This fixes an issue when migrating a VM with an unused volume with format
qcow2 or vmdk. Since 'snapshots' wasn't set, 'storage_migrate' wanted to
export/import with format 'raw+size' instead. Therefore it used
'qemu-img convert', which fails when its output leaves through a pipe. Upon
importing a second error was present, since the volid didn't match the format.
Signed-off-by: Fabian Ebner <f.eb...@proxmox.com>
---
Here is the error messages:
2020-01-08 10:34:47 found local disk 'myzfsdir:111/vm-111-disk-0.vmdk' (via
storage)
2020-01-08 10:34:47 copying local disk images
qemu-img: /dev/stdout: error while converting raw: Could not resize file:
Invalid argument
command 'qemu-img convert -f vmdk -O raw
/myzpool/myzfsdir/images/111/vm-111-disk-0.vmdk /dev/stdout' failed: exit code 1
cannot import format raw+size into a file of format vmdk
send/receive failed, cleaning up snapshot(s)..
PVE/QemuMigrate.pm | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/PVE/QemuMigrate.pm b/PVE/QemuMigrate.pm
index 1de1540..a1e2dea 100644
--- a/PVE/QemuMigrate.pm
+++ b/PVE/QemuMigrate.pm
@@ -316,6 +316,9 @@ sub sync_disks {
my ($volid, $sid, $volinfo) = @_;
$local_volumes->{$volid}->{ref} = 'storage';
+
+ # always set 'snapshots' for qcow2 and vmdk volumes
+ $local_volumes->{$volid}->{snapshots} = ($volinfo->{format} =~
/^(?:qcow2|vmdk)$/);
});
}
@@ -370,8 +373,9 @@ sub sync_disks {
if !$owner || ($owner != $self->{vmid});
my $format = PVE::QemuServer::qemu_img_format($scfg, $volname);
- $local_volumes->{$volid}->{snapshots} = defined($snaprefs) ||
($format =~ /^(?:qcow2|vmdk)$/);
if (defined($snaprefs)) {
+ $local_volumes->{$volid}->{snapshots} = 1;
+
# we cannot migrate shapshots on local storage
# exceptions: 'zfspool' or 'qcow2' files (on directory storage)
--
2.20.1
_______________________________________________
pve-devel mailing list
pve-devel@pve.proxmox.com
https://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
