Post

In regards to this I think @matrix.org should offer stable and long-running alternatives in their setup process, most users just go next next next without thinking of anything, and having a place where they are obligated to choose from some long-standing servers would in my opinion the way.

Except that with mastodon your home server doesn’t bind you if you later decide to move on to another instance or even self-host. Migration is painless and you don’t lose any contacts either. That alone takes a lot of stress out of picking where to put your account initially.

And what do you propose exactly? Because I think the Foundation would be very happy not having to maintain a huge flagship server as Mastodon does. But the reality is that many people still need this centralized “step“ toward decentralization.

I doubt they are using btrfs, and I think that's wise. Many people (myself included) have had a bad time with btrfs in production environments, especially with snapshots. ZFS for Linux may be a better bet but comes with a performance penalty.

Yeah I was just responding that it's possible for many "rm" cases. Aware of ZFS license restrictions (shame!). I've had good results with BTRFS but I'm sure ZFS is more stable. I would assume they're using a db partition. That said, I run VMs on BTRFS with nocow and performance is good for my needs.

btrfs is pretty cool, and I use it a lot, but both btrfs - and other filesystems or storage with differential snapshotting - have a performance hit that really hurts databases. On updates, small chunks of data need to be moved, causing fragmentation plus metadata overhead from tracking each piece.

Sorry, but it's bad news: we haven't been able to restore the DB primary filesystem to a state we're confident in running as a primary (especially given our experiences with slow-burning postgres db corruption).

So we're having to do a full 55TB DB snapshot restore from last night, which will take >10h to recover the data, and then >4h to actually restore, and then >3h to catch up on missing traffic. Huge apologies for the outage. Again, folks using their own homeservers are not impacted.