Hello everyone,
We have just migrated our datacenter to an all(well primarily) nexus environment with roughly 30 9Ks running in NX-OS standalone mode. I've currently been using an SSH script that I wrote to automate some of my regular tasks. As you all know, NX-OS provides a much better way to get this data through the NX-API using HTTP calls but there are some serious security implications by enabling the API for use. For example, the API when enabled becomes available through every SVI on the router which if publicly exposed is a big problem from a security perspective.
So I'm curious to see what others are doing, if anything, to secure the API connections. Ideally the requirements I have are locking down the ability to manage the device to specific source IP addresses. Similar to adding an access-class to VTY line for restricting SSH access. Second, I need to enable the API ONLY on the management interface.
I certainly understand using the API requires authentication and can be restricted to using HTTPS so it's encrypted, but in order for me to gain approval to enable the API I need to come up with a list of compensating controls.
Thanks all for any time you put into this thread. It is much appreciated.