Specifications

AMF Introduction and Configuration
New and Enhanced Features in AlliedWare Plus 5.4.4 Major and Minor Versions 111
Key Benefits of AMF
The key benefits of AMF include its unified command-line, simple configuration backup
and recovery process, and time-saving rolling firmware upgrade.
Unified Command-Line
The conventional means of configuring and controlling AlliedWare Plus (AW+) switches is
to use their text-based command-line interface (CLI). In existing networks, the CLI is
available via a serial console port and also to remote login sessions such as SSH.
AMF extends this capability from managing either a single switch to managing a whole
network by using a single (unified) CLI session. Using the unified CLI, a network
administrator can nominate all nodes or a subset of nodes within the AMF network to
comprise an entity known as a “working-set”. Commands can then be executed
concurrently across all switching nodes within the defined working-set as if they were a
single unit. Any existing configuration or diagnostic actions can thus be applied to
multiple devices using a single command sequence, thus reducing maintenance costs and
configuration complexity, while still retaining complete flexibility in network design and
control.
Multiple AMF networks can exist side by side across a single physical network. Note that
AMF treats a Virtual Chassis Stack (VCStack) as a single node.
Configuration Backup and Recovery
The master nodes use external storage to automatically backup the complete
configuration information for all their member nodes, including boot configuration,
firmware, licenses, and user scripts.
If an AMF member node should fail, the AMF process will automatically recognize and
reconfigure an unconfigured replacement (standby) unit, completely recreating the
stored configuration of the failed unit into the replacement unit. The new unit will then
reboot and resume service, without any need for user intervention beyond physical
hardware replacement and cable connection. In this way AMF provides a complete zero-
touch recovery solution. For more information, see “Configuring Multiple Nodes at the
Same Time: the Unified CLI” on page 131.
Rolling-Reboot Upgrade
Installing Firmware upgrades on a production network is typically an infrequent but
sensitive and labor-intensive process. AMF is able to roll-out upgrades to a user-selected
subset of nodes. All that needs to be entered is the target group of nodes, and the location
where the new firmware is stored; AMF will then take care of the rest. Nodes are upgraded
in a serial fashion, with each node tested before continuing the upgrade on the next node.
If an upgrade fails on a particular node, the upgrade process is automatically terminated
and that node will revert to its previous firmware version. In this way firmware updates are
almost completely hands-free, whilst also providing confidence that a bad update will not
result in loss of service. For more information, see “Performing a Rolling-Reboot
Upgrade” on page 153.