Posted on December 21, 2020 4:38 am
 |  Asked by Andrew Coleman
Print Friendly, PDF & Email

I apologize if this has been asked already (I did a quick forum search, and looked through the blog, and didn’t see anything).

Do anyone know how to import vEOS-lab into CML-P  (aka VIRL2) from Cisco, without overwriting one one the existing VMs?.   I can convert the .vmdk file to .qcow2 format, and as it’s a single disk image (I can’t use Juniper vMX, for example, since the control-plane VM needs 3x files),   but here’s having vEOS completely replace one of the other nodes I use in CML-P  iis suboptimal at best.   You have to pick from a predefined list of templates,  so if I select the one for IOSvL2,  those multilayer swiches will always be vEOS now.


I can’t use vEOS either, since it seems like CML doesn’t support creating your own Docker containers.


I mean,  sure,   I *could* run vEOS in VMware or something, but I’ve only got WS Pro  (VMware won’t support any consumer CPUs,  so no ThreadRipper 3000 based workstation for me),  and WS Pro only supports 18 or 19 vmnets.  The more non-Cisco VMs I run that way,  the fewer interfaces they’ll each have.


I’m pretty much done with GNS3 at this point (long drawn out story, and i don’t feel like burning bridges), but I’d love to be able to drop it completely, instead of need to keep it around for non-Cisco VMs  (Juniper, Arista, Extreme, Aruba, etc..)



Answered on January 8, 2021 4:46 pm

I'm not too familiar with CML-P - does it allow you to define custom images the same way as full CML?

If so I've a sample definition here, so you could create vEOS as a custom node.

Post your Answer

You must be logged in to post an answer.