vPC best Practice in Cisco Datacenter Environment
Today I am going to talk about the quick facts on the vPC best practices in the datacenter environment.
Basics about vPC, A virtual port channel (vPC) allows links that are physically connected to two different Cisco Nexus 7000 Series devices to appear as a single port channel to a third device. The third device can be a switch, server, or any other networking device that supports link aggregation technology.
 |
| Fig 1.1- Cisco DC vPC Best Practice |
Best Practices in vPC best Practices
- Manually define VPC primary and VPC secondary switch
- Form L2 channel with diverse 10GE modules for VPC peerlink ports in dedicated mode.
- Form L3 channel with diverse GE modules for VPC peer keepalive link (non-default VRF).
- Enable BA on VPC peerlink interface (default)
- Enable UDLD aggressive on VPC peerlink interface
- Align STP root, HSRP active router and PIM DR with VPC primary switch
- Align STP secondary root, HSRP standby router with VPC secondary switch
- By eliminating L2 loops with VPC, BA and Loopguard are not needed on access switches
- Enable spanning-tree port type edge on host ports
- Enable spanning-tree BPDU guard globally
VPC With Single 10GE I/O Module
- Implement L3 GEC between VPC peer switches to protect against single point of failure
- Enable IGP routing on the dedicated interface
- The amount of bandwidth depends on customer requirement
Common Layer-3 Features
- Use default IGP timers in a dual supervisor system to avoid unnecessary network convergence during supervisor failover
- Reduced IGP timers (1 sec hello and 3 sec hold timer) improve IGP convergence over L2 cloud.
- Implement consistent auto-cost reference bandwidth. IOS default is 100M and NX-OS default is 10G.
- With large number of FHRP groups, limit the control plane load and FHRP timers should not be tuned to less than 1 sec hello and hold 3 sec hold timer.
- Other L3 best practices includes passive-interface, routing authentication, route summarization and preemption delay
