Bug 1024692

Summary: yum doesn't work after dnf upgrade
Product: [Fedora] Fedora Reporter: Petr Spacek <pspacek>
Component: yumAssignee: Packaging Maintenance Team <packaging-team-maint>
Status: CLOSED INSUFFICIENT_DATA QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 19CC: admiller, akozumpl, ffesti, firas.alkafri, packaging-team-maint, pnemade, rholy, zpavlas
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-10-31 10:54:02 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
transcript from console none

Description Petr Spacek 2013-10-30 09:24:20 UTC
Created attachment 817368 [details]
transcript from console

Description of problem:
DNF leaves rpmdb in inconsistent state so YUM refuses to work with the database.

Version-Release number of selected component (if applicable):
dnf-0.3.11-3.git7bdc9e1.fc19.noarch

How reproducible:
?

What I did:
1. $ dnf upgrade
2. $ yum upgrade

Actual results:
YUM upgrade failed:
$ yum upgrade
Loaded plugins: auto-update-debuginfo, langpacks, refresh-packagekit, versionlock
Error: Package tuple ('ghostscript-cups', 'x86_64', '0', '9.07', '15.fc19') could not be found in rpmdb

Expected results:
It should work...


Additional info:
$ yum clean all
solved the problem.

See console.log for all the gory details.

Comment 1 Ales Kozumplik 2013-10-30 12:40:25 UTC
Hello, as far as I can see from the output, ghotscript-cups was obsoleted and properly removed during the first transaction (performed by DNF in your case). So Yum shouldn't be looking for it---it is probably a problem of invalidating the Yum's cache when the actual state of RPMDB has changed.

Comment 2 Zdeněk Pavlas 2013-10-31 10:54:02 UTC
Although it seems to be the most likely reason, I doubt dnf could run and finish a rpm transaction in less than 1 second after yum has saved the cache. Might be a FS issue ("version" file was updated, but "pkgtups-checksums" somehow was not). yum clean all" runs "yum clean rpmdb", so now we can't investigate the root cause.

If it happens again, please reopen and attach a tar.gz of the /var/lib/yum/rpmdb-indexes/ directory (it should be quite small, under 100k)