Setting failover policy at the port group level vs. at the virtual switch level

[StageCoach20111]

How does setting a failover policy at the port group level differ from setting it at the virtual switch level?

For example, I am setting an active / passive failover policy.  When would I set it at the port group level, and when would I set it at the virtual switch level?

If I am setting active/passive at the port group level, would I then use active / active at the vswitch level?  Or would I set both the port group and vswitch to active/passive?

Thanks!

[vGuy]

By default, policies configured at the vSwitch are propagated to all the Portgroups underneath it. You can configure a different failover policy for a portgroup if you want to override the vSwitch policies. An example use case would be, configuring a Portgroup to enable Promiscous mode or configuring different active nics for each portgroups.

With dvSwitch you are only able to configure these policies at the PG layer.

...hth!

[rickardnobel]

A R wrote:

How does setting a failover policy at the port group level differ from setting it at the virtual switch level?

For example, I am setting an active / passive failover policy.  When would I set it at the port group level, and when would I set it at the virtual switch level?

It does not matter where you set the failback setting. If you set it at vSwitch level it will be inherited by all portgroups, but only be actually be used by portgroups having an active/standby configuration and ignored for other portgroups.

I do think it makes more sense to actually set it on the actual portgroup where you want to use both active/standby and the special failback setting, then it would be easier to note that some specific configuration has been done.