Stack Forced Mode - Avaya 4524GT Configuration Manual

Large image file
If the agent image size exceeds 6 Mb, the switch cannot perform the DAUR. The switch sends
an error message to the base unit. You must perform a manual image upgrade or downgrade
in this situation for both the diagnostic and agent images.

Stack Forced Mode

Stack Forced Mode allows one or both units to become stand-alone switches if a stack of two
units breaks. The Stack Forced Mode allows you to manage one of the stand-alone devices
from a broken stack of two with the previous stack IP address.
If you enable Stack Forced Mode on a stack, you enable Stack Forced Mode on all units in the
stack. Stack Forced Mode becomes active only if the stack fails.
You can configure Stack Forced Mode through ACLI.
See Setting Stack Forced Mode
on a switch.
Stack Forced Mode applies to a stand-alone switch that is part of a stack of two units. When
functioning in this mode, the stand-alone switch keeps the previous stack IP settings (IP
address, netmask, gateway). That allows an administrator to reach the device through an IP
connection by telnet or EDM.
If one unit fails, the remaining unit ( base or non-base unit) keeps the previous stack IP settings.
The remaining unit issues a gratuitous ARP packet when it enters Stack Forced Mode, in order
for other devices on the network to update their ARP cache.
If the stack connection between the two units fails (a stack cable failure, for example), both
stand-alone units retain the IP settings. To detect if the other stack partner is also using the
previous stack IP settings, each device issues an ARP request on the IP address.
When a failure occurs in a stack of 2 units when forced stack mode is enabled, the previous
non-base unit sends out a gratuitous ARP onto the management network. The purpose of
sending out this gratuitous ARP is so that the non-base unit of a failed 2 unit stack can
determine if the base unit is still operational and using the stack IP address. Such a failure
situation in which both the base unit and non-base unit were operational, but not part of a stack
could be possible if the 2 units in a stack were connected by a single stack cable and that stack
cable were then removed or failed. If the previous non-base unit receives a reply from the
previous base unit of the stack, the previous non-base unit knows that the previous base unit
is still operational and does not take over ownership of the stack IP address, but instead will
use the local switch IP address if configured. If on the other hand the previous non-base unit
does not receive a response from the previous base-unit; the previous non-base unit will now
take over ownership of the stack IP address and issue a gratuitous ARP with it's own MAC
Configuration — System
on page 109 for procedures to set the Stack Forced Mode
Stack Forced Mode
April 2011 29