When to use TFTP Server Option?

[Claudia de Luna]

So during project creation there is an option to enter a TFTP server and path.

Under what circumstances is that option used?

In our Proof of Concept we've just been using the APIC-EM VM for both image and configuration storage and distribution and that is working very well.  If I understand the underlying architecture of the APIC-EM to scale up, say we were provisioning 50 devices at a time, as more resources are needed the platform can spin up more LXCs/Containers so why would we not take advantage of that (vs a separate TFTP server?)

Thanks!

Tftp server address	Tftp server Path

[ngoldwat]

Hey Claudia,

No other reason than what is in the Configuration Guide for 3.1.

"When you do not have the option to download the configuration and the image file from the Cisco APIC-EM server, you can deploy the configuration and Cisco device Image file from an external TFTP."

Nick

[Claudia de Luna]

Thanks Nick,

Can you share any examples of the use cases for which that was intended?  I'm trying to wrap my head around why you could not use the controller? 

is it something like, Security has said that images will only come from *this* authorized TFTP server, etc...

[sairasan · Accepted solution]

The other use case is when your APIC-EM controller and device are across geographies and the connectivity between them might be slow. Having a local TFTP server could help speed up the image provisioning process.

[Claudia de Luna]

Sairam, Thank you...that is just the type of use case i was looking for to better understand this capability!