Cajun P550/P880/P882 Switch User Guide
Managing the Address Forwarding Table
6-3
3. Enter a new value in the Super Age Time field. The Super Age
Timer marks all invalid table entries, then checks to see if
they remain invalid for the specified super age interval. This
clears the table of entries that are no longer used. The default
value is 7 days.
4. Click APPLY to save your changes, or CANCEL to restore
previous settings.
Configuring the
Age Timer and
Super Age
Timer Using the
CLI
To change the aging values for all instances of the address table
using the CLI, enter one of the following commands from
Configure mode:
> set aft agetime <age-time-value>
> set aft super-agetime <age-time-value>
Refer to the Avaya™ P550™/P880/P882 Command Line Interface
Reference Guide for Version 5.0 for details about this command.
Controlling Reconfiguration of AFT Sizes
Each VLAN you define creates a separate version of the switch AFT.
When these address tables become large, they may begin to use
address space inefficiently.
By default, the hash table (AFT Instance) has auto-increment (hash
table size) on. This means that when a hash table either can't add a
new entry or the utilization of the space in the hash table was poor,
that the hash table would automatically increment its size (thereby
deleting all learned entries). This new size is kept in NVRAM so that
when the box reboots, it doesn’t have to go through the resizing
process again, instead the new size is used.
You may specify that auto-increment be disabled on a per-AFT
Instance basis if they wish for the hash table to remain a constant
size. The problem is how to determine that you have "poor"
utilization. Bucket Utilization is the number of AFT entries divided
by the total AFT bucket space (times 100 for a percentage). Buckets
only come in sizes that are multiples of 2 from 1 to 128. Therefore, if
you have 5 entries to put in a bucket, the best you can do is put it in
a bucket of size 8 with 62.5% utilization for that bucket. It may be
the case, however, that you can only fit those 5 entries in a bucket
of size 16 (because of hashing difficulties), yielding 31.25%
utilization for that bucket. If you have low total utilization, you can
almost always get better utilization by increasing the size of the hash
table (simply because there will be less hash collisions to resolve in
buckets).