You must connect the
leaf interface and accomplish the global configuration using the below step:
Set the speed of
the interface connected to compute node correctly, by entering the following.
command is required when the server port of 1G is connected to Switch's 10G
fabric forwarding port-tracking
in Cisco Prime DCNM
For DHCP configuration
in Cisco Prime DCNM:
configuration to work, the profile's DHCP address needs to be modified manually
in Cisco Prime DCNM. This is because Openstack does not populate the DHCP
address to Cisco Prime DCNM, since it has no way of knowing this information.
For example, under a profile, you will see a config like,
ip dhcp relay
address $dhcpServerAddr use-vrf $vrfDhcp.
Modify this to the
actual IP address of the DHCP server:
ip dhcp relay address 10.10.2.86 use-vrf management
Figure 1. Auto-Configuration Profile Tab
Fill the Nexus
Fabric backbone VLAN subnet value in the Settings of the Cisco Nexus Fabric
section under the Admin tab in the Cisco Prime DCNM console.
Figure 2. Admin
Primary (Backbone) subnet with Use local DHCP for Cisco Nexus Fabric checked is
consistent with the backbone VLAN subnet for the leaf and spine nodes in the
Cisco Nexus Fabric. If the fabric is brought up from PoAP in Cisco Prime DCNM,
the configuration screen will look like the following:
Figure 3. PoAP