(In reply to Carlos Robinson from comment #91) > 2018-06-06T16:27:33.192040+02:00 Eleanor-423 kernel: [14171.371792] > lvm2-activation[29061]: segfault at e0 ip 00007fed2151d006 sp > 00007ffe20b2e460 err > or 4 in liblvm2app.so.2.2[7fed2150b000+ef000] > 2018-06-06T16:27:33.680334+02:00 Eleanor-423 systemd-coredump[29070]: Due to > PID 1 having crashed coredump collection will now be turned off. Well it seems that systemd-coredump assumed that PID1 has crashed whereas it was simply reloading its config... which made the crash of lvm2-activation not recorded. But let's focus on the initial bug. > > zypper dup started the rpm transactions about 2018-06-06T14:15:12, finished > 2018-06-06T17:01:25 > > As you can see, about five hours, very slow even without crashes. Well it's only 3 hours (not 5 hours) ;) > I will try, this afternoon, to apply (undo?) the snapshot and redo dup with > the test repo added. Even if it runs slow, I don't have to tend to it much > :-) Perhaps make sure to reproduce without the test repo first ?