up to 10 in advance, so it could be a long wait. And even then, you wouldn’t have a copy of the database. It is quite possible the actual hosting of the instance has not been pre-paid for anywhere near that long (or might fail in any number of ways, or fall out of date and get compromised, or need some kind of manual intervention following a problem), so it could go down and not come back up a long time before the domain name expires.

If the admin has made same arrangements in advance for takeover of the instance in case they are unable to continue, the picture can be a lot better. They might, for example, if they created a legal structure (company, not-for-profit organisation) for the instance, and that is the registrant and owner of cloud resources etc…, then maybe other members of the organisation could call a Special General Meeting (or whatever similar procedure the org’s rules set up in advance say), appoint a new president, and then the new president can prove their authority to any providers involved to get access to the organisation’s resources (e.g. hosting server). Or maybe they could set up a dead-man’s switch system to automatically email credentials for all the resources if they don’t check in for a couple of weeks; or give a few trusted people the credentials (possibly encrypted with something like Shamir’s Secret Sharing to ensure n of m trusted friends - e.g. any 7 of 10 or something, need to combine their secrets to get the credentials) to take over the instance. But any of those things would require they have planned for that eventuality in advance.

Otherwise, of course, all users of the instance could just chose another instance (possibly posting a last message with a public key to that instance, to establish the link to their new account).