exchange 2010 CAS Array and WNLB

Dear All ,

I have an issue in Exchange 2010 Cas Array configuration with WNLB

I have two physical Hyper-v server's and each one hosts virtual  HUP CAS and one MB DAG as follow :

SRV1: physical server and host virtual HUPCAS1

NIC HUPCAS IP:
192.168.20.48
255.255.255.0
192.168.29.12

NIC of NLB  
 
192.168.20.39
255.255.255.0
192.168.29.12

SRV2: physical server and host virtual HUPCAS2

NIC HUPCAS2 IP:
192.168.20.47
255.255.255.0
192.168.29.12

NIC of NLB  
 
192.168.20.48
255.255.255.0
192.168.29.12

the NLB NIC's in both hupcas1 and hupcas2 are virtual and on the virtual switch in hyper-v and also on the separate adapter .

I have CAS array record in  DNS named casarray.domain.local IP :192.168.20.50

A record in DNS : 192.168.20.50

my WNLB cluster name also casarray.domain.local IP :192.168.20.50

my problem , I'm not able to reach the cluster from clients side even I have assigned default gateway in the private NLB NIC for both hosts hubcas1 and hubcas2

but I can reach the private NLB NIC from client 192.168.20.39
clients are on the same subnet and we have no issue in Vlans

note : a while ago a cluster  IP was reachable from all clients and the CAS Array was working fine but suddenly the request time out shows up and now I can't reach it anymore ,have u face same issue before ? we are also checking in the switching side , we have HP core switch

the cluster configuration is multicast and also I tried IGMP multicast but still not reachable ,?