Red Hat Bugzilla – Bug 23813
anaconda traceback, Device or resource busy
Last modified: 2007-04-18 12:30:39 EDT
During a text Local CD-ROM install, received this dialog message partway
through installing packages:
The file /mnt/source/RedHat/RPMS/traceroute-1.4a5-25.i386.rpm cannot be
Pressed OK, and it appeared to retry it. But then later, when installing
another RPM, anaconda crashed with the attached traceback. Machine is a
486DX-100 with 48MB RAM and an IDE CD-ROM drive.
Created attachment 7445 [details]
Assigning to a developer.
Did you use VC2 for anything? For some reason the installer can't unmount the
first cdrom to eject it.
What's VC2? I did a "custom" install and I can provide a list of the options I
checked if that will help.
How easily can you reproduce this? (VC 2 is virtual console 2, the Alt-F2
This defect is considered MUST-FIX for Florence Beta-3
No, I've never switched virtual consoles during Flo beta-2 installs.
With regard to reproducibility, I've tried several times in the past few days,
and it dies in different ways at different places, leading me to believe this
may be hardware. I haven't been able to reproduce the anaconda traceback.
So, I swapped out the Mitsumi 8X IDE CD-ROM with a newer Mitsumi 32X IDE
CD-ROM. During my first attempted install, after a few CD swaps, I get "install
exited abnormally -- received signal 11". It's worth pointing out that a
network-based install went fine on this machine, so I'm suspecting the CD-ROM or
IDE bus as the culprit.
Let me know if there's more information I can give or if there are other things
you want me to try with this hardware setup.
This is probably flaky media, but it's hard to know for sure. The sig11 popping
up occasionaly is normally a media or hardware problem though, esp as nobody
else is reporting this.