For hours today, yesterday, and intermittently since I’ve been on here (a month or so) lemmy stops loading using jerboa, liftoff, and a web browser.
The site isn’t listed on any down detector I know of, and each app gives different errors but ultimately just won’t load.
It’s often enough that recently when I consider jumping on here I just don’t because there will probably be an issue. Ranging from not loading, JSON errors, or just blank screens and my comments not working…
What’s going on? Is there a status page for these places?
It’s not that we can not handle the load caused by users or by the amount of communities. It’s because of DDOS attacks and even with cloudflare some of these attacks are challenging due to the way lemmy works. No instance is safe from these attacks but the bigger instances get targeted. We weren’t the only instance that went down today.
I don’t think we do anything wrong here? None of us are being paid, we all put in a lot of time and effort to keep things running. You don’t know how many passionate people are involved “behind the scenes” seeing you call out the admins of lemmy.world.
You can’t please everyone, and some people will always find a stick. But I still think a lot of people believe in our team, our policies and what we are trying to do here. If that’s not your thing, fine, you can look elsewhere.
I, for one, am glad to be on a server that isn’t run by tankies, nazis or some other crazy fuckos.
Is there a post or blog somewhere that goes into detail about why DDOS is such an issue with lemmy or activitypub? Or are you saying that about DDOS in general?
Then don’t work to become a big instance.
If not you, someone is profiting from this
I’d love to, except lemmy.world went on to a huge land grab, cloned every possible popular community on reddit and is not giving any signs that will stop. Almost 50% of the user base is unreasonable and it goes against the ethos of federation and decentralization. An instance going down should not be newsworthy, but because it’s so big (relative to the others) it introduces systemic risk and approaches “too big to fail” status.
I shouldn’t be the one telling you.
I am sure that it has nothing to do with you having any financial gains hosting instances. And that you do it all for the good of federation: https://lemmy.world/comment/1510374
Yes you would do things differently if you were in our position.
@ruud@lemmy.world is very open about what comes in from donations in his monthly blog posts. He even links those in !lemmyworld@lemmy.world. Have a look: https://blog.mastodon.world/june-2023
He has the trust of every one involved. We are very thankful for the community supporting us as they do and it’s not because income is more than expenses now that it will stay like that. We expect donations to drop off at one point. But whatever happens we will always be open about this. Even you linked to information that is freely available.
Facts. Let’s not entertain the troll, everyone. They’re just trying to make money off you, just check their instance.
Yes for both, without sarcasm. I don’t think that the donation-based model is healthy or sustainable and I would rather see more service providers like mine.
Actually, I like to see more providers that can make real money and prove that this is feasible. I’ve been running communick for more than 3 years already, and it has been nothing but a small money pit. The managed hosting side of things is just barely breaking even.
You just sound salty now that you quit your job to start a fediverse hosting company and it’s not working out the way you want. Donation-based models have been used for ages and it worked for mastodon.world so why shouldn’t it for lemmy.world? If donations and interest decreases we can always downscale.
Sorry you’re not breaking even, seems like running a managed hosting service for lemmy is not feasible
Maybe I wasn’t clear on the blog post. There are two objections to donation-based funding:
You might not see it that way, but my argument is that relying on donations hides the true costs of running the server from the users and (like in ad-funded business) distorts the “market” in a way that makes the overall system less efficient.
The instance itself did not do a “big land grab,” users on the instances made the communities. And, as you should know, the fact that there’s a community on one instance doesn’t mean that the same topic can’t be on another, there are several of those kinds of duplicates.
I signed up for .world because I liked the policies, it didn’t seem to be heavily communist or hosted in an authoritarian country, and it seemed to be robust. Nobody told me I should make my account there; I saw zero advertising. I’m not sure what you think the admins did to make other people settle there.
And the fact that some people are donating to it in no way means they’re making anything like profit. The admins didn’t make a plea for me to donate anywhere that I saw, other than having the link in the sidebar, like many/most instances.
You seem to be taking frustrations out on people who don’t deserve it. If the stability problems become an issue, people will just make accounts elsewhere.
They opened the gates and let people come in without knowing if they were able to handle the influx of people. By presenting themselves as a place that could welcome everyone, they end up robbing the opportunity for other instances to share the load and to absorb part of the user base. This is what I mean about “land grab”.
A more sensible approach would be to have a feedback loop where they open up a limited number of spots, fill them, see how their instance and the overall fediverse behaves and adjust based on that new information.
You leave out the fact that @ruud was already running mastodon.world before all this. So he does have experience running a big instance. He had a team of moderators from mastodon.world that helped from the start.
The influx of people was never a problem, if you choose the right hosting provider you are prepared for these things. And the hosting company we use provides all those tools to help us grow. We started with a small server at Hetzner.de and gradually upgraded when it was required. They have no limits on bandwidth so that is also something Ruud looked at.
Anyway, you have a lot of say about how you would do things but you had a 3 years head start…
It’s not a technical problem, but a systemic one. Getting way too big relative to the rest of the fediverse paints a target on your back. There is a reason your instance is being DDOS’d while so many others aren’t and one instance being DDOS’d shouldn’t be have such an impact on the overall system.
My point is that the sensible thing to do would’ve been to limit growth of .world and let others catch up. This is what the lemmy devs did with their instance, this is what Hugo from masto.host did to his service (stopped accepting new customers when he got close to 50% of the users) and this is even what Eugen did with mastodon.social and mastodon.online in the beginning.
I agree with your point of view but there is nothing that can be done about it.
It does feel sad to see one giant instance have almost all users and all traffic for me too. I was hoping it would become a proper decentralized platform with hundreds of islands of different servers filled with people and communities.
But fine, we don’t always get what we want. I’m disappointed but will keep using Lemmy anyway. It’s not a big tech service at least which is wonderful, and most people are nice.
Yes, from my comments it seems that I am criticizing the people working on lemmy or trying to paint them in a bad light. I am not saying that what they are doing is wrong, just misguided.
And I totally agree, at least this is not (and will not be) owned by Big Tech.