Cisco Cisco Nexus 5010 Switch Libro bianco

White Paper

Page 49 of 56

Explicit Failover Order based load balancing:

Explicit failover order routes all vNIC traffic to the highest priority pNIC. In case of a failover event on the highest

priority pNIC, the user can designate a prioritized list of alternate pNICs to take its place.

Network Failure Detection provides two different methods for determining whether a link to a physical NIC has failed

via (a) Link Status Only or (b) Beacon Probing.

Link Status Only:

In Link Status Only failure detection, ESX relies solely on the pNIC to signal that the link has failed. This will detect

some failures, such as cable pulls and physical switch power failures, but cannot detect many configuration errors.

Beacon Probing:

Beacon Probing failure detection transmits and listens for Ethernet broadcast frames known as beacon probes that

are sent by other Ethernet adapters. ESX relies on beacon probing results, as well as link status, to identify when a

link has failed. This will capture both link failures as well as configuration errors.

The following example shows how to create an IP hash based load balancing team with beacon probing failure

detection using vSphere client. The same steps can be followed to create other modes of teaming on a virtual switch

or port groups. The QLE81xx VMware drivers support all the above load balancing and network failure detection

modes. In the example below, it is assumed that the port group “FCoE SAN” template is inherited from existing

vSwitch properties.

Click on the properties menu to bring up the vSwitch1 properties dialog box. The “Virtual Switch” pane is found

under the “Networking” option of the “Configuration” tab.