On Thu, 23.04.15 19:00, arnaud gaboury (arnaud.gabo...@gmail.com) wrote: > On Thu, Apr 23, 2015 at 4:47 PM, Lennart Poettering > <lenn...@poettering.net> wrote: > > On Thu, 23.04.15 14:18, arnaud gaboury (arnaud.gabo...@gmail.com) wrote: > > > >> > Pick one: > >> > > >> > a) download the raw image and use that, but it will be a loopback file > >> > with its own file system inside > >> > > >> > or: > >> > > >> > b) do the dnf/yum install root thing, and install it into a directory > >> > tree. > >> > >> I installed yum package on Arch but couldn't manage to do the install. > >> > >> # yum -y --releasever=22 --nogpg --installroot=/var/lib/machines/enl > >> --disablerepo='*' --enablerepo=fedora install systemd passwd dnf > >> fedora-release-server > >> Error getting repository data for fedora, repository not found > >> # yum repolist all > >> repolist: 0 > >> > >> In fact, /etc/yum/repos.d is empty, so I am not surprised. > > > > Of course, you could use the .raw image, mount the extern btrfs volume > > into it via nspawn's --bind= switch, then use yum inside of that > > container to install into the btrfs volume. Then get rid of the .raw > > image again, and you still have the btrfs volume that should be > > bootable. > > > > A bit complex, but you almost were there already... ;-) > > Wunderbach > > # systemd-nspawn -M Fedora-Cloud-Base-22_Beta-20150415.x86_64.raw -- > bind=/var/lib/machines/enl:/mnt > [root@Fedora-Cloud-Base-22_Beta-20150415 ~]#dnf -y --releasever=22 > --nogpg --installroot=/mnt > --disablerepo='*' --enablerepo=fedora install systemd passwd dnf > ................ > Complete! > > ------------------------------------------------------------------------- > $ ls /var/lib/machines/enl > boot/ etc/ media/ opt/ root/ srv/ tmp/ var/ lib@ sbin@ > dev/ home/ mnt/ proc/ run/ sys/ usr/ bin@ lib64@ > ------------------------------------------------------------------------------------- > > But now booting but hanging: > > #systemd-nspawn -bD /var/lib/machines/enl > Spawning container enl on /var/lib/machines/enl. > Press ^] three times within 1s to kill container. > systemd 219 running in system mode. (+PAM +AUDIT +SELINUX +IMA > -APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL > +XZ -LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD +IDN) > Detected virtualization 'systemd-nspawn'. > Detected architecture 'x86-64'. > Running with unpopulated /etc. > > Welcome to Fedora 22 (Twenty Two)! > > Initializing machine ID from random generator. > Populated /etc with preset unit settings. > Unit etc.mount is bound to inactive unit dev-sdb1.device. Stopping, too. > Unit var.mount is bound to inactive unit dev-sdb1.device. Stopping, too. > Cannot add dependency job for unit display-manager.service, ignoring: > Unit display-manager.service failed to load: No such file or > directory. > Startup finished in 51ms. > > ------------------------------------------------------------------------ > > Maybe my btrfs story ? > > Unit etc.mount is bound to inactive unit dev-sdb1.device. Stopping, too. > > Unit var.mount is bound to inactive unit dev-sdb1.device. Stopping, too. > Will investigate. > > --bind was certainly the most easy trick once the raw image is downloaded.
Hmm, my guess is that you somehow lost the /etc and /var directories half way, probably because of the weird mounting your are doing. --bind= should normally be recursive, but maybe this didn't work? Lennart -- Lennart Poettering, Red Hat _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel