Description of problem: Autopart on two virtio disks, confirmed delete-all, Anaconda crashed. Version-Release number of selected component: anaconda-23.17-1 The following was filed automatically by anaconda: anaconda 23.17-1 exception report Traceback (most recent call first): File "/tmp/updates/blivet/devices/storage.py", line 697, in _setFormat raise errors.DeviceError("cannot replace active format", self.name) File "/tmp/updates/blivet/devices/lvm.py", line 646, in _setFormat super(LVMLogicalVolumeDevice, self)._setFormat(fmt) File "/tmp/updates/blivet/devices/storage.py", line 742, in <lambda> lambda d,f: d._setFormat(f), File "/tmp/updates/blivet/deviceaction.py", line 669, in apply self.device.format = None File "/tmp/updates/blivet/devicetree.py", line 323, in registerAction action.apply() File "/tmp/updates/blivet/devicetree.py", line 275, in recursiveRemove self.registerAction(ActionDestroyFormat(leaf)) File "/tmp/updates/blivet/blivet.py", line 611, in recursiveRemove self.devicetree.recursiveRemove(device) File "/tmp/updates/pyanaconda/ui/gui/spokes/lib/resize.py", line 422, in _recursiveRemove self.storage.recursiveRemove(device) File "/tmp/updates/pyanaconda/ui/gui/spokes/lib/resize.py", line 447, in _scheduleActions self._recursiveRemove(device) blivet.errors.DeviceError: ('cannot replace active format', 'fedora_dhcp45-root') Additional info: addons: com_redhat_kdump cmdline: /usr/bin/python3 /sbin/anaconda cmdline_file: initrd=/test/mk/rawhide/initrd.img inst.stage2=http://cobra02/trees/mk/rawhide/ nokill updates=http://cobra02/ks/mk/master_updates.img resolution=1024x768 ks=http://cobra02/ks/mk/testing.ks BOOT_IMAGE=/test/mk/rawhide/vmlinuz dnf.rpm.log: Aug 10 14:38:20 INFO --- logging initialized --- executable: /sbin/anaconda hashmarkername: anaconda kernel: 4.2.0-0.rc5.git3.1.fc24.x86_64 product: Fedora release: Cannot get release name. type: anaconda version: Rawhide
Created attachment 1061120 [details] File: anaconda-tb
Created attachment 1061121 [details] File: anaconda.log
Created attachment 1061122 [details] File: dnf.log
Created attachment 1061123 [details] File: environ
Created attachment 1061124 [details] File: ks.cfg
Created attachment 1061125 [details] File: lsblk_output
Created attachment 1061126 [details] File: nmcli_dev_list
Created attachment 1061127 [details] File: os_info
Created attachment 1061128 [details] File: program.log
Created attachment 1061129 [details] File: storage.log
Created attachment 1061130 [details] File: syslog
Created attachment 1061131 [details] File: ifcfg.log
Created attachment 1061132 [details] File: packaging.log
*** Bug 1252450 has been marked as a duplicate of this bug. ***
Okay, so I did a little bit of debugging and here's what happens: 1. blivet tries to verify that it is not removing an active format 2. format is considered active if it's mounted 3. to check if a format is mounted the mountsCache is used 4. to prevent issues with device having multiple symlinks, blivet tries to resolve the device spec to the raw device spec (e.g. /dev/dm-4 for /dev/fedora/root) using udev 5. the root LV is not activated, so udev doesn't know about it 6. the resolved device spec is None 7. mount points that have no device are returned for the None device from mountsCache 8. blivet thinks the root LV is mounted 10. exception raised
https://github.com/rhinstaller/blivet/pull/201
updates.img: http://vpodzime.fedorapeople.org/1252052_updates.img
Proposed as a Blocker for 23-beta by Fedora user vpodzime using the blocker tracking app because: This clearly violates the following criterion: Guided partitioning When using the guided partitioning flow, the installer must be able to: Cleanly install to a disk with a valid ms-dos or gpt disk label and partition table which contains existing data and sufficient unpartitioned space for a Fedora installation Complete an installation using any combination of disk configuration options it allows the user to select Remove existing storage volumes to free up space, at the user's direction Reject or disallow invalid disk and volume configurations without crashing.
According to Vratislav, this probably affects all installations with pre-existing LVM/LVMthinp layout.
Yeah, I saw what was probably this when installing my new Fedlet image over an existing Fedlet install the other day. Couldn't file because Fedlet is debranded.
bug #1212920 is probably a duplicate, please check.
Discussed at 2015-08-20 blocker review meeting: http://meetbot-raw.fedoraproject.org/fedora-blocker-review/2015-08-20/f23-blocker-review.2015-08-20-16.05.log.txt . Accepted as a blocker per criteria cited in #c18.
*** Bug 1212920 has been marked as a duplicate of this bug. ***
This does not appear to be fully fixed. I hit it (or a variant) when testing on a live image with anaconda and blivet updated. See https://bugzilla.redhat.com/show_bug.cgi?id=1255969 (which has the updated logs).
anaconda-23.19.1-1.fc23, python-blivet-1.12.1-1.fc23 has been pushed to the Fedora 23 testing repository. If problems still persist, please make note of it in this bug report.\nIf you want to test the update, you can install it with \n su -c 'yum --enablerepo=updates-testing update anaconda python-blivet'. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2015-13831
anaconda-23.19.1-1.fc23, python-blivet-1.12.1-1.fc23 has been pushed to the Fedora 23 stable repository. If problems still persist, please make note of it in this bug report.
David Shea just reported in passing that he'd seen this with a current Rawhide compose, which should have blivet 1.13. I'm really not sure this is fully fixed, given https://bugzilla.redhat.com/show_bug.cgi?id=1255969 . <davidshea> cannot replace active format * davidshea tries again <adamw> wait a minute, you hit that too? with current anaconda/blivet? <davidshea> yeah I guess "current" being the last rawhide compose, which looks like 8/28 <adamw> it was supposed to be fixed...
never mind, sounds like he had 1.12.
8/28 was a lie, apparently. .buildstamp says 8/20. Sorry for the noise.