- Preface
- Hardware and Software Requirements
- Prerequisite
- OpenStack for Nexus Fabric Overview
- Before You Start
- Servers Topology
- Nexus Fabric Enabler Installation
- Post Installation
- Configuration in the Leaf and Cisco Prime DCNM
- Create Project and Launch VM
- Scalability
- Known Limitations and Caveats
- Technical Support Model
Servers Topology
Required Resources
-
DCNM needs to be running in the entire system. For more information, see Cisco DCNM Installation Guide.
-
One server is dedicated for the OpenStack Controller where the dashboard Horizon is running for doing orchestration. Optionally, control server node can also act as a compute node to host VM.
-
Many compute nodes can be used in the setup, per general OpenStack scaling guidelines.
Topology
A sample topology is shown below. This is a critical step as it lays down your foundation so it is highly recommended that you follow the same wiring scheme.
The interface in the topology that is connected to the cluster should be operationally up (sudo ifconfig eth0 up).
The DCNM is also connected to the OpenStack control node via an IP network. You can optionally use the same server to run both control and compute node. In that case, you need to connect that server node to the Nexus Fabric leaf switch as well.