Brocade Multi-Service IronWare Administration Guide (Supporting R05.6.00) - Manual (Page 45)

Multi-Service IronWare Administration Guide 27
53-1003028-02
SNMP scalability optimization
2
20 is the default.
You cannot change the maximum ifIndex per module to a number less than the number of ports.
After this command is issued the following are generated:
• “System: IfIndex assignment was changed” is logged in the Syslog.
• The snTrapIfIndexAssignmentChanged trap is generated.
Configuration notes for the Brocade NetIron XMR and Brocade MLX series
Note the following if you are upgrading the software on the Brocade NetIron XMR and Brocade MLX
series:
• If you are running an earlier version of the software and you will not be installing the
NI-MLX-1Gx48-T module, you do not need to change your ifIndex allocation scheme. The
current definition is maintained. The maximum ifIndex per module can remain at 20 or 40.
• If you are running an earlier version of the software and you will be installing the
NI-MLX-1Gx48-T module on you Brocade MLX series, you must configure the maximum ifIndex
per module to 64. You must change the ifIndex allocation before installing the NI-MLX-1Gx48-T
module; otherwise, the module status remains in the Offline state.
• If you have a new Brocade MLX series (no previous software installed), but will not be installing
an NI-MLX-1Gx48-T module, it is recommended that you configure the maximum ifIndex per
module to 64 to avoid future ifIndex problems in case an NI-MLX-1Gx48-T module is installed
in the future.
• If you have a new Brocade MLX series (no previous software installed), and you will be
installing an NI-MLX-1Gx48-T module, you must configure the maximum ifIndex per module to
64; otherwise, the module remains in the Offline state.
SNMP scalability optimization
To ensure that SNMP requests are responded to promptly and that SNMP loads do not impact
other device activities, the Brocade device speeds SNMP tasks and limit their effects on the CPU by
a combination of throughput optimization and load throttling.
Configuring SNMP throughput optimization
SNMP throughput is optimized on the Brocade device through a combination of SNMP value
caching, conditional yielding by the SNMP agent, and acceptance of incoming packets during
queue processing.
Configuring SNMP value caching
To accelerate SNMP communication between the management module (MP) and interface module
(LP), the LP will return values for multiple ports when it receives a request from the MP. These
values will be cached on the MP to reduce MP/LP communication time.
The SNMP value caching mechanism is enabled by default. To disable the caching mechanism, use
the following command at the MPLS configuration level of the CLI.
Brocade(config-mpls)#snmp-server cache disable