GarrettCom S-Ring™ for Magnum 6K Switches Software for Redundancy Management

Call us for Pre-sales and discount pricing on
Garrett S-Ring™ for Magnum 6K Switches
Software for Redundancy Management

301-924-7400 Support

Features

• Fast fault recovery in Industrial Ethernet LANs using a ring structure
• Operates with extensions to industry-standard Spanning Tree
• Multi-vendor switches and hubs supported in the rings
• Supports large rings over long distances using fiber media
• User controlled set-up for ring management via software commands

Magnum™ S-Ring (patent pending) Redundancy Manager software product, built upon networking industry standards including IEEE
802.1d Spanning Tree Protocol (STP), enables Magnum 6K Managed Switches to simplify and speed up recovery from faults in Industrial Ethernet LAN configurations that use a ring structure. With S-Ring, fault tolerant LANs finally can have both standards and speed.

The S-Ring product supports multi-vendor Ethernet switches and hubs in the rings. While fault recovery times will vary with the particular ring topology and ring member devices selected, almost any ring using standard Ethernet products will experience recovery
speed improvements over STP using S-Ring. In addition, S-Ring makes Ethernet ring-topology LANs more reliable because fault
recovery is less complex and much faster. Industrial LANs often extend over great distances using fiber media, and a ring structure provides fault tolerance at low cost because it minimizes cabling expenses. The number of ring members or “drops” may be 10 or 20, even 50 or more in some cases such as in energy and transportation systems. Pipelines, railroads, windmill farms, oil and gas producing fields, waterways and canals, tunnels, highways and city traffic control systems are all good examples of redundant ring applications covering large areas and long distances.

Other industrial facilities that benefit from large rings include water treatment plants, mines and quarries, forest product mills, agricultural buildings, and warehouses. With S-Ring, fault recovery speed is as fast for large rings as it is for small rings.
S-Ring software operates in a Magnum 6K Switch from specifically defined port pairs that have ring-topology Ethernet devices attached. It builds upon the foundation of STP, but offers an additional option related specifically to ring topologies. S-Ring acts to
recover the ring traffic from faults at sub-second speed, over-riding the STP analysis delay but without conflicting with standard STP. The user configures and controls the S-Ring set-up as part of the 6K Switch’s managed networks software, MNS-6K. A Ring-Scan feature simplifies ring definitions and makes reliable installation easy.

Ring-Closed:
The two ring-control ports, interconnected by the ring members in a
daisy-chain, form an otherwise-illegal redundant path. Standard STP
causes one of these two ports to block traffic in order to enable normal
Ethernet traffic flow. All ring traffic goes through the non-blocking
port for normal LAN operation. Meanwhile, there is a regular flow
of status-checking multi-cast packets (called BPDUs or Bridge Protocol
Data Units) sent out by STP at fixed intervals of time that move
around the ring to show that things are functioning normally. This
normal status is designated as RING_CLOSED. For the RING_CLOSED
state, STP and S-Ring operate the same.
A fault occurs, Ring-Open:
A fault anywhere in the ring will interrupt the flow of standard STP
status-checking BPDU packets, and will signal to STP that a fault has
occurred. According to the standard STP-defined sequence, protocol
packets are then sent out, gathered up and analyzed to enable STP to
calculate how to re-configure the LAN to recover from the fault. After
the standard STP reconfiguration time period, the STP analysis will
conclude that recovery is achieved by changing the blocking port of
the ring port-pair to the forwarding state. With this action, the fault is
effectively bypassed and there is a path for all LAN traffic to be handled
properly. This abnormal status is designated RING_OPEN. For the
RING_OPEN state, STP and S-Ring operate the same, except that SRing
acts faster.
The STP time interval for BPDU packets may be set by the user in the
range of 2 seconds to 15 seconds. This is defined as part of the STP
standard, IEEE 802.1d. Today, the minimum value of 2 seconds is
almost always chosen. S-Ring provides for fault recovery to be triggered
by either the BPDU packet interruption or by Link-Loss, whichever
occurs first. This enables S-Ring to recover fast, even less than
200 milliseconds, See timing data below.
Ring Restoration upon Fault Repair:
Ring restoration, i.e., moving from RING_OPEN back to the normal
RING_CLOSED state, is a mirror image reversal of the fault recovery
process, and normally takes the same amount of time. (The difference
is that restoration is a planned event while a fault incident is
usually a surprise). After restoration, it is necessary for ring operation
to be maintained for 4 to 5 seconds in order to stabilize timers and
buffers, synchronize BPDUs, and reset Link sense circuits.
Initiate Ring-Learn:
This command causes the scanning of all ports in the 6K Switch for
the presence of rings. This command can be a handy tool in setting up
the S-Ring feature for correct initial operation. During a ring-learn
scan, if any port receives a BPDU packet that was also originated by
the same Switch, the source and destination ports are a ring port-pair
and are automatically added to the S-Ring port-pair list. The user can
also manually enable or disable port pairs that are on the S-Ring list.
Definitions:
Path Recovery (or “ring recovery”) is defined as the operating
state such that a new node can come on and find a working
path enabling use of the ring elements to communicate with
another new node.

Fault Recovery is defined as the operating state such that all
existing nodes that previously communicated using the ring elements
can communicate again.
Fault Timing Measurements:
Lab measurements using Fault Time Analysis (FTA) software
with random faults manually generated. A minimum of 10
sample points are taken in each instance.
STP only Timing Data:
Standard 10Mb hubs in the ring: 48.3 sec (47.0-49.4) qty 3
Standard 100Mb hubs in the ring: 48.3 sec (46.9-49.6 ) qty 3
Unmanaged 100Mb switches in the ring, quantity 5:
Path Recovery: 48.5 seconds (47.5 to 49.8 range)
Fault Recovery, P62F: 424 seconds (308 to 570 range)
Fault Recovery, QS: 186 seconds (130 to 217 range)
Fault Recovery, 4K24: 423 seconds (299 to 566 range)
Managed switches in the ring, quantity 5:
Path Recovery, 100Mb: 48.7 seconds (48.0-49.8 range)
Fault Recovery, mP62, 100Mb: 504 seconds (319 to 573 range)
Fault Recovery, 6K, 100Mb: 389 seconds (299 to 531 range)
All switches in ring with Link-Loss-Learn (LLL) enabled, qty 7:
Path Recovery: 48.4 seconds (42.8 to 49.6 range)
Fault Recovery: 48.5 seconds (45.1 to 50.1 range)
Fault Recovery, 6K, Gigabit*: 48.6 seconds (45.6 to 50.7 range)
Quantity 50 mP62 Fault Recovery: 48.8 sec (48.1-50.2)
S-Ring Timing Data:
Standard 10Mb hubs in the ring: 0.853 sec. (0.792-1.70) qty 3
Standard 100Mb hubs in the ring:1.12 sec. (0.979-1.85 ) qty 3
Unmanaged 100Mb switches in ring, quantity 5:
Path Recovery: 0.936 seconds (0.140 to 2.21 range)
Fault Recovery, P62F: 424 seconds (308 to 570 range)
Fault Recovery, QS: 186 seconds (130 to 217 range)
Fault Recovery, 4K24: 423 seconds (299 to 566 range)
Managed switches in the ring, no LLL, quantity 5:
Path Recovery, 100Mb: 0.936 seconds (0.140 to 2.21 range)
Fault Recovery, mP62, 100Mb: 504 seconds (319 to 573 range)
Fault Recovery, 6K, 100Mb: 389 seconds (299 to 531 range)
All switches in ring with Link-Loss-Learn (LLL) enabled, qty 7:
Path Recovery: 0.182 seconds (0.140 to 0.210 range)
Fault Recovery: 0.214 seconds (0.145 to 0.275 range)
Fault Recovery, 6K, Gigabit*:0.447 sec. (0.280-0.490 range)
Quantity 50 mP62 Fault Recovery: 0.208 sec. (0.200-0.215)
*: Fiber GBICs with Gigabit Ethernet protocol, IEEE 802.3z