These install methods will lead to crashes (python exceptions) if there are files in the /RedHat/RPMS directory which are not valid RPM files, or are misspelled files. Installer should really handle this a little better. Best case of this is when retrieving files from CDROM and getting the TRANS.TBL file, or when using WSFTP and getting the ftp log in the directory.
I am seeing the same crashes as per 5618, 5875, 5546, 5551, and 5615. My system has only formatted FAT32/VFAT and formatted ext2 partitions. The rpms directory is on one of the FAT32 partitions. Running "ls -a | grep -v rpm" shows that the only non-RPM entries in this directory are "." and "..". ------- Additional Comments From 11/07/99 16:02 ------- This is a real nuisance if you downloaded redhat form a mirror site that automatically creates ls-lR files. Please, when writing utilities like anaconda, include diagnostics. If there was a filename with the error, I could have figured it out.
*** Bug 6808 has been marked as a duplicate of this bug. *** I upgraded my Redhat 6.0 system from local hard drive. Somehow, when I was ftping all the RPMs a few went missing, and so my copy of Red Hat was incomplete. The upgrade script crashed in a confusing way, throwing incomprehensible Python stack trace. It would have been more helpful to have an English error message. I believe this bug is in the same general class as #6159: a corrupted RPM repository not handled gracefully by the installer.
I think that this is something that should be mentioned on the 6.1 gotchas page. It certainly is one!
This issue is resolved in the latest beta.