Heartbeat Network for Cluster and Clustering Failure time question

Hello,

Do I still need a heart beat network? Is it suppose to connect from Node A to Node B? Below is our configuration below:

EMC SAN connected to Cisco Switch

Node A:

Hyper-V 2012 R2

Clustering enabled

4- 10 gig ISCSI (multi path) connections to Cisco Switch (2 separate switches)

2-solarflare 10 gigs nics but only have 2 interfaces - one from each nic TEAMed(both are connected to different 10 gig switches). Should I configure all 4 or just current setup?

1-1 GIG Command and Control interface

1-ILO 

CSV configured

Node B:

Hyper-V 2012 R2

Clustering enabled

4- 10 gig ISCSI (multi path) connections to Cisco Switch

2-solarflare 10 gigs nics but only have 2 interfaces - one from each nic TEAMed(both are connected to different 10 gig switches). Should I configure all 4 or just current setup?

1-1 GIG Command and Control interface

1-ILO 

CSV configured

Failure (live migration) works on the Host A <> Host B and Host B <> A. I'm also doing guest clustering with 2 sql guest VM's. Failure works, but it takes up to 30 seconds to complete. Is that normal? This is all one datacenter. Any input would be greatly appreciated.