[Bug 1200744] New: MD-RAID1 refuses to assemble while "mdadm -E" says the disks are fine (v1.2 superblock)
http://bugzilla.opensuse.org/show_bug.cgi?id=1200744 Bug ID: 1200744 Summary: MD-RAID1 refuses to assemble while "mdadm -E" says the disks are fine (v1.2 superblock) Classification: openSUSE Product: openSUSE Distribution Version: Leap 15.3 Hardware: x86-64 OS: Other Status: NEW Severity: Major Priority: P5 - None Component: Basesystem Assignee: screening-team-bugs@suse.de Reporter: Ulrich.Windl@rz.uni-regensburg.de QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- I migrated a Xen PVM to VMware. The VM also had an MD-RAID1 using metadata version 1.2. I copied the blocks from the old disks to the new ones using dd via SSH. When examining the disks on the new VM using "mdadm -E /dev/sdc", it reports a "clean" state and a "correct" checksum. For the other disk as well. Using "mdadm -v -A /dev/md1" it found both members for md1 and the corresponding slots, but fails to assemble saying "Invalid argument". Looking in syslog I see the kernel complains "md: sdc does not have a valid v1.2 superblock, not importing!" My guess is that new new disks are a bit larger than the original ones, but still: Shouldn't "mdadm -E" complain if the kernel would not accept the disk? Or should the kernel accept what mdadm accepts? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1200744 http://bugzilla.opensuse.org/show_bug.cgi?id=1200744#c1 --- Comment #1 from Ulrich Windl <Ulrich.Windl@rz.uni-regensburg.de> --- (In reply to Ulrich Windl from comment #0)
My guess is that new new disks are a bit larger than the original ones, but still:
To correct myself: The disks were not larger, but smaller (mixing up blocks with kBs, once again), but still the error messages are not very helpful! -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1200744 http://bugzilla.opensuse.org/show_bug.cgi?id=1200744#c3 --- Comment #3 from Ulrich Windl <Ulrich.Windl@rz.uni-regensburg.de> --- (In reply to Neil Brown from comment #2)
Can you please show me the "mdadm -E" output.
In the meantime I had repaired the images of the RAID legs (the RAID assembled OK), so probably the output won't help you much. In fact the new disks had a size that was too small to hold all the data of the original disks. Currently the output line you are interested in reads (copied by hand, so spacing may be incorrect): Unused Space: before=133040 sectors, after=0 sectors -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@suse.com