GE PACSystems RX3i User Manual page 14

Bumpless Operation with MRP
The RX3i CEP Carrier supports bumpless operation with the GE Automation & Controls PROFINET IO-
Controller if specific conditions are met. Bumpless operation means that a single break in an MRP ring will not
cause the PROFINET connection to be lost and there is no observed loss and addition of PROFINET I/O Devices
while the ring network recovers.
Without bumpless MRP, when a device is lost, it must be re-acquired by the IO-Controller; a typical recovery
time is on the order of seconds. There are two ways an MRM detects a break in the ring:
• Message from an MRC that provides LinkUp/LinkDown detection
• Test packet timeout interval
A network using Media Redundancy Protocol recovers from a ring failure within 80 milliseconds when running
at 100/1000 Mbps full duplex with default Media Redundant Manager (MRM) test packet values. Actual failover
time depends on the device responsiveness to network disconnection and reconnection, number of devices in
the ring, media speed, length of media, and frequency of sending test frames over the network. Network
recovery time is shorter with fewer devices, faster media speed, and shorter media lengths. Third-party devices
in the MRP ring may introduce additional network recovery time. Network recovery time is limited by the ring
participant with the slowest ring failure recovery time. Devices that do not provide LinkUp/LinkDown detection
should be taken into account when calculating network recovery time.
For bumpless network recovery (without disturbing I/O communications to an IO-Device), the I/O Update Rate
for the IO-Device should be configured to be greater than 1/3 of the network recovery time. This permits the
ring to be disconnected or reconnected without timing out the communication connection between the IO-
Device and its IO-Controller.
MRP Operation for I/O Update Rates of 16ms or Greater
The RX3i PROFINET IO-Controller supports bumpless operation at 16ms with the requirement that the MRP
Manager be configured for an MRP Test Packet Interval and MRP Test Packet Count that is faster than the
fastest IO timeout possible.
For example, assume the worst-case scenario when a ring break occurs immediately after the RX3i CEP
received a test packet and immediately before the RX3i CEP was scheduled to receive an I/O packet. In the
case of a 16ms I/O Update Rate, a timeout will occur after three consecutive missed I/O packets, which can
occur in slightly over 32ms. Assuming that the Test Packet Interval is configured for 10ms and the Test Packet
Count is configured for 2ms, we can detect a ring break in just less than 30ms. This ring recover scenario would
look like the following timeline.
To ensure a successful ring recovery using the Test Packets at a 16ms or greater I/O Update rate, make sure
the following statement is true:
Test Packet Interval × (Test Packet Count + 1) < I/O Update Rate × 2
GFK-2883E March 2018
Chapter 1. Overview
7