On Fri, 2006-04-28 at 16:07 -0700, Robinson Tiemuqinke wrote:
Hi,
Just encounts an odd problem with yum 2.4 on Fedora
Core 4: obsoleted rpms are not cleanup by yum when the
latter is killed and runs again.
I tried to update the system the newest patch level,
it works at the beginning but suddently yum hangs --
most packages were updated with a few left.
I had to kill the yum process and restart again, this
time it runs to the end and upgrade all the left
packages.
But when I tried to see what's in my rpm database, I
found that there are duplicate entries for all
packages upgraded before hang -- Yum doesn't cleanup
the old version info from rpm database because it
hanged and then be killed.
Any one know how to deal with this problem
effectively? I hate to list tens of obsoleted rpms one
by one and feed them to "rpm -e --justdb" command.
I am not sure how to do a cleanup other than by hand.
The cause was killing the yum process before it had performed the
cleanup phase (as you already indicated)
You might try redoing the update one package at a time with a force to
see if that has any effect. Otherwise you are stuck with manually doing
the clean up.
/var/log/yum.log will list those packages that were updated with
date/time so it should be easy to identify the ones of concern.
A simple bit of scripting would automate the processing of multiple
packages for cleanup.
You would need 2 things. A list of the packages of concern complete
with the version that rpm can understand, and a simple script with a for
loop to walk through the list one package at a time and do the rpm -e on
it.
If I don't clean up the obsoleted rpms from rpm
database, then next upgrade will definitely falls into
trouble.
Please help.
Thanks.
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com