[Bug 1193762] New: libstorage-ng fails with 'device not found' even though the device exists
https://bugzilla.suse.com/show_bug.cgi?id=1193762 Bug ID: 1193762 Summary: libstorage-ng fails with 'device not found' even though the device exists Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: YaST2 Assignee: yast2-maintainers@suse.de Reporter: snwint@suse.com QA Contact: jsrain@suse.com Found By: --- Blocker: --- Created attachment 854604 --> https://bugzilla.suse.com/attachment.cgi?id=854604&action=edit yast logs + lvm + lsblk output Running 'yast2 storage' throws an exception [libstorage] BlkDeviceImpl.cc(find_by_any_name):393 THROW: \ device not found, name:/dev/mapper/vg_b-tpool_tdata_rimage_1 even though the device clearly exists. If I mount the corresponding device with UUID 2f3d4144-0cb4-499a-a183-dd012f4c9dd6 and run the probing again, it works. The setup is an admittedly slightly convoluted LVM test setup. YaST logs for both failing and working runs are attached. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 Steffen Winterfeldt <snwint@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|yast2-maintainers@suse.de |aschnell@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c1 Arvin Schnell <aschnell@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |aschnell@suse.com, | |snwint@suse.com Flags| |needinfo?(snwint@suse.com) --- Comment #1 from Arvin Schnell <aschnell@suse.com> --- Please explain the LVM setup. LVM thin pool on a LVM RAID? -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c2 Steffen Winterfeldt <snwint@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(snwint@suse.com) | --- Comment #2 from Steffen Winterfeldt <snwint@suse.com> --- Yes; vg_b is a VG over sdc & sdd; lvs are created with "lvcreate --type raid1 -m 1 ..." some thin, some normal; then a snapshot of a thin volume and a snapshot of that snapshot are added. It's weird, but the PR was about showing nested LVM stuff. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c3 --- Comment #3 from Steffen Winterfeldt <snwint@suse.com> --- It could also be related to the btrfs on top of that or the btrfs subvolumes there. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c4 --- Comment #4 from Arvin Schnell <aschnell@suse.com> --- AFAIR btrfs and snapshots on the block device layer do not work together. lvs reports tpool_tdata_rimage_0 as being private (thus libstorage does not add it to the devicegraph). Also nothing else should be using it. But btrfs reports the filesystem on /dev/mapper/vg_b-tpool_tdata_rimage_1. In the good logs btrfs reports to use /dev/mapper/vg_b-tv_02 (maybe due to the filesystem being mounted). So at first sight it looks like a problem with btrfs. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c5 Arvin Schnell <aschnell@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Component|YaST2 |Kernel QA Contact|jsrain@suse.com |qa-bugs@suse.de --- Comment #5 from Arvin Schnell <aschnell@suse.com> --- I was able to reproduce the problem even without snapshots, so in a setup that should be supported. Here are the steps: # pvcreate /dev/sdc1 # pvcreate /dev/sdd1 # vgcreate test /dev/sd[cd]1 # lvcreate --type raid1 -m 1 --size 2g --name foo test # lvconvert --type thin-pool test/foo # lvcreate --type thin test/foo --virtualsize 1G --name thin1 # mkfs.btrfs /dev/test/thin1 # btrfs filesystem show shows path /dev/mapper/test-foo_tdata_rimage_0 # mount /dev/test/thin1 /mnt/ # btrfs filesystem show shows path /dev/mapper/test-thin1 I tried with a different name that "foo" and it worked - maybe due to btrfs selecting another of the devices? -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 Arvin Schnell <aschnell@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|aschnell@suse.com |kernel-bugs@opensuse.org -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c6 Takashi Iwai <tiwai@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |rgoldwyn@suse.com, | |tiwai@suse.com --- Comment #6 from Takashi Iwai <tiwai@suse.com> --- Adding filesystem team to Cc. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1193762 https://bugzilla.suse.com/show_bug.cgi?id=1193762#c7 Jiri Slaby <jslaby@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|kernel-bugs@opensuse.org |nborisov@suse.com --- Comment #7 from Jiri Slaby <jslaby@suse.com> --- It looks like a btrfs bug? If not, please reassign appropriately. -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠suse.com