Strange behaviour from http install

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]



The Symptoms;
=============

While watching a CentOS5 install hammer my webserver ( i used "askmethod' to point the machine to the webserver) i noticed a bunch of 404s;

10.1.1.119 - - [22/May/2007:15:01:53 +1200] "GET /cd/1/CentOS/termcap-5.5-1.20060701.1.noarch.rpm HTTP/1.1" 404 319 "-" "urlgrabber/3.1.0"

and the installation failed to get started (the same file re-requested over and over)

The Cause:
==========

I had set my webserver up to serve the iso for centos 5 on /cd/0 and the sme_server7 iso on /cd/1 and the centos 4 iso on /cd/2 (/cd/3 is there, but not mounted to anything)

Even though i used
server : 10.1.1.8
path : cd/0
the centos 5 installation (anakonda?) seems to have gone up a folder, seen the 1,2,3 dirs and asked for .discinfo on the following paths;
/cd/0
/cd/1
/cd/2
/cd/3

Upon finding .discinfo in folders /2 and /3 it has then got all confused (not enough to crash as both sme server and centos 4 are familiar file structures i guess?) and then asked the server for wrong versions of files (again and again in an endless loop)

Removing the confusion ( umount /var/www/html/cd/1 ;umount /var/www/html/cd/2 ) solved the problem and everything went smooth after that

Just thought i would let people know :)
Out of curiousity, does anyone know the cause of this? I imagine it is some kind of fuzzy logic to do its best to recover from paths that are one folder too short/long, or maybe to make sure it is installing the highest version numbered folder?


MrKiwi






_______________________________________________
CentOS mailing list
CentOS@xxxxxxxxxx
http://lists.centos.org/mailman/listinfo/centos

[Index of Archives]     [CentOS]     [CentOS Announce]     [CentOS Development]     [CentOS ARM Devel]     [CentOS Docs]     [CentOS Virtualization]     [Carrier Grade Linux]     [Linux Media]     [Asterisk]     [DCCP]     [Netdev]     [Xorg]     [Linux USB]
  Powered by Linux