Hide Forgot
Created attachment 1140745 [details] dnf output Description of problem: Using Cent7 I can use "yum --installroot=/path/to/container/root --releasever=7 -y install @core" to bootstrap a cent rootfs. Doing a "dnf --installroot=/path/to/container/root --releasever=23 -y install @core" on Fedora23 results in the attached dnf output. The bootstraped system is broken. The output looks similiarly broken when I install and try it with yum on fedora23. Note that the systemd-nspawn man page suggests this kind of bootstrapping: Example 2. Build and boot a minimal Fedora distribution in a container # dnf -y --releasever=21 --nogpg --installroot=/srv/mycontainer --disablerepo='*' --enablerepo=fedora install systemd passwd dnf fedora-release vim-minimal # systemd-nspawn -bD /srv/mycontainer Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. dnf --installroot=/path/to/container/root --releasever=23 -y install @core Actual results: See attached file. Expected results: A bootstraped fedora rootfs. Additional info:
Probably rpm team will have a clue.
Works for me on my local F23 installation. But it sounds familiar. We had similar results with an issue of not setting the SELinux context properly when running with SELinux as permissive. Can you please try with SELinux in enforcing, permissive and disabled and check the SELinux logs.
Created attachment 1142666 [details] enforcing
Created attachment 1142667 [details] permissive
Created attachment 1142668 [details] disabled
The file names are self explaining, I guess. There are only a couple of non fatal errors when I try it with selinux disabled. It does not work with selinux in permissive mode though.
Created attachment 1142685 [details] selinux log - permissive
I hope that's sufficient, ask if you need anything else :3
These two SIGABRTs from log seem to be interesting: type=ANOM_ABEND msg=audit(1458383364.495:3849): auid=1000 uid=0 gid=0 ses=40 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=33649 comm="dnf" exe="/usr/bin/python3.4" sig=6 type=ANOM_ABEND msg=audit(1458383410.956:3851): auid=1000 uid=0 gid=0 ses=40 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=33686 comm="dnf" exe="/usr/bin/python3.4" sig=6 Do you have an idea what / when happened?
Which version of rpm do you use? This problem should be fixed since version 4.13.0-0.rc1.4.
I able to reproduce this only with rpm-4.13.0-0.rc1.3 and with that version of rpm I get almost exactly the same output from dnf as the one attached to this bug. But there are no problems with higher version of rpm so I will close this bug as currentrelease. If you have still this problem with the current rpm version then please reopen this bug and specify the version of rpm with which you encountered the problem.