HamSwagwich
It’s showeq.com
I appreciate you checking into it
I just tried this, but it doesn’t seem like it’s really ready for anything but a basic test environment.
When your system creates the service, it does so with the default elestio domain and there is no way to change it from within KBin, therefore your are stuck with a huge security hole and a nonsense domain name that’s impossible for people to remember.
While you can indeed use your own domain name to resolve it, it doesn’t appear that the domain is editable once KBin is setup (which is done automatically, and understandably on the federation side, you can’t have the domain name changing)… so when you set up a KBin on Elestio, you are forever suck with “kbin-???-u5400.vm.elestio.app” as your server name in the Fediverse, which sucks and is really a non-starter.
I don’t want to be @HamSwagwich@kbin-mynewkbininstance-u5400.vm.elestio.app
This appears to have the added effect of making it impossible to use Cloudflare as your proxy, since you get a bunch of 301 redirects bouncing between your resolved domain and the elestio domain, since KBin thinks it’s name is the elastio domain and rediredts you, then our browser thinks it’s going to the resolved domain and redirects you. Boing boing boing
That’s correct and that’s the problem. If a given community server goes down, that community basically just becomes an archive. It really needs to be able to continue without the host instance, similar to how a mesh works. Each remaining server routes around the dead node.
There is also the problem of search engine indexing… If a given server goes down, that information is lost to the search engine, even though it’s still on other nodes.
Which also leads to duplicate content problem for search engines, as ECU m each node of a given community contains the same information for a given post, making it crappy to index and search.
400C is like 650K. Not even close to absolute zero