Dell PowerSwitch S4820T Configuration Guide for the S4820T System 9.14.1.0 - Page 708

The port monitoring or mirroring function when applied to VLT devices works as expected except with some restrictions. You can configure RPM or ERPM monitoring between two VLT peers. As VLT devices are seen as a single device in the network, when a fail over occurs, the source or destination port on one of the VLT peers becomes inactive causing the monitoring session to fail. As a result, Dell EMC Networking OS does not allow local Port mirroring based monitoring to be configured between VLT peers. However, you can create local Port mirroring monitoring sessions separately on individual devices that are a part of the VLT configuration. NOTE: For more information on configuring VLT, see Configuring VLT. VLT Non-fail over Scenario Consider a scenario where port monitoring is configured to mirror traffic on a VLT device's port or LAG to a destination port on some other device (TOR) on the network. When there is no fail over to the VLT peer, the VLTi link (ICL LAG) also receives the mirrored traffic as the VLTi link is added as an implicit member of the RPM vlan. As a result, the mirrored traffic also reaches the peer VLT device effecting VLTi link's bandwidth usage. To mitigate this issue, the L2 VLT egress mask drops the duplicate packets that egress out of the VLT port. If the LAG status of the peer VLT device is OPER-UP, then the other VLT peer blocks the transmission of packets received through VLTi to its port or LAG. As a result, the destination port on the device to which the packet analyzer is connected does not receive duplicate mirrored packets. VLT Fail-over Scenario Consider a scenario where port monitoring is configured to mirror traffic on the source port or LAG of a VLT device to a destination port on an other device on the network. A fail-over occurs when the primary VLT device fails causing the secondary VLT device to take over. At the time of failover, the mirrored packets are dropped for some time. This time period is equivalent to the gracious VLT failover recovery time. RPM over VLT Scenarios This section describes the restrictions that apply when you configure RPM in a VLT set up. Consider a simple VLT setup where two VLT peers are connected using VLTi and a top-of-rack switch is connected to both the VLT peers using VLT LAGs in a ring topology. In this setup, the following table describes the possible restrictions that apply when RPM is used to mirror traffic: Table 70. RPM over VLT Scenarios Scenario RPM Restriction Recommended Solution Mirroring an Orphan Port on a VLT LAG - In this scenario, the orphan port on a VLT device is mirrored to the VLT LAG that connects a top-of-rack (TOR) switch to the VLT device. The packet analyzer is connected to the TOR switch. The bandwidth of the VLTi link is unnecessarily used by mirrored traffic if max rate limit value is configured in the RPM mirror session. Use ERPM session instead of RPM. Mirroring an ICL LAG to Orphan Port - In this scenario, the ICL LAG is mirrored to any orphan port on the same VLT device. The packet analyzer is connected to the local VLT device through the orphan port. No restrictions apply. If the packet analyzer is directly connected to the VLT device, use local Port mirroring session instead of RPM. Mirroring an ICL LAG to the VLT LAG - In this scenario, the ICL LAG is mirrored to the VLT LAG on the same VLT device. Packet analyzer is connected to the TOR. No restrictions apply. None. 708 Port Monitoring