Engineering Encyclopedia: Troubleshooting Token Ring Networks

Engineering Encyclopedia
Saudi Aramco DeskTop Standards
Troubleshooting Token Ring Networks
Note: The source of the technical material in this volume is the Professional
Engineering Development Program (PEDP) of Engineering Services.
Warning: The material contained in this document was developed for Saudi
Aramco and is intended for the exclusive use of Saudi Aramco’s
employees. Any material contained in this document which is not
already in the public domain may not be copied, reproduced, sold, given,
or disclosed to third parties, or otherwise used in whole, or in part,
without the written permission of the Vice President, Engineering
Services, Saudi Aramco.
Chapter : Communications For additional information on this subject, contact

File Reference: CDA20303 J.S. Phillips on 873-0228
Engineering Encyclopedia Communications
CONTENTS PAGES
IDENTIFYING THE RING INITIALIZATION SEQUENCE OF A TOKEN

RING........................................................................................................................ 1
Ring Insertion ............................................................................................... 1
Lobe Media Check............................................................................ 2
Physical Insertion.............................................................................. 3
Monitor Check .................................................................................. 7
Duplicate Address Verification......................................................... 8
Neighbor Notification ....................................................................... 8
Request Initialization ...................................................................... 12
Counters and Timers Used to Monitor Protocol Activity ............... 15
Counters.......................................................................................... 16
Timers ............................................................................................. 17
AMP Frames (Active Monitor Present) and SMP Frames.............. 19
Standby Monitor Timer (TSM)....................................................... 24
Latency Buffer............................................................................................ 25
No Token Timer.............................................................................. 28
PC Transmission of Claim Token............................................................... 28
MAC Purge Frame ..................................................................................... 31
Duplicate Address Test ................................................................... 33
IDENTIFYING NORMAL TOKEN RING PROTOCOL OPERATION .............. 36
Active Monitor Present (AMP) .................................................................. 36
Purpose of the Active Monitor Present Frames .............................. 37
Protocol Sequence .......................................................................... 38
Two Active Monitors...................................................................... 39
Protocol Analyzer Display of AMP Frame Protocol ...................... 40
No Active Monitor Present on the Network ................................... 41
Standby Monitor Present (SMP) ................................................................ 43
Purpose of the Standby Monitor Present Frames............................ 43
Standby Monitor Present (SMP)..................................................... 44
No AMP Frame Received by Standby Monitor.............................. 45
Upstream and Downstream Neighbor Addresses............................ 46

Protocol Analyzer Display of Upstream and Downstream

Addresses ........................................................................................ 47
Token Transmission ................................................................................... 48
Purpose of Token Transmission...................................................... 49
Token Frame Byte and Bit Composition .................................................... 53
Protocol Analyzer Display of Token Transmission ........................ 54
Early Token Release ....................................................................... 56
IDENTIFYING NETWORK FAULTS.................................................................. 58
Types of Equipment Available ................................................................... 58
Copper Wire Cable Test Sets .......................................................... 58
Fiber Optic Test Sets....................................................................... 59
Power Line Disturbance Analyzers................................................. 60
Analog Test Set............................................................................... 60
Protocol Analyzers.......................................................................... 61
Attaching a Token Ring Protocol Analyzer to a Network.......................... 62
What Types of Problems the Protocol Analyzer will Detect...................... 63
Real-Time Protocol Analysis...................................................................... 66
Benefits to Performing Real-Time Protocol Analysis..................... 67
Recognizing Normal and Abnormal Protocol Operations .............. 67
Traffic Statistics Analysis........................................................................... 71
Purpose ........................................................................................... 71
Neighbor Notification Incomplete .................................................. 74
Active Monitor Errors..................................................................... 75
Soft Errors....................................................................................... 75
Hard Errors ..................................................................................... 78
Ring Failure................................................................................................ 79
Beacon ............................................................................................ 79
Neighbor Notification (NN)............................................................ 83
Determining the Upstream NeighborÕs Address............................ 84
Error Counters ............................................................................................ 85
Line Error........................................................................................ 86
Internal Error................................................................................... 87
Congestion Error............................................................................. 87

Frame Copied Error ........................................................................ 87

Multiple Tokens ......................................................................................... 87
Definition ........................................................................................ 88
Problems Created by Multiple Tokens............................................ 88
Identifying Multiple Tokens ........................................................... 89
WORK AID 1: GUIDE FOR IDENTIFYING A NORMAL RING
PROTOCOL INITIALIZATION SEQUENCE...................................................... 90
WORK AID 2: GUIDE FOR IDENTIFYING A NORMAL RING
PROTOCOL SEQUENCE ..................................................................................... 91
WORK AID 3: GUIDE FOR IDENTIFYING NETWORK PROTOCOL
FAULTS .............................................................................................................. 102
GLOSSARY ........................................................................................................ 109

IDENTIFYING THE RING INITIALIZATION SEQUENCE OF A TOKEN RING
To attach a host PC to a Token Ring network, an NIC must transmit and receive a series of
MAC frames. MAC frames that perform a specific function must be transmitted in a specific
order for an NIC to gain access to the network. This section explains the ring initialization
sequence that is used by an NIC to access the network and includes the following topics that
are pertinent to that discussion:
• Ring Insertion
• Active Monitor Present (AMP) Frames and Standby Monitor Present (SMP) Frames
• Standby Monitor Timer (TSM)
• Latency Buffer
• No Token Timer (TNT)
• PC Transmission of Claim Token
• MAC Purge Frame
• Duplicate Address Test
Ring Insertion
The protocol operating sequence that is transmitted by a PC after it accesses the ring is called
the Ring Insertion Process. For an NIC to become active on a ring, the following five
protocol transactions must occur in a specific sequence:
• Lobe Media Check/Physical Insertion

• Monitor Check
• Duplicate Address Verification
• Neighbor Notification
• Request Initialization
Saudi Aramco DeskTop Standards 1

Lobe Media Check
Any application program that is in operation in a host PC that wants to establish

communications with another PC on the network issues a software command to the NIC to
initiate the Ring Insertion Process. Prior to physical connection of the NIC and the host PC to
the ring, a lobe media check is performed.
The lobe media check verifies that the lobe cable that connects the NIC to the Lobe Access
Module (LAM) has continuity and that no bit errors occur as the frame travels through the
loopback path. The Destination Address (DA) field of the Lobe Media Test MAC frame is set
to a null address. A null address consists of all 0 bits. An arbitrary arrangement of data bits is
inserted in the information field of the frame. The data bits are used to determine the error
rate.
As shown in Figure 1, to test the lobe cable, the NIC transmits a Lobe Media Test MAC
frame. The test frame travels from the NIC that is installed in the host PC through the lobe
cable to the LAM, where the frame is passed through a loopback. The frame then travels
back through the lobe cable to the NIC. If the frame is received back as error-free, the NIC
sends a station-attach signal, which is a phantom dc voltage.
Lobe Media Test

Figure 1

Physical Insertion
If the lobe test is unsuccessful, the NIC reports the failure to the host PC. The application
program that is in operation in the host PC notifies the user. If the lobe test is successful, the
NIC proceeds with the Ring Insertion Process.
Prior to the NIC insertion in the ring, the NIC interface, lobe cable, and LAM interface are in
a bypass mode. Figure 2 is a schematic diagram of the lobe cable, the NIC interface, and the
LAM interface in a bypass mode. No voltage is applied to the -V and +V voltage connection
points, and, therefore, no DC current flows through the transmit pair and the receive pair and
the relay is not activated. Because the relay is not activated, the relay contacts are set so that
the signal from the ring enters the Ring In (RI) path, bypasses the lobe cable, and then exits
the LAM through the Ring Out (RO) path. In the bypass mode, the NIC is not inserted into
the ring.

LAM Bypass Mode

Figure 2

The ring insertion mode is shown in Figure 3. The physical insertion of the NIC into the ring
is performed by the application of a dc voltage between +V and -V on the transmit wire pair
and the receive wire pair in the lobe cable. The dc current activates the relay in the LAM.
The contacts on the relay switch the signal path so that the signal from the ring enters the
Ring In (RI) path and then passes through the relay contacts to the receive interface. The
signal then travels down the receive pair of the lobe cable to the NIC interface. Electronics on
the NIC process the signal and then send the signal to the transmit pair. The signal travels
down the transmit pair to the LAM transmit interface through the relay contacts and, then, to
the Ring Out (RO) path. The signal leaves the LAM through the RO and travels through the
ring wires to the downstream NIC.
According to the IEEE 802.5 standard, insertion occurs with a dc voltage of 4.1 to 7.0 volts
for a current that is less than 1 milliamp and with a dc voltage of 3.5 to 7.0 volts for a current
that is between 1 and 2 milliamps. Some NIC manufacturers apply 12 volts at 30 milliamps.

LAM Insertion Mode

Figure 3

Monitor Check
After the NIC has activated the relay in the LAM, the NIC monitors the ringÕs protocol
transmissions to determine its operational status. The monitor function identifies the various
error situations that can occur on a ring. During the monitor check phase, the station (NIC)
looks for one of three MAC frames:
• Active Monitor Present (AMP)

• Standby Monitor Present (SMP)
• Ring Purge
The presence of both AMP and SMP frames indicate that the ringÕs operation is normal. The
presence of ring purge frames indicates that the active monitor is purging the ring before it
transmits a new token.
The NIC, dependent on the results of the monitor check, will either continue with the ring
insertion process or will engage in an error detection and recovery procedure to restore
network operation.
To complete the ring insertion process, three more protocol exchanges must occur. The three
protocol exchanges include address verification, neighbor notification, and request
initialization.
Figure 4 shows a protocol analyzer layer 2 display of the following three protocol exchanges:
• The Duplicate Address (Addr) Test that is sent as part of the address verification
process.
• The Report NAUN Change frame that performs the neighbor notification function.
• The Request Initialization frame.
A layer 2 display is an industry term for a protocol analyzer display that is a descriptive
interpretation of data link layer protocol transactions. Layer 2 refers to the OSI ModelÕs
second layer, which is the data link layer.
The ÒSecsÓ field is the time in seconds after the start of the last minute that the frame was
monitored. ÒBytesÓ indicates the quantity of octets in the frame. ÒPhysÓ identifies the
IEEE standard that defines the frames structure. ÒDestinationÓ indicates the destination
address of the frame. ÒSourceÓ indicates the source address of the NIC that originated the
frame. The ÒTypeÓ column indicates that MAC frames have been transmitted. Both
destination and source Service Access Points (SAP) are shown in the display. ÒDescriptionÓ
provides detailed information about the type of frame that was analyzed. The display
includes two ÒStandby Monitor (Mon) PresentÓ frames that are used in the protocol
sequence to identify the Upstream NeighborÕs Address.

Secs Bytes Phys Destination SAP Source SAP Type Description

20.281 22 802.5 400000000001 RS 400000000001 RS MAC Duplicate Addr Test
20.380 36 802.5 CfgReportSvr CR 400000000001 RS MAC Report NAUN Change
20.400 36 802.5 Broadcast RS 400000000001 RS MAC Standby Mon Present
20.400 36 802.5 CfgReportSvr CR 40005494E000 RS MAC Report NAUN Change
20.401 66 802.5 RingParamSvr RP 400000000001 RS MAC Req Initialization
Protocol Display of the Ring Insertion Sequence

Figure 4
Duplicate Address Verification
The Duplicate Address Test (DAT) frame verifies that the address that is assigned to the NIC
is unique to the network and that no other NIC has the same address. The NIC sends two
Duplicate Address Test MAC frames that are addressed to itself. If another NIC has the same
address, that NIC then sets the address recognized bit (A-bit) in the Frame Status (FS) field of
the frame to a 1. If the frame returns with the A-bit set to a binary 1, another NIC on the ring
has the same address. The NIC with the duplicate address may also set the Frame Copied (C-
bit) to a 1 if the frame was copied. The NIC that recognizes the DA and copies the DAT
frame from the network will ignore the frame. If duplicate addresses exist on the network, the
NIC returns to the bypass state.
The NIC assumes that no other NIC on the network has a duplicate address when the NIC
receives two duplicate address test MAC frames with both the A-bit and C-bit set to 0Õs.
Such settings indicate that the test was passed.
Neighbor Notification
After the Duplicate Address Test has been passed, the NIC enters the ring poll phase of the
ring insertion process. In this phase, the NIC waits to receive an Active Monitor Present
(AMP) frame or a Standby Monitor Present (SMP) frame. The AMP frame or SMP frame is
transmitted by an upstream neighbor NIC as part of the neighbor notification process. The
AMP and SMP frames also perform other protocol functions.

Figure 5 shows the network configuration and the A-bit and C-bit values for an SMP frame.
The SMP frame contains all of the fields that are found in a normal MAC frame. The SMP
frame includes the following: Starting Delimiter (SD), Access Control (AC), Frame Control
(FC), Destination Address (DA), Source Address (SA), Information (INFO), Frame Check
Sequence (FCS), Ending Delimiter (ED), and Frame Status (FS). NICs that receive an AMP
or an SMP frame that have both the A-bit and C-bit equal to a 0 know that the frame was
transmitted from the upstream neighborÕs NIC. The A-bit and C-bit are in the Frame Status
(FS) field of the SMP frame. The Upstream Neighbor NIC in the host PC transmits the SMP
frame with the A-bit and C-bit in the FS field set to 0Õs. The frame travels to the LAM and
around the ring to the next active NIC.
Upstream Neighbor Sends Address

Figure 5
The NIC that receives the SMP frame from the Upstream Neighbor with the A-bit and C-bit
both set to 0Õs copies the frame into memory. The NIC then retransmits the SMP frame and
sets the C-bit to a 1 to tell the Downstream Neighbor NIC to ignore the frame. Figure 6
shows the binary bit configuration of the A-bit and C-bit after the frame has been transmitted
by the NIC to the downstream neighbor. The NIC that copied the frame then stores the
Upstream NeighborÕs Address in memory. At this stage in the Ring Insertion sequence, the
NIC that copied the frame from the network and set the C-bit to a 1 has determined the
Upstream NeighborÕs Address.

NIC Sends Address to Downstream Neighbor

Figure 6
In Figure 7, PC#1 is the Upstream Neighbor. The address of PC#1 is 40 00 00 00 00 01.

PC#2 has broadcast an SMP frame. The address of PC#2 is 40 00 54 94 E0 00. In the SMP
frame are Medium Access Control Vectors (MAC Vector) that contain data fields that
describe the purpose of the frame. Embedded in the MAC Vector is the address of the
Nearest Active Upstream Neighbor to address 40 00 54 94 E0 00. The address that is
embedded in the MAC Vector is 40 00 00 00 00 01.

Network Address Relationship

Figure 7
The exchange of addresses between the NICs can be monitored by a protocol analyzer and
shown on a level 2 interpretative display. The Upstream NeighborÕs address can be seen in
the protocol analyzer level 2 interpretative display that is shown in Figure 8. A description of
the information that is contained in a level 2 interpretative display was covered in Module
CDA 203.02. The level 2 interpretative display that is under the heading ÒMedium Access
Control VectorsÓ indicates that the Nearest Active Upstream Neighbor address is 40 00 00 00
00 01. The address of the NIC that originated the frame is shown in the top line of the level 2
interpretative display under the title source. The address of the NIC that originated the SMP
frame is 40 00 54 94 E0 00.
Through the sequential analysis of SMP frames, the addresses of Upstream Neighbors and
Downstream Neighbors can be obtained and a complete description of the ring topology can
be developed.

Secs Bytes Destination SAP Source SAP Type Description

37.377 36 Broadcast RS 40005494E000 RS MAC Standby Mon Present
IEEE 802.5 Medium Access Control (MAC)
Type: MAC Length: 36 Address: 40005494E000 Broadcast
Time Stamp: 00:01:37.377208
Token Priority: 0 Monitor: 1 Reservation: 0
Frame Status: Destination address recognized, copied.
Medium Access Control Vectors
Type: Standby Mon Present
Physical Location: 00000000
Nearest Active Upstream Neighbor: 400000000001
Hexadecimal Frame
18 06 C0 00 FF FF FF FF 40 00 54 94 E0 00 00 12
00 06 06 0B 00 00 00 00 08 02 40 00 00 00 00 01
Protocol Analyzer Display of Upstream Neighbor Address

Figure 8
Request Initialization
To complete the Ring Insertion process, the NIC sends a Req (Request) Initialization MAC
frame to the Ring Parameter Server (RingParamSvr or RPS). The frame informs the RPS that
an NIC has been inserted into the ring. The frame contains descriptive information about the
NIC. The Request Initialization MAC frame requests the RPS to send additional operational
parameters. The parameters that are received in this process replace the default parameters
that were set at the start of the ring insertion process.
As shown in the protocol analyzer level 2 display in Figure 9, the NIC that is engaged in the
Ring Insertion Process sends a series of Req Initialization MAC frames to the ring parameter
server functional address (C0 00 00 00 00 02). The Request Initialization frames in the
example are transmitted by Source address 40 00 00 00 00 01. The other frame types that are
shown on the screen and the different Destination Addresses and Source Addresses can
present a problem to the protocol analyst. The problem that the protocol analyst must
overcome is to visually separate the frame types and the addresses to ensure that the frame
sequence that is decoded relates to the correct Source Address.
The first two frames that are transmitted by Source Address 40 00 00 00 00 01 are Duplicate
Addr (Address) Test frames. The next frame is a Report NAUN Change. The Report NAUN
Change is sent by the NIC to the Configuration Report Server (CRS) when a change is made
in the Upstream NeighborÕs Address as a result of the neighbor notification process. A
Standby Mon (Monitor) Present frame is then sent by Source Address 40 00 00 00 00 01.
The next frame is a Report NAUN Change; however, the frame is generated by a different
Source Address than the address that is being analyzed. The next four frames are the Req
Initialization frames that are transmitted by Source Address 40 00 00 00 00 01.


20.380 36 802.5 CfgReportSvr CR 400000000001 RS MAC Report NAUN Change
20.400 36 802.5 CfgReportSvr CR 40005494E000 RS MAC Report NAUN Change
Request Initialization Protocol Sequence

Figure 9
If one of the four Req Initialization frames is returned with the A-bit or C-bit set to a 1, an
RPS is present on the ring. To determine whether the A-bit or C-bit is a 1, the level 2
interpretative display for each frame must be examined. If an RPS is present on the ring, the
NIC waits for an Initialize Ring Station MAC frame from the RPS or a Change Parameter
MAC frame from the Network Manager.
Figure 10 shows a typical protocol analyzer level 2 interpretative display. Only the pertinent
part of the level 2 display is shown for each frame. Each frame must be analyzed to
determine whether the RPS recognized the functional address and copied the frame from the
network. If four Request Initialization MAC frames are received back after the frames have
circled the ring and the A-bit and C-bit in each frame are 0Õs, there is neither an RPS or a
Network Manager in operation on the network. Examination of the Frame Status field for
each of the four frames (the bottom line of the frame display) indicates that the functional
address (40 00 00 00 00 02) was not recognized; therefore, the level 2 display indicates
ÒDestination address not recognized...Ó In all four frames, the Frame Status field indicates
the frame was Ò... not copiedÓ from the network.
After four attempts to Request Initialization, the NIC assumes that no Ring Parameter Server
or Network Manager is active on the network, and it will use the default operational
parameter values. For this protocol sequence, the Ring Insertion Process is complete.


Type: MAC Length: 66 Address: 400000000001 RingParamSvr
Time Stamp: 00:00:20.401190
Token Priority: 0 Monitor: 1 Reservation: 0 Express Buffered: n
Frame Status: Destination address not recognized, not copied.
Frame 1

Time Stamp: 00:00:20.401322
Frame 2

Time Stamp: 00:00:20.401454
Frame 3

Time Stamp: 00:00:20.402227
Frame 4
No Response to Request Initialization Frames

Figure 10

However, if an Initialize Ring Station or a Change Parameters MAC frame is received, the
NIC will reconfigure and replace the default parameters with the parameters that are
embedded in the frame. If this protocol sequence occurs, the Ring Insertion Process also is
complete.
However, if the Request Initialization MAC frame circles the ring and the A-bit and C-bit are
set when the frame is received back, the NIC waits two seconds for a Response MAC frame.
If no Response MAC frame is received, the NIC retransmits the Request Initialization MAC
frame. After four unsuccessful retries, the NIC terminates the Ring Initialization Process.
When the Ring Insertion Process is complete, the NIC is both physically and logically
attached to the ring. A logical attachment to the ring means that the NIC is actively engaged
in the MAC ring protocol.
Counters and Timers Used to Monitor Protocol Activity
Counters and timers are used to monitor network protocol activity. A counter is incremented
and an elapse timer is started for each protocol sequence that occurs and for each frame that is
transmitted. Counters tabulate the number of events that occur. Elapse timers monitor the
time intervals between events. The term Òreset,Ó when applied to timers, means reset the
timer to its initial value and then restart the timer.
Counters and elapse timers manage and control the protocol operation. Counters and timers
check to ensure that once a protocol task is started, the task will conclude within a given
period of time. The task may conclude with a positive result or a negative result, but the
counters and timers ensure that the task ends. If the timer did not stop the function and force
the task to reset, a task that encounters a problem could go on forever. If a counter did not set
a threshold for the number of times a task can be performed, the task can be constantly reset
and retried and it also will never end.
Counters and timers work together. Each time that the task is tried, a counter is incremented.
If a task is started but it is not completed within the time that is allotted, the application
program can make a decision to reset the timer and to retry the task. After a given number
of attempts are made to complete a task, a threshold is established. When the counter
threshold is exceeded, the program stops the task. The application program determines the
steps that are to be performed after a task has been halted.

Counters
Counters are used by the Token Ring protocol to count the number of times that a particular
protocol event occurs. Isolating error counters isolate errors to the NIC that transmitted the
frame, to the NIC that received the frame, and to the cable and other hardware that connect
the two NICs together. The isolating errors are counted only by the first NIC that detects the
error. Other downstream NICs also detect the error, but because the first NIC already set the
E-bit to 1, the downstream NICs ignore and do not count the error. Isolating errors detect that
an error occurred on a ring between a particular upstream NIC and its corresponding
downstream neighbor.
Nonisolating error counters count errors that could have been caused by any other NIC that is
on the ring. The fault cannot be isolated to two specific NICs. The specific counters that are
implemented within the Token Ring protocol are described in the sections that follow.
Line Error - The Line Error counter is incremented when a frame in which the E-bit is a zero
and in which either one of the following conditions exist is copied or repeated by an NIC.
• A non-data symbol (J-bit or K-bit) appears in the frame between the Starting Delimiter
(SD) and the Ending Delimiter (ED).
• A Frame Check Sequence (FCS) error exists in the frame.
Internal Error - The Internal Error counter is incremented when an NIC detects a
recoverable internal error. An internal error indicates that the NIC is marginal and may be
defective.
Burst Error - The Burst Error counter is incremented each time that an NIC detects the
absence of received bit transitions for five half-bit times. This error is classified as a burst-
five order.
AC Error - The AC Error counter is incremented when an NIC receives an AMP or an SMP
frame in which both the A-bit and the C-bit in the Frame Status are set to zero, followed by a
second SMP frame in which the A-bit and C-bit are also set to zero. If an AMP frame is
received before the second SMP frame then, no error is counted.
Abort Delimiter Transmitted - The Abort Delimiter Transmitted counter is incremented

when an NIC transmits an abort delimiter during a frameÕs transmission.
Lost Frame Error - The Lost Frame Error counter is incremented each time that an NIC fails
to receive back a fame that is transmitted.

Receive Congestion Error - The Receive Congestion Error counter is incremented when an
NIC recognizes a frame that is addressed to its Destination Address, but the NIC has no
available buffer space to store the frame. The lack of buffer space indicates the NIC is
congested.
Frame Copied Error - The Frame Copied Error is incremented when an NIC recognizes a
frame that is addressed to its Destination Address and in which the frameÕs Address
Recognized (A-bit) in the Frame Status field is already set to a 1.
The Frequency Error is incremented when the incoming signal frequency differs by more than
0.01% from the expected frequency.
Token Error - The Token Error counter is incremented when an NIC that is acting as the
active monitor recognizes an error condition that requires a new token frame to be
transmitted. The token frame will be transmitted when the TVX timer that is in the active
monitor expires.
Frame Copied - The Frame Copied counter is incremented each time that an NIC originates a
frame and it is decremented each time that the frame is stripped from the network. The Frame
Copied counter indicates the number of frames that the NIC originated and that are still on the
ring.
Timers
Different types of timers are defined for the various protocol functions that occur within the
Token Ring protocol. The timers that are implemented within the Token Ring protocol are
defined in the sections that follow. Some of the timeout values have been changed as newer
versions of the IEEE 802.5 standard have been adopted. Equipment that complies with two
different versions of the IEEE 802.5 standard are now in use in networks. The two standards
are the IEEE 802.5 - 1989 version and the IEEE 802.5 - 1992 version. The timeout values
have been given for both standards.
Timer, Return to Repeat (TRR) - Each NIC has a TRR to ensure that the NIC will return to
a repeat state. The TRR timeout value is from 2.5 to 4 ms (IEEE 802.5 - 1989 defines the
default as 2.5 ms. IEEE 802.5 - 1992 defines the implementation as 4 ms). In a repeat state,
the bits that are received by the NIC are repeated on the wires of the ring to the next NIC.
Certain bits and fields that are in the repeated bit stream may be modified and actions may be
taken based on the content of the repeated bit stream. The maximum length of time that a
frame may be held by an NIC before the frame is transmitted to the next NIC is the TRR
timeout value. The TRR timer tracks how long the NIC has held the frame. When the TRR
timeout value is reached, the NIC is required to transmit the frame. The frame is transmitted
even if the frame content has not been fully processed by the NIC and actions have not been
taken by the NIC to revise the frame content.

Timer, Holding Token (THT) - Each NIC has a THT to control the maximum period of time
that the NIC may transmit frames after the Token has been captured. An NIC may transmit a
frame if the transmission can be completed before the THT timeout is exceeded. The THT
timeout value is between 8.9 ms and 10 ms. (IEEE 802.5 - 1989 defines the default as 10 ms.
IEEE 802.5 - 1992 defines the implementation as 8.9 ms.).
Timer, Queue PDU (TQP) - Each NIC has a TQP. The TQP is started after the NIC receives
an AMP frame or an SMP frame that has the A-bit and C-bit that are equal to 0. The TQP
timeout determines the maximum length of time that the NIC can wait after the AMP or an
SMP frame has been received before an SMP frame is transmitted. The TQP timeout is from
10 ms to 20 ms. (IEEE 802.5 - 1989 defines the default as 10 ms. IEEE 802.5 - 1992 defines
the implantation as 20 ms.).
Timer, Valid Transmission (TVX) - Each NIC has a TVX. The timer is activated only when
an NIC is the active monitor. The TVX timeout is used to detect the absence of valid
transmissions on the network. The timer verifies that data frames or token frames constantly
circle the network. The TVX timeout value, according to the IEEE 802.5 - 1989 standard, is
the sum of the timeout value of THT plus the timeout value of TRR. The IEEE 802.5 - 1992
standard defines the timeout value as 10 ms.
Timer, No Token (TNT) - If this timer does not correctly operate, it indicates that there is no
token circuiting the network and, because there is no token circling the network, no NICs can
access the ring. Each NIC that is a standby monitor has a TNT timer. A token frame or data
frame must constantly circle the network. If a valid token frame is not received every 1 to 2.6
seconds, the NIC enters the monitor contention process. (IEEE 802.5 - 1989 defines the
default as 1 second. IEEE 802.5 - 1992 defines the implantation as 2.6 seconds).
Timer, Active Monitor (TAM) - The TAM is used by the active monitor. The TAM default
timeout is 3 seconds. The expiration of the TAM timeout instructs the active monitor to
transmit an AMP frame. If no data frames or Token frames have been transmitted on the ring
in the last 3 seconds, the active monitor sends an AMP frame. Most NIC manufacturers set
the TAM timeout to 7 seconds. (IEEE 802.5 - 1989 defines the default as 3 seconds. IEEE
802.5 - 1992 defines the implantation as 7 seconds).
Timer, Standby Monitor (TSM) - Each NIC that is in the Standby Monitor mode has a
TSM. The TSM measures the elapsed time between AMP frames. The TSM default timeout
value is from 7 to 15 seconds. If an AMP frame is not received within 7 to 15 seconds, the
NIC enters the monitor contention mode and attempts to take over the function of active
monitor. (IEEE 802.5 - 1989 defines the default as 7 seconds. IEEE 802.5 - 1992 defines the
implementation as 15 seconds).

Timer, Error Report (TER) - Each NIC has a TER that is used to report error conditions
that are detected by the NIC. When the TER timeout expires, if any error counter is not zero,
the NIC sends a Report Error MAC frame to the network Error Monitor. The timeout value is
2 seconds. (The timeout value is the same for the IEEE 802.5 - 1989 and IEEE 802.5 - 1992
standards).
Timer BCN (Beacon) Transmit (TBT) - Each NIC has a TBT that is used to specify the
length of time that an NIC can remain in a Beacon (BCN) transmit mode. When the TBT
expires, the NIC enters the Bypass mode. The timeout value is 16 to 26 seconds. (IEEE
802.5 - 1989 defines the default as 26 seconds. IEEE 802.5 - 1992 defines the
implementation as 16 seconds).
Timer, BCN Receive (TBR) - Each NIC has a TBR that is used to control the length of time
that an NIC will continue to receive BCN frames from the downstream neighbor before the
bypass mode is entered. The timeout value is 160 milliseconds. (The timeout value is the
same for the IEEE 802.5 - 1989 and IEEE 802.5 - 1992 standards).
AMP Frames (Active Monitor Present) and SMP Frames
The purpose of AMP frames within the networkÕs protocol operation is to indicate to the
other NICs that an Active Monitor is present on the network. The purpose of the SMP frame
is to participate in the neighbor notification process of the Token Ring protocol operation.
AMP Frames (Active Monitor Present) and SMP Frames (ContÕd)
Any NIC on the ring can function as either an Active Monitor or a Standby Monitor. Only
one Active Monitor at a time can be present on a ring. The other NICs that are on the ring
serve as Standby Monitors. The Active Monitor ensures normal token operation and provides
the crystal-controlled master data clocking signal for data transmission. The Standby
Monitors ensure that the Active Monitor is properly functioning and is still inserted on the
ring.
The role of the Active Monitor is to ensure normal token operation. Any NIC may become
the Active Monitor by active participation in the monitor contention process. The NIC that
wins the contention process becomes the Active Monitor. The monitor contention process
will be discussed in detail later in this section.
The Active Monitor transmits an AMP frame whenever the TAM timeout expires. As shown
in Figure 11, AMP frames are typically transmitted every seven seconds and at the end of the
ring purge process. A ring purge will be discussed later in this section. Examination of the
Secs field in Figure 11 shows that if there is no other activity on the ring, the AMP frames are
transmitted every seven seconds. The following procedure is applied to every AMP frame or
SMP frame that is received by an NIC:

• If the Active Monitor transmits an AMP frame that is later received back with the A-bits
and C-bits in the FS field being both set to 0, the Active Monitor is the only NIC in
operation on the ring.
• If an AMP or SMP frame is received by an NIC and the A-bit and C-bit are not 0, the
Active Monitor is not the only NIC on the ring.
• If a Standby Monitor receives an AMP or SMP frame with the A-bits and C-bits set to 0,
no other NIC has copied the frame, and the Source Address in the frame contains the
Upstream NeighborÕs Address.
The NIC copies the Upstream NeighborÕs address from the frame and stores the address in
memory. The NIC compares the Source Address of the frame to the Upstream NeighborÕs
previously saved address. If the two addresses are different, the NIC saves the new address.
The NIC then queues an SMP frame for transmission. The SMP frame propagates the ring
poll to the next downstream neighbor.
The AMP and SMP frames engage in the ring poll process that enables each NIC to acquire
the address of its Upstream Neighbor NIC. In this protocol sequence, each NIC transmits its
address and physical drop number (4 octets) to the next downstream NIC. Each NIC stores
the Upstream NeighborÕs address in memory. The address field is constantly updated.
Updating of the Upstream Neighbor Addresses is known as the neighbor notification process.

03.484 36 802.5 Broadcast RS 41000001010 RS MAC Active Mon Present
03.516 36 802.5 Broadcast RS 40000000A400 RS MAC Standby Mon Present
Protocol Analyzer Display of AMP and SMP Frames

Figure 11

Through the examination of Protocol Analyzer Level 2 interpretative displays, the details of
the neighbor notification process can be more closely examined. Figure 12 shows a Protocol
Analyzer Level 2 interpretative display for the five frames that are involved in the neighbor
notification process. The Hexadecimal Frame decode has been eliminated from each
display. The Source field indicates the address of the NIC that originated the frame. Under
the heading ÒMedium Access Control VectorsÓ is the title ÒNearest Active Upstream
Neighbor.Ó The address of the Nearest Active Upstream Neighbor (NAUN) is displayed
adjacent to the title.
Frame #1 is the AMP frame. The Active Monitor NIC address is 41 00 00 00 10 10. The
nearest active upstream neighbor is 41 00 00 00 00 01.
Frame #2 is the SMP frame that is transmitted by the first Standby Monitor that is
downstream from the Active Monitor. The Source Address of the NIC is 41 00 00 00 00 03.
The display indicates that the Nearest Active Upstream Neighbor is azure 001010 (the Active
Monitor).
Frame #3 is the Standby Monitor frame that is transmitted by the second Standby Monitor
that is downstream from the Active Monitor. The Source Address of the frame is 40 00 00 00
A4 00. The display indicates that the Nearest Active Upstream Neighbor is 41 00 00 00 00
03.
Frame #4 is the Standby Monitor frame that is transmitted by the third Standby Monitor that
is downstream from the Active Monitor. The Source Address of the frame is 40 00 00 00 00
00. The display indicates that the Nearest Active Upstream Neighbor is 41 00 00 00 A4 00.
Frame #5 is the Standby Monitor frame that is transmitted by the forth Standby Monitor that
is downstream from the Active Monitor. The Source Address of the frame is 40 00 00 00 00
01. The display indicates that the Nearest Active Upstream Neighbor is 41 00 00 00 00 00.


Type: MAC Length: 36 Address: azure001010 Broadcast
Time Stamp: 00:00:03.484002
Token Priority: 0 Monitor: 0 Reservation: 0 Express Buffered: y
Type: Active Mon Present
Frame 1

Type: MAC Length: 36 Address: 410000000003 Broadcast
Time Stamp: 00:00:03.497228
Nearest Active Upstream Neighbor: azure001010
Frame 2

Type: MAC Length: 36 Address: 40000000A400 Broadcast
Time Stamp: 00:00:03.516687
Frame 3

Figure 12


Time Stamp: 00:00:03.530228
Nearest Active Upstream Neighbor: 40000000A400
Frame 4

Time Stamp: 00:00:03.543534
Frame 5

Figure 12 (ContÕd)

If the new address is different from the previously saved address, the NIC transmits a Report
SUA (stored upstream neighborÕs address) Change MAC frame to the network manager.
When all of the address relationships are reported to the network manger, the network
manager can develop a ring topology map.
Standby Monitor Timer (TSM)
The purpose of the Standby Monitor Timer (TSM) is to trigger the first PC that is attached to
a network to initialize the network protocol operation. Each NIC has a TSM. The TSM is
used by the Standby Monitor to ensure that there is an Active Monitor on the ring. The
Standby Monitor expects to receive a continuous flow of token frames, AMP frames, or Purge
frames. Each time that a frame of this type is received, the TSM timer is reset and the elapse
timer is started. If the correct type of frame is not received within seven seconds, the timer
expires and triggers the NIC to initiate a protocol sequence that may lead the NIC to become
the Active Monitor.
The TSM timer triggers the first PC that is attached to a network to initialize the networkÕs
protocol operation. In this scenario, no other NICs are attached to the network. In this
example, an NIC has just completed the physical and logical ring attachment process. At this
point, the NIC is the only device that is active on the ring. After the NIC is logically attached
to the ring, the NIC activates the TSM. The timer starts at the moment that the logical ring
attachment is complete. The TSM timer is reset each time that a new Token frame, AMP, or
Purge frame is received by the NIC; however, in this scenario, no other NIC is attached to the
ring; therefore, no frames are circling the ring. Because no frames appear on the ring, the
timer is never reset. After seven seconds, the TSM elapse time expires. Because the TSM
time has expired, the NIC assumes that no Active Monitor is present on the network.
Expiration of the TSM elapse time triggers the next step in the protocol sequence.
The next step in the protocol sequence is called monitor contention. The monitor contention
process involves all of the NICs that are in the ring. The process is used to establish an NIC
as an Active Monitor. The monitor contention process is started when any NIC on the ring
detects that no Active Monitor is in the ring or that the operation of the Active Monitor is
incorrect.

Latency Buffer
In order for the 24 bits of the Token frame to properly circle the ring when the bits are
repeated by all of the NICs that are on the ring, the ring must have a minimum of a 24-bit
propagation delay. The delay guarantees that the Token frame will not return to the sending
NIC before the Token frame is completely transmitted. The purpose of the Latency Buffer is
to introduce a 24-bit propagation delay into the bit stream that circles the ring. The delay
prevents the last bits that were transmitted by the NIC from being overlapped by the first bits
as they are received by the NIC that serves as the Active Monitor. A ring must be about
4,500 ft (1,371.6m) long to introduce enough propagation delay to ÒholdÓ the entire Token
frame. Because most rings are not 4500 ft long, the propagation delay of the ring is
artificially increased through the use of a first-in-first-out (FIFO) memory that is called a
Latency Buffer.
The time that is required for a signal to propagate once around the ring is the ring latency.
Ring latency is measured in bit times. Ring latency includes the signal propagation delay
through the main ring cable, the lobe cables, plus the NICs.
For a Token Ring network to correctly operate, Token frames must continuously circle the
ring even if there are no other frames to be transmitted. When no other frames are circling the
network, the Active Monitor defines the bit structure for a Token frame and then transmits
each bit of the Token frame through use of its own clock signal.
There are 24 bits (three octets) in the Token frame. The ring must have enough latency or
delay to hold the 24 bits of the Token frame. Only after the 24th bit of the Token frame has
been transmitted by the Active Monitor into the network may the first bit of the Token frame
be passed by the network back to the NIC that is the Active Monitor. The network must hold
the 24 bits of the Token frame. If the bit rate of the ring is 4 Mbps., the ring must have a
latency of 24 bits/4Mbps., which is equal to 6 microseconds. At 16 Mbps., the ring must have
a latency of 1.5 microseconds. If sufficient latency is not added to the Active MonitorÕs
receive FIFO, the transmit bit stream will be overlapped by the receive bit stream and the
Token frame will be destroyed.
Figure 13 is a picture that shows the way in which the Latency Buffer is used in the network.
The Active Monitor provides to the ring a master clock signal that is integrated with data bits.
The clock signal is propagated around the ring as a reference to indicate the location of each
data bit in the serial bit stream. The clock signals are produced by a crystal controlled
oscillator that is in the Active Monitor.

The Active Monitor transmits the master clock signal. All of the Standby Monitors (every
other NIC that is attached to the ring) derive their time base from the master clock signal that
is placed on the ring by the Active Monitor. The NICs that operate in a Standby Monitor
mode derive their time base by phase-synchronizing a voltage-controlled oscillator to the
incoming bit stream. A phase-locked-loop (PLL) circuit is used to derive the clock signal
from the ring. The clock that is derived from the received bit stream is the Latency Buffer
clock. The Latency Buffer clock provides the time pulses that are used to locate data bits in
the receive bit stream. The clock then is used to identify the line signal transition that is a data
bit to read the binary bit from the line and to latch the data bit into an elastic bit buffer. Next,
the Latency Buffer clock signal is used to read each bit from the elastic buffer and to transmit
the data bit to each network node. A node is a host PC that has an NIC that is active on the
ring.
The signal that is received by the Active Monitor must be specially treated because the bit
stream is no longer phase-aligned to the crystal controlled clock that is used by the Active
Monitor to generate the master clock. The loss of phase alignment is caused by propagation
delays and phase delays that are imposed on the signal as the signal is repeated by each NIC.
The misalignment is referred to as jitter.
To compensate for jitter, the Active Monitor NIC clocks the received serial bit stream into a
first-in-first-out (FIFO) elastic buffer with the PLL clock signal that is derived from the
received bit stream. The data is clocked out of the elastic buffer through use of the master
clock.

Latency Buffer Operation

Figure 13

No Token Timer
The function of the No Token Timer (TNT) is to time a Token related error so that the
protocol fault that was created by the error condition can be eventually cleared. The TNT
timer is reset each time that a valid Token frame is received. If the ring encounters a protocol
fault and no Token frames are on the network, the TNT timer will eventually expire.
The default timer range for TNT is one second; however, the IEEE 802.5 specification
indicates that the TNT timeout should be calculated as follows:
TNT = TRR timeout value (2.5 milliseconds default) + n times THT (where n is the
maximum number of stations on the ring). THT default timeout is 10
milliseconds.
A Token Ring network that has 72 active NICs will have a TNT timeout value of
approximately 722 milliseconds. A network with 260 active NICs will have a TNT timeout of
approximately 2.6 seconds.
As a result of the TNT timeout, the proper error recovery procedures are started to return the
ring to normal protocol operation. The TNT timeout initiates the Ring Purge Process. The
Ring Purge Process returns the ring to a normal protocol operation that allows frames to be
transmitted through use of the token protocol.
PC Transmission of Claim Token
The protocol sequence that causes a PC to transmit a Claim Token occurs when the TSM
(Timer, Standby Monitor) timeout expires in a Standby Monitor. If an AMP frame or a Purge
frame is not received before the TSM timeout, the NIC is programmed to assume that there is
no Active Monitor in the ring. The NIC resets the TNT timeout and enters the monitor
contention process.
The monitor contention process is used to establish an NIC as an Active Monitor. The
process is started when any NIC on the ring detects that no Active Monitor is in the ring or
that the Active Monitor protocol exchanges are not correct.
The monitor contention process starts when an NIC transmits a Claim Token MAC frame.
Other NICs on the ring will join the monitor contention process when a Claim Token MAC
frame is received. The NICs can join the monitor contention process in an active contention
mode or passive contention repeat mode. If one of the following conditions exist, an NIC
may actively enter the monitor contention process:
• The TSM timer in the NIC has expired and the NIC has transitioned to the monitor
contention process. The NIC has not received a Claim Token frame.

• The TSM timer in the NIC has expired, the NIC has transitioned to the monitor
contention process, and a Claim Token frame has been received. The Source Address in
the Claim Token frame has a lower address than the NIC that received the Claim Token.
The other NICs on the ring enter the monitor contention process in a passive role that is
known as the repeat mode. The NIC with the highest Source Address among the contenders
is established as the Active Monitor.
The counters and timers that interact with the Claim Token protocol sequence are the
following:
• Timer, Standby Monitor (TSM)

• Timer, No Token (TNT)
No counters are used in the Claim Token protocol sequence. Both the TSM and TNT timers
must expire before the monitor contention process can be initiated.
Figure 14 shows the Claim Token protocol sequence as it appears on the screen of a Protocol
Analyzer. A detailed explanation about the content of this display was presented in CDA
203.02.
Under the heading ÒDescriptionÓ in Figure 14, the type of frame that is analyzed is shown.
The first three frames are Beacon frames. Four Claim Token frames are shown after the
Beacon frames. A Ring Purge frame follows the Claim Token frame. The Source address for
the Beacon, Claim Token, and Ring Purge frame is ibm7BA7A9. Source Address
ibm7BA7A9 is actively engaged in the monitor contention process. The protocol analyzer
display indicates that Source Address ibm7BA7A9 entered the monitor contention process
because of a network fault that created a Beacon condition. The NIC with Source Address
ibm7BA7A9 detected the fault and transmitted a series of Beacon frames.
When a ring is in a Beacon condition, Token frames do not circle the ring. Because no Token
frames are detected, the TSM (Timer, Standby Monitor) and TNT (Timer, No Token) expire.
More information about a Beacon frame is presented later in this section.
When the TSM and TNT timers expire, the NIC with Source Address ibm7BA7A9 transmits
a series of Claim Token frames at 20 millisecond intervals. Claim Token frames are
immediately transmitted without waiting for a free Token frame.

When a downstream NIC receives the Claim Token frame, the NIC exits the Beacon repeat
mode and then the NIC examines the Claim Token frame Source Address field. If the Source
Address is higher than the NICÕs own address, the NIC retransmits the Claim Token frame
and then enters the contention repeat mode.
When the NIC that originated the Claim Token frames receives three successive Claim Token
MAC frames back, the NIC wins the monitor contention process. The NIC must receive back
the Claim Token frames that it originated in order to win the monitor contention process. The
NIC becomes the new Active Monitor. The NIC then exits the monitor contention process
and performs the ring purge process. Next, the NIC transmits a Report New Act (Active)
Mon (Monitor) frame and sends it to the CfgReportSvr (Configuration Report Server). At this
point, the Active Monitor sends an AMP frame. Downstream Standby Monitors respond and
send SMP frames.

36.982 40 802.5 Broadcast RS ibm7BA7A9 RS MAC Beacon
37.023 36 802.5 Broadcast RS ibm7BA7A9 RS MAC Claim Token
37.083 36 802.5 Broadcast RS ibm7BA7A9 RS MAC Ring Purge
37.084 56 802.5 CfgReportSvr CR ibm7BA7A9 RS MAC Report New Act Mon
37.084 36 802.5 Broadcast RS ibm7BA7A9 RS MAC Active Mon Present
37.097 36 802.5 Broadcast RS azure001004 RS MAC Standby Mon Present
37.117 36 802.5 Broadcast RS ibm3A4D3A RS MAC Standby Mon Present
37.135 36 802.5 Broadcast RS ibm3A61F8 RS MAC Standby Mon Present
Claim Token Protocol Display

Figure 14
The nodes that receive a Claim Token transmission will respond in the following ways:
• When an NIC enters the contention repeat mode, the NIC resets the TNT timer. TNT
serves as a watchdog timer to ensure that the monitor contention process does not
continue indefinitely. If the monitor contention process cannot be resolved in the 1
second (default) timeout value that is set by TNT, the NIC enters the Beacon
transmission process when the TNT timeout expires.

• An NIC that is not actively engaged in the monitor contention process when it receives a
Claim Token frame takes the following action:
- If the Source Address in the Claim Token frame is less than the Source Address of
the NIC that has received the frame, the NIC enters the monitor contention process.
If the condition is not met, the NIC enters the monitor contention repeat mode.
- If the Source Address in the frame is equal to the address of the NIC that received
the frame, the frame is ignored. The NIC enters the monitor contention repeat
mode.
• If an Active Monitor receives a Claim Token frame, it performs the following functions:
- The Active Monitor deactivates the active monitor functions and assumes the
status of Standby Monitor.
- Once the NIC is in the Standby Monitor mode, it enters the monitor contention
repeat mode.
- The NIC queues an error report for transmission. The report is sent to the network
error monitor. The error report is a report monitor error MAC frame with an error
code subvector value that is equal to >0001. The error report indicates that a
Standby Monitor detected an error in the Active Monitor. The error report is
transmitted after the monitor contention process is completed.
MAC Purge Frame
The purpose of a MAC Purge frame is to put the ring into a normal condition that allows
frames to be transmitted through the use of the token protocol. The Ring Purge process is
started by the Active Monitor when one of the following conditions occurs:
• A token error condition is detected by the Active Monitor.

• An NIC has just won the monitor contention process.
The protocol sequence that causes a Purge Frame to be transmitted was described in the
previous section, titled ÒPC Transmission of Claim Token.Ó
The following protocol transactions occur as a result of a Purge Frame transmission. When
the Active Monitor enters the Ring Purge process, it transmits a Ring Purge MAC frame and
resets the TNT (Timer, Not Token). TNT has a default timeout period of one second;
therefore, the Active Monitor has one second to complete the Ring Purge process.

The Active Monitor does not have to be in possession of the Token to be able to transmit a
frame. The frame is transmitted before a Token frame is received. No Token frame is
released after the Ring Purge MAC frame is removed from the ring. After the Ring Purge
MAC frame is transmitted, the NIC sends continuous idles (binary 0s).
When the Active Monitor receives the Ring Purge MAC frame back, it checks the frame for
errors. The errors can be transmission errors that are detected through the examination of the
FCS field or the errors can be frame format errors. The Active Monitor continues to transmit
Ring Purge MAC frames until a frame is received without errors or until the TNT timer
expires.
The TNT timer functions as a watchdog timer to limit the time that the NIC may continue to
transmit Ring Purge MAC frames. The TNT timer is reset when the Active Monitor receives
one error-free Ring Purge MAC frame.
After the Active Monitor receives an error-free Ring Purge MAC frame, it transmits a Token
frame with the priority bits equal to the reservation priority of the last ring purge MAC frame
that was received by the NIC.
When a Ring Purge MAC frame is received by an NIC that is in the monitor contention mode,
it returns to normal operation and implements the Standby Monitor functions.
If an NIC that believes it is the network Active Monitor receives a Ring Purge MAC frame,
the NIC checks the Source Address of the frame to determine if it transmitted the frame. If
the Source Address does not belong to the NIC, then the NIC deletes the Latency Buffer it
installed in the ring, resets the TNT, TSM, and TER timers, and transitions to the Standby
Monitor mode.
Figure 15 shows how the Ring Purge MAC frame protocol sequence appears in a Level 2
Interpretative display format on the screen of a protocol analyzer. Information that is relative
to the Level 2 Interpretative display format was presented in Module CDA 203.02. In the
example, a series of Ring Purge frames was transmitted one after the other until the last frame
was received error-free. When a Ring Purge frame is received error-free, a Token frame is
transmitted followed by an AMP frame. Downstream NICs respond with Standby Monitor
frames.
When the Active Monitor transmits the first Ring Purge MAC frame, it resets the TNT timer.
If the Active Monitor is not able to complete the Ring Purge process before the TNT timer
expires (the default time is one second), the NIC reports a ring failure condition to the
application software that is in operation in the PC and disconnects from the ring. The user
must then request the NIC to establish a new ring connection.


Ring Purge MAC Frame Protocol Display

Figure 15
Duplicate Address Test
The purpose of the Duplicate Address Test (DAT) is to verify that each NIC address is unique
to the ring. The DAT ensures that the Source Address that is used by an NIC to transmit
frames is not used by another NIC that is inserted onto the ring.
If the Duplicate Address Test was not performed and two NICs are allowed to simultaneously
operate on the ring, the frames sent by one NIC may end up at the correct address; however,
the frames may end up on the wrong PC because the NICs in two different PCs have the same
address. If two NICs had the same address, the origination point and destination point of
frames would become confused. Because of the way that the MAC layer protocol operates,
the duplicate addresses would go undetected, and frames would appear to be correctly
delivered to the Destination Address. The higher layer protocols would report that some of
the Information fields were correctly received and that other Information fields were not;
therefore, some application program tasks would complete and other tasks would never
complete. As a result, the task would timeout and the PC would end up off line. If the
network has a problem that is similar to the one that was just described, a Duplicate Address
Test should be run on each NIC that is on the ring to determine whether a duplicate address
exists on the ring.

The Duplicate Address Test is performed only on a ring basis. The test is not transmitted to
every ring segment in a network each time that an NIC attaches to the network. If there are
multiple rings in a network and if all of the rings are connected together through Bridges or
Routers, the Duplicate Address Test will be conducted only on the ring onto which an NIC
has attached. The Duplicate Address Test is not broadcast network wide; therefore, an NIC
on another network ring may have the same address. There is no test by which duplicate
addresses can be identified on a multiple ring (multiple segment) network.
The Duplicate Address Test is performed during the Ring Insertion sequence. The Ring
Insertion sequence occurs when an NIC attaches to the ring, and it consists of several phases:
Lobe Media Check, Physical Insertion, Address Verification, Neighbor Notification, and
Request Initialization. The Duplicate Address Test is performed during the Address
Verification phase of the Ring Insertion sequence.
The protocol sequence that occurs as a result of the DAT frame transmission depends on the
way in which the A-bit (Address Recognized) and C-bit (Frame Copied) are set when the
DAT frame is received back by the NIC that transmitted the frame. When the DAT frame is
transmitted, the NIC inserts its own address in both the Destination Address and Source
Address fields. If another NIC on the ring has an address that matches the Source Address in
the DAT frame, then it will set the A-bit (Address Recognized) to a 1. The frame copied bit,
C-bit, may also be set to a binary 1 if the frame was copied by the NIC. If the DAT frames
are received with either the A-bit or the C-bit set to 1Õs, the NIC removes itself from the ring
and terminates the Ring Insertion sequence. The application program that is in operation in
the PC is notified that the Ring Insertion sequence cannot be completed because the Duplicate
Address Test has failed.
The NIC that originated the DAT frame assumes that another NIC has the same address when
it receives two Duplicate Address Test MAC frames with the A-bit and C-bit set to 0Õs.
If the Duplicate Address Test sequence is not completed within 18 seconds, the NIC
terminates the Ring Insertion process and removes itself from the ring. The error condition is
reported to the application program.

The Duplicate Address Test protocol sequence consists of a series of Duplicate Address Test
(DAT) MAC frames that are transmitted by the NIC. The Duplicate Address Test protocol
sequence, as it is shown on the screen of a protocol analyzer, is presented in Figure 16. A
detailed description of the information that is shown in a Protocol Analyzer Level 2
Interpretative display was covered in Module CDA 203.02. The Duplicate Address Test is
transmitted after the Lobe Test. The Lobe Test verifies the continuity of the lobe cable. After
the lobe cable is tested, the Physical Insertion sequence occurs where the NIC is inserted into
the ring. Next, the Address Verification phase of the Ring Insertion process occurs. Address
Verification is performed when the NIC sends the Duplicate Addr Test (DAT) frame. Upon
the successful completion of the Duplicate Address Test, the NIC sends a Report NAUN
Change MAC frame (also called a Report SUA Change) to the CfgReportSvr (Configuration
Report Server, also known as the CRS). The Report NAUN Change MAC frame notifies the
Configuration Report Server that address 40 00 00 00 09 99 has a new upstream neighbor.
The Report NAUN Change MAC frame contains the address of the new upstream neighbor.
The NIC then transmits an SMP frame that is followed by a series of Req Initialization MAC
frames.

37.377 36 802.5 Broadcast RS 40005494E000 RS MAC Standby Mon Present
37.422 36 802.5 CfgReportSvr CR 400000000999 RS MAC Report SUA Change
Duplicate Address Test

Figure 16

IDENTIFYING NORMAL TOKEN RING PROTOCOL OPERATION
To understand the normal Token Ring Protocol operation, a protocol analyzer must be
installed in the ring network to monitor the protocol transactions. LAN protocol analyzers are
used to selectively eavesdrop on all LAN transmissions. The protocol analyzer captures all of
the frames and then analyzes each frame to determine the reason that the frame was
transmitted.
This section explains the normal Token Ring Protocol operation and includes the following
topics:
• Active Monitor Present (AMP)

• Standby Monitor Present (SMP)
• Token Transmission
Active Monitor Present (AMP)
The Active Monitor times various protocol transactions. If a protocol error occurs, the
timeout value of the timer will expire and trigger the Active Monitor to initialize a protocol
sequence that will allow the network to recover from the protocol fault.
The Active Monitor is the NIC that has the highest Source Address and that wins the monitor
contention process. The Active Monitor performs the following functions:
• Provides the master clock from which all other NICs in the ring derive their timing.
• Monitors the network to ensure that a frame or a Token frame circles the ring every 10
milliseconds.
• Removes Token frames that have a priority that is greater than zero from the ring after
the Token has circled the ring once.
• Removes all data frames that circle the ring more than one time.
• Inserts a latency buffer in the ring to guarantee that a Token frame constantly circles the
ring.
• Transmits an AMP frame to notify the other NICs that an Active Monitor is active on
the ring.

The Active Monitor is not an NIC with special network hardware. Any NIC on the ring can
become the Active Monitor. All of the other NICs on the ring function as Standby Monitors.
The choice of which network node (NIC) assumes the function of the Active Monitor is
realized through the monitor contention process. The monitor contention process was
described in the previous section, titled ÒIdentifying the Ring Initialization Sequence of a
Token Ring.Ó
Purpose of the Active Monitor Present Frames
The AMP frame serves several purposes:
• Notifies the Standby Monitors that the Active Monitor is present in the ring.
• Initiates the Neighbor Notification Process (also called the Ring Poll Process).
The Active Monitor will transmit an AMP MAC frame in the protocol sequence every seven
seconds to initiate the Ring Poll Process. The AMP frame is also transmitted at the end of the
Ring Purge process.
The presence of the AMP frame on the ring resets the TSM (Timer, Standby Monitor) in the
Standby Monitor NICs. The AMP frame must be received by each Standby Monitor every
seven seconds so that the Standby Monitor knows that an Active Monitor is in operation.
Examination of the ÒSecsÓ column in the Level 2 Protocol Analyzer Display that is shown in
Figure 17 reveals that the AMP frames are transmitted every seven seconds. The description
of the manner in which a Level 2 Protocol Analyzer Display is decoded was covered in CDA
203.02.
The first AMP frame was transmitted by Source address azure 001010 at 03.484 seconds.
The next AMP frame was transmitted by the same Source address seven seconds later at
10.473 seconds. The third AMP frame was sent at 17.462 seconds and the last AMP frame
was sent at 24.451 seconds. Transmitted between the AMP frames are SMP frames. The
AMP frames and SMP frames are the protocol transactions that are initiated by the Active
Monitor and Standby Monitors that are engaged in the Ring Poll Process.


03.484 36 802.5 Broadcast RS azure001010 RS MAC Active Mon Present
Protocol Analyzer Display of Active Monitor Present Frames

Figure 17
If the TSM timeout expires in a Standby Monitor, the Standby Monitor initiates the Claim
Token process and attempts to take over the task of Active Monitor.
Protocol Sequence
The protocol sequence that occurs after the AMP frame is transmitted will vary dependent on
the setting of the bits that are in the Frame Status (FS) field of the MAC frame after the frame
has circled the network and has arrived back at the Active Monitor.
• If an AMP frame is received by an Active Monitor, the Active Monitor checks the
Source Address field of the frame to verify that it originated the frame. The A-bit and
C-bit in the FS field are examined to determine whether another NIC received the frame.
If the A-bit and C-bit are both 0Õs, the Active Monitor is the only NIC on the ring. If
the bits are not both 0Õs, the NIC is not the only device on the ring.
• If an AMP frame is received by a Standby Monitor, the TSM timeout is reset to zero.
• When an AMP or an SMP frame is received that has the two A-bits and the two C-bits
set to different binary values in the Frame Status (FS) field, a poll failure has occurred.
The NIC saves the Source Address in the frame as the latest poll address, and when the
next Report Error MAC frame is transmitted to the Error Monitor, the failure is reported.
• If an NIC receives an AMP frame or an SMP frame with the A-bits and C-bits all set to
0Õs, the NIC knows that no other NIC has copied the frame. The NIC compares the
Source Address in the frame to the upstream neighborÕs address that is stored in its

memory. If the Source Address does not equal the Stored Upstream NeighborÕs
Address (SUA), the NIC saves it as the new upstream neighborÕs address. A Report
SUA Change MAC frame is then queued for transmission to the Configuration Report
Server.
Two Active Monitors
When two network nodes are transmitting AMP frames at the same time, a protocol fault
exists. If the Active Monitor receives an AMP frame that contains the Source Address of
another NIC, then two NICs think that they are the Active Monitor. Only one NIC may be the
Active Monitor on a ring.
The Active Monitor that receives the AMP frame that contains the Source Address of a
different NIC resets its active monitor functions and sends a Report Active Monitor Error
MAC frame to the Error Monitor. An indication that a duplicate monitor was detected is
placed in the MAC frame subvector. The NIC then configures as a Standby Monitor. When
the NIC receives the next AMP frame or SMP frame, it responds as a Standby Monitor and
transmits an SMP frame.
If both of the devices that think they are the Active Monitors detect that another NIC has
assumed the task of Active Monitor, both NICs may reset and assume the duties of a Standby
Monitor. In this case, the ring has no Active Monitor. When this condition occurs, the TSM
(Timer, Standby Monitor) timeout period will expire in one of the ring NICs. The timeout
will trigger the NIC to enter the monitor contention process. The NIC will transmit a Claim
Token and bid to take over the Active Monitor functions. If the NIC wins the bid, it becomes
the Active Monitor.

Protocol Analyzer Display of AMP Frame Protocol
A normal AMP frame protocol sequence, as displayed on a Protocol Analyzer, is shown in

Figure 18. Figure 18 contains an example of a Level 2 Protocol Display that includes many
of the protocol exchanges that have been discussed in this Module. Figure 18 shows that
source address 40 00 54 94 E0 00 transmitted a Duplicate Addr Test while it was engaged in
the Ring Insertion process. The frame was sent twice. Source Address azure 00 10 10 is the
Active Monitor. The Active Monitor transmitted an AMP frame to start the Ring Poll
process. The Standby Monitors participated in the Ring Poll process through the transmission
of SMP frames. When Source address 40 00 54 94 E0 00 first seized the token after the Ring
Insertion process had been completed, it transmitted a Report SUA Change MAC frame. The
frame was sent to the Configuration Report Server (CfgReportSvr). The frame reported that
Source address 40 00 54 94 E0 00 has a new Stored Upstream NeighborÕs Address (SUA).
The new address is included in the MAC Vector field of the frame. Source address 40 00 54
94 E0 00 then participated in the Ring Poll process through the transmission of an SMP
frame. After the SMP frame was stripped from the network, Source Address 40 00 54 94 E0
00 sent a single Req Initialization frame.
Source Address 41 00 00 00 00 01 seized the Token frame next and transmitted a Report
SUA Change frame to the CfgReportSvr. The Token was then released. Source Address 40
00 54 94 E0 00 again seized the Token frame and completed the transmission of the three
additional Req Initialization MAC frames that are addressed to the Ring Parameter Server
(RingParamSvr).

24.099 22 802.5 40005494E000 RS 40005494E000 RS MAC Duplicate Addr Test
24.099 22 802.5 40005494E000 RS 40005494E000 RS MAC Duplicate Addr Test
24.144 36 802.5 CfgReportSvr CR 40005494E000 RS MAC Report SUA Change
24.155 36 802.5 Broadcast RS 40005494E000 RS MAC Standby Mon Present
24.156 66 802.5 RingParamSvr RP 40005494E000 RS MAC Req Initialization
24.156 36 802.5 CfgReportSvr CR 410000000001 RS MAC Report SUA Change
24.156 66 802.5 RingParamSvr RP 40005494E000 RS MAC Reg Initialization
Ring Poll Sequence

Figure 18

An abnormal AMP frame protocol sequence, as displayed on a Protocol Analyzer, is shown in

Figure 19. Figure 19 shows a Ring Poll sequence that has two AMP frames. Source address
40 00 00 00 00 08 and 40 00 00 00 00 48 both think that they are the Active Monitor. This
protocol sequence is abnormal. The frames were analyzed 42 seconds after the start of the
last minute as indicated in the Secs column. The other NICs that are on the ring are in normal
operation and respond to the Active Monitor frame through the transmission of SMP frames.

42.997 36 802.5 Broadcast RS 00008381896B RS MAC Standby Mon Present
Abnormal Ring Poll Process

Figure 19
No Active Monitor Present on the Network
When two NICs that think they are the Active Monitor detect that another NIC has assumed
the task of Active Monitor, both NICs may reset and assume the duties of a Standby Monitor.
When this condition occurs, no active monitor will be present on the network. In the Standby
Monitor mode, the NICs reset the TSM timeout and wait for an AMP frame. When no node
assumes the position of Active Monitor, all of the NICs function in the Standby Monitor
mode. The network is in a failure mode. In order to trigger the network to recover and to
return to normal protocol operation, the TSM timer that monitors the Active Monitor
operation must expire. The TSM timeout expires in seven seconds.

When the timeout expires, the NIC declares that the ring has no Active Monitor and the NIC
enters the monitor contention process. As shown in Figure 20, every NIC that is on the ring
entered the monitor contention process. In the example, many of the NICs on the ring have
transmitted Claim Tokens. The NIC with the lowest Source address transmitted the first
Claim Token. As each Claim Token frame was processed by the other NICs that are on the
ring, each NIC with a higher Source address value than the address that appeared in the Claim
Token frame issued a new Claim Token frame.

58.625 36 802.5 Broadcast RS 400000000001 RS MAC Claim Token
NICs Respond to Abnormal Ring Poll Sequence

Figure 20
The NIC with the highest Source Address wins the Claim Token process and the monitor
contention. As shown in Figure 21, Source Address 40 00 00 00 00 59 won the monitor
contention process and became the Active Monitor. When the monitor contention process
ended, Source Address 40 00 00 00 00 59 transmitted a Ring Purge frame and then an AMP
frame. The other NICs that are on the ring responded with SMP frames. The ring protocol
has now returned to normal.


42.895 36 802.5 Broadcast RS 400000000059 RS MAC Ring Purge
NIC Wins Monitor Contention and Becomes Active Monitor

Figure 21
Standby Monitor Present (SMP)
The Standby Monitor is an NIC that has not assumed the task of Active Monitor. The
Standby Monitor monitors the network protocol transmissions for AMP frames or SMP
frames. The Standby Monitor monitors the ring protocol for the following:
• A good Token frame must be received at least once every 2.6 seconds. A good Token
frame is defined as a Token with zero priority, or a Token frame with a priority that is
greater than zero that is followed by a frame with a priority field that is greater than
zero.
• A good AMP frame must be received every seven seconds.
• The clock frequency that is derived from the received ring signal must be acceptable. If
the frequency deviates too much from an acceptable value, the deviation indicates that
there is no Active Monitor on the ring or that the clock signal that is transmitted by the
Active Monitor is not acceptable. If a frequency error is detected, the Standby Monitor
engages in the monitor contention process.
• If the NIC is no longer receiving a signal from the ring, it enters the monitor contention
process.
When the Standby Monitor determines that all three conditions are met, it then queues an
SMP frame for transmission.
Purpose of the Standby Monitor Present Frames

The purpose of the SMP frame is to participate in the Ring Poll process and to notify the other
NICs that are on the ring that the Active Monitor is engaged in normal protocol activity.
Transmission of the SMP frame by an NIC satisfies the following protocol functions:
• The NIC is engaged in the neighbor notification process.
• The NIC detected an AMP frame or an SMP frame.
• The transmission of an SMP frame is an acknowledgement that the NIC is receiving

Token frames.
• The clock frequency that is received by the NIC is within an acceptable range.
Any NIC that is not the Active Monitor but has completed the Ring Insertion process
performs the functions of a Standby Monitor, and it is responsible for the transmission of
SMP frames. The function of the Standby Monitor is to monitor the events on the ring to
determine the operational status of the Active Monitor.
The Standby Monitor functions are disabled while the NIC is in the ring insertion process, in
the monitor contention process, or in a beacon state.
When an NIC functions as a Standby Monitor, the NIC derives its clock from the received bit
stream. The same clock signal then is used to time the transmission of the bits that are
transmitted by the NIC.
Protocol Sequence
The SMP frame is transmitted in the protocol sequence after an AMP frame or SMP frame is
received. The A-bit and C-bit in the FS field of the MAC frame must be set to 0Õs. After an
AMP frame or an SMP frame is received, the Timer Queue PDU (TQP) is reset. When the
TQP timeout expires, an SMP frame is queued for transmission when the next available
Token frame is acquired.

The transmission of the SMP frame is delayed by the TQP timer for ten milliseconds. The
purpose of the delay is to ensure that many SMP frames are not transmitted one after the
other, thereby using all of the network bandwidth. The TQP timer is an attempt by the
protocol designers to limit the bandwidth that is used by the transmission of SMP frames to a
maximum of 1%.
An NIC that receives an AMP frame or an SMP frame that has the A-bits and C-bits set to 0
knows that the frame was originated by its upstream neighbor. The upstream neighborÕs
address is copied from the Source Address of the frame into memory and is retained as the
Stored Upstream neighbor Address (SUA). When the NIC transmits its own SMP frame, the
SUA address is placed in a MAC subvector field.
No AMP Frame Received by Standby Monitor
When a Standby Monitor does not receive an AMP frame prior to the TSM timeout, the
Standby Monitor assumes that the Active Monitor has failed. The Standby Monitor responds
to the failure condition through the monitor contention process. The Standby Monitor
transmits three Claim Token frames in an attempt to win the monitor contention process and
become the Active Monitor. The Standby Monitor stops sending Claim Token frames if one
of the following conditions occur:
• A Claim Token frame is received and the Source Address in the frame is greater than the
address of the NIC that has received the frame. If two or more NICs transmit Claim
Token frames, the NIC with the higher address becomes the Active Monitor.
• If a Beacon MAC frame is received, the NIC will stop the transmission of Claim Token
frames. The reception of a Beacon MAC frame indicates that a major ring failure has
occurred.
• If a Purge frame is received, the NIC knows that another NIC has won the monitor
contention process and has taken on the task of Active Monitor.
When the Standby Monitor that originated the Claim Token frames receives three successive
Claim Token MAC frames back, the NIC wins the monitor contention process and becomes
the new Active Monitor.

Upstream and Downstream Neighbor Addresses
The IEEE 802.5 Token Ring specification includes built-in diagnostics that are used by NICs
to isolate problems with the ring. Many of the diagnostics use the concept of failure domain
or fault domain to indicate the location of a fault in the ring.
Because the flow of data in a ring is unidirectional (the signal flows in one direction), every
NIC has an upstream neighbor and a downstream neighbor, as shown in Figure 22. The
active upstream neighbor is the NIC from which data is received. The active upstream
neighbor is called the Nearest Active Upstream Neighbor address (NAUN) or the Stored
Upstream Neighbor Address (SUA). Both the NAUN and SUA abbreviations are used in
protocol analyzer displays to identify the nearest active upstream neighbor address. Data is
transmitted to the active downstream neighbor. The use of the word Active implies that the
NAUN may not be the nearest physical upstream neighbor because the physically closest
neighbor may not be physically and logically attached to the ring.
Each NIC that is active in the ring knows the address of its nearest active upstream neighbor.
The NAUN is determined by the neighbor notification process that is activated by the
transmission of the AMP frame.
Active Upstream and Active Downstream Neighbors

Figure 22

Protocol Analyzer Display of Upstream and Downstream Addresses
The NAUN is displayed on most protocol analyzers in the Level 2 Interpretative Display.
Figure 23 shows a Level 2 Interpretative Display of an SMP frame. A description of the
information that is contained in a Level 2 Interpretative Display was covered in Module CDA
203.02.
When an SMP frame is transmitted, the NAUN address is included in the MAC subvector. In
the example that is shown in Figure 23, the Nearest Active Upstream Neighbor address is
shown under the heading ÒMedium Access Control Vectors.Ó The SMP frame indicates that
Source address 40 00 00 00 A4 00 reports that the Nearest Active Upstream Neighbor address
is 41 00 00 00 00 03.

03.516 36 Broadcast RS 40000000A400 RS MAC Standby Mon Present
Type: MAC Length: 36 Address: 40000000A400 Broadcast
Time Stamp: 00:00:03.516687
Standby Monitor Present (SMP) Frame

Figure 23
The downstream neighbor address is not included in the SMP frame. To obtain a complete
map of the addresses that appear on the ring and the relationship of one address to another
address, the network manager program must be accessed. Many of the network manager
programs print ring topology maps such as the one that is shown in Figure 24. The Active
Monitor is represented by the box that is labeled AM. Each box contains the Source Address
of an active NIC. Address 400000001010 is the Active Monitor. The arrow indicates the
direction of transmission. The Active Monitor (AM), address 400000001010, transmits to the
nearest active downstream neighbor 410000000003. The entire topology of the ring can be
seen in the map. Address 410000000003 transmits to the nearest downstream neighbor
address 40000000A400. Address 40000000A400 transmits to the nearest downstream
neighbor address 4100000000000. Address 410000000000 transmits to the nearest
downstream neighbor address 40005494E000. Address 40005494E000 transmits to the
nearest downstream neighbor address 410000000000. Address 410000000000 completes the
ring and sends data to the AM. The ring topology map provides a complete description of the
location of active NICs.

Ring Topology Map

Figure 24
Token Transmission
A Token Ring network is constructed of a set of PCs that are serially connected by a pair of
wires. The PCs connect to the ring wires through an interface card that is called a Network
Interface Card (NIC). All of the information is serially transferred bit-by-bit from one active
NIC to the next active NIC. In the ring, each NIC receives, regenerates, and then repeats each
bit to the next active downstream neighbor.

Purpose of Token Transmission
The purpose of Token Transmission in the network protocol is to control access to the ring.
Conceptually, the Token is a symbol of authority that is passed between NICs. The Token is
merely a specific bit sequence that circulates among the NICs and that gives the NICs
permission to transmit. Because of the way that the Token sequentially moves around the
network, Token Ring networks are sometimes described as Òdistributed polling
environments.Ó The Active Monitor ensures that a Token frame constantly circles the
network. When an NIC receives the Token frame, it may hold the Token frame and replace
the Token frame with data.
The NIC that currently possesses the Token has access to the ring and may transmit
information. All of the other active NICs on the ring must be silent. The Token is a data
frame that is three octets long (24 bits). When the NIC that currently holds the Token has
completed the transmission of data, it creates a new Token frame and transmits the Token
frame into the network. The new Token frame provides the other NICs with the opportunity
to gain access to the ring.
A token holding timer (THAT) controls the maximum period of time that a station may
transmit data before the token is released. The THT default is ten milliseconds.
Protocol Sequence
A Token frame is transmitted in the protocol sequence when one of the following conditions
exist:
• After an NIC has won the monitor contention process and has completed the Ring Purge
process.
• When the THT timeout expires.
• When the NICÕs transmit buffer is empty and before the THT timeout expires.
• The Active Monitor ensures that a Token frame constantly circles the ring every 2.5
milliseconds.

When an NIC queues a MAC or LLC frame in the transmit buffer, it begins to search for a
Token frame. The NIC searches for Starting Delimiters. The unique bit configuration of a
Starting Delimiter allows the NIC to easily identify a new frame. When a new frame is
detected, the NIC examines the first octet after the Starting Delimiter. The octet is called the
Access Control. The fourth bit is the Token indicator bit. This bit is set to 0 for Token frames
and to a 1 for MAC and LLC frames.
If the Token indicator bit is a 0, the frame is a Token frame. Once the NIC has determined
that the frame is a Token frame, the NIC examines the priority bits that have been set in the
Access Control field. Eight priority levels can be set: 000 is the lowest and 111 is the highest.
If the MAC or LLC frame that the NIC has queued for transmission has a priority level that is
equal to or greater than the priority level of the Token frame, the NIC may seize the Token
frame and remove it from the network. The NIC now has access to the ring. An NIC may
remove the Token frame from the ring only if the priority conditions are met.
If the Token indicator bit is a 1, the frame is a MAC or LLC frame and not a Token frame. In
this case, the NIC examines the Reservation bits that appear in the Access Control field. If
the Reservation bits are 000, the NIC executes a program to set the reservation bits to reserve
the next Token frame. If the MAC or LLC frame that is queued for transmission has a higher
priority level than the reservation bits that are set in the AC octet, the NIC may change the
bits to reflect the reservation level of the frame. If the priority level of the frame that is
queued for transmission is lower, the reservation bits cannot be changed.
When the NIC that originated the frame receives the frame back, the Reservation bits are
copied from the received frame into the Priority bits of the new Token Frame. The
Reservation bits in the new Token Frame are set to 000.
The protocol sequence that occurs after a Token frame is transmitted requires the NICs that
have MAC or LLC frames in queue to detect whether the received frame is an information
frame and to determine the frameÕs priority level. The Token Frame now has the priority
value requested by the NIC that placed the reservation for the Token. The new Token Frame
with a high priority is transmitted around the ring. Only NICs with higher or equal priority
can capture the Token. The NIC that reserved the Token Frame has an equal priority level to
that of the Token; therefore, the NIC can capture the Token and transmit information frames
on the ring.

Priority Bits have been assigned by the IBM Corporation for the different tasks that are
performed by NICs that are attached to a Token Ring network. The Priority Bit assignments
are as follows:
Priority Bit Function
000 Normal user priority. For use by Media Access Control (MAC)
Frames that do not need to possess a token for transmission, such as
a Beacon.
001 Normal user priority.
010 Normal user priority.
011 Media Access Control (MAC) Frames that require a token for
transmission.
100 Used by a Bridge or Router when it wants a token.
101 Reserved for future use.
110 Reserved for future use.
111 Used for Specialized Station Management

If the high priority token is not used after one pass around the ring, the Active Monitor resets
the Priority bits to 000. Zero is the lowest priority level. With this priority level, any NIC on
the ring can use the token.
After the transmission of the frame has been completed, the NIC checks the Source Address
to verify that the frame has entirely circled the ring. If the frame has not been entirely
transmitted around the ring, the NIC transmits fill bits until the Source Address has been
received. The NIC then transmits a Token frame. Fill bits may be either a series of 0 bits or 1
bits, or any combination. After the Token frame has been transmitted, the NIC remains in the
transmit state until all of the frames that the station originated have been removed from the
ring.
If an NIC detects any one of the following conditions, an Abort Delimiter is transmitted:
• The Token frame does not end with an ED.

• A Beacon is received.
• A Purge frame is received.
• A Claim Token frame is received.
• A logic error has occurred within the NIC.
When an NIC detects any one of the previously listed conditions, it transmits an Abort
Delimiter. The NIC also increments the abort transmitted error counter. The NIC then
transmits fill bits until the Abort Delimiter has circled the network and has been removed.
The NIC then issues a new Token frame.
The structure of an Abort Delimiter is shown in Figure 25. The Abort Delimiter consists of a
Start of Frame Delimiter and an End of Frame Delimiter. Fill bits are transmitted after the
Abort Delimiter.

Abort Delimiter Format

Figure 25
Token Frame Byte and Bit Composition
The Token frame byte and bit composition is shown in Figure 26. The ring transmission
order requires that the Starting Delimiter (SD) be sent first, followed by the Access Control
(AC) octet and the Ending Delimiter (ED). The Most Significant Bit (MSB) of each octet is
transmitted on the ring wire first.
The SD consists of the J-bit and K-bit non data symbols. The concept of the J-bit and K-bit
nondata symbols was covered in Module CDA 203.02. Zero bits are interspersed in the octet.
The AC is composed of the Priority (P) bits, Token (T) bit, Monitor (M) bit, and the
Reservation (R) bits. To identify the Token Bit within a Token frame, the fourth bit in the AC
octet must be examined. The fourth bit is the Token Bit (T). When the Token Bit is set to a
0, the frame is a Token. When the Token Bit is set to a 1, the frame contains Medium Access
Control (MAC) data or Logical Link Control (LLC) data. The meaning of all of the other bits
that are in the frame depend on the setting of the Token Bit.
The ED is made up of the J-bit and K-bit nondata symbols followed by a 1 bit. The last two
bits in the ED are the Intermediate frame (I) bit and the Error Detected (E) bit. The I-bit is set
to a 0 when a Token frame is transmitted. The E-bit is transmitted as a 0. If a transmission
error is detected as the Token frame is repeated, the NIC that detected the error sets the E-bit
to a 1 to indicate that an error occurred. The Active Monitor will detect the error and transmit
a new Token frame.

Figure 26
Protocol Analyzer Display of Token Transmission
Many different types of Token Ring Protocol Analyzers are manufactured. Some of the
analyzers gather only statistical data; other analyzers gather statistical data and decode MAC
and LLC frames. Most of the protocol analyzers do not decode every individual Token
frame; instead, they decode only the MAC and LLC frames.

Protocol analyzers that do not analyze Token frames do gather error statistics about Token
frames. Token frame statistics are displayed on the screen of a Protocol Analyzer in a
manner that is similar to the information that is shown in Figure 27. Under the heading
ÒReport Active Monitor Errors,Ó the number of frame errors that have occurred that relate to
the Active Monitor are tabulated. Most Protocol Analyzers will indicate whether a Dup
(Duplicate) Address exists on the network and will note the number of times that the
condition has occurred. In the example, 0 errors were counted. The number of times that
Dup (Duplicate) Active Monitors appear on the network is also tabulated. Under the heading
ÒReport Soft Errors,Ó the example shows the number of Abort Delim (Delimiters) that were
transmitted. Adjacent to the title ÒToken,Ó the example shows the number of times that a
Token error occurred that required a new Token frame to be transmitted. The other
information that is shown in the example is not relevant to Token frames. The other error
counts will be discussed in the next section of the Module.
Error Statistics
Error Types and Totals Phys Loc Station Addr Last Neighbor
Neighbor Notification Incomplete: 0
Report Active Monitor Errors: 0
Dup Address: 0 Dup Monitor: 0
Report Soft Errors: 3 00000000 401024801021 400000000048
Code/FCS: 0 Internal: 0 00000000 400000000008 400000000031
Burst: 9 A/C : 0 00000000 400000000016 400000000059
Abort Delim: 0 Lost Frames: 0
Congestion: 0 Frame Copied: 0
Frequency: 0 Token: 1
Beacon Frames (Hard Errors): 0
Recovery: 0 Signal Loss: 0
Streaming: 0 Claim Token: 0
Token Frame Errors

Figure 27

Early Token Release
The Early Token Release concept, incorporated in the IEEE 802.5 standard, enhances the
network performance of the ring by allowing multiple frames to be on the ring at the same
time. The Early Token Release (ETR) option increases the available bandwidth and improves
the data transmission efficiency. ETR is an option that the network management may choose
to install. ETR allows an NIC to release a Token frame as soon as it completes a frame
transmission. The Token frame can be transmitted even though the original frame has not
been stripped off the ring. The priority for Token frames that are released prior to reception
of the frameÕs header is the priority of the most recently received frame.
The use of Early Token Release may impact the network protocol operation and network
throughput. The repeated use of the same priority level may delay priority traffic when an
ETR ring is heavily loaded with short frames. Short frames may disable the use of the priority
reservation system that is used by Token Ring. The priority system is affected because, when
a short frame is transmitted, the next Token frame may be released before the frameÕs header
returns to the NIC that originated the frame. If a series of short frames are transmitted, the
same priority level will be maintained until all of the frames have been stripped from the ring.
NICs that implement the ETR option are compatible and interoperable with those NICs that
do not implement the ETR option. ETR is primarily implemented on 16-Mbps Token Ring
networks; however, the IEEE 802.5 standard allows ETR to be implemented on 4-Mbps
Token Ring.
Figure 28 shows how ETR operates. The NIC in PC #1 sends a data frame and then it sends a
Token frame. The NIC in PC #2 captures the Token frame and transmits a data frame that is
followed by a Token frame. The NIC in PC #1 receives the frame that it transmitted back and
strips the frame from the network. The NIC in PC #1 does not issue a new Token frame
because the NIC in PC #2 already transmitted a new Token frame through the use of ETR.

Early Token Release

Figure 28

IDENTIFYING NETWORK FAULTS
To identify Token Ring network faults, the proper test equipment must be used to monitor
protocol transactions and to test the network cable and equipment. The use of the proper test
equipment ensures that network problems will be easily detected. This section describes the
type of test equipment that should be used to test a Token Ring network, explains how to
identify Token Ring protocol faults, and includes the following topics that are pertinent to that
discussion:
• Types of Equipment Available

• Attaching a Token Ring Protocol Analyzer to a Network
• What Types of Problems the Protocol Analyzer will Detect
• Real-Time Protocol Analysis
• Traffic Statistics Analysis
• Ring Failure
• Error Counters
• Multiple Tokens
Types of Equipment Available
Many different types of equipment are available for use in the maintenance and
troubleshooting of a Token Ring Local Area Network. The following are the typical test sets
that are found in a Token Ring network troubleshooterÕs tool kit:
• Cable Testing Tools (Copper Wire and Fiber Optic Test Sets)
• Power Line Disturbance Analyzers
• Analog Test Set
• Protocol Analyzers
Copper Wire Cable Test Sets
The following are typical pieces of copper wire test equipment:
• Cable Tracers
• Cable Continuity Testers
• Time Domain Reflectometer (TDR)
• Cable Impairment Tester

Cable Tracers - For years, the telephone industry has used cable tracers to trace twisted pair
cables that are run between cross-connect blocks in telephone equipment rooms. One of the
problems that must be faced when twisted wire pairs are re-used for a different application is
the identification of the present location of the wire pair. A cable tracer can be used to track
the route of a cable through a building. Hidden cables can be traced through the ceiling,
behind walls, and through a computer room. Some manufacturers produce cable tracers that
can transmit a signal along a cable for distances of up to 1,000 ft.
To identify a wire pair, a tone generator is connected to one end of the cable path. At the
other end of the cable, an inductive amplifier is used to search through a field of wires to
identify the particular wire pair to which the tone source is connected. The cable tracers are
used to identify an individual wire pair out of a field of many possible candidates.
Cable Continuity Testers check the continuity of the wire pair to verify that a signal can be
passed from one end of the cable to the other end of the cable. Cable Continuity Testers
check for shorts between the wire pairs, opens (a break in one wire of the wire pair), and, in
some cases, a short between either of the wire pairs and the cable shield.
A Time Domain Reflectometer (TDR) is often called a radar test set. The TDR can measure
the distance from one end of the cable to a fault such as a broken wire or a shorted wire pair.
A TDR operates through transmission of a short electrical pulse into the wire pair and
measurement of the length of time that elapses before reflected energy arrives back at the
source. The amplitude of the original pulse and the reflected pulse are measured. Through a
simple calculation that is performed by the TDR, the distance in feet or in meters to the cable
fault can be determined. The TDR can be used to identify open circuits, short circuits,
impedance mismatches, cable crimps, knicks in the cable, and any other cable fault that
creates an impedance mismatch.
The Cable Impairment Tester is a comprehensive test set that performs many different tests on
copper wire pairs. The types of problems that can be identified by this test set include open
circuits, short circuits, impedance, noise level, and crosstalk. The Cable Impairment Tester
also can measure the cable length and may contain a rudimentary TDR function that will
measure the distance to a cable fault. A printed record of the test results also can be
produced.
Fiber Optic Test Sets
The following are typical pieces of fiber optic test equipment:
• Optical Power Source

• Radiometer (Optical power meter)
• Optical Time Domanin Reflectometer (OTDR).

The Optical Power Source is a test set that generates a laser signal or an LED signal. The
level of the signal that is transmitted by the Optical Power Source is controlled. The signal
from the test set is inserted into the transmit end of the fiber optic cable. The Optical Power
Source is used with a Radiometer to identify signal level problems.
The Radiometer (Optical Power Meter) measures the optical fiber attenuation. The
Radiometer measures the overall signal loss in a fiber optic cable. The Radiometer is attached
to the cable at the receiver-end of the fiber. The problem that the Radiometer identifies is
excessive signal loss through the cable.
The Optical Time Domain Reflectometer (OTDR) measures the amplitude of reflected, or
reverse, power in a fiber optic cable. The types of problems that the OTDR can identify relate
to cable discontinuities. The OTDR also measures the distance to the discontinuity that
caused the signal to be reflected. A discontinuity may be caused by a cable splice, a
connector, or the end of the cable. The OTDR displays the signal trace on an oscilloscope.
Power Line Disturbance Analyzers
Momentary power interruptions, voltage drops, electrical transients, and other power faults
can create many problems in mainframe computers, personal computers, bridges, and routers.
Power monitors are used to determine whether a power problem exists in a LAN environment.
These devices monitor and record the main power voltage level, detect voltage spikes and
surges, measure the power frequency, and detect electrical transients. Most power monitors
print a report that identifies power anomalies. Some units will sound an audible alarm when
severe power problems occur.
Analog Test Set
VOMÕs (volt-ohm-milliammeter) and Oscilloscopes are analog test sets that can be used to
identify LAN problems. A VOM can be used to measure the main power voltage level to a
PC, to a server, and to other computer devices. The VOM also can be used to check the
continuity of a lobe cable or a main ring cable and to measure the DC voltage level that is
applied by an NIC to the Lobe Cable wires to activate the relay in the LAM.
An oscilloscope can be used to measure the noise level on the main ring cable or on a lobe
cable. The transmitted and received signal levels also can be measured with an Oscilloscope.

Protocol Analyzers
Protocol Analyzers monitor the bits that are transmitted by the NICs on the ring, compile the
bits into MAC and LLC frames, decipher the content of the frame, and determine the frameÕs
meaning to the protocol transaction. LAN Protocol Analyzers can capture, record, and
analyze frames that are transmitted on the LAN.
A LAN Protocol Analyzer can be operated in a monitor mode or in an active mode. In the
monitor mode, the unit watches the protocol transactions and gathers information about the
protocol events that occur on the ring. In the active mode, the Protocol Analyzer transmits
frames to test the ring operation.
Protocol Analyzers monitor frame transmissions and gather the Error MAC frames that are
transmitted by the NICs to the network Error Monitor. The information that is encoded in the
subvectors of the Error MAC frames is decoded. The decoded information is used by the
Protocol Analyzer to determine the ring domain in which the problem occurred. Error
conditions are tabulated by NIC Source Address and ring domain. The tabulations are shown
as counts that indicate the number of times that a specific NIC Source Address experienced a
particular error. Ring error conditions can be easily identified through the analysis of the
error counts.
The Protocol Analyzers also can operate in an active mode and generate test frames to check
the ability of the network to handle different types of traffic and different levels of load. Load
relates to the amount of traffic that a network can handle without errors.
The hardware components that comprise a LAN Protocol Analyzer typically are circuit cards
that plug into the bus of a Personal Computer (PC). Portable PCs normally are used as a
Protocol Analyzer so that the test set can be moved from one location to another. The
Protocol Analyzer hardware uses the resource of the PC to process, store, and display data.
Some manufacturers produce a software Types of Equipment Available (ContÕd)
program that can be implemented on any PC that is equipped with an NIC. The software
program will gather the data that is encoded in the subvectors of the Error MAC frames, and it
will use the data to produce statistics. Other manufacturers produce an integrated hardware
and software-based Protocol Analyzer. The combination of the hardware and software allow
the Protocol Analyzer to monitor all of the frames that are transmitted on the LAN. Analyzers
of this type also gather data from the MAC Error Frame subvector.

Attaching a Token Ring Protocol Analyzer to a Network
LAN Protocol Analyzers interface to the ring through NICs and attach to the network in the
same way as another NIC and host PC. The connection arrangement is shown in Figure 29.
Figure 29 shows that a Protocol Analyzer is attached to a Type 3 and a Type 1 Lobe Access
Module (LAM). The Type 3 LAM uses a Modular Connector and unshielded twisted wire
pair cable. The Type 1 LAM uses a Data Connector and shielded twisted wirepair cable. The
Protocol Analyzer is composed of both hardware and software. The NIC provides the
interface to the LAM.
Protocol Analyzer Connection to Token Ring Network

Figure 29

What Types of Problems the Protocol Analyzer will Detect
The types of problems that can be identified by a Protocol Analyzer are shown below. The
problem is listed in the ÒFailureÓ column; a brief description of the cause of the problem and
a possible solution to the problem are listed in the ÒPossible CauseÓ column.
Failure Possible Cause
AC Error This error indicates that two SMP frames were received in a row
and that the A-bits and C-bits in the FS field were 0Õs in both
frames. The AC Error indicates that a problem occurred during
the Ring Poll process and that neighbor notification may not
have been correctly completed. The network protocol will
continue to operate undisturbed. Any problem that is associated
with the neighbor notification function will be corrected within
seven seconds when the Ring Poll process is completed.
Abort Delimiter Error This error indicates that a transmission was aborted. If a high
quantity of Abort Delimiter Errors are associated with a
particular Source Address, the NIC that is at that address may
have a receive buffer problem. To correct the problem, the PC
must be rebooted. The problem may be corrected when the NIC
resets and reloads its program.
Abort Delimiter Errors also may indicate that the Upstream

Neighbor NIC has transmitted frames that have defective SD or
ED fields. In this case, the Upstream Neighbor NIC needs to be
replaced.
Beacon Frames An NIC transmits a Beacon frame when it no longer receives an

inbound signal from the Upstream Neighbor. The fault domain
is the transmission path that is between the Upstream Neighbor
NIC and the NIC that transmitted the Beacon Frame. The
problem can be caused by any one of the following conditions:
• A defective transmitter in the Upstream Neighbor NIC.

Beacon Frames - A broken Lobe Cable between the Upstream Neighbor NIC and
the LAM.
- A broken main ring cable.
- The local LAM may be broken.
- The Lobe Cable that attaches the LAM to the NIC that
transmitted the Beacon frame may be defective.
- The receive circuitry in the NIC may be broken.
The Source Address of the Beacon frame and the NAUN in the
subvector of the frame must be used to identify the fault domain.
Burst Error This error indicates that the NIC detected no bit transitions for at
least five bit times. Every time that an NIC enters and leaves the
ring, a Burst Error occurs. A high quantity of Burst Errors at the
start and at the end of the work day is normal. At such times of
the day, users are connecting to and disconnecting from the ring.
A high Burst Error count from a particular NIC indicates a clock
recovery problem. In this case, the NIC is defective.
Frequency Error This error indicates that the Upstream NeighborÕs NIC may
have a clock recovery problem or that the NIC that is reporting
the Frequency Error may not be able to recover the received line
clock. The crystal oscillator in the NIC that reported the
Frequency Error may be defective.
FCS Error This error indicates that a line transmission problem exists that
has created errors in MAC or LLC frames. The problem is in the
fault domain between the NIC that has reported the FCS error
and the Upstream Neighbor.
Internal Error An NIC has an internal malfunction from which it can recover.
A high quantity of Internal Error counts for a specific Source
Address indicates that the NIC is in a marginal condition.

Line Error A high Line Error count indicates that the J-bit or the K-bit in the
SD or ED have been incorrectly coded or an FCS error has
occurred. The Source Address that reported the Line Error must
be identified. The fault domain is located between the Upstream
Neighbor and the NIC that reported the Line Error.
Lobe Test Failure Lobe Test Failure indicates that the lobe cable, the DB-9
connector that attaches the lobe cable to the NIC, or the LAM
connector is open.
If it is the first time that the lobe cable has been placed in
service, the connectors may be incorrectly wired.
Lost Frame Error This error indicates that the NIC that transmitted a frame cannot
recognize the Source Address after the frame has circled the
network. This problem typically occurs when the network has a
high FCS error rate or a high Line Error rate. Correction of the
line error rate problem will decrease the Lost Frame error count.
However, if the Lost Frame Error count is high and the FCS and
Line Frame Error count is low, the problem is the NICÕs
inability to identify and strip frames from the network.
Receive Congestion A Receive Congestion (or just Congestion) error occurs when an
NIC detects that a frame is addressed to itself, but the NIC does
not have enough memory space to copy the frame from the
network. The NIC that reported the problem is congested.
The problem may be caused by one of the following conditions:
• Insufficient I/O ring buffer space.
• The PC clock is too slow to handle the network data throughput

requirements. The slow PC must be replaced with a faster unit.
Signal Loss A signal loss will cause the NIC that detects the fault to send a
Beacon frame. Beacon Frames previously were discussed in this
section.

Streaming The following types of streaming problems may exist.
• If an NIC continues to send the same frame repeatedly, the NIC

is streaming. The NIC may not remove the frame from the
network. The Active Monitor will strip the frame from the
network and will issue a new Token frame. The Active Monitor
will report the streaming condition to the Error Monitor. The
address of the streaming device is contained in the Report Error
MAC Frame Subvector. If a Source Address is constantly
sending frames with M-bit errors, it is streaming. The NIC is
defective.
• If an NIC continues to send a random series of 0Õs and 1Õs, it

also is in a streaming condition. To isolate the condition, each
lobe cable must be unplugged from the LAM while the protocol
analyzer is watched. When the lobe cable for the defective NIC
is unplugged, the streaming condition will end and the protocol
operation will be restored.
Token Error The Token Error counter increments when a new Token frame
must be sent by the Active Monitor after it stripped a lost frame
from the network. The network may have a high line error rate.
Real-Time Protocol Analysis
Protocol Analyzers analyze and display MAC and LLC frames in either a real-time mode or
an offline mode. Real-time means that the Protocol Analyzer displays the frame data as the
frame analysis occurs. As a frame appears on the ring, the Protocol Analyzer receives the
bits, divides the bits into the appropriate Token Ring fields, decodes the frame content, and
displays the information on a CRT.
In the offline mode, the Protocol Analyzer monitors the network protocol transactions and
records the frames to a hard disk. After the data has been collected, the Protocol Analyzer
stops the monitor function. The data is then analyzed and displayed on the CRT.

Benefits to Performing Real-Time Protocol Analysis
The following are the benefits to performing real-time Protocol Analysis:
• Trouble analysis can be immediately performed. As events occur, the protocol analyst
can see their affect on the network.
• The protocol analyst can trigger events to see how the network hardware and software
responds.
• Error conditions can be monitored as they occur. Changes in the data patterns can be
seen and analyzed.
Real-time protocol analysis allows the operator to flip through a series of screens to look for
obvious problems. As a protocol problem starts to occur, the pattern of events that created the
problem can be clearly seen and the results of the fault can be monitored. In the offline mode,
trouble analysis occurs after the fact. The ability to see events as they occur is lost.
Recognizing Normal and Abnormal Protocol Operations
Several pictures of real-time Protocol Analyzer displays are shown in Figure 30. A
description of the information that is contained in the Protocol Analyzer Level 2 Interpretative
display was provided in Module CDA 203.02.
Display #1 shows a normal protocol sequence. A series of AMP and SMP frames are shown.
The frames are engaged in the Ring Poll process. Every seven seconds, the Ring Poll process
is repeated. Source Address ibm7BA7A9 is the Active Monitor.
Display #2 shows an abnormal protocol sequence that has started to occur. A series of Ring
Purge frames have been transmitted by Source Address ibm7BA7A9. The Ring Purge is an
attempt to remove all data bits from the ring and to return the ring to a normal protocol
operation. When the Ring Purge process is complete, the Active Monitor then transmits an
AMP frame. The downstream NICs respond with SMP frames.
Display #3 shows the continuation of the Ring Poll process. A new AMP frame is transmitted
every seven seconds. Downstream NICs participate in the Ring Poll process and neighbor
notification process by sending Standby Monitor frames. Source Address ibm7BA8A9 (the
previous Active Monitor) detects a transmission fault and begins to send Ring Purge frames.

Display #4 indicates that Source Address ibm7BA7A9 (the previous Active Monitor) is no
longer receiving Token frames and has transmitted a series of Claim Token frames in an effort
to become the Active Monitor. The downstream neighbor Source address azure001004 is not
receiving the Claim Token frames from Source Address ibm7BA7A9 and begins to transmit
Beacon frames.
Display #5 shows that the Beacon frame transmission from Source Address ibm7BA7A9
continued for several seconds.

Display #1

41.332 36 802.5 CfgReportSvr CR ibm7BA7A9 RS MAC Report NAUN Change
Display #2
Real-Time Protocol Analyzer Displays

Figure 30


43.347 52 802.5 RingErrorMon RE ibm7BA7A9 RS MAC Report Soft Error
Display #3

58.977 40 802.5 Broadcast RS azure001004 RS MAC Beacon
Display #4

Figure 30 (ContÕd)


Display #5

37.084 56 802.5 CfgReportSvr CR ibm7BA7A9 RS MAC Report New Act Mon
Display #6

Figure 30 (ContÕd)

Display #6 reveals that the Beacon frames were transmitted until 37.022 seconds after the
start of the last minute, at which time Source Address ibm7BA7A9 started to receive the
Beacon frames back without any errors. At that point, Source Address ibm7BA7A9 began to
transmit Claim Token frames in an attempt to become the Active Monitor. Four Claim Token
frames circled the network and were received back by Source Address ibm7BA7A9. Upon
reception of the fourth Claim Token, frame Source Address ibm7BA7A9 assumed the task of
Active Monitor and transmitted a Ring Purge frame. The Active Monitor sent a Report New
Act (Active) Mon (Monitor) frame to the CfgReportSvr (Configuration Report Server). Next,
the Active Monitor sent an AMP frame and started the Ring Poll process. The downstream
NICs responded with SMP frames.
Traffic Statistics Analysis
Traffic statistics are analyzed to ensure that the performance of the network is at peak
efficiency. A continuous analysis of traffic statistics helps to identify long term network
trends.
Many different types of traffic statistics are gathered and include the following:
• LAN Utilization
• Average MAC Frames per second
• Average Frame size
• Average Errors per second
• Station Utilization
• Protocol Channel Utilization
• Response Time
Purpose
LAN Utilization is a measurement of the total number of data bits that are transmitted over the
network per second. The measurement includes all Token, MAC, and LLC frames. All of the
transactions are included in the calculation. The purpose of LAN utilization is to show the
raw amount of bandwidth that is used each second as compared to the maximum bit rate that
may appear on the ring. Figure 31 shows an example of a typical LAN Utilization graph.
The percentage of average utilization is shown on the x-axis and the time in seconds is shown
on the y-axis.

LAN Utilization
Figure 31
LAN Statistics that are shown in Figure 32 include Average MAC Frames per second,
Average Frame size, and Average Errors per second. Statistical information is presented as a
bar graph. The exact numeric values are typically displayed under each bar graph as the max
(maximum), avg (average), and now (present value). The x-axis of the graph shows numeric
values.
Average MAC Frames per second indicate how many Token frames and MAC frames are
transmitted each second to manage the ring operation. This statistic indicates the amount of
overhead on the ring.
Average Frame size indicates the average number of octets in a frame. For a 4-Mbps Token
Ring network, the maximum frame size is 4,500 octets. Older NICs have a maximum frame
size of 2,048 octets. A 16-Mbps. network has a maximum frame size of 18,000 octets.
Average Errors per second indicate the number of Report Error MAC frames that appear on
the ring each second. The frames report error statistics to the Ring Error Monitor.

LAN Statistics
Figure 32
Station Utilization of the network indicates how much of the network bandwidth is in use by a
specific Source Address. Figure 33 shows an example of a Station Utilization graph. This
statistic is used to identify NICs that use an excessive amount of bandwidth. If too many
heavy users are on the same ring, the network response time will decrease. The graph must
be examined to determine whether one of the heavy users needs to be moved to a different
ring in an effort to balance the utilization load. The utilization is presented in the x-axis by
LAN Station (NIC) address. The y-axis indicates the percentage of utilization.
Protocol Channel Utilization identifies how much of the ring bandwidth is used by a
particular higher layer protocol, such as SNA, TCP/IP, NetWare, and other protocols.
Response Time is a time lapse measurement that indicates how much time is required for a
user request to be processed. A Response Time measurement indicates how much time a user
must wait for a file to be delivered that was requested from a server. Most users expect a
response time of about one second or less.

Station Utilization Graph

Figure 33
Neighbor Notification Incomplete
The purpose and meaning of the neighbor notification incomplete statistic is to identify the
number of times that the Ring Poll process was initiated but never completed. The Ring Poll
process is the protocol sequence that causes the Active Monitor to transmit an AMP frame
and the downstream Standby Monitors to transmit SMP frames. The transmission of the
AMP and SMP frames performs the neighbor notification process.
The Active Monitor initiates the neighbor notification process through the transmission of an
AMP frame. Downstream Standby Monitors respond to the AMP frame through the
transmission of SMP frames. If the neighbor notification process does not propagate through
the network so that an SMP frame is received by the Active Monitor before the TAM timer
expires, the neighbor notification process is incomplete.

The TAM timeout is used by the Active Monitor to verify that the neighbor notification
process was propagated through the network. If the TAM timeout occurs before the Active
Monitor receives an SMP frame, an error exists in the network that prevented the completion
of the neighbor notification process.
When the neighbor notification process is incomplete, the Active Monitor sends a Report
Ring Poll failure MAC frame for transmission to the Ring Error Monitor.
Active Monitor Errors
Purpose - The NIC that assumes the task of Active Monitor plays an important part in the
operation of the Token Ring protocol and the error free performance of the ring. If an NIC
that has internal logic faults or a congestion problem becomes the Active Monitor, the NICÕs
performance can severely impact the ringÕs performance. Through the observation and
analysis of the different Active Monitor error counters, a marginal NIC that has assumed the
task of Active Monitor can be isolated.
Duplicate Address - A duplicate address exists when two NICs have been assigned the same
Source Address. A Duplicate Address Test (DAT) is conducted during the Ring Insertion
process by each NIC that wishes to access the ring. If another NIC has the same Source
Address as the NIC that is engaged in the Ring Insertion process, the DAT will fail.
A DAT test failure is reported to the application program that is resident in a PC that has
attempted to access the ring. No other action is taken.
Duplicate Monitor - If the Active Monitor receives an AMP MAC frame with a Source
Address that is not its own, two adapters are performing the function of the Active Monitor.
The Active Monitor that detected the duplication resets its active monitor functions and
assumes the tasks of a Standby Monitor. The device then sends a Report Error MAC frame to
the network Error Monitor.
Soft Errors
A soft error is an intermittent error that causes data to have to be transmitted more than once
to be received. A soft error affects the networkÕs performance but does not affect the overall
reliability of the network; however, an excessive number of soft errors may affect the
networkÕs reliability.

Purpose - Soft errors allow the ring recovery protocols to restore normal Token Ring protocol
operation. The following are soft errors:
• Code/FCS
• Burst
• Abort Delimiter
• Congestion
• Internal
• A-bit/C-bit
• Lost Frame
• Frame Copied
• Line Error
• Token Error
Code/FCS - A Code/FCS error indicates that a line transmission problem exists that has
created errors in MAC or LLC frames. The problem is in the fault domain that is between the
NIC that has reported the Code/FCS errors and the Upstream Neighbor.
Burst - When a Burst Error is counted, the NIC has detected no bit transitions for at least five
bit times. The NIC that detects the burst-five error repeats the frame as a burst-four error to
notify the downstream NICs that the Burst Error has been detected and reported to the Error
Monitor.
Each time that an NIC enters and leaves the ring, a Burst Error occurs. A high quantity of
Burst Errors at the start and at the end of the work day is normal. At such times of the day,
users are connecting to and disconnecting from the ring. A high Burst Error count from a
particular NIC indicates a clock recovery problem.
Abort Delimiter - The transmission of an Abort Delimiter indicates that a transmission was
aborted. If a high quantity of Abort Delimiter Errors are associated with a particular Source
Address, the NIC that is at that address may have a receive buffer problem.
Congestion - The Congestion Error (also called Receive Congestion) counter is incremented
when an NIC recognizes the Destination Address as its own but the NIC does not have
enough memory space to copy the frame from the network. The Receive Congestion Error
indicates that the NIC does not have enough computer memory to process any additional
information frames.
Internal - The Internal Error counter is incremented when an NIC recognizes a recoverable
internal error. Internal errors indicate that a malfunction has occurred in the NIC.

A-bit/C-bit - The AC Error counter is incremented when an NIC receives an AMP or SMP
frame that has an A-bit and a C-bit that are equal to 0. In the protocol sequence, the next
frame that is received also has an AMP or SMP frame that has an A-bit and a C-bit that are
equal to 0. No AMP or SMP frame was received between the previous two frames with the
A-bit and C-bit set to 1. The protocol sequence that was just described is an AC error. An
AC Error indicates that the Upstream Neighbor NIC is unable to set the A-bit and C-bit in the
frame status field of a frame that the Upstream Neighbor copied.
Lost Frame - The Lost Frame Error counter is incremented when an NIC is transmitting a
frame at the time that the TRR timer expires. A Lost Frame is transmitted by an NIC, but it is
not stripped from the network by the NIC that originated the frame. The frame is lost. The
Lost Frame Error counter increments each time that an NIC fails to remove a frame from the
network after the frame has circled the network once. When a frame is not removed from the
network by the NIC that originated the frame, the Active Monitor must remove the frame and
issue a new Token frame.
Frame Copied - The Frame Copied Error counter is incremented when an NIC detects its
address in the Destination Address of a frame, and the A-bit (Address Recognized) of that
frame is already set to a 1. Because the A-bits are already set to a 1, another NIC on the ring
also detected the Destination Address as its own and set the A-bits to a 1. When the
Destination Address copies a frame into memory, the C-bits in the frame are set to a 1Õs.
When both the A-bits and the C-bits are set to 1 at the time that the frame arrives at the
Destination address, the NIC assumes that there is an NIC with a duplicate address in the ring
or that a line transmission problem occurred that erroneously set the A-bits and C-bits to 1Õs.
Line Error - The Line Error counter is incremented when a data frame or Token frame is
copied or repeated by an NIC. The E-bit in the MAC frame is transmitted as a 0. The Line
Error counter is incremented when one of the following conditions exists:
• A non-data symbol (J-bit or K-bit) is between the SD and ED of a frame.
• An FCS error exists in a frame.
The first NIC that detects the line error increments the appropriate error counter and sets the
E-bit in the ED of the frame to a 1. The 1 in the E-bit indicates to the downstream NICs that
the problem has already been detected and reported. The downstream NICs do not log or
report the error.

Token Error - The Token Error counter is incremented when the Active Monitor
detects an error condition that requires a Token frame to be transmitted. A Token
Error occurs when the TVX timeout expires. The Token Error counter is also
incremented when one of the following conditions occurs:
• A Token frame is received with a priority that is not equal to zero and the Monitor Bit
(M-bit) is set to 1.
• A frame is received with a Monitor Bit that is equal to 1.
• No Token frame is received within a ten millisecond window.
• The Starting Delimiter (SD) field has a code violation in a bit field. A code violation is
a non-data symbol that appears in a bit position in which it is not allowed.
Hard Errors
A hard error produces an isolating error. Hard errors result in the entrance of the NICÕs
token ring into the beaconing process for ring recovery. Hard errors take the network out of
service. Hard errors produce isolating error counts that can be used to identify the fault
domain in which the error occured. A hard error on a network requires that the cause of the
error be removed from the network or that the network be reconfigured before reliable
operation can be restored.
Purpose - Hard errors prevent the restoration, through use of ring recovery protocols, of
normal Token Ring protocol operation. Hard errors include Beacon frames, streaming,
frequency, and signal loss.
Beacon Frames - Beacon frames are transmitted as a result of a major ring failure such as a
ring break. The Beacon frame is used to locate and isolate the ring fault. A Beacon frame
error counter is implemented in each NIC.
Streaming - The Streaming Error counter is not required by the IEEE 802.5 standard;
however, it is built into most NICs. The following are the two forms of streaming: bit
streaming and frame streaming.
• Bit streaming occurs as an NIC repeats a frame. Token frames and data frames are
destroyed by the NIC when it replaces the bits in a frame with a random stream of 1Õs
and 0Õs.
• Frame streaming consists of the continuous transmission of Token frames and an abort
sequence. Frame streaming also occurs when the same data frame is constantly
retransmitted.

Frequency - The Frequency Error counter is incremented when the frequency of the crystal
clock on the NIC and the signal frequency that is decoded from the receive wires differ by an
excessive amount. Detection of a Frequency Error causes the NIC to enter the monitor
contention process.
Signal Loss - The Signal Loss counter is incremented when an NIC ceases to receive an
incoming signal and the TVX timer expires. A signal loss counter is implemented in each
NIC. Each time that a signal loss occurs, the NIC that detected the condition sends a Report
Error MAC frame to the network Error Monitor. The Signal Loss counter value is placed in
the MAC subvector of the frame for transmission to the Error Monitor.
A signal loss is detected when the received signal has insufficient signal energy or the
received signal is grossly out-of-phase with the local phase locked loop (PLL). When either
condition exists for 200 milliseconds, a signal loss condition exists.
Ring Failure
A ring failure condition exists when an NIC no longer receives a signal from the upstream
neighbor. A Token frame can no longer be transmitted around the ring; therefore, the ring is
inoperable. The ring failure may have occurred because of a broken transmitter in the
upstream neighborÕs NIC, a cable break, a defective MAU port, or a broken receiver in the
local NIC.
The protocol sequence that occurs as a result of a ring failure is called the Beacon process.
The Beacon process is started when a TNT or a TVX timeout occurs. An NIC that initiates
the Beacon process transmits a Beacon MAC frame.
Beacon
When an NIC sends a Beacon frame, all of the other NICs on the ring enter the Beacon
transmit or Beacon repeat mode. An NIC does not have to have a Token to access the ring to
transmit a Beacon frame. Beacon frames can be transmitted at any time.
In a Beacon transmit mode, Beacon MAC frames are transmitted at 20-millisecond intervals.
Zero bits are transmitted between the frames. An NIC that is engaged in the beacon transmit
mode signals that it is no longer receiving a signal from the upstream neighborÕs NIC.

An NIC that is not in the Beacon process and that receives a Beacon MAC frame enters the
beacon repeat mode. In this mode, Beacon frames are received and repeated. An NIC that is
in the beacon repeat mode does not originate its own Beacon frames.
An NIC that is in the Ring Insertion process will not enter the beacon process; instead, the
NIC will terminate the Ring Insertion process and will notify the application program that the
NIC could not access the ring.
A Protocol Analyzer display of a ring failure is shown in Figure 34. Display #1 shows a
Protocol Analyzer Level 2 display. Display #2 shows a Protocol Analyzer Level 2
interpretative display. A detailed description about how to interpret a Protocol Analyzer
display was presented in Module CDA 203.02.
Display #1 indicates that Source Address ibm7BA749 transmitted a series of Claim Token
frames. The nearest active downstream neighbor Source Address azure001004 did not
receive the Claim Token frame and entered a Beacon process. Examination of the ÒSecsÓ
column indicates that the Beacon frame transmissions were spaced 20-milliseconds apart.
Display #2 describes the Beacon frame content. All Beacon frames contain a beacon
subvector that identifies the reason that the NIC transmitted the Beacon frame. Under the
heading ÒMedium Access Control Vectors,Ó the reason for the Beacon frame is explained.
The ÒBeacon TypeÓ describes why the Beacon frame was transmitted. In this case, the
signal that was received from the upstream neighbor was a ÒStreaming Signal, Claim
Token.Ó A streaming condition indicates that the NIC repeatedly transmitted the same frame.
The address of the ÒNearest Active Upstream NeighborÓ is shown as ibm7BA7A9. Display
#1 indicates that this is the Source Address of the NIC that sent the Claim Tokens.


58.977 30 802.5 Ring Status Adapt Ring Beaconing
Display #1

58.997 40 Broadcast RS azure001004 RS MAC Beacon
Type: MAC Length: 40 Address: azure001004 Broadcast
Time Stamp: 00:00:58.997392
Type: Beacon
Beacon Type: Streaming signal, Claim Token.
Nearest Active Upstream Neighbor: ibm7BA7A9
Hexadecimal Frame
11 02 c0 00 FF FF FF FF 00 01 7F 00 10 04 00 16
00 02 04 01 00 04 06 0b 00 00 00 00 08 02 10 00
5A 7B A7 A9
Display #2
Protocol Analyzer Display of Beacon Process

Figure 34

The problem in the example is that a streaming signal has been received by Source Address
azure001004 from Source Address Òibm7BA7A9.Ó The NIC that created the streaming
condition may be defective, or the NIC software program may have a processing error that
requires the NIC hardware to be reset.
The following network conditions create a Beacon condition:
• Signal loss
• Streaming
• Monitor contention could not be resolved
The protocol sequence that occurs after the Beacon process has started triggers the Timer
BCN Transmit (TBT) to reset. The TBT timeout is used to restrict the length of time that an
NIC may remain in the Beacon process. The default timeout value of TBT is 26 seconds.
If two consecutive Beacon frames are received from the nearest upstream neighbor, the
problem that created the beacon condition has been cleared. Once the problem is cleared, the
NIC enters the monitor contention process and transmits a Claim Token frame.
A Beacon transmit auto removal test is a process for automatic withdrawal of a faulty NIC
from the ring. If the fault is in the NIC that transmitted the Beacon frame, the process corrects
a fault. If the NIC does not leave the Beacon transmit mode within 26 seconds, the NIC
executes the Beacon transmit auto removal test sequence and the following events occur:
• The NIC turns off the dc voltage to the relay in the MAU and disconnects itself from the
ring.
• The NIC initiates the Ring Insertion process.
• If the Lobe Test is successful, the NIC reinserts itself into the ring.
• If any MAC frame is received, the NIC enters the Beacon transmit mode and discards
the MAC frame.
• If a signal loss is detected, the NIC enters the Beacon transmit mode and resumes the
transmission of Beacon frames.
• If the Lobe Test is unsuccessful, the NIC notifies the application program that is in
operation in the PC. The NIC remains disconnected from the ring.

The Beacon receive auto removal test corrects ring faults in the lobe cable of the NIC that is
upstream of the NIC that has transmitted the Beacon frame. If a fault occurs in an NICÕs
transmit path, the NICÕs downstream neighbor will send a Beacon frame. The upstream NIC
that has the faulty transmitter will receive the Beacon frames. After eight Beacon frames are
received in sequence, the upstream NIC performs the Beacon receive auto removal test as
follows:
• The NIC turns off the dc voltage to the relay in the LAM and disconnects itself from the
ring.
• The NIC initiates the Ring Insertion process.
• If the Lobe Test is successful, the NIC reinserts itself into the ring.
• The NIC transmits Beacon frames.
Neighbor Notification (NN)
The purpose of neighbor notification is to identify the upstream neighborÕs address so that
the fault can be isolated to a particular section of the ring. The section of the ring that has a
failure is called a fault domain.
When a ring fault is detected, the cause of the ring fault must be isolated to the proper fault
domain so that the fault can be repaired. Figure 35 describes the fault domain. The fault
domain consists of the following:
• The NIC that reported the failure.

• The NIC that is upstream of the NIC that transmitted the Beacon frame.
• The lobe cables that connect the two NICs to the LAMs.
• The main ring cable that interconnects the LAMs.
To identify the cause of a fault, all of the elements of the failure domain must be known at the
time that the failure is detected. In other words, at any given time, each NIC must know the
Source Address of its upstream neighbor. The process by which the upstream neighborÕs
address is determined is called neighbor notification.

Fault Domain
Figure 35
Determining the Upstream NeighborÕs Address
Each network node obtains the upstream neighborÕs address through the use of the neighbor
notification protocol sequence and the Address Recognized (A-bits) and the Frame Copied
(C-bits). The A-bits and C-bits are in the Frame Status (FS) field of the MAC frame. The A-
bits and C-bits are transmitted as 0Õs. If an NIC recognizes the DA of the frame as its own,
the NIC sets the A-bits to 1Õs. If an NIC copies the frame from the ring, the C-bits are set to
1Õs.
When the AMP frame is broadcast by the Active Monitor, the DA field is set to indicate that
the frame is a broadcast. When a frame is broadcast to all NICs on a ring, the first NIC that is
downstream from the broadcaster will detect that the A-bits and C-bits are all 0Õs. The first
NIC to receive the broadcast frame sets the A-bits and C-bits to 1Õs. All of the other NICs
that are further downstream recognize that the A-bits and C-bits are set to 1Õs, and they know
that the broadcast frame address already has been recognized and that the frame has been
copied from the network.

Upon reception of the AMP frame, the first NIC that is downstream from the Active Monitor
takes the following action:
• The NIC resets the TSM timeout.
• The NIC copies the AMP frame from the ring and stores the Source Address in the
Upstream Neighbor Address (UNA) memory field of the NICÕs microprocessor.
• The NIC repeats the frame downstream to the next NIC and, in the process, sets the A-
bits and C-bits to 1Õs.
• The AMP frame is repeated by every downstream NIC until it circles the ring and
arrives back at the Active Monitor.
• The Active Monitor strips the AMP frame from the network and transmits a Token
frame.
• The first NIC downstream from the Active Monitor seizes the Token frame and
broadcasts an SMP MAC frame. The NIC places its address in the Source Address field
of the SMP frame.
One by one, each NIC receives an SMP frame with the A-bits and C-bits set to 0Õs. The NIC
stores the Source Address that is in the frame in the UNA memory location and continues the
neighbor notification process through the origination of its own SMP frame and the broadcast
of the frame to the next downstream NIC.
Error Counters
The purpose of Error Counters is to provide statistics that can be used to identify faults that
may affect the operation of the Token Ring protocol. Error Counters are implemented in every
NIC. When an NIC detects an error condition on the ring, the NIC increments the appropriate
error counter The error counts are reported on a periodic basis to the Error Monitor.
Persistent errors of one particular type that occur in a particular fault domain indicate that a
fault exists in that section of the ring.
As discussed previously in this Module, Token Ring networks experience either hard errors or
soft errors. A hard error may be caused by a protocol fault or an internal fault in the NIC.
When a hard error occurs, the NIC transmits a Beacon frame. A Protocol Analyzer is used to
discover the NIC that caused the hard error. At the conclusion of the Beacon process, the
NIC that created the fault is removed from the ring.

Soft errors are error conditions from which the NIC or the network protocol process may
recover without any human intervention. Each NIC has a series of soft error counters. Each
NIC counts soft errors. Once every two seconds, the TER (Timer, Error Report) timeout
expires and the NIC transmits a Report Error MAC frame to the network Error Monitor.
Line Error
The Line Error counter is contained on all NICs. Each NIC counts Line Errors. A Line Error
is defined as follows:
• The J-bit or the K-bit in the SD or ED have been incorrectly coded.
• An NIC detects an FCS error in a frame as the NIC repeats the frame.
Every frame that has one or both of the conditions that are listed above has a Line Error.
Even if the frame has multiple Line Errors, a Line Error is counted once per frame. The fault
domain is between the Upstream Neighbor and the NIC that reported the Line Error.
The protocol sequence that occurs as a result of a Line Error is as follows:
• The NIC checks every frame that is copied into memory or that is repeated on the ring
for a valid FCS to ensure that no coding errors exist in the J-bits and K-bits of the SD
and ED.
• An NIC that detects an error sets the E-bit (Error Detected) in the ED field of the MAC
frame.
• The NIC increments the Line Error counter.
• The errored frame is repeated to the downstream neighbor NIC and around the ring back
to the Source Address.
• The Source Address strips the errored frame from the network.
• The Source Address seizes a Token frame and resends the frame a second time.
• When the TER timeout expires, the NIC sends a Report Error MAC frame to the
network Error Monitor.

Internal Error
The Internal Error counter is incremented when an NIC recognizes an internal error. An NIC
detects an internal hardware error through the background diagnostics run on the NIC.
Internal errors indicate that a malfunction has occurred in the NIC. The fault may be an
improper hardware configuration such as the I/O Address, Interrequest (IRQ) line, Direct
Memory Access (DMA) line, slot setting in the PC, or NIC microcode level.
Token Ring NIC cards can fit into 8-bit, 16-bit, and sometimes 32-bit bus slots in a PC. If the
wrong adapter card is inserted into an interface, or if the correct adapter card is inserted into
the wrong interface, an internal error may occur. In most cases, a problem of this type will be
detected prior to the Ring Insertion process; however, sometimes the NIC will appear to
operate normally. The NIC works slower then it should and, as a result, internal errors occur.
An Internal Error causes the NIC to remove itself from the ring. After the NIC has
disconnected from the ring, the NIC reports the internal error to the application program.
Congestion Error
The Congestion Error (also called Receive Congestion) counter is incremented when an NIC
recognizes the Destination Address as its own, but the NIC does not have enough memory
space to copy the frame from the network. The Receive Congestion Error indicates that the
NIC does not have enough computer memory to process any additional information frames.
Frame Copied Error
The Frame Copied Error counter is incremented when an NICÕs address is in the Destination
Address of a frame and the A-bit (Address Recognized) is set to a 1. A Frame Copied Error
may indicate that a possible line hit has occurred or that an NIC with a duplicate address is in
the network.
Multiple Tokens
The error count is transmitted to the Error Monitor in the Report Error MAC frame. The
count is transmitted as a field in the subvector The error information is retrieved from the
Network Management application program that is in operation on a server.
Under normal operation, multiple Token frames may not appear on a network; however,
multiple Token frames are allowed on the network when the Early Token Release option has
been enabled.

Definition
The difference between Multiple Tokens and Early Token Release (ETR) is the
implementation of the ETR option. Under normal operation, when multiple Tokens appear on
a token Ring network, a protocol fault has occurred. The Early Token Release option must be
enabled for multiple Tokens to appear on the ring.
Multiple Tokens - Multiple Tokens may appear on a ring when a protocol fault has occurred
and two Active Monitors are in operation on the ring. The multiple Token condition, in most
cases, will correct itself the next time that an AMP frame is transmitted by either Active
Monitor. The maximum time period that multiple Active Monitors will appear on a ring is
seven seconds. Within seven seconds, the normal protocol operation will be restored.
When either Active Monitor transmits an AMP frame, the frame will circle the ring as part of
the Ring Poll process. When the AMP frame is received by the other Active Monitor, that
Active Monitor will examine the Source Address in the AMP frame and realize that the frame
was originated by a different device. The Active Monitor then will reset and assume the
duties of a Standby Monitor. One Active Monitor then will be in operation on the ring.
Early Token Release - Early Token Release (ETR) is an option that may be implemented on
either 4-Mbps or 16-Mbps Token Ring networks. When the option is enabled, an NIC may
transmit a new Token as soon as it has completed the frame transmission. The NIC does not
have to wait to strip the frame from the ring before it transmits a new Token. Multiple
Tokens are not allowed on a ring unless the ETR option has been enabled. When the ETR
option is enabled, the ring has only one Active Monitor.
Problems Created by Multiple Tokens
When multiple Token frames appear on a ring, two NICs could transmit data frames at the
same time. The problems that may occur, if any, are somewhat unpredictable.
Long data frames that are transmitted by a downstream NIC may be overwritten by the long
data frames that are transmitted by an upstream NIC. If the data frames that are transmitted
by both NICs are short, the data frames may normally circle the network and not have any
impact on each other. In the worse case, two Active Monitors will appear on the ring for a
maximum of seven seconds. After seven seconds, the Ring Poll process will occur and one of
the Active Monitors will detect that a second Active Monitor is in operation, and it will
remove itself from the ring.

When an Active Monitor receives an AMP MAC frame with a Source Address that is not its
own, two NICs have assumed the task of Active Monitor. The Active Monitor that detected
the duplication resets its active monitor functions and assumes the tasks of a Standby
Monitor. The device then sends a Report Active Monitor Error MAC frame to the network
Error Monitor to report the fault. At this point, the problem has been cleared.
Identifying Multiple Tokens
Multiple Token transmission can be identified through the examination of ring statistics. The
ring Error Monitor will receive from the duplicate Active Monitor a Report Active Monitor
Error MAC frame. The report is sent to the Error Monitor when an Active Monitor detects an
AMP frame on the ring that it did not transmit. The duplicate Active Monitor problem is
reported to the Error Monitor in the subvector of a MAC frame. The statistic is typically
displayed on the statistics screen of a Protocol Analyzer as ÒDuplicate MonitorÓ or ÒDup-
Monitor.Ó Figure 36 shows a section of a Protocol Analyzer display that includes a ÒDup-
MonitorÓ report field. The number of times that the duplicate Active Monitor condition
occurs is reported as a numerical value. ÒPhys LocÓ (Physical Location) indicates the ring
number or network segment number. ÒStation AddrÓ (Station Address) is the address of the
station that sent the Report Active Monitor Error MAC frame to the Error Monitor. The
ÒStation AddrÓ indicates the Source Address of the NIC that was the duplicate Active
Monitor. In the example, the duplicate Monitor was ÒStation AddrÓ 400000101001.
The fault domain is identified as the hardware and transmission path that connects the NIC
address shown under the heading ÒStation AddrÓ and the ÒLast Neighbor.Ó The address of
the NIC that reported error counts to the error monitor is shown under the heading ÒStation
AddrÓ in the display.
After the error condition is reported to the Error Monitor, the duplicate Active Monitor is
removed from the ring and the protocol operation is restored to normal.
Error Statistics
Error Types and Totals Phys Loc Station Addr
Report Active Monitor Errors:

Dup Address: 0 Dup Monitor: 1 400000101001
Error Statistics
Figure 36

WORK AID 1: GUIDE FOR IDENTIFYING A NORMAL RING PROTOCOL

INITIALIZATION SEQUENCE
The five phases of the Ring Initialization sequence are listed below.
• Lobe Media Check to verify the continuity of the lobe cable. The NIC sends one or
more Lobe Media Test MAC frames.
• Physical Insertion is the process in which the NIC applies a DC current to the wire pairs
in the lobe cable to activate the LAM relay.
• The Duplicate Address (Addr) Test to verify that no other NIC on the ring has the same
Source Address. The NIC sends two Duplicate Address Test MAC frames.
• The Neighbor Notification process is performed so that the NIC may learn the nearest
active upstream neighborÕs address. The NIC sends a Report NAUN Change frame to
the Configuration Server.
• The NIC requests the Ring Parameter Server for operational parameters. The NIC
transmits a Request Initialization frame to the Ring Parameter Server.

WORK AID 2: GUIDE FOR IDENTIFYING A NORMAL RING PROTOCOL

SEQUENCE
This Work Aid is designed to assist the Participant in performing Exercise 2. This Work Aid
contains seven figures (Figure 47 through Figure 53) that contain flowcharts that illustrate all
of the normal protocol sequences that can occur in the MAC sublayer. The portions of the
protocol sequences that can be viewed on a protocol analyzer display are the frames that are
transmitted or received, and these frames are identified by the boxes that are outlined in bold
face type in the flowcharts.
Bypass Mode
Figure 47

Ring Insertion Process

Figure 48

Ring Initialization Sequence

Figure 49

Claim Token Process

Figure 50



Standby Monitor Mode

Figure 51

Transmit Beacon State

Figure 52



Active Monitor Operation

Figure 53

WORK AID 3: GUIDE FOR IDENTIFYING NETWORK PROTOCOL FAULTS
ERROR STATISTICS
In an Error Statistics display, the fault domain is identified as the hardware and transmission
path that connects the NIC address that is shown under the heading ÒStation AddrÓ and the
ÒLast Neighbor.Ó The Source Address (SA) of the NIC that reported error counts to the
error monitor is shown under the heading ÒStation AddrÓ in the display.
Error Statistics displays normally are divided into sections that include the following: Report
Active Monitor Errors, Report Soft Errors, and Beacon Frames (Hard Errors). The remainder
of this Work Aid consists of a table that can be used to identify network protocol faults
through an analysis of the Error Statistics display.







GLOSSARY
A-bit The A-bit is the address recognized indicator bit. The A-bit is in
the Ending Delimiter of the MAC frame. If the A-bit is set to a
1, at least one NIC on the ring has received the frame and has
recognized the Destination Address.
bridge Equipment that allows multiple segments of a Token Ring

network to be interconnected. The Bridge allows
communication between devices on separate networks that use
the same protocol.
buffer A memory device that is used to temporarily hold data to be

transmitted to the ring or data received from the ring.
burst-4-error A physical signal that consists of 4 bits without any high-to-low

or low-to-high transitions and that is applied to the rings of a
Token Ring network.
burst-5-error A physical signal that consists of 5 bits without any high-to-low

or low-to-high transitions.
C-bit A bit in the Ending Delimiter of a MAC frame that indicates that
at least one ring has received the frame and has copied the frame
into memory.
contention A dispute between two or more NICs over the right to transmit
on the ring at the same time.
I/O Input/Output.
impedance The total resistance to current flow in a circuit or cable.

Impedance is comprised of three components: resistance of the
cable, inductive reactance, and capacitive reactance.
lobe cable The cable that physically connects the NIC to the wire
concentrator (LAM). Nearest Active Upstream Neighbor.
NIC Network Interface Card.
node A node consists of an NIC and a host PC.
propagation delay The time that is required for a signal to travel from one point to
another.

SUA Stored Upstream NeighborÕs Address.
retry The process of resending a frame a defined number of times until

the frame is received and accepted.

Engineering Encyclopedia: Troubleshooting Token Ring Networks

Uploaded by

Copyright:

Available Formats

Engineering Encyclopedia: Troubleshooting Token Ring Networks

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Engineering Encyclopedia: Troubleshooting Token Ring Networks

Uploaded by

Copyright:

Available Formats

What is the purpose of the Active Monitor Present (AMP) frames?

What is the purpose of the Active Monitor Present (AMP) frames?

What is the purpose of the Standby Monitor Present (SMP) frames?

What is the purpose of the Standby Monitor Present (SMP) frames?

Engineering Encyclopedia

Saudi Aramco DeskTop Standards

Troubleshooting Token Ring Networks

Chapter : Communications For additional information on this subject, contact

IDENTIFYING THE RING INITIALIZATION SEQUENCE OF A TOKEN

Saudi Aramco DeskTop Standards

Protocol Analyzer Display of Upstream and Downstream

Saudi Aramco DeskTop Standards

Frame Copied Error ........................................................................ 87

Saudi Aramco DeskTop Standards

IDENTIFYING THE RING INITIALIZATION SEQUENCE OF A TOKEN RING

• Standby Monitor Timer (TSM)

• No Token Timer (TNT)

• PC Transmission of Claim Token

• MAC Purge Frame

• Duplicate Address Test

• Lobe Media Check/Physical Insertion

Saudi Aramco DeskTop Standards 1

Lobe Media Check

Any application program that is in operation in a host PC that wants to establish

Lobe Media Test

Saudi Aramco DeskTop Standards 2

Saudi Aramco DeskTop Standards 3

LAM Bypass Mode

Saudi Aramco DeskTop Standards 4

Saudi Aramco DeskTop Standards 5

LAM Insertion Mode

Saudi Aramco DeskTop Standards 6

• Active Monitor Present (AMP)

Saudi Aramco DeskTop Standards 7

Secs Bytes Phys Destination SAP Source SAP Type Description

Protocol Display of the Ring Insertion Sequence

Duplicate Address Verification

Saudi Aramco DeskTop Standards 8

Upstream Neighbor Sends Address

Saudi Aramco DeskTop Standards 9

NIC Sends Address to Downstream Neighbor

In Figure 7, PC#1 is the Upstream Neighbor. The address of PC#1 is 40 00 00 00 00 01.

Saudi Aramco DeskTop Standards 10

Network Address Relationship

Saudi Aramco DeskTop Standards 11

Secs Bytes Destination SAP Source SAP Type Description

Protocol Analyzer Display of Upstream Neighbor Address

Saudi Aramco DeskTop Standards 12

Secs Bytes Phys Destination SAP Source SAP Type Description

Request Initialization Protocol Sequence

Saudi Aramco DeskTop Standards 13

Secs Bytes Phys Destination SAP Source SAP Type Description

Secs Bytes Phys Destination SAP Source SAP Type Description

Secs Bytes Phys Destination SAP Source SAP Type Description

Secs Bytes Phys Destination SAP Source SAP Type Description

No Response to Request Initialization Frames

Saudi Aramco DeskTop Standards 14

Counters and Timers Used to Monitor Protocol Activity