mirror.inode.at

This is a mirror-server located in Vienna, Austria, at the Interxion Data Center in the 21st district.

It currently hosts the following projects/files:

clamav:	http
cpan:	http ftp rsync
debian:	http ftp rsync
exim:	http ftp
fedora:	http ftp rsync
gentoo:	http ftp rsync-source rsync-portage
linux kernel:	http ftp rsync
linuxpackages:	http ftp rsync
packman:	http ftp rsync
opera:	http ftp rsync
slackware:	http ftp rsync
suse:	http ftp rsync
ubuntu:	http ftp rsync
random stuff :	http ftp (grml, gnustep, etc)

Bandwidth and rackspace is sponsored by ~~inode~~ UPC Austria

Stats about the server: munin and the /server-status.

Running a server yourself? Be sure to tune your TCP settings for maximum efficency.

In case of problems contact us at mirror@inode.at.

Current bandwidth utilization 66.13 Mbit/s

Log:

23.01.2009: Over a week of downtime due to very undeterministic failure behaviour of the server and lack of troubleshooting ressources. The mainboard was eventually replaced and everything is back in service again.
29.11.2008: New hardware, new luck. A new homepage with more detailled information is coming too. Eventually. Possibly. Maybe. (Volunteers wanted!)
27.03.2006: Increased outbound bandwidth to 300mbit.
17.03.2006: Disabled deeplinking of the wow-patches since the resulting traffic got intolerable.
17.03.2006: Dropped hurd-i386, s390 and hppa from the debian archive due to diskspace constraints. Sorry if this affected you, I'll readd them as soon as I've got new hardware.
28.02.2006: Replaced the drives in the volatile array with new ones. Doubled it's capacity from 36 to 72GB.
05.01.2006: Switched from boa/apache to lighttpd. Finally a clean, fast and extensible webserver which is actively maintained!
28.10.2005: Added a raid0 reiserfs scsi partition to ease the strain on the server due to the idi^H^H^Htraditional build system of gentoo. I lovingly dubbed it "volatile", lets see how long it'll last ;).
12.10.2005: Link upgraded to 200mbit. Things should be a bit faster now ;).
08.02.2005: Slackware 10.1 was released. Terror ensued. Server had a few performance problems (see hotsanic); there will be hardware upgrades eventually.
02.12.2004: Excessive downtime yesterday and today was caused by a multiple disk failure in one of the raid 5 arrays. The exact reason of the crashes was not clear at first, because the software-raid5-stack in 2.6.9 decided to go on a suicide loop when it hit a bad sector while resyncing disks, leaving me without log messages (/ got remounted readonly) and flooding the kernel log buffer with errors. For the curious: the drives in question were two Samsung SP1614N in an array of four disks. Given the fact that I already replaced one of the disks in this array a few months ago, I can only strongly suggest that you stay clear from samsung drives at the moment.