Since we updated from SP1 to SP2 multicast imaging is extremly slow: an
image restore which takes about 1 hour with unicast imaging needs about 10
hours with multicast.

Before SP2 we had no problem, used multicasting heavily (and rely on it).

Imaging server is an Windows satelite on VMware ESXi 5.

In novell-pbserv.log I see messages like this:

mcastBufferOutput: sent all 200 packets, calling mcastMasterResend
Sending session: 1114 , pkt:sequence:0
, data:SRV_DATA
Sending session: 5529 , pkt:sequence:0
, data:SRV_DATA
Sending session: 1114 , pkt:sequence:0
, data:SRV_DATA
Time spent while wating : 893
mcastSendMcastWithRetry:after MCAST_RETRY_COUNT, ccode=1, returned after
retryCount=1 for client:1
mcastSendMcastWithRetry:response form client id:1 :: 0x66
mcastSendMcastWithRetry:Received session no. matches with current :51908
client no. matches for client 1, returning
mcastMasterResend:client id:1, response: 0x66
mcastMasterResend:response form client id:1 :: 0x66
mcastMasterResend:received SLV_RESEND_CANCEL from clie 1

Any ideas?