Red Hat Bugzilla – Bug 111906
anaconda error with existing Solaris partion on disk
Last modified: 2007-11-30 17:10:34 EST
Description of problem: Attempting to install Fedora Core 1 onto a disk with multiple existing operating systems, one of which is Solaris. e.g. /dev/hda1 - Windows 2K /dev/hda2 - Solaris /dev/hda3 - extended container /dev/hda5 - RedHat 7.3 /dev/hda6 - swap /dev/hda7 ... other Linux /dev/hda15 - free space for Fedora Anaconda detects the Solaris partition as an unformated Linux Swap (they both share the same #) and offers to format it. If you say no, install continues. Then in DiskDruid, that partition is STILL selected to be reformatted. If you change the selection to Leave as is and exit DiskDruid, you receive an error box: "Error: Informing the kernel about modifications to partition /dev/hda2 - Device or Resource Busy. Click continue and a little while later, anaconda dies. The abbreviated traceback is: /usr/bin/anaconda line 1120 in ? /usr/lib/anaconda/text.py line 474 in dispatch.gotoNext dispatch.py line 236 packages.py line 527 partedUtils.py 697 ... Version-Release number of selected component (if applicable): Fedora Core 1 Install Disk How reproducible: Steps to Reproduce: 1. Configure system with multiple OS (Solaris as /dev/hda2) 2. Start install of Fedora 3. Indicate NOT to reformat 'unformated swap partition' /dev/hda2 4. In disk druid, set file options 'leave unchanged' for /dev/hda2 and /dev/hda15 for mount point / for Fedora 5. Bombs Actual results: Install fails Expected results: Install succeeds Additional info:
can you record the entire traceback and post it here?
Nope - unfortunately, I worked around the problem by installing to an empty disk, and then ghosting that over. However, I have to tell you, the only thing in that traceback was those .py scripts at those line #s. No data, nothing... -----Burton
Does this still happen wiht Fedora Core 2?
Closing due to inactivity. Please reopen this bug if you have further information to add to the report.