Stratix redundancy setup?

I have a network with a Stratix8300 switch and a bunch of Stratix5700 switches, all connected via fiber in a ring topology. If this was a Siemens network, a simple flick of a dipswitch would set a switch to be the Redundant Manager an all would be OK. Unfortunately it seems that with the Statix range of switches is not that easy.

I think the equivalent of RM here is REP? If so, can anyone tell me how to set up a redundant ring with these switches or point me in the direction of a LAB or document for dummies? The Rockwell literature is not as clear as I would like it to be.

Seems like no-one ever worked with these switches before.

Anyways, I was going to get Rockwell out to site to help me with the redundancy setup. In the time waiting for a quote, as they don't provide any assistance for free, I scratched around a bit more. After testing various settings, I found a combination that works a charm. Below is some screenshots of the setup. The network is a fiber ring connected to the fiber ports. there is one Startix 8300 switch in the PLC room and 17 Stratix 5700 switches in various RIO panels and MCCs.

First of all, to enable REP(Resilient Ethernet Protocol) on these switches, the firmware has to be on the version as on pictured first screen grab. That is applicable to all switches as you have to set the port type to TRANSIT, which is not available on earlier SW versions.





To set up REP, input the following settings on your switches. If you have them on a test bench, then in no particular order. If in a running plant, start somewhere in the middle of the network furthermost from your switch that you have connected your laptop to. Best is to go to each switch and do it locally, but I went the network route for firmware upgrade and setup. By default the port type is set to NONE, but if you change it to TRANSIT, None and TRANSIT does not communicate with each other, so you lose comms.

Leave the ports on the switch you will use as your redundant manager on TRANSIT till the end. I used the 8300 for that. Here is the setup of REP on the switches.






I'm not going to go into the theory behind this, as it is pretty well documented by Rockwell and Cisco. This is more a real world quick-setup guide with pictures. This info I could not find anywhere. Once you are done setting all the switches, you can set up the RM as follows.




Once done, you should have a redundant network that works well. I have pulled fibers all over without any equipment failures.

When monitoring REP, you can see exactly where which port goes to and which port was opened by the RM.



Hope this helps someone in future.

A little late but the Stratix by default handles switch redundancy by using Spanning-Tree. This is enabled out of box but has a recovery time of 1-5 seconds.

REP has a recovery time of about 200 ms depending on the segment.

Davkul said:

A little late but the Stratix by default handles switch redundancy by using Spanning-Tree. This is enabled out of box but has a recovery time of 1-5 seconds.

REP has a recovery time of about 200 ms depending on the segment.

Click to expand...

1-5 Seconds is just not good enough. 200ms I can live with, as it is not noticeable.

I have used 8000 series switches on REP protocol a couple of years ago. I recall having problems with the Web interface. I had to use an old 8.0 version of Internet Explorer, on a virtual XP.

The new interface works fine on Chrome and FF. Rarely bother with IE though.

Thanks for the info.
Saved me a lot of heartache.
Now to find a way to get that diagnostic info out of the Stratix into the PLC.

Alan Case said:

Thanks for the info.
Saved me a lot of heartache.
Now to find a way to get that diagnostic info out of the Stratix into the PLC.

Click to expand...

If it's a Rokwell Logix Based PLC they have an AOP and Faceplate for the switch.