Just your overly smol dwagon that serves as the lead network engineer for The MFN Project. Mostly friendly but bitey! le nomf

IT geek, ham radio hobbyist and VR grimlin!

  • 0 Posts
  • 5 Comments
Joined 1 year ago
cake
Cake day: June 18th, 2023

help-circle

  • Better now that I got the upgrade to 0.18.1 to run correctly. Blew up the VM twice last night, restored it again this morning, manually shut down docker and run the playbook again and it just worked this time. Got around to imaging the other parts of the network, broke a non-important router on upgrade. Fix that mess later, brain is fry.

    But I have been given freeze dried strawberries and cheesy crackers from Big Dwagon, so all is good I suppose.



  • I have some headroom for growth set aside. Since my instance is virtualized, its not too hard to scale it a bit. But there are hard limits due to other projects on the host.

    For a lot of smaller instances that are currently running on cheaper VPS instances, they most likely have an upper limit to what their willing to pay for scaling up as growth happens. The only way to balance that is getting tooling in place to purge older data, but that isn’t really a good idea either.

    Really though, any web platform that hits the public eye is going to face these issues over time. But allowing a large company to federate with a smaller instance will accelerate the issues. You also need to keep in mind that you don’t have all the control of these instances, as your users will cause you to federate with more and more content. Sure, you can purge and defederate, but that is a cat and mouse game.

    Also, I cannot speak for the goals of others; but lemmon bar isn’t run with the goal of replacing reddit. It is meant to be a point of access to the fediverse. No more, no less.


  • One of the things that I feel isn’t being thought about much, is that it isn’t just Meta’s ideology and policies that will harm smaller instances and the fediverse itself; but the volume of data that their userbase will generate.

    For smaller instances like mine running on six vcores, 4GB of memory, 512GB storage and a 120Mbps network…I feel like all it would take is a handful of users federating with them and the data flow alone would destroy our resources at the network if not disk level.

    No, I don’t plan on allowing my instance to see or interact with theirs; but the point applies to all small instances and part time hobby servers. We don’t have the means to take on the data they could throw out into the federated network.