No announcement yet.

P3K network fail over using Stride SE2 managed switches

  • Filter
  • Time
  • Show
Clear All
new posts

  • P3K network fail over using Stride SE2 managed switches

    i have a project that involves having some IO and an HMI on a platform that moves back and forth. I plan on using constant flex Ethernet cable but would like some redundancy. i see that the SE2 dose trunking but will that work with V-LAN separating the Remote expansion port and the main port Ethernet. the goal is to just have 2 managed switches and 2 Ethernet cables in the cat track and if a cable fails it will detect it and still run.

  • #2
    If I am understanding your question right, you want to have your external port for EtherNet/IP and TCP traffic connected to the stride switch, and the Remote IO connected to the same switch?
    You probably *could* do that, but I doubt it is advisable. Remote IO is usually meant to be faster and more direct. Having it on the same switch and network as external traffic would cause a lot of collisions and both speed and timing of remote IO will be come more chaotic and not as guaranteed.

    If you want the same traffic to flow over both ports, it wont. Connecting them both to the same switch with RSTP or similar wont work as they are not for the same types of traffic or types of instructions.


    • #3
      Thanks I wasn't sure what a manged switch could do. Is there a better way to implement some kind of fail over system that does not require 4 managed switches.

      Another idea I just had is having 2 processors and the one on the platform Io moped to word and use network writes and reads to it.

      I am probably over thinking this and a continuous flex cable won't fail as long as I protect it in continuous flex conduit


      • #4
        Having continuous flex cable in continuous flex conduit should last quite a while. Eventually it will need to be replaced as a wear item simply because it is a moving part, but I wouldnt expect the wear to be much worse than normal power cable or other signal cables running through there as well. Having a second cable for failover in the same place wont really add much reliability either because they will both be wearing at the same time even if one isnt used for communication. Maybe im misunderstanding what you are talking about though.


        • #5
          I was thinking if the managed switch was set up wit 2 cables connecting them to each other that the switch would detect the failure of one cable and be able to notify some one about it. Then we could replace both cables at our convenience without causing down time.

          This grinder has a casting date before ww1 and has been in our plant for 25+ years and I don't think we have ever had to replace the power cables since it got here so I am probably way over thinking this. The machine is so old for speed control there using ac motors cupled to DC generators and reostats for speed adjustments I am trying to bring it up to the 21 centry controls.


          • #6
            If you only want failover between two switches you can do that with spanning tree protocol, or rapid spanning tree protocol, or other proprietary protocols if you have two switches that support those same protocols. It will failover between switch-switch communications only, and if a cable between the switch and P3k fails then the p3k will not have any kind of failover. There is only 1 port for each type of traffic on the p3k, so it is impossible right now to do failover on it. People have been asking for the ability to turn the remoteIO port into an external port for a while now though, so if we ever get the ability to do that in the future then you could implement failover on the p3k itself.


            • #7
              I am not all that concerned about the p3k to the switch. It was switch to switch that I was worried about. I was hoping hopping a managed switch would allow both remote IO and regular Ethernet traffic travel over the same wires but that is not in the cards. Most likely I will order some regular switches and not worry about failover


              • #8

                on section 3-32 it was doing what i was talking about.

                I wish there was a good class i could take on industrial networking. something that goes over layer 2 and layer 3 switches and best practices on integrating plant to machine networks.