I should put one of these on my desk… Let’s see how many questions I still get then 🙂

Got my Fosdem LPI exam results earlier today… 620 points! (500 required to pass). That’s an average of just a little over 75%. Yay ! 🙂

Internet Explorer thought Seamonkey’s icon was a lot nicer it seems… 😉 

A note to employers .

Nice reading, beginning of the post is quite similar to "Time management for System administrators " intro text. Geeks live and breathe "geek". 🙂 

If you’re running Debian Sarge, have a >2TB ext3 volume and your fsck.ext3 exits with signal 11 leaving the damn thing all dirty and stuff, get 1.39+1.40-WIP from backports.org . This fixed the problem for me on my home box (2.4T) and one of the servers here at work (4.7T). Takes a long while to fsck, but at least it gets through instead of crashing somewhere half way!

– ext2 doesn’t have a journal 

– ext3 is so "default". AND it gets b0rken just as well; on my home fileserver there were errors and fsck.ext3 exited with a SIGSEGV. Yikes. Upgrading to the 1.39-1.40 release from Sarge backports made that go away, luckily, but it’s still icky that it just broke like that and kept on humming along at the same time.

– XFS turned out to be really, really slow with many small files.

– ReiserFS only gave me corruption, headache and a murderer as an author 😉

– ZFS? Quoting Andrew Morton, "I haven’t seen a patch for that yet, but it’s very welcome!" 

What to do, what to do… 

Help this friendly man get his bike back . 🙂

Yummy 🙂 This is Dedibox ‘s colocation facility. They put 2 Via 2000+ CPU boards in custom 1U rack chassis, and lots and lots of them; and offer el cheapo dedicated servers.

It’s Jan Becaus !

Right after my previous datacenter power loss blogpost … Read this then this 🙂 Save one datacenter, forget the other.

NIKHEF powerdown cause appears to be that they didn’t check the utilisation of their circuits and just blew all the breakers when tripping them. Oops.