Allen Bradly EN2T card

EN2T card is capable of 256 Connection. What exactly a connection means?
I am going to use a Generic Module communication with 128bytes input and 128 bytes output. how can i estimate how many devices i can connect with an EN2T card.
Any help will be appreciated

Thanks in advance

They are referring to CIP connections. One TCP/IP connection may consume many CIP connections. Take a look at this manual for a starting place. https://literature.rockwellautomation.com/idc/groups/literature/documents/um/enet-um006_-en-p.pdf

The web page of the EN2T card will show the amount of connections currently in use. Take a snap shot of the current situation, add one device then see what you have. I would guess 1 or 2 CIP connections per module.

Ken Moore said:

They are referring to CIP connections. One TCP/IP connection may consume many CIP connections. Take a look at this manual for a starting place. https://literature.rockwellautomation.com/idc/groups/literature/documents/um/enet-um006_-en-p.pdf

The web page of the EN2T card will show the amount of connections currently in use. Take a snap shot of the current situation, add one device then see what you have. I would guess 1 or 2 CIP connections per module.

Click to expand...

+1 on the method Ken describes. Sometimes it's just easier to do it and monitor the results!



My experience has been:
- two connections for each device/rack for digital type I/O without timestamps (rack optimized)

- two connection for each device with it's own config (analog, rtd, Timestamped digital inputs, etc)
- four connections for each scanner type device - Devicenet, Controlnet, Profibus, or another ethernet (which is not good networks design but it happens)

thingstodo said:

...


- four connections for each scanner type device - Devicenet, Controlnet, Profibus, or another ethernet (which is not good networks design but it happens)

Click to expand...

That is quite often seen as the preferred method when using Ethernet I/O. You deploy it as a separate network so that I/O data (which wants to be as determinist as possible) is kept away from the general chit-chat on a corporate network, or even just a SCADA or HMI network. You are also protecting the real-time data transfer on the I/O network (which is the most important!) from any wrong-doings on the corporate/SCADA networks. Most often any infrastructure devices on an I/O Ethernet network are not in the IT departments domain, so any IT "tweaking" of infrastructure devices cannot affect the running of the plant.

Pushing everything onto one network will inevitably lead to a performance hit. And, of course, you are "sharing" the connection resource across multiple comms adaptors.

daba said:

That is quite often seen as the preferred method when using Ethernet I/O. You deploy it as a separate network so that I/O data (which wants to be as determinist as possible) is kept away from the general chit-chat on a corporate network...

Click to expand...

In my opinion, you are correct. As I am hearing from *EVERYONE* involved in IT, cyber security, intrusion detection, et al ... well, you can't protect what you can't see. They want it put onto a separate VLAN ... which is only separate until you hack the cisco switch.


I'm willing to have no intrusion detection on our IO network in exchange for any cyber attack having to cross the PLC backplane on CIP. There are tools for doing this, but not nearly as many as for IP stacks. You lose attacks from the 'script kiddies' at least


I'm told that this is 'old thinking'. It may be old, but it has worked so far.


If an entity with sufficient resources wants to get into my network, they will. All I'm trying to do is keep out the less determined groups. If I make it a bit harder to crack, maybe they will look for easier targets elsewhere ...

Thank you so much
I could able to see the connection. Normally the device is taking one connection but at some point it is opening 2 connection.