Yum & RPM: Recovering from a crash

[Janina Sajka]

RPM Update: Cleaning up from the crash
Reply-To: 
X-Operating-System: Linux concerto.rednote.net 2.6.8-1.541.rootsmp 
Organization: Capital Accessibility LLC (http://www.CapitalAccessibility.com)
X-PGP-Key: http://www.CapitalAccessibility.com/JaninaSajka_gpg_key.html
I was in the middle of a massive 'yum update' when the process simply
hung. It just sat and sat, with about 1,000 rpm packages in and some 300
to go. Yes, this was perhaps an abuse of yum--I realize using yum to go
from 2 to test 2.91 is unsupported.

But, is it possible this could happen in another circumstance when many
packages are to be updated? What's the correct response when this does
happen.

I'm pleased to say that we recovered by manhandling rpm to --nodeps
remove a few packages that prevented subsequent update, then putting
them back. Things seem mostly OK now, except that there are a
significant number of multiple rpm versions in the rpm db. Indeed, rpm's
cleanup didn't really have a chance to run.

Rather than trying to clean up the dups by hand, is there some other
way? Is there some tmp file that maintains the clean data that should
have been set aside?

Any thoughts on this most welcome.

-- 
	
				Janina Sajka, Chair
				Accessibility Workgroup
				Free Standards Group (FSG)

janina at freestandards.org	Phone: +1 202.494.7040