Foundry NetIron MLX Series Installation and Basic Configuration Guide
7-34
© 2008 Foundry Networks, Inc.
December 2008
changes to the secondary port.
Preparing the Configuration File
A configuration file that you create must follow the same syntax rules as the startup-config file the device creates.
•
The configuration file is a script containing CLI configuration commands. The CLI reacts to each command
entered from the file in the same way the CLI reacts to the command if you enter it. For example, if the
command results in an error message or a change to the CLI configuration level, the software responds by
displaying the message or changing the CLI level.
•
The software retains the running-config that is currently on the device, and changes the running-config only
by adding new commands from the configuration file. If the running config already contains a command that is
also in the configuration file you are loading, the CLI rejects the new command as a duplicate and displays an
error message. For example, if the running-config already contains a command that configures ACL 1, the
software rejects ACL 1 in the configuration file, and displays a message that ACL 1 is already configured.
•
The file can contain global CONFIG commands or configuration commands for interfaces, routing protocols,
and so on. You cannot enter User EXEC or Privileged EXEC commands.
•
The default CLI configuration level in a configuration file is the global CONFIG level. Thus, the first command
in the file must be a global CONFIG command or “ ! ”. The ! (exclamation point) character means “return to
the global CONFIG level”.
NOTE:
You can enter text following “ ! “ as a comment. However, the “ !” is not a comment marker. It returns
the CLI to the global configuration level.
NOTE:
The CLI changes to the global CONFIG level if you load the configuration as a startup-config file
instead of the running-config (using the
copy tftp startup-config
<ip-addr> <filename> command or
ncopy
tftp
<ip-addr> <from-name>
startup-config
command).
NOTE:
If you copy-and-paste a configuration into a management session, the CLI ignores the “ ! “ instead of
changing the CLI to the global CONFIG level. As a result, you might get different results if you copy-and-paste
a configuration instead of loading the configuration using TFTP.
•
Make sure you enter each command at the correct CLI level. Since some commands have identical forms at
both the global CONFIG level and individual configuration levels, if the CLI’s response to the configuration file
results in the CLI entering a configuration level you did not intend, then you can get unexpected results.
For example, if a trunk group is active on the device, and the configuration file contains a command to disable
STP on one of the secondary ports in the trunk group, the CLI rejects the commands to enter the interface
configuration level for the port and moves on to the next command in the file you are loading. If the next
command is a spanning-tree command whose syntax is valid at the global CONFIG level as well as the
interface configuration level, then the software applies the command globally. Here is an example:
The configuration file contains these commands:
interface ethernet
4/2
no spanning-tree
The CLI responds like this:
NetIron
(config)# interface ethernet 4/2
Error - cannot configure secondary ports of a trunk
NetIron
(config)# no spanning-tree
NetIron
(config)#
•
If the file contains commands that must be entered in a specific order, the commands must appear in the file
in the required order. For example, if you want to use the file to replace an IP address on an interface, you
must first remove the old address using “no” in front of the
ip address
command, then add the new address.
Otherwise, the CLI displays an error message and does not implement the command. Here is an example:
