Solved: Netconfig job behaves unexpectedly

orsonjoon · ‎01-16-2009

Hello,

I just created a netconfig job with an ad-hoc command on a interface of a cat 6506 switch.

The first command was a description command on the interface like "description test"

The second command was "switchpoort access vlan 19" which is obviously wrong, type error poort (2 oo's)

The job executed without any error, even the CLI output of the job didn't show any error.

When eexcuting this command from the switch directly this error "cisco6506(config-if)#switchpoort access vlan 19

^

% Invalid input detected at '^' marker."

appears.

In the job I have selected "stop on failure", but the job executed anyway without any error. The vlan wasn't modiyfied on the switch though.

How is this possible, is this a bug or normal operation?

Joe Clarke · ‎01-16-2009

I assume you have TFTP selected as your primary protocol. If this is the case, then what you're seeing is expected. The TFTP deployment method has no way of checking for errors. The commands are loaded completely asynchronously. If you want to do more aggressive error checking, switch to using one of TELNET or SSH.

View solution in original post

john.wright · ‎01-16-2009

I tried the same thing and I got a failure notice on my CW running LMS 2.6.

And it did not apply the change.

Joe Clarke · ‎01-16-2009

I assume you have TFTP selected as your primary protocol. If this is the case, then what you're seeing is expected. The TFTP deployment method has no way of checking for errors. The commands are loaded completely asynchronously. If you want to do more aggressive error checking, switch to using one of TELNET or SSH.

orsonjoon · ‎01-20-2009

You were right again, thanks for your answer Joe.