I was being glib in my first post, so I'd like to explain.
Controlling an AC drive on a communications network from a ControlLogix may, on paper, look the same as controlling it from a MicroLogix 1100/1400. The differences are very important.
The ControlLogix uses cyclic messaging that is performed by a dedicated network module with reserved resources that allow the control program to remain in constant and reliable connection with the drive.
MicroLogix uses non-cyclic messaging that is performed by the main CPU. The reliability of the connection with the drive depends on factors that are out of the control of the MicroLogix firmware; factors like retries and the reaction of the controller to communication failure are entirely up to the user program. The Ethernet port and CPU don't dedicate resources to this communication, so it can be affected by functions like messaging to other controllers, online monitoring with RSLogix, or the traffic from HMI computers or operator interfaces.
This isn't to say it can't be done, or can't be done with a good degree of reliability. There are sample programs in the RA "Connected Components" CD and website that do precisely this. I wrote a reliable and well-tested SLC-5/05 version to run PowerFlex 40 drives and filed it on the MrPLC.com Downloads section.
I've become a little jaded over the years with the ability of many users to write clear and reliable communications code, and I don't have the time to prevent them from mangling the examples I write. This becomes especially true when the user is a novice, though I've seen atrocious comms code from seasoned engineers.
My standard recommendation is that the MicroLogix should use discrete control for the drive's Start and Stop (at the very least the Stop) signals, and that Reference, Feedback, and Parameter Access can all be done perfectly happily across the network.