I'll be kicking the tires of the bsd.network instance in about 15 minutes, so we will have some downtime.
hopefully, this will address some delays in delivering of messages to other instances.
bsd.network is going down now, we'll see you from the new datacenter.
the server that bsd.network is on will be moved to another datacenter tonight between 10pm-2am Pacific time (5am-9am UTC, 7am-11am CEST).
I'll be powering off the system about 15 minutes before the maintenance window starts, and will bring us back up when the server is reachable again.
Olrite, the bsd.network #instance is back up.
We were offline for about 12 hours, so there is a *long* series of queues to fill our series of tubes.
Database migrations were totally broken, and was a nightmare to deal with. As was the usual "node uses C++ in horrific ways".
In around 1 hour, we will be bringing down the bsd.network instance for upgrades and maintenance.
If things go well, we're offline for 2 hours. If things go less well, 8 hours.
/me puts Admin Hat on.
The bsd.network instance has a new Code Of Conduct, viewable at https://bsd.network/terms or via the "Terms and Conditions" link that is sprinkled around a few places.
The TL;DR is "be cool unto others". However, since society at large has demonstrated they can't be cool, we had to write out a bunch of things.
I'm bringing down the bsd.instance now to do some upgrades and shuffle disks around. See you on the other side!
Good morning party people.
The server that the bsd.network instance is on, will be moving datacenters today. Expect some downtime this evening (European time), or morning/early afternoon (California).
We've mostly recovered from the bsd.network outage.
Looks like the last few actions before the server stopped responding were lost , but nothing else was.
The firehose is currently loading the Fediverse that happened. Toots and actions may arrive out of order.
Your normal Tooting schedule will resume at the toot.
I just committed this to #OpenBSD -current:
Back in the history of time, IPv4 had classes of addresses. This was widely acknowledged as a failure. At the same time IPv4 classes were declared a
failure, #IPv6 decided to add
them back because using a mac address for IP address configuration was easy.
Now that we have RFC7217 support we can remove this artificial limitation: allow non-/64 prefixes to be configured by SLAAC.
Taking down bsd.network in ~10 minutes, to try to fix things. Hope the downtime isn't too bothersome.
OK, I'm moving off mastodon.social to make room for more folks.
Follow me to @mwlucas!