On 19.10.2014 14:08, Andre Robatino wrote: > Andre Robatino <robatino <at> fedoraproject.org> writes: > >> >> Fairly often in the last few weeks, in both my F21 and Rawhide VirtualBox >> guests, I've had a shutdown while doing a yum distro-sync, shortly before >> the transaction starts (so I've never actually had to clean up a >> transaction). I can then always reboot and redo the distro-sync without >> incident. The following messages always appear in /var/log/messages: >> >> Oct 12 08:02:34 localhost dnf: not found updateinfo for: Fedora - Rawhide > - Deve >> lopmental packages for the next Fedora release >> Oct 12 08:03:31 localhost dnf: Metadata cache created. >> Oct 12 08:03:34 localhost systemd: Started dnf makecache. >> Oct 12 08:03:53 localhost systemd: Starting Cleanup of Temporary > Directories... >> Oct 12 08:03:54 localhost systemd-tmpfiles: > [/usr/lib/tmpfiles.d/man-db.conf:1] >> Unknown user 'man'. >> Oct 12 08:03:55 localhost systemd: systemd-tmpfiles-clean.service: main > process >> exited, code=exited, status=1/FAILURE >> Oct 12 08:03:55 localhost systemd: Failed to start Cleanup of Temporary > Director >> ies. >> Oct 12 08:03:55 localhost systemd: Unit systemd-tmpfiles-clean.service > entered f >> ailed state. >> Oct 12 08:03:55 localhost systemd: systemd-tmpfiles-clean.service failed. >> Oct 12 08:04:01 localhost systemd: Startup timed out. >> Oct 12 08:04:01 localhost systemd: Forcibly powering off as result of failure. >> Oct 12 08:04:01 localhost systemd: Shutting down. >> >> I can't find any reported bugs or any other sign that anyone else is seeing >> this. Am I actually the only one? It happened in F21 just yesterday. If I >> need to report it, what component should it be? > > Obviously no one else is seeing this. It just happened again in F21, and for > the first time, it happened in the middle of a transaction, so there's > nothing preventing that (but yum-complete-transaction was able to finish it, > at least). I'm guessing it has to do with my having slower hardware than > most other testers, like with ... core: introduce "poweroff" as new failure action types http://cgit.freedesktop.org/systemd/systemd/commit/src/core/failure-action.c?id=f07756b man 5 systemd-system.conf /StartTimeoutSec /StartTimeoutAction -> /etc/systemd/system.conf man 5 systemd.service /StartLimitAction poma -- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test
