Description of problem: This command fails in Fedora 23: sudo dnf -y --releasever=22 --nogpg --installroot=/data/src/blah --disablerepo='*' --enablerepo=fedora install systemd passwd dnf fedora-release vim-minimal This sort of command is used heavily by mock. It is also documented on the systemd-nspawn page. Version-Release number of selected component (if applicable): dnf-1.1.2-4.fc23.noarch rpm-4.13.0-0.rc1.3.fc23.x86_64 How reproducible: Every time Steps to Reproduce: 1. Run above command Actual results: RPMs fail to install, warnings are printed, including: systemd-219-13.fc22.x86_64 was supposed to be installed but is not! error: %prein(systemd-219-13.fc22.x86_64) scriptlet failed, exit status 126 Error in PREIN scriptlet in rpm package systemd Error in PREIN scriptlet in rpm package systemd error: systemd-219-13.fc22.x86_64: install failed Expected results: No failures
Created attachment 1081864 [details] DNF command output DNF command output, including: ... warning: %posttrans(fedora-release-22-1.noarch) scriptlet failed, exit status 126 Non-fatal POSTTRANS scriptlet failure in rpm package fedora-release Non-fatal POSTTRANS scriptlet failure in rpm package fedora-release warning: %posttrans(filesystem-3.2-32.fc22.x86_64) scriptlet failed, exit status 126 Non-fatal POSTTRANS scriptlet failure in rpm package filesystem Non-fatal POSTTRANS scriptlet failure in rpm package filesystem warning: %posttrans(shared-mime-info-1.4-2.fc22.x86_64) scriptlet failed, exit status 126 Non-fatal POSTTRANS scriptlet failure in rpm package shared-mime-info Non-fatal POSTTRANS scriptlet failure in rpm package shared-mime-info warning: %posttrans(nss-3.18.0-1.fc22.x86_64) scriptlet failed, exit status 126 Non-fatal POSTTRANS scriptlet failure in rpm package nss Non-fatal POSTTRANS scriptlet failure in rpm package nss warning: %posttrans(rpm-4.12.0.1-9.fc22.x86_64) scriptlet failed, exit status 126 Non-fatal POSTTRANS scriptlet failure in rpm package rpm Non-fatal POSTTRANS scriptlet failure in rpm package rpm ... coreutils-8.23-9.fc22.x86_64 was supposed to be installed but is not! ... ca-certificates-2015.2.4-1.0.fc22.noarch was supposed to be installed but is not! ... pinentry-0.9.1-1.fc22.x86_64 was supposed to be installed but is not! ... systemd-219-13.fc22.x86_64 was supposed to be installed but is not! ... dbus-1:1.8.16-1.fc22.x86_64 was supposed to be installed but is not! ... libutempter-1.1.6-6.fc22.x86_64 was supposed to be installed but is not! Failed: ca-certificates.noarch 2015.2.4-1.0.fc22 coreutils.x86_64 8.23-9.fc22 dbus.x86_64 1:1.8.16-1.fc22 libutempter.x86_64 1.1.6-6.fc22 pinentry.x86_64 0.9.1-1.fc22 systemd.x86_64 219-13.fc22
Tried with selinux both enforcing and permissive.
Created attachment 1081866 [details] dnf.log
Created attachment 1081867 [details] dnf.rpm.log
Created attachment 1081868 [details] dnf.librepo.log
Created attachment 1081869 [details] hawkey.log
This blocks Cockpit development on Fedora 23
Proposed as a Freeze Exception for 23-final by Fedora user stefw using the blocker tracking app because: This bug prevents development using mock. It prevents teams like the Cockpit team from finishing up delivery of software for Fedora 23.
as DNF reported, the installation of these packages failed: ca-certificates.noarch 2015.2.4-1.0.fc22 coreutils.x86_64 8.23-9.fc22 dbus.x86_64 1:1.8.16-1.fc22 libutempter.x86_64 1.1.6-6.fc22 pinentry.x86_64 0.9.1-1.fc22 systemd.x86_64 219-13.fc22 Looking at dnf.rpm.log: The scriplets of most of the packages were not executed and thats probably the cause of the problem. Reassigning... Possible dup of bug 1262953
It seems to be already fixed in rpm-4.13.0-0.rc1.4.fc23 because I am able to reproduce this with rpm-4.13.0-0.rc1.3.fc23 but not with rpm-4.13.0-0.rc1.4.fc23. Can you confirm that it is fixed?
Discussed at 2015-10-15 freeze exception review meeting: https://meetbot-raw.fedoraproject.org/fedora-blocker-review/2015-10-15/f23-blocker-review.2015-10-15-20.11.log.txt . There doesn't seem to be a clear explanation of why this needs a freeze exception: i.e., why does the fix need to be in the frozen release tree and images - wouldn't just having it in the updates repository be sufficient? If there's a reason this can benefit from a freeze exception, please explain. Thanks!
Somehow my response was not recorded by bugzilla. It appears that with rpm-4.13.0-0.rc1.4.fc23 does in fact fix this issue.
> If there's a reason this can benefit from a freeze exception, please explain. Thanks! The reasoning: In my opinion RPM bugs should be fixed before release, because RPM will be heavily used by people upgrading to the new release. RPM should be rock solid at release time.
Discussed at 2015-10-19 freeze exception review meeting: https://meetbot.fedoraproject.org/fedora-blocker-review/2015-10-19/f23-blocker-review.2015-10-19-16.00.html . There is still no convincing rationale for breaking the freeze for this. Note that upgrades are performed with the RPM (and dnf and so on) from the *source* release, not the target: when you upgrade from F22 to F23, F22's rpm is used, not F23's.
As there is a bug already ON_QA: Closing as duplicate *** This bug has been marked as a duplicate of bug 1268883 ***