HSR and PRP Network Redundancy

Posted by Mike Mahana
We're all familiar with the various industrial proprietary ring redundancy protocols: HiperRing, TurboRing, S-ring, N-ring, etc. These protocols typically offered recovery times in the 20 to 300 ms range depending on implementation, total number of switches, etc. But what if your application can't tolerate ANY loss? High availability Seamless Redundancy (HSR) and Parallel Redundancy Protocol (PRP) are ideal for applications where zero communication loss is critical.

Both HSR and PRP operate on the same zero convergence time principle that other redundancy protocols simply cannot provide. To accomplish this, packets are sent out both links and are evaluated by the receiving node. The receiving node accepts the first frame and discards the second. This is in contrast to other redundancy protocols that only use the primary link, and blocks the secondary link, unless a break occurs. Below is a description of the features of the HSR and PRP protocols:

HSR Example Diagram

  • HSR is typically in a ring or multi-ring topology.
  • Frames are sent out both ports (A and B) by the sending node to the destination node. the destination node will accept the first frame it receives and will discard the second.

 PRP Example Diagram

            • DAN - Dual Attached Node
            • SAN - Single Attached Node
            • VDAN - Virtual DAN
  • PRP is typically in a double star topology utilizing two separate local area networks.
  • Frames are duplicated at the "Redbox" and sent through the two networks (LAN A and LAN B) to the DAN's. The DAN will accept the first frame it receives and discards the second.

In summary, HSR and PRP provide the following advantages:
  • Zero convergence time when a node or link is down due to error or maintenance
  • No single point of failure 
  • HSR is ideal for ring network configurations.
  • PRP is ideal for double star network configurations

Links of Interest 
Labels: , , ,