White Papers

Stacking and VLT
Dell EMC Technical White Paper
2 Overview of stacking
In this technique, network switches are connected to operate as single unit called a stack. Such configuration
can be used to quickly increase the capacity of the network.
Stacking makes it easier for users to expand their network without introducing the complexity of managing
multiple devices. Stackable switches can be added or removed from the stack without disturbing the overall
performance of the stack. Even in scenarios where one link or unit fails, data transfer continues uninterrupted.
These are some of the reasons that make stacking a flexible, effective and scalable solution to add network
capacity.
The maximum number of units that can be added to a single stack depends on the switch series. For further
information see the Appendix.
2.1 Stacking architecture
The stack elects the management units for the stack management.
• Stack master – The primary management unit, also called the master unit.
• Standby – Secondary management unit.
• Stack units – The remaining units in the stack, also called stack members
• Stack group – Each set of four 10G ports, or each individual 40G port correspond to a stack-group.
The master holds the control plane and the other units maintain a local copy of the forwarding databases. From
the stack you can configure both the system-level features and interface-level features that apply to all stack
members.
The master synchronizes the stack unit topology, stack running configuration, and logs. In the event of switch
failure, inter-switch stacking link failure, switch insertion or switch removal, the standby replaces it as the new
master, and the switch with next highest priority.
2.2 Features of stacking
• Single IP management: When connected multiple switches form a stack with larger port counts. The
stack is managed as a single entity. As one of the switches, in the stack acts as a master, the entire
stack is managed through the management interface. Web, CLI, SNMP of the stack master.
• Master failover with transparent transition: In the case of stack master failure, the standby unit
assumes the stack master role. As soon as the stack master fails, the standby unit initializes the control
plane and enables all other stacks units with current configurations.
• Nonstop forwarding on stack: The non-stop forwarding (NSF) feature allows the forwarding plane of
stack units to continue to forward packets even when the control and management planes restart as a
result of a power failure, hardware failure, or a software fault on the stack master, and allows the
standby switch to quickly take over as master.
• Hot add/delete and firmware synchronization: Units can be added to and deleted from the stack
without power-cycling the stack. The units that are to be added to the stack must be powered off before
they are cabled into the stack to avoid election of a new master unit and a possible downgrade of the
stack. When the newly-installed unit in a stack is powered on, the stack firmware synchronization
feature is enabled. This feature automatically synchronizes the firmware version with the version
running on the stack master. Synchronization can either cause an upgrade or a downgrade of the