it is possible to use a self-defined container object for the documentation of a modular switch system. But i think, there is a smarter way. Use the network category to document modular switching systems. In our network model you have 'Interfaces', 'Ports' and 'logical Interfaces' as optional elements.

An interface is like a NIC or like a module in a modular system. On those interfaces you can define any number of ports, based on your own numbering scheme. With that you already have a physical representation of your (modular) switch. Let's assume you have a model with four slots, you would create 4 different interfaces with the number and characteristics of ports, as they are on the specific module.

Next, and here at the latest you will see the difference to a container based object, you can define logical interfaces. This could be a VLAN, combining ports from different slots, a management interface for accessing the switch (not bound to a particular port), your STP structure et al. As you see, this could not happen if you use different "switch" objects coupled by a container.

P.S.: Mail about support will reach you soon…