Model 922-MES
– User Manual
Focal Technologies Corporation
Page 8-2
A Moog Inc. Company
Document No. 922-0662-00, Rev 1.0
8.2 Redundancy Configuration
The redundancy protocols supported by the 922-MES are presented in the following subsections.
8.2.1 Media Redundancy Protocol (MRP)
MES Core FW P/N:
922-1008-50
SFP1 and SFP2 Only
The Media Redundancy Protocol (MRP) is one of the standard solutions for redundancy in networks with a ring
topology. It is designed to react deterministically on a single failure of an inter-switch link or switch in the network,
under the control of a dedicated media redundancy manager (MRM) node. Ring redundancy allows the
implementation of high-availability ring-shaped network structures, and this protocol enables converting a line
structure to a redundant ring structure.
The MRP defines a media redundancy manager (MRM) and media redundancy client (MRC). An MRP ring
consists of one switch in MRM mode and one or more switches in MRC mode. The MRM sends MRP test frames
in both directions of the ring. These test frames contain a special MAC address and are forwarded by the MRC
switch(es) in the ring. The ring is detected as closed if the test frames arrive on both ends back to the MRM in
certain period of time and in this case the MRM opens the ring by changing one of its port states to
blocked
and
the other to
forwarding
. This blocked port only sends test and management MRP frames. Data frames are sent
via the port that is in the
forwarding
state.
The MES supports the MRP protocol in the role of a media redundancy manager (MRM) or the role of a media
redundancy client (MRC). Both protocols are implemented in the MES switch core at the MAC layer. Only the
two SFP ports on the MES support the MRP protocol and when this function is enabled, these two ports are
used as the ring ports. Note that by default the MES cards are shipped from the factory with MRP enabled in
MRC mode which allows the use of the two SFP ports at the same time. When connecting two MES card ports
via both SFP ports for redundancy purposes, ensure that one card is configured as MRM (typically the topside
unit) and the other as MRC (typically the subsea unit).
During normal operation, the network is in a ring-closed state where an actual physical ring topology exists and
one of the MRM ring ports is in a
blocked
state while the other is in a
forwarding
state (loop avoidance). The
ring ports of the MRC (or MRCs) in this case would be in a
forwarding
state. A ring-opened state is entered
when a segment of the ring has a failure and, for example if the failure is between two MRCs, then both ring
ports of the MRM would enter a
forwarding
state. In both ring-closed and ring-opened states the network logical
topology is a stub.
In MRM mode the card performs the following functions:
Controls the ring state by sending MRP test frames at a configured time period in both directions of the
ring.
Sets one ring port to
forwarding
state and the other port to
blocked
state if it receives its own MRP
test frames.
Sets both ports to
forwarding
if it does not receive its own MRP test frames on both ring ports within a
configured time.
Transmits MRP topology change frames to indicate changes in the ring state to the MRCs.
The ring ports take one of the following port states:
Blocked
: Ring ports drop all received frames except MRP control frames.
Forwarding
: Ring ports forward all received frames.
In MRC mode the card performs the following functions:
Detects a failure or recovery of a ring port link and notifies the change via both ring ports.
Processes the topology change frames and clears its filtering data base (FDB) if requested in a given
MRP time interval.
Forwards MRP frames such as: test, topology change, link change, in-test, in-link change, in-topology
change and in-link status poll between its ring ports only.
The ring ports take one of the following port states:
