We have got a problem with ZDM7 Imaging on an OES 2 Cluster (based on Linux).
Because it's a two node cluster, DHCP and the Preboot Services must run on one resource.
The problem is, that the PXE Clients are not able to get their boot file downloaded.

What is configured:
1. One virtual server with a shared NSS volume.
2. DHCP cluster resource is configured as described in the OES SP2 manual.
3. ZDM7SP1IR2 installed with the cluster install-set on both nodes.
4. Changed LocalDHCPFlag parameter in novell-proxydhcp.conf.
5. Checked IP address entries in novell-proxydhcp.conf, novell-tftp.conf, novell-zmgprebootpolicy.conf and novell-zimgserv.conf: all are configured with IP address of the virtual server.
6. DHCP option 60 enabled with "PXEClient" for the respective subnet.

1. DHCP: DHCP clients are working.
2. PXE:
A trace shows that the first step is working fine: DHCP discover, offer, request, ack are sent as designed. Then the PXE client sends the next DHCP request packet to request the boot image file, but this is answered with an ICMP (destination unreachable) from the server.
As far as we can see, the problem is, that the the DHCP server acts with the IP address of the physical node (not the virtual one). Therefore, when the the client sends the second DHCP Request packet this is addressed to the IP address of the physical node to port 4011. Netstat shows, that port 4011 is open for the virtual server (as expected), but not the physical one - that's why the ICMP comes back.

So, it looks more like a DHCP problem than a ZDM problem - nevertheless we hope we have chosen the right forum...

Thanks for any ideas and hints!