The exam procedure changed recently, and contains now only one single session of 3.5 hours. The topics by themselves are not hard, and should be easily mastered by every sysadmin with some decent experience. Some things are Red Hat specifics of course, so don’t try it if you never worked on a Red Hat (or obviously CentOS) box. What makes it hard though, is that you get a lot of topics you need to handle within a relatively short period. If you need to start diving in man pages to figure out how to configure things, better prepare yourself a bit more.

Oh, and yes, I passed . The RHCE part was pretty good actually, but I was in luck to just have slightly over the needed 70% for the RHCT part though. Thanks to one topic I didn’t master well enough.. oh well.

(If someone needs RHCE training books, send me a note!)

Get twitter or get a life.

Make me a sandwich v2

Tested in bash only… feel free to propose patches for other shells

Ze hebben ook te kort op mijn toestel gestaan op er een backup van te hebben. Ik werk met een sync systeem die ieder uur mijn data van mijn desktop naar mijn server versast, en de laatste versie van die dag komt dan in de normale server backup terecht, waar met rsnapshot een historiek wordt van bijgehouden. Blijkbaar heb ik die foto’s dus geïmporteerd en gewist op dezelfde dag.

Uiteraard heb ik mijn 3 maandelijkse “Empty Trash” net in de voorbije maand gedaan. Of had ik die foto’s commandline en niet via mijn desktop file manager gewist.

Maar gezien dat mijn $HOME volledig in de backup zit, zit dus ook ~/.local/share/Trash/files . En in de daily.5/daily.6/weekly.0/weekly.1 versies van die Trash zaten wel nog mijn foto’s

Handig die backups waar je niet moet naar omzien en die je dus niet kan vergeten.

I am so confused. If I have to pay you that much each year, I might as well go with the solution from IBM or Microsoft like I always have.

(Via Kris)

My daily example: try to sell a Zimbra license for an in house server to a small business, and compare that with a Microsoft Small Business Server offering.

Often I praise the good things when stuff is neatly integrated, and different software talks nicely through different API’s.
It’s something where Microsoft gets things done, where OSS keeps changing the API.

But as allways, it gets ugly when you realise you only have this (proprietary) API to talk through. you only get to patch something with binary patches you select on filtering by a specific error message.

Error and solution conditions which sometimes do not match the exact environment and versions you behave in. This problem was patched in the lateest service pack of the preceding version. D’uh. And that often is not even a regression.

It’s that bad that even with lots of experience, you don’t get to fix the problem by gut feeling – short of logic as usual. You’re really on your own, ’cause the software version is that new that even google isn’t helpfull. Even all those years of experience don’t give the assurance of where to look, what kind of problem you’re facing, and how fast (or not) you will solve the problem.

Just googling one specific error message giving 3 hits in return (at least from here), pointing to some not publicly released hotfix is one of the many possible trial and error solutions.

Luckily it helped.

And now, ladies and gentlemen, cheer, party and dance. My customer has again backups, the first since one week.

Until it will again suddenly stop working that is.

Windows really is full of binary closed crap. And don’t ask how often I needed to reboot. Even installing the bloody hotfix took twice: one to be able to start the patch process (it hang because some windows update process already hung before), and one after the setup, of course.

Windows still needs too many reboots. It even sometimes asks for it after an ip address change.

This work is covered by a Creative commons Attribution-Non Commercial-Share Alike 3.0 licence.