boombatower and i spent quite a bit of time trying to get his testing framework running on the testing.drupal.org server, and ran into numerous problems:

  1. the server is using some custom httpd server package that i've never seen before
  2. we are unable to install php_curl because yum can't find it, even though this should be a standardly available package
  3. $PATH is screwed up -- it was missing /sbin and /usr/sbin, so many standard programs were harder to find. i ended up adding those in .bashrc for root temporarily
  4. yum kept locking up, and at one point we needed to reboot to get past that. i don't think that fully fixed the problem, either.

i found all those issues inside of 20 minutes, and simply gave up trying to do anything else. if possible, i'd love to see us simply back up what we need, and do a clean install of the latest CentOS.

fwiw, i run all of my business machines on CentOS.

Comments

hunmonk’s picture

Title: testing.drupal.org server hosed » testing.drupal.org down, Centos repair needed
boombatower’s picture

Interestingly traceroute to the box fails and the apache errorlog shows no issues.

Talked with amazon to confirm backup procedure:

  • Drupal 5 database being used for testing.drupal.org
  • Related files, while ignoring the large number of testing instances.

I'll try to get this done tonight and figure out where to put backups, but I'm currently working on setting up an internal business server as well...so we'll see how time plays out.

chx’s picture

The tarballs Jimmy gave me are sitting safely on NowPublic's rsync.net account. It does not get safer than that.

boombatower’s picture

Next phase, someone needs to re-install box.

Amazon’s picture

Traceroute to http://testing.drupal.org no longer resolves to the IP: 88.208.239.80

Amazon’s picture

Assigned: Unassigned » Amazon

With ps |grep httpd, we found two apache servers conflicting. Chx is removing the OS server in favor of Apache2.

chx’s picture

Nope. There was one http running but that was not properly configured. I killed that and went to /usr/local/apache2 and started the proper Apache which was unmistakenly set up to run this Drupal.

Amazon’s picture

Status: Active » Fixed
boombatower’s picture

Server ran a patch successfully...I need rest...this is great!

http://drupal.org/node/321962
http://testing.drupal.org/new/pifr/node/1/321962

Will make detailed blog post on what is left to get reporting working, etc...plus analysis of how server handles.

Anonymous’s picture

Status: Fixed » Closed (fixed)

Automatically closed -- issue fixed for two weeks with no activity.