Introduction to ITU G. 8032 Standard Ethernet Ring Protection

Summary

By Lantech

Background Industrial Ethernet has been migrating as the main stream technology, with the advantage of high speed, low cost, long range, PHY independence and easy deployment. The market is shifting toward Ethernet technology, which is a proven and matured technology where industrial communication is required. The most renowned redundancy protocol “RSTP” or “MSTP” in Ethernet enjoys the freedom of topology that connects several switches in a ring forming a “closed loop.”  In this topology, the RSTP or MSTP will calculate the closest backup path to reconnect the path if broken - this is commonly called “Ethernet Ring Switching.” But the protection scheme from generating broadcast storm is essential. You must negotiate which link is the most appropriate one to use, and thus avoid the loop among all switches. The calculation time and response time will be greater along with the growth of network which could be 2~120 seconds to as long as 5 minutes. This irrevocable recovery time is intolerable in some critical network infrastructure like sub-station, aviation, traffic light control, etc. where even a 1 second lost connection is fatal.  The proprietary ring is commonly seen in the market for each brand to improve the ring recovery time to the range of 10~50ms. However, lacking of inter-operability from brand to brand results in less resiliency of infrastructure and higher TOC. Also the test measurement is unlikely the same to tell which product is actually better. The same phenomena is happening in the Telcom industry where many ISPs are adapting carrier Ethernet to replace expensive SDH/Sonet/ATM.  As a result, big Ethernet system players are launching their proprietary recovery protocol like CISCO PRP (Resilient Packet Ring), Extreme Network EAPS (Ethernet Automatic Protection Switching) to have the same failure recovery as the SDH/Sonet ring. The controversial interoperability is affecting the ISPs with the most evidential example of CISCO PRP that has a different MAC header than Ethernet standard. All those ring telco switches are having the following issues in the real site deployment: 1.    Command setting is minute and complicated --Each ring switch must complete the parameters correctly so that the redundancy will operate normally.  If any switch is missing out, the backup will fail. 2.    Status quo is missing --There are many elements to form a ring redundancy, but the current available products only can inform of ring fail without warning of where it has gone wrong. 3.    Multicast exclusion--When the fail over is occurring, the current redundancy recovery is only valid within data packets while leaving out the multicast packets causing the blockage of video stream until it rebuilds the IGMP table. This is the discrepancy of current ring. Solution To solve interoperability issue, ITU (International Telecommunication Unit) has drafted a protocol, G.8032 Ethernet Ring Protection Switching (ERPS), to be compatible with the Ethernet standard and enhance the recovery scheme and protocol to be less than 50ms.  The G.8032 v2 draft (almost final) is complete enough that many big Ethernet system players are rushing to market with conforming equipment. G.8032 will be deem as main stream technology soon.  The benefits of G.8032 are: 1.    <50ms recovery time when failover 2.    G.8032 is the standard protocol which can interact among different brands without being tied up by the same supplier, ensuring the best investment 3.    G.8032 has defined the protocol scheme, parameters, functions, and test measures to be unified, so that the users can evaluate the possible network infrastructure without literally testing each brand in large scale.