pvmove with nbd

hi

i have a bit of a problem with pvmove. it may not be pvmove's fault, we'll see.
i have two systems, sysA and sysB. on sysA i have a VG with 3 PVs (say sda,sdb,sdc). i've exported these two block devices thru nbd, and i've imported them on sysB as nbd0, nbd1, nbd2.

on sysB i can see the "exported" VG, i can mount the LVs, it's working fine. now i add a 4th PV local on sysB, so in the VG i have it like this:

nbd0, nbd1, nbd2, sda (this one is the local PV).

now i want to move everythin from nbd* to this new local sda, and then remove all the nbd*s from the VG. practically this would mean that i've moved the VG from one machine to another through the network.

but the pvmove is not really working. it just hangs, and stays in D when i look at it with 'ps'. and no disk activity whatsoever. i cant kill the process, i cant do anything, sort of a reboot, after which i can do a pvmove --abort, but nothing more, because whatever i want to do, the LVM related stuff just hangs again (im guessing that's because pvmove wants to resume the moving operation). one CPU always stays in 100% i/o wait during this.

the LVs in the VG are NOT mounted and are unused during the move, noting is mounted. so basically this would be a remote pvmove, from one system to another through nbd export.

obviously i'm missing something, or this can't be done.

any other ideas on how i could do this without physically moving the storage from sysA to sysB ?

[root@emailro nbd-2.8.8]# pvmove -v -n log /dev/nbd2 /dev/sdd
Finding volume group "zerg"
Archiving volume group "zerg" metadata (seqno 12). Creating logical volume pvmove0
Moving 1920 extents of logical volume zerg/log
Found volume group "zerg"
Creating zerg-log
Loading zerg-log table
Resuming zerg-log (253:2)
Updating volume group metadata
Creating volume group backup "/etc/lvm/backup/zerg" (seqno 13). Found volume group "zerg"
Found volume group "zerg"
Suspending zerg-log (253:2)
Found volume group "zerg"
Creating zerg-pvmove0
Loading zerg-pvmove0 table
Resuming zerg-pvmove0 (253:3)

'ps output, filtered:

11650 pts/0    D<L+   0:00 pvmove -v -n log /dev/nbd2 /dev/sdd
11666 ?        S<     0:00 [kcopyd]

sysA is centos 4.4 with kernel 2.6.9-42.0.8.plus.c4smp, nbd 2.8.8 and LVM version

[root@email ~]# lvm
lvm> version
LVM version: 2.02.06 (2006-05-12)
Library version: 1.02.07 (2006-05-11)
Driver version: 4.5.0

and sysB is centos 5.0 with kernel 2.6.18-8.1.6.el5.centos.plus, nbd 2.8.8 and LVM version

[root@emailro ~]# lvm
lvm> version
LVM version: 2.02.16-RHEL5 (2006-12-14)
Library version: 1.02.13 (2006-11-28)
Driver version: 4.11.0

the systems have fibrechannel connected SAN storages, but would guess that's not so important. between the two systems there is at least a 200Mbps connection, with no network problems.

any help greatly appreciated.