On Wed, Mar 16, 2016 at 09:06:50AM +0100, Kashyap Chamarthy wrote:
Running this:
$ virt-builder fedora-23 -o vm1.qcow2 --format qcow2 \
--root-password password:123456 --selinux-relabel --size 100G
And, importing vm1.qcow2 into libvirt results in these "corrupted
metadata buffer" errors, in a continuous loop, on the serial console:
[ 51.717069] XFS (vda3): Metadata corruption detected at
xfs_agf_read_verify+0x70/0x120 [xfs], block 0x29b
I can also reproduce this.
It's interesting that it happens on x86_64. I have seem the same
problem, but it was on aarch64:
https://bugzilla.redhat.com/show_bug.cgi?id=1315895
That bug is rather inconclusive. Although I found a way to reproduce
the problem (involving creating disk images on aarch64 and copying
them over to x86_64), I don't know if the problem could be caused by
the way we run xfs_growfs or by XFS itself, nor if the problem is
pagesize / architecture specific or not.
Anyway, it needs deeper investigation.
Rich.
--
Richard Jones, Virtualization Group, Red Hat
http://people.redhat.com/~rjones
Read my programming and virtualization blog:
http://rwmj.wordpress.com
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW