As ajax quite elegantly summed up, due to a series of catastrophic power failures at PSU, where fd.o is hosted, we were down for a good chunk of yesterday. Despite the machines having redundant power supplies, being connected to separate power rails in the rack, which were hooked up to independent, UPS-backed, power supplies, we still (like a good chunk of Portland, and certainly everyone in the PSU machine room) lost our power.

As far as we can tell, when annarchy.fd.o (websites, people.fd.o, cgit, anongit, et al) came back up, power was again interrupted while the ext3 journal was being replayed. When it came up the n'th time, fsck dumped almost the entire filesystem in lost+found, then started saying increasingly unhappy things about the state of the filesystem on its second pass. In the end, we just went with mkfs, and now we have a brand new and shiny filesystem.

It's worth pointing out that even if this was another filesystem, such as /srv, which hosts all project data, we would've been fine, as they're all backed up. But, unfortunately for some, we made a decision a while ago to not back /home up, and didn't advertise that as widely as we should have. So, if you had stuff in annarchy:/home, it's now gone, and I hope you have backups.

Sorry about that. On the upside, I got to see PSU's new and really very nice machine room this morning, thanks to XDC being about 250m away from the PSU machine room, and fd.o is otherwise running fine. We've been talking this week about replacing our ageing hardware, which would also allow for more redundancy as well as better performance from those machines. But we still have no plans to back up /home, so if you put stuff there, please, please keep your own backups (or make sure the Wayback Machine knows about it).

A quick FAQ: the reason all DSA keys have been removed from fd.o and we aren't accepting any new ones is that they are vulnerable to man-in-the-middle attacks if they have ever been used (not just generated) on a system with a predictable RNG: see Steinar's summary of the maths. We're going with precedent of debian.org rejecting DSA keys, and a general desire to be safe rather than sorry. RSA keys are the default in OpenSSH anyway, so I'm not really sure why you'd want to generate DSA.

It's just not our week, I guess. For the past couple of days, gabe.freedesktop.org (www.fd.o, ns1.fd.o, lists.fd.o) has been playing up: it will just mysteriously die with no explanation. It appears to still be alive, but we can't access it at all. Bringing it back up yields nothing suspicious before it dies, so we're suspecting some bad hardware. This wouldn't be too major a problem, bar the loss of the above-named services and dri.fd.o, openclipart.org, and a few other services, if it weren't for other things conspiring against us.

All revision control services (git/CVS/SVN committing and checkout, web access to same), Bugzilla, xorg.fd.o, people.fd.o and planet.fd.o, are on separate machines which are unaffected by this failure. But, unfortunately, we got less than 24 hours notice that our secondary DNS server at MIT was being moved, and would be offline for a while. Two DNS servers gone. Our third DNS server, hosted in the UK, is still servicing requests just fine, but apparently our registrar never noticed the update. So the third DNS server according to the glue records is home.keithp.com, which is also off the air right now for some reason.

So, ns0.bluelinux.co.uk is still serving names, and you can still access some services if you use that, but we're not sure how long until normal DNS is back up, let alone how long it will take us to get gabe fully fixed.

Sorry for the outage, please bear with us. If you think it's frustrating not being able to check www.fd.o, try checking the remote console and watching /proc/meminfo hopelessly spew into the serial console as you are unable to log in ...

fd.o's CVS migration is finally complete. CVS has moved from gabe.freedesktop.org (formerly 'the freedesktop.org machine') to kemper.freedesktop.org, a new secure host. Anonymouse CVS and ViewCVS now run on annarchy.freedesktop.org.

This means that, technically, you should now be committing to:
username@cvs.freedesktop.org:/cvs/reponame
and checking out anonymously from:
:pserver:anoncvs@anoncvs.freedesktop.org:/cvs/reponame

Anything using '@freedesktop.org' or '@pdx.freedesktop.org' is incorrect and may, at some point in the future, stop working. Anything using '@cvs.freedesktop.org' is incorrect, but will probably keep working for the time being. The canonical URL for ViewCVS is now http://webcvs.freedesktop.org; cvs.freedesktop.org will probably keep issuing 302s, but that too may disappear at some point in the future, so update your URLs.

Some code has moved to GIT, too. I can't remember exactly what the process for committing there is, but the process for checking out is:
git-clone git://anongit.freedesktop.org/modulename

The pserver redirection to anoncvs.fd.o wasn't working for a while. I'll post what I did, and see if you can figure out why:
root@kemper:~% iptables -t nat -A PREROUTING -d kemper.freedesktop.org -p tcp --dport cvspserver -j DNAT --to-dest 131.252.208.36:2401

Just as a heads-up, gabe.fd.o, its primary MX, is now running postgrey. Some mails may take a while to get through as MXes back off too quickly, so don't be alarmed. If your mailer throws messages away after getting a 451, that's its fault. Remain seated. Please do not panic.

Aaron writes about the stupidity of DConf. To be honest, I haven't been following the DConf discussions very closely at all, so I can't really comment in any useful way on the proposed standard, the people pushing it, whatever. All I can say is that fd.o is currently two machines with a couple of people have root on, that happens to run some reasonably popular lists. Anyone can subscribe and post, so what one person says, or a few people say, on xdg, might not be reflective of the opinion of the people who will be implementing and integrating whichever technology finally comes about. I don't doubt that a useful technology will come out of this -- if DConf is a failure, so be it. I'm sure that if it is, then something useful will be borne of its ashes.

That being said, I think if any good standard comes about, it will be a very good thing. I think the goal is worthwhile enough to persist with any bumps that may come along the way. But then again, I'm just hot-airing on a couple of planets, not writing any code, so take my views with the appropriate grain of salt. :)

Linux/OSS Technical Project Manager (GNOME/fd.o).

Just looking at the title of that job ad makes me happy.

I have suspended all SSH access to fd.o (well, except my own), thanks to the nasty little DoS without a proper patch circulating (yes, I've seen the one on lkml, and I'm kind of hesitant to apply it). Sorry guys.

There are four new faces in important positions at fd.o. In lexicographical order:

• Adam Conrad - admin. Adam is a 'dedicated' admin - it is the only task he currently performs at fd.o, and he has pledged to take ultimate responsibility for the machine. This should help keep it more secure and eliminate some of the more funky issues with it. Adam earns his food and shelter by being an admin, so hopefully he's alright.
• Chris Lee - release team member. Chris is the other member (along with myself) of the release team, and will work with me on the platform, et al.
• Colin Charles - listmaster. Colin's main work is in moderation of all the fd.o-operated lists, but also does other list-related tasks. This should cut turnaround time down immensely.
• Pascal Hakim - listmaster. Pasc is one of Debian's listmasters, and is currently mainly working on creative spam-reduction techniques, so Colin has to do less work.