Due to the use of
VMware virtual switch (vSwitch) and the nature of the fault and recovery
mechanism, Cisco Systems requires VMware NIC Teaming to connect to redundant
physical switches for the Packaged CCE public and private networks. See the
Installation and Upgrade Guide
for more on this design.
designs other than that detailed in the
Installation and Upgrade Guide may expose the applications to unexpected
fault scenarios that impact the recovery and/or operation of the Contact
This fault tolerant
design specifies use of two VMware vSwitches, one each for the Packaged CCE
public and private network VLANs. Each vSwitch has one Active and one Standby
vmnic. Each Active/Standby pair should be split across physical NIC adapters
(for example, Active onboard/Standby on add-in card) and alternate Active paths
to two data center switches. These two data center switches must be configured
for both public and private network communications. The intention of the
uplinks is to ensure that no single failure (physical NIC card, cable, physical
switch) causes the loss of both the public and private network communications
at the same time, or the loss of both the Active and Standby for the same
vSwitch at the same time.
VMware and server
management vSwitch should also be configured separately for enterprise
management separate from either the public or private network communications
paths, following VMware best practices.
Campus LAN, Metro
Area Network (MAN) and Wide Area Network (WAN) (all considered WAN from here)
split Packaged CCE deployments require a highly available Visible network
communications between those data centers. These data centers must also have
dedicated or shared Packaged CCE Private network communications path in a
topology such that no single point of failure within that data center and WAN
infrastructure results in the loss of both the Packaged CCE visible and private
network communications for more than 500ms.
||Maximum Round Trip Time