Today the infrastructure in data centers is distributed. With layers of virtual and physical devices and systems, the data center infrastructure has become more complex to troubleshoot. The accelerating adoption of IP-based services like cloud computing, virtual desktop infrastructure, and unified communications has further compounded this complexity, massively increasing the scope of the network. As a result, network or application problems could have a major impact on business operations.

Network administrators will need better visibility into network traffic across the data center infrastructure. Traffic visibility will improve troubleshooting and help you find the root cause of network problems. Because network performance has become more critical, if network performance drops even slightly, IP-based services will start to fail. Consequently, poor network performance has the potential to cause outages across the data center, with unpredictable and unpleasant results.

Network administrators can use traffic monitoring to detect, diagnose, and resolve network performance issues. The Cisco UCS Manager offers traffic monitoring that copies traffic from one or more source ports and sends the copied traffic to a dedicated destination port for analysis by a network analyzer. This feature is also known as Switched Port Analyzer (SPAN).

There are two types of monitoring sessions:

Ethernet

Fibre Channel

The type of destination port determines what kind of monitoring session you need. For an Ethernet traffic monitoring session, the destination port must be an unconfigured physical port. For a Fibre Channel traffic monitoring session, the destination port must be a Fibre Channel uplink port, except when you are using Cisco UCS 6454 Fabric Interconnects and 6300 Series Fabric Interconnects.

Note

For Cisco UCS 6332, 6332-16UP, and 6454 Fabric Interconnects, you cannot choose Fibre Channel destination ports. The destination port must be an unconfigured physical Ethernet port.