3com 4005 Installation Instruction

C

HAPTER

 7: S

WITCH

 F

ABRIC

 M

ANAGEMENT

 M

ODULES

 

AND

 I

NTERFACE

 M

ODULES

If you install two Switch Fabric Management Modules (in slot M1 and M2), your 
system has redundancy in both management and system controller functions. 
When you power on or reboot the system with two SFMMs installed, the module 
in slot M1 becomes the primary SFMM and the module in slot M2 becomes the 
secondary SFMM. However, if you install an SFMM while the system is powered 
on, the first module installed becomes the primary SFMM and second module 
installed becomes the secondary SFMM.

The system categorizes the two SFMMs as primary and secondary management 
entities. The secondary module operates in hot standby mode, which means that it 
is constantly kept informed about the dynamic state of the management activities 
that are occurring on the primary SFMM.

The system treats both SFMMs as a single logical device. The primary SFMM and 
the secondary SFMM become synchronized after redundancy is established. When 
any configuration or non-volatile data is modified on the primary SFMM, the data 
is automatically modified on the secondary SFMM. Thus, if the primary SFMM fails 
for any reason, the secondary SFMM immediately takes over all primary functions. 

If you remove (deinstall) the primary SFMM or if the module fails in some way, the 
following process occurs automatically:

1 The system initiates the fail-over mechanism (after, for example, the primary 

module fails or is removed).

2 The system reboots and the secondary SFMM becomes the primary SFMM. 

Because it learned all the configuration settings from the primary SFMM, it 
continues to provide all the management functions.

3 If you remove a failed SFMM that used to be primary and install a new SFMM in 

that slot, the new module remains secondary and starts duplicating all 
configuration information.

If, after the failover occurs, the failed SFMM recovers to a normal operating 
condition, it will remain in the secondary state. The failover mechanism is 
non-revertive. Even if the problem that caused the failover is resolved, the failover 
process does not switch the primary state back to the original primary SFMM 
unless the system is rebooted. 

The following connectivity rules apply after you establish SFMM redundancy:

■

You cannot access the secondary SFMM through its console port. Console 
management is only available through the primary SFMM.

■

You cannot Telnet to the secondary SFMM from an external source because 
both the primary SFMM and the secondary SFMM share the same IP address 
for the Switch. Therefore, when you attempt to Telnet to the shared IP address, 
you always access the primary SFMM.