|
|
A Frame Relay connection can be setup either from Cisco WAN Manager (CWM) or via the MGX 8220 command line interface.
Setting up a Frame Relay connection is normally performed from CWM using the Connection Manager graphical user interface (GUI). An example of the CWM screen used for making an MGX 8220 Frame Relay connection is shown in Figure 5-1. For full details of how to set up a connection, refer to the Cisco WAN Manager Operations Guide.
The command-line interface (CLI) provides the capability to set up a variety of Frame Relay connections.
The following sections describe how to establish an end-to-end Frame Relay connection with network interworking and MGX 8220 FRSM end points.
Figure 5-2 shows two BPX nodes in a BPX network in which each of these two nodes is connected to an MGX 8220 shelf via a BNI card. User Frame Relay equipment, located at "A," is attached to one of the MGX 8220 shelves via a port on the shelf's FRSM card. User Frame Relay equipment, located at "B," is attached to the other MGX 8220 shelf. This chapter describes how a Frame Relay connection can be established to permit bidirectional communication between the Frame Relay equipment at "A" and "B."
To make the connection, the path from "A" to "B" is made up of three segments as shown in Figure 5.2. When using the CLI, each segment must be established and configured separately.
Two segments span from the FRSM to the BNM on the MGX 8220 shelves. These segments are part ATM and part Frame Relay with the conversion being made in the MGX 8220 shelves.
There is also an ATM trunk segment that spans the BPX backbone network from one of the BPX nodes to the other BPX node, this segment terminates on a BNI feeder trunk in each node. This segment may include intermediary BPX nodes (not shown in the diagram).
The links between the segments must be configured properly so that the three segments make up one complete end-to-end connection from "A" to "B." This process consists primarily of ensuring that the VPI between the MGX 8220 shelf and its co-located BPX switch must contain the MGX 8220 slot number of the FRSM and the VCI must contain the logical channel number assigned to the virtual circuit.
This procedure must be performed on the MGX 8220 at BOTH ends of the connection ("A" and "B").
To establish an end-to-end Frame Relay connection, perform the following steps:
Step 2 If not enabled, enable the T1 line to be used for the Frame Relay connection by entering the addln command using the physical FRSM connector number (1 to 4) connected to the T1 line.
Step 3 If not configured, configure the T1 line to the Frame Relay equipment by entering the cnfln command. Specify parameters as appropriate.
Step 4 If not enabled, enable the port to the Frame Relay equipment by entering the addport command using the parameters as follows:
For port number, specify an unused port number (1 to 96).
For line number, specify the FRSM line used to connect to the Frame Relay equipment
(1 to 4, with 1 being the top line).
For DS0 speed, specify either 1 for 56 kbps or 2 for 64 kbps.
For the beginning timeslot, specify the beginning timeslot in the T1 or E1 line.
For number of timeslot, specify the number of consecutive T1 or E1 timeslots to be used for the connection.
Step 5 Enter the addchan command to enable the Frame Relay channel.
Select network interworking or service interworking in the chan_type parameter. 1 is for network interworking.
For channel number, enter a value between 16 and 271. THIS WILL BE THE SAME NUMBER SPECIFIED IN THE VCI FIELD TO/FROM THE BPX SWITCH.
For port number, enter the port number previously enabled.
For DLCI, enter a DLCI number to be used in communicating with the Frame Relay equipment.
 |
Note Once a service module channel is started in the MGX 8220 shelf, the T3 line to the BPX switch is automatically up, configured, and started by the MGX 8220 shelf and no action is required by the operator. |
|
Note Remember that this process must be repeated at the remote end to establish the segment at that end. |
At this point the MGX 8220 shelf segment is up with default parameters.
The following steps should be performed to establish the required BPX-to-BPX segments.
1. Enter the addcon command at one of the BPX nodes (not both) as follows.
For slot number and port number, specify slot and port of the BNI port connected to MGX 8220.
For VPI, specify the slot number in the MGX 8220 shelf that contains the FRSM attached to the BPX.
For VCI, specify the logical channel number (LCN) of the Frame Relay connection configured on the local MGX 8220 shelf.
For Nodename, specify the nodename of the BPX at the other end of the connection.
For Remote Channel, specify the BNI slot and port number of the BNI port attached to MGX 8220 at the remote end. Specify the VPI as the slot number of the remote MGX 8220 FRSM connected to the BPX and specify VCI as the LCN of the Frame Relay connection at the remote MGX 8220.
Specify the type of connection. ATFR when ForeSight is not being used and ATFST when ForeSight is being used.
2. Enter the other addcon parameters of bandwidth, and so on.
Minimum Cell Rate (MCR) is only used with ForeSight (ATFST).
MCR and Peak Cell Rated (PCR) should be specified according to the following formula:
MCR = CIR * 3/800 cells per second
PCR = AR * 3/800 cells per second but less than or equal to 6000
AR = Frame Relay port speed in bps.
The above MCR and PCR formulae assume a fairly pessimistic frame size of 100 octets, however even smaller frame sizes can result in worst-case scenarios. For example,
The % Util should normally be set to the same value as that used for the Frame Relay segments of the connection.
FRSM service interworking connections are made in the same manner as the network interworking connections except that chan_type in the MGX 8220 addchan command is specified as service interworking (transparent or translation) and the connection end that is remote from the MGX 8220 is an ATM UNI.
These services are setup the same as Frame Relay except the for port_type in the addchan command, which is set as 2 for FUNI or 3 for frame forwarding.
AUSM connections can be setup either through CWM or via the CLI.
Setting up an AUSM connection is normally performed from Cisco WAN Manager (CWM) using the Connection Manager graphical user interface (GUI). An example of the CWM screen used for making an MGX 8220 ATM-to-ATM connection is shown in Figure 5-3. For full details of how to set up a connection, refer to the Cisco WAN Manager Operations Guide.
Use the following sequence of commands to establish an ATM UNI/NNI connection using the AUSM card. The connection is between a T1 or E1 ATM UNI on the AUSM card and an ATM service interface elsewhere in the IPX/BPX network.
2. Enter the addln command. Specify the line and port number (between 1 and 4 on a 4-port card; between 1 and 8 on an 8-port card), 1 being the top line and port.
3. If required, enter the cnfln command specifying line code, line length, and clock source.
4. Enter the upport command specifying the port to be upped.
5. Enter the cnfportq command to setup egress queues. Other defaults you need to specify:
port number (1-4 on a 4-port card, 1-8 on an 8-port card)
queue number (1-16)
queue priority
0 = disable queue
1 = high priority, always serve
2 = best available
3 = Min. guaranteed bandwidth
4 = Min. guaranteed bandwidth with max. rate shaping
5 = CBR with smoothing
service sequence number (1-16)
max. queue depth (1-8000)
CLP low threshold (1-8000)
CLP high threshold (1-8000)
EFCI threshold (1-8000)
6. Enter the addcon command to add the connection, specifying
logical connection (LCN 16-271)
connection type (1 = vpc, 2 = vcc)
port number (1-4 on a 4-port card, 1-8 on an 8-port card)
VPI (0-255)
VCI (0-65535)
service type (1 = cbr, 2 = vbr, 3 = abr)
queue number (1-16)
7. For configuring UPC, use one
cnfupc cbr
cnfupc vbr
cnfupc abr
8. Enter the cnfchanfst command to configure ForeSight.
9. If queue depths need to be changed, enter the cnfchanq command.
The AUSM 8T1/E1 has a similar command sequence for adding ATM connections for ATM ports.
1. addln—all constituent links
3. addimagrp—to add the IMA port follow the command sequence for the ATM ports
The AUSM 8T1/E1 LCN range is 16 to 1015.
For the BPX segment, set up the connection in the same manner as that for FRSM. The connection type should be specified as ABR, CBR, or VBR to match the connection type used at the connection endpoint (for example, AUSM). The parameter values map directly from those specified at the connection endpoint.
Use the following procedure to setup a CESM connection.
Setting up a CESM connection is performed through the command-line interface (CLI). The procedure is to first add the line entering the add line command (addln) and then add and configure a channel entering the add channel (addchan) command. In the addchan command the channel number, the cell delay variation, the cell loss integration period, and buffer size are all specified. The command sequence is
2. addchan "chan_num CDV CellLossIntegPeriod bufsize"
chan_num -- value ranging from 16 to 19
16 - line 1, 17 - line 2, 18 - line 3, 19 - line 4
CDV -- Cell delay variation: Range 1000-65535 microseconds
CellLossIntegPeriod -- Cell loss integration period:
Range 1000-65535 milliseconds
bufsize -- egress bufsize: Min value: 0.6*CDV-T1, 0.7*CDV-E1.
Max value: 65535, 0 to auto-compute
Use the following procedures to setup an FRASM connection.
Setting up an FRASM connection is performed through the CLI. There are three basic types of connections.
For more information about these types of connections, see Chapter 4, "MGX 8220 Service Modules," "Frame Relay Access Service Module" section.
For all three types of connections, the procedure is to first establish a physical line for the connection entering the add line command (addln) in which the physical back card port is specified and then to establish and configure ports on that line entering the add port (addport) command for each port.
In the addport command, the port number, line number, line speed, time slot, the port type, the encoding type, and the interface type are specified.
To complete the connection, additional commands are used that are specific to the type of connection being set up (FRAS BNN, STUN, or BSTUN).
To complete the FRAS BNN connection, proceed as follows:
1. Enter the addln command to specify the physical port number (from 1 to 8) that is to be used for the FRAS BNN connection.
2. Enter the addport command to specify a port number for the connection and to specify its parameters. Specify the port speed, the slot number, the encoding to be used, and the type of DS0 interface. The port type field is used to specify the connection as a FRAS BNN connection. The role (primary, secondary, and so forth) of the FRASM port to be used in the link protocol must be specified.
addport <port_num> <line_num> <line_speed> <begin_slot> <port_type> <port_role> <encoding> <interface>
3. Once the line and port have been specified, use the add link station command (addls) to specify the FRASM port link station address and xid (exchange ID). A link consists of two link stations and the connecting transmission medium. In order to start an SNA session, a link between the two nodes needs to be established. In an FRAS BNN connection the SNA part of the connection is terminated at the FRASM; a session requires that the FRAS BNN port act as a link station. The link station name and the xid are used during the process of establishing a link.
addls <port_num><lsaddress><xid>
4. The add channel (addchan) command is used to specify the Frame Relay end of the connection. This consists of specifying the DLCI and committed information rate (CIR) for the channel.
addchan <chan_num> <port_num> <dlci_num> <cir>
5. Enter the add FRAS BNN route command (addfrasbnnroute) to establish a connection between the Frame Relay channel number and the link station.
addfrasbnnroute <port_num> <lsaddress> <chan_num> <lsap> <rsap>
port number of the FrasBNN connection to be routed in the range 1-192 |
|
channel number of the connection to be routed in the range 16-1015 |
|
To complete the STUN connections, proceed as follows:
1. Enter the addln command to specify the physical port number (from 1 to 8) that is to be used for the STUN connection.
2. Enter the add STUN group command (addstungroup) command to create a STUN group. This command is used to specify the group number and the protocol type. When a STUN connection is made, the connection is assigned to a group in the add STUN port command.
3. Enter the addport command to specify a port number for the connection and to specify its parameters. Specify the port speed, the slot number, the encoding to be used, and the type of DS0 interface. The port type field is used to specify the connection as a STUN connection. The role (primary, secondary, and so forth) of the FRASM port to be used in the link protocol must be specified.
addport <port_num> <line_num> <line_speed> <begin_slot> <port_type> <port_role> <encoding> <interface>
4. Enter the add STUN port command (addstunport) to establish a group number for the connection.
addstunport<port_num><group_num>
5. Once the group, line, and port have been specified, enter the add link station command (addls) to specify the FRASM port link station address and xid. A link consists of two link stations and the connecting transmission medium. In order to start an SNA session, a link between the two nodes needs to be established. Since in a STUN connection the SNA can be terminated at the FRASM, a session requires that the FRAS BNN port act as a link station. The link station name and the xid are used during the process of establishing a link.
addls <port_num><lsaddress><xid>
6. Enter the add channel (addchan) command to specify the Frame Relay portion of the connection. This consists of specifying the DLCI and committed information rate (cir) for the channel.
addchan <chan_num> <port_num> <dlci_num> <cir>
7. Enter the add STUN route command (addstunroute) to establish a connection between the Frame Relay channel number and the link station.
addstunroute <port_num><lsaddress><chan_num><lsap>
port number of the STUN connection to be routed in the range 1-192 |
|
channel number of the STUN connection to be routed in the range 16-1015 |
|
To complete the BSTUN connections, proceed as follows:
1. Enter the addln command to specify the physical port number (from 1 to 8), which is to be used for the BSTUN connection.
2. Enter the add BSTUN group command (addbstungroup) command to create a BSTUN group. This command is used to specify the group number and whether local acknowledge is to be implemented. When a BSTUN connection is made, the connection is assigned to a group in the add BSTUN port command.
3. Enter the addport command to specify a port number for the connection and to specify its parameters. Specify the port speed, the slot number, the encoding to be used and the type of DS0 interface. The port type field is used to specify the connection as a BSTUN connection. The role (primary, secondary, and so on) of the FRASM port to be used in the link protocol must be specified.
addport <port_num> <line_num> <line_speed> <begin_slot> <port_type> <port_role> <encoding> <interface>
4. Enter the add BSTUN port command (addbstunport) to establish a group number for the connection.
addbstunport<port_num><group_num>
5. Enter the add channel (addchan) command to specify the Frame Relay portion of the connection. This consists of specifying the DLCI and committed information rate (cir) for the channel.
addchan <chan_num> <port_num> <dlci_num> <cir>
6. Enter the add BSTUN route command (addbstunroute) to establish a connection between the Frame Relay channel number and the 3270 control unit.
addbstunroute <port_num><cuaddress><chan_num><lsap>
port number of the BSTUN connection to be routed in the range 1-192 |
|
channel number of the BSTUN connection to be routed in the range 16-1015 |
|
Add the lines before doing the clock configuration.
Add the line by entering addln "line no." Enter the dsplns command to check all the added lines on the cards.
On the sample display shown above, line number 1, 2, 3, 4 are out of alarm. Lines 5, 6, 7, 8 are in alarm.
Enter the following command on the command line at IMATM prompt to check the alarm on ds1 line#6:
To check the Out of Frames, the number of RAIs and so on, enter the following command on the command line:
Always clear up all the alarms before checking up for dspalmcnt. To clear up all the alarms enter the following command on the command line:
This command will display any HEC errored cell received on a particular line (which is part of the AIM group).
Make sure the ds3 is out of alarm. There should not be any alarms. To display the ds3 alarms, enter the following command on the command line:
The dspalm display shows that there is an UAS statistical alarm on ds3. To see exactly which statistical alarm occurred, enter the following command:
AXIS18.1.9.IMATM.a > dspalmcnt "-ds3 1"
Type <CR> to continue, Q<CR> to stop:
The display above shows that there was an AIS 34 times in last 24 hours. The best way to see the alarm is clear the alarm and see it.
To see the dsx3 line parameter, enter the following command:
LineAIScBitsCheck: Check C-bits
LineRcvFEACValidation: 4 out of 5 FEAC codes
line number -- value of 1 is accepted, for IMATM-T3T1/E3E1
The command to configure plcp for ds3 is xcnfln. Syntax follows:
The command to see the plcp configuration is dspplcplns.
PLCP CellFraming PayloadScramble PlcpLoopback
---- ----------- --------------- ------------
Make sure that Payload Scramble is configured correctly on IMATM as well as on the other side of dsx3. Otherwise, there will be an alarm on dsx3.
|
Note If you are connecting dsx3 to IGX, make sure that LSS is disabled. The command to do that is xcnfalm "-plcp 1 -lsen 2". |
IMATM has a T3/E3 interface and multiple T1/E1 interfaces. It is supposed to replace a physical long distance T3/E3 ATM trunk by a group of long distance T1/E1 lines. Had there been a physical T3/E3 line, the clock sync info automatically reaches from one end to other along with T3/E3 data traffic. Since we are breaking this continuation and replacing it with a group of T1/E1 lines, we need to have some mechanism in IMATM to pass the clock across.
There are two commands in IMATM that can be used to change the clock configuration.
1. The cnfclksrc command lets you configure a primary clock source and a secondary clock source. An on-board PLL generates a clock, phase locked to the primary clock source (If primary has gone bad then phase is locked to the secondary. It switches back to primary when primary clock becomes OK. Should both secondary and primary becomes unusable, PLL switches to backplane). This phase-locked o/p then drives the entire card (for example, both T3 and N T1s on that card).
2. With the cnfln command you can configure the individual T1 lines as either Loop Clock or Local Clock. When Local Clock is configured, it uses the clock selected using the cnfclksrc command, whereas, Loop Clock is simply the clock recovered from T1/E1 receive. (Even when the T1/E1s are configured in Loop Clock, the T3 transmit is still driven by the primary/secondary clock source configured by cnfclksrc.)
3. The user needs to select which 10 sources should be used to drive all T1 and T3 lines on that card, and then configure the clock source entering the cnfclksrc command.
Possible clock sources are 8-T1/E1 lines, T3/E3 line, or backplane 8 kKHz.
|
Note Primary from one of the DS1 lines and secondary from T3 does not make sense, since these two mean different directions of passing clock sync. |
T3 --- N* T1 --------------- N*T1 ---- T3
In order to pass clock info from END A to END B:
END A T1 lines need to derive the clock derived from the END A T3.
END B derives clock from one of the incoming N T1s and drives the T3 transmit using this derived clock.
cnfclksrc -pri T3 -sec BP8K (or T3) -cur PRI ; at END A
cnfclksrc -pri DS1_1 -sec DS1_2 -cur PRI ; at END B
Thus, when we configure clock source at END B as being DS1_1 or DS1_2, and so on, we are assuming that the incoming DS1 to the END B card has a valid clock present on its T1 lines inserted by a clock derived from the END A's T3).
T3 --- N* T1 --------------- N*T1 ---- T3
More accurate and useful clock-
typically provided by teleco. network.
If we have a configuration like this, where the T1 lines are retimed to a STRATUM clock in the network, then we are expected to derive sync from the T1 lines only at both ends. We then need to configure pri/sec sources from the DS1 lines at both ends and control Xmit of the T3 as well as T1s using this derived clock.
If the T1 lines are not retimed in network then we need to insert proper clock at one end on T1s (END A in the previous example) and take it out from T1 lines at other end (END B).
1. T1s can take the transmit clock from Local clock or Loop clock. This is configured entering the cnfln command.
2. The "Local clock" for T1s and T3 is taken from one source. This source is configured entering the cnfclksrc command.
3. The correct way of configuring the clocking depends on the setup. The idea is to sync up the network to the most accurate and useful clock available.
The command to do that is addaimgrp. Make sure that lines are added and out of alarm.
Syntax: addimagrp (or addaimgrp) "group_num port_type list_of_lines"
IMA group number—Value ranging from 1 to 8
Port type—1-UNI, 2-NNI, 3-STI, 4-Virtual trunks UNI (STI in UNI)
List of lines—List of lines separated by periods
In the following example, we are adding AIM Group number 1, which uses the UNI port type and includes lines 1, 2, and 3.
After adding the AIM group, add the channel route entry (with set of VPI values), so as to route the cells, with configured VPIs to a particular AIM group. Without channel route entry, all the cells will be discarded as unknown vpi_vci cells.
Enter the dspchrtes command to display all the channel route entries currently configured for all the AIM groups.
Enter the cnfchrte command to configure a channel route entry.
The command to configure AIM group is cnfaimgrp.
cnfimagrp (or cnfaimgrp) grp max_diff_delay n_res_lns
IMA group number—Value ranging from 1 to 8
Max diff delay—Value between 0 and 50 for Model A IMATM T1/E1;
between 0 and 275 for Model B IMATM T1;
between 0 and 200 for Model B IMATM E1
# resilient links—Value between 0 and MAX_PHYS_LINKS
Resilient Link—Maximum number of T1/E1 links within the AIM group that can go down and the AIM group will still remain active.
In the following example, the number of resilient link is changed to 2 and the differential delay to 100 ms. Enter the following command:
The command is dspaimgrp group number.
Port Speed (cells/sec) : 13476
LcpCellsPeriodicity(cells) : 128
LcpDelayTolerance (IMA frames) : 1
NumResiliency : 2 /* Resilient links */
MaxTolerableDiffDelay (msec) : 100 /* Diff delay tolerable */
Local IMA id : 0x33 /* The aim group is
Observed Diff delay (msec) : 0 in loopback, and hence
Syntax : dspimagrp (or dspaimgrp) "imagroup_number"
IMA group number -- value ranging from 1 to 8
Explanation of some of the parameters:
Local AIMUX ID—Indicates the IMA-ID in use at the local end.
Used in AIM state machine to communicate with the other side entity.
Remote AIMUX ID—Indicates the IMA-ID in use at the remote end.
Observed Diff delay—The observed differential delay is between the two different physical links in the IMA group.
This command can be used to configure the maximum queue depth and the EFCI threshold for the QueueNo. (1...8)
The Maximum Qdepth of the queue should be less if the AIM group is carrying delay-sensitive traffic (for example, voice, video).
The maximum Qdepth could be maximum possible, and EFCI threshold could be less (ForeSight can kick in quickly, if Q starts building up) for AIM group carrying nondelay-sensitive data traffic.
PortNo could be 1...8. QueueNo can be only 1. This command will display the queue configuration for the corresponding queue. It will also show the number of cells discarded due to Qfull condition on this queue.
|
Note The "EgressQFullDiscardedCells" is cleared with the clrportcnt command. |
This command will clear the EgressQFullDiscardedCells.
This command is useful to see Egress Received cells (from T3) for this particular port and ingress transmitted cells to group of T1s and vice versa.
This command is useful to see cells Rx. from dsx3 and cells Tx to dsx3. The last unknown vpi_vci from dsx3, if there is one, is also reported.
Posted: Thu Nov 20 21:43:20 PST 2003
All contents are Copyright © 1992--2003 Cisco Systems, Inc. All rights reserved.
Important Notices and Privacy Statement.
