From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; Q312461) Description of problem: This may be related to the bug 74726, as I had that problem happen quite a few times during manual upgrades (ie rpm -Fvh rpmname). Here is an example list of RPMS: LPRng-3.8.9-6 MAKEDEV-3.3.1-2 Omni-0.5.1-3 Omni-0.7.0-6 Omni-foomatic-0.5.1-3 Omni-foomatic-0.7.0-6 PyXML-0.7.1-6 SysVinit-2.84-5 VFlib2-2.25.6-8 XFree86-Mesa-libGL-4.2.0-72 XFree86-font-utils-4.2.0-72 XFree86-libs-4.2.0-72 XFree86-truetype-fonts-4.2.0-72 XFree86-xfs-4.2.0-72 a2ps-4.13b-24 alchemist-1.0.24-4 anacron-2.3-23 apache-1.3.23-14 aspell-0.33.7.1-16 at-3.1.8-23 at-3.1.8-31 authconfig-4.2.12-3 authconfig-4.2.8-4 autoconf-2.13-17 autoconf-2.53-8 automake-1.4p5-4 automake-1.6.3-1 basesystem-8.0-1 bash-2.05b-5 bc-1.06-10 bdflush-1.5-21 bind-9.2.1-1.7x.2 bind-9.2.1-9 bind-utils-9.2.1-1.7x.2 bind-utils-9.2.1-9 binutils-2.11.93.0.2-11 binutils-2.13.90.0.2-2 bison-1.35-1 bison-1.35-4 byacc-1.9-19 byacc-1.9-22 bzip2-1.0.2-2 bzip2-1.0.2-5 bzip2-libs-1.0.2-2 bzip2-libs-1.0.2-5 chkconfig-1.3.5-3 chkconfig-1.3.6-3 Version-Release number of selected component (if applicable): How reproducible: Didn't try Additional info:
Did you ^C while updating? In rpm-4.1 packages are erased at the end of a transaction, what's confusing is that the (install) progress bar says 100%.
For the most part I was unable to use ^C, I either had to use ^Z and kill the process or login on another terminal and kill -9 the process. Also, when it was hung doing what it was doing, I was able to rpm -Fvh another rpm file and it would install.
Did you kill rpm while progress bar was at 100%? If so, then please see bugzilla #73200, multiple copies of headers will be left on abnormal termination, clean up by erasing (with --allmatches) and reinstalling packages.
Chris, did you kill the upgrade process? If so this report can be closed a dup.
Closing due to inactivity. Please reopen if you have further information to add to this bug report.