Administrator Guide

Table Of Contents
show tdr tengigabitethernet slot/port
Splitting QSFP Ports to SFP+ Ports
The platform supports splitting a single 40G QSFP port into four 10G SFP+ ports using one of the supported breakout cables
(for a list of supported cables, refer to the Installation Guide or the Release Notes).
NOTE: When you split a 40G port (such as fo 1/4) into four 10G ports, the 40G interface configuration is still available in
the startup configuration when you save the running configuration by using the write memory command. When a reload of
the system occurs, the 40G interface configuration is not applicable because the 40G ports are split into four 10G ports
after the reload operation. While the reload is in progress, you might see error messages when the configuration file is
being loaded. You can ignore these error messages. Similarly, such error messages are displayed during a reload after you
configure the four individual 10G ports to be stacked as a single 40G port.
To split a single 40G port into four 10G ports, use the following command.
Split a single 40G port into four 10G ports.
CONFIGURATION mode
stack-unit stack-unit port number portmode quad
stack-unit: enter the stack member unit identifier of the stack member to reset. The range is from 0 to 11
number: enter the port number of the 40G port to be split. The Z9000 range is from 0 to 31.
Important Points to Remember
Splitting a 40G port into four 10G ports is supported on standalone and stacked units.
You cannot use split ports as stack-link to stack a system.
To verify port splitting, use the show system stack-unit 0 fanout {count | configure} command.
The quad port must be in a default configuration before you can split it into 4x10G ports. The 40G port is lost in the
configuration when the port is split; be sure that the port is also removed from other L2/L3 feature configurations.
Converting a QSFP or QSFP+ Port to an SFP or SFP+
Port
You can convert a QSFP or QSFP+ port to an SFP or SFP+ port using the Quad to Small Form Factor Pluggable Adapter (QSA).
QSA provides smooth connectivity between devices that use Quad Lane Ports (such as the 40 Gigabit Ethernet adapters) and
10 Gigabit hardware that uses SFP+ based cabling. Using this adapter, you can effectively use a QSFP or QSFP+ module to
connect to a lower-end switch or server that uses an SFP or SFP+ based module.
When connected to a QSFP or QSFP+ port on a 40 Gigabit adapter, QSA acts as an interface for the SFP or SFP+ cables. This
interface enables you to directly plug in an SFP or SFP+ cable originating at a 10 Gigabit Ethernet port on a switch or server.
You can use QSFP optical cables (without a QSA) to split a 40 Gigabit port on a switch or a server into four 10 Gigabit ports.
You must enable the fan-out mode in order for this mechanism to work.
Similarly, you can enable the fan-out mode to configure the QSFP port on a device to act as an SFP or SFP+ port. As the QSA
enables a QSFP or QSFP+ port to be used as an SFP or SFP+ port, Dell Networking OS does not immediately detect the QSA
after you insert it into a QSFP port cage.
After you insert an SFP or SFP+ cable into a QSA connected to a 40 Gigabit port, Dell Networking OS assumes that all the four
fanned-out 10 Gigabit ports have plugged-in SFP or SFP+ optical cables. However, the link UP event happens only for the first
10 Gigabit port and you can use only that port for data transfer. As a result, only the first fanned-out port is identified as the
active 10 Gigabit port with a speed of 10G or 1G depending on whether you insert an SFP+ or SFP cable respectively.
NOTE:
Although it is possible to configure the remaining three 10 Gigabit ports, the Link UP event does not occur for these
ports leaving the lanes unusable. Dell Networking OS perceives these ports to be in a Link Down state. You must not try to
use these remaining three 10 Gigabit ports for actual data transfer or for any other related configurations.
NOTE: Trident2 chip sets do not work at 1G speeds with auto-negotiation enabled. As a result, when you peer any device
using SFP, the link does not come up if auto-negotiation is enabled. Therefore, you must disable auto-negotiation on
platforms that currently use Trident2 chip sets (S6000 and Z9000). This limitation applies only when you convert QSFP to
SFP using the QSA. This constraint does not apply for QSFP to SFP+ conversions using the QSA.
282 Interfaces