Catalyst 5000 and Catalyst 3900 Frequently Asked
QuestionsPrintable Pdf
Document ID: 12270
This document provides information in the form of an FAQ for the Catalyst 3900 and 5000 family. It contains
troubleshooting information, new features with latest software revisions and some design and connectivity
guidelines.
Questions
Can ISL be used in parallel to ATM between two Catalyst 3900 switches?
Why do the messages QTP_FSM: Port x: QMAC Not Responding.... and QTP_Task(): CmdDone
not expected here appear in the message log on the Catalyst 3900?
What is VTP pruning and is it supported on the Catalyst 3900?
Where can I find a document with most of the Token Ring Switching acronyms?
How can I connect a Catalyst 5000 to a Catalyst 3900 using ISL?
What is DRiP and how does it work?
How can I use HSRP with Token Ring switches?
What is the most common cause for a Catalyst 3900 Stack to split?
Can the Catalyst 3900/3920 and Catalyst 5000/5500 detect soft errors on a per port basis
and isolate the offending station?
Is there a way to reduce Explorers on the Catalyst 5000 and 3900 switches?
How do I design a redundant switched backbone?
How does RI-RO support work on the switch?
Why can't we configure the typical parallel bridge scenario?
How do you configure redundancy for ISL in a Token Ring network?
Why does the entire TokenChannel go down if a single link fails in the Catalyst 3900?
What type of errors cause the Catalyst 3900 to switch to store-and-forward?
Can Ethernet be tunneled over the Catalyst 3900 ISL links?
What Catalyst 5000 Fast Ethernet and Gigabit modules support TRISL?
What is the minimum IOS release to support routing of Token Ring VLANs on an RSM and MLS for
Fast Ethernet on the same RSM/Catalyst 5000?
What Routing Modules of the Cat5000/6000 family support Token Ring VLANS?
Is there a command on the Catalyst switches that can display a listing of the order of
stations on the monitored rings?
Q. Can ISL be used in parallel to ATM between two Catalyst 3900 switches?
A. Because the Catalyst 3900 supports the propagation of LAN turning
information via ILL connections only, it is important that the ILL connection be the active
path in an ISL-ATM parallel connection. If the ISL module is configured in parallel connections
with ATM or Token Ring, the STP allows only one active port at a time. When using the default
Catalyst 3900 STP values, the path cost is calculated based on a 200-Mbps connection that results
in a path cost of 5 and causes the STP to place the ISL port in forwarding mode and the ATM port
or the Token Ring port in blocked mode.
However, modifying the Catalyst 3900 port STP values or using devices from other vendors
that use different STP values can block the ISL port. If an ISL port becomes blocked in an
ISL-ATM parallel connection, traffic passes via the ATM link, but VLAN trunking data is
not passed. Also, if your STP configuration makes an ATM or Token Ring port the
forwarding path to the root switch instead of the ISL link, the switch on the other end of a
blocked ISL port might incorrectly limit AREs to the incoming TrCRF. Therefore, when
modifying STP values, always ensure that the STP port path costs are configured so that the
ISL port is the preferred path. In an ISL parallel configuration, a Token Ring or ATM link
should never have a lower cost to the root bridge than the ISL link.
Q. Why do the messages QTP_FSM: Port x: QMAC Not Responding.... and, QTP_Task():
CmdDone not expected here appear in the message log on Catalyst 3900?
A. The above messages are caused by an error condition that occurs most
likely when the switch is connected to a ring which is reporting a lot of errors. Some ports
may not be operational after this message is displayed. This was fixed in version 3.0(6)--
of Catalyst 3900 and later.
Q. What is VTP pruning and is it supported on the Catalyst 3900?
A. VTP pruning is used to prevent unnecessary flooding of broadcast information
from one vlan across all trunks, since the switches don't know which vlans exist in a remote switch.
VTP pruning permits switches to negotiate which vlans are assigned to ports at the other end
of a trunk and hence, prune the vlans that aren't assigned remotely. Pruning is disabled by
default. Pruning is supported in Catalyst 3900 version 4.1(1) and later.
Q. Where can I find a document with most of the Token Ring Switching acronyms?
A. Token Ring Switching Acronyms can be found at
http://www.cisco.com/warp/customer/482/TRacronyms.htm
Q. How can I connect a Catalyst 5000 to a Catalyst 3900 using ISL?
A. The Catalyst 3900 switch can be connected to the Catalyst 5000 via the
dual 100-Mbps ISL expansion module. The Catalyst 3900 Token Ring switch doesn't support any mode
other than ISL, so it will always be trunked. The Catalyst 3900 ISL modules also only support a
100 Mb connection and defaults to full duplex.
Be very careful when connecting a Catalyst 3900 and a Catalyst 5000 switch via the ISL link.
The main problem is that the Catalyst 3900 doesn't support Fast Ethernet media negotiation.
For this reason if the Catalyst 5000 is configured for AUTO it will default to 100Mb half
duplex. This will cause problems like the port going from trunk to non trunk and packet loss.
If you want to attach the Catalyst 3900 ISL port to the ISL port of a Catalyst 5000, you must
manually configure the ISL port on the Catalyst 5000 for
100 Mbps using the set port speed command:
Usage: set port speed < mod/port>
and full-duplex using the set port duplex command:
Usage: set port duplex < mod/port> < full|half>
Q. What is DRiP and how does it work?
A. DRiP is Cisco Duplicate Ring Protocol and its job is to ensure the proper
configuration of Token Ring vlans and to create explorer reduction. One of the most important
features of DRiP is to enforce TrCRF distribution. In the Token Ring world, distributing any VLAN
other than 1003 is very dangerous because of spanning issues. For this reason, if a TrCRF
other than vlan 1003 is distributed, all ports to which that vlan is associated are disabled by
DRiP. See more information.
Q. How can I use HSRP with Token Ring switches?
A. HSRP (Hot Standby Routing Protocol uses a multicast destination address in
the network. Since nobody in the network actually sources packets with this multicast address,
the switches never learn these MAC addresses and hence, flood frames throughout the network. To
overcome this problem, there is a need to use a MAC address that can actually be used as a
SMAC by the routers in the HSRP hellos. This allows the switches to learn this address and
switch the packets appropriately. This is done by configuring a new "virtual" MAC address in
the routers. Clients need to send packets to the DMAC of this new virtual address. Here is the
output of show standby.
vdtl-rsm#sh stand
Vlan500 - Group 10
Local state is Active, priority 100
Hellotime 3 holdtime 10
Next hello sent in 00:00:01.224
Hot standby IP address is 1.1.1.100 configured
Active router is local
Standby router is unknown expired
Standby virtual mac address is 0000.0c07.ac0a
From the above display a standby group 10 (standby ip 10 1.1.1.100) has been created. The
MAC address (0000.0c07.ac0a) is basically the new virtual MAC address and the last byte is
the group (0xA = 10). Now because the router is sourcing packets with the DMAC of the
HSRP virtual MAC, the switches will learn this MAC address and only forward the packets to
the active HSRP router. In the case of the active HSRP router failing and the standby going
active, the new active router will initiate sending HSRP hellos with the same SMAC, causing
the switch MAC address tables to switch over their learned entries to the new switch
port/trunk.
More information see: HSPR with Token Ring Switches
Q. What is the most common cause for a Catalyst 3900 Stack to split?
A. The two most common causes for a stack split are: one switch in
the stack has been reset,or very heavy traffic is going through the stack backbone.
The former can be caused by aswitch crash or manual reset. In either case, further
investigation is required in order to prevent future reoccurence. The latter problem
is caused by very heavy broadcast traffic, which causes the heartbeat signal, (used as a
keepalive for the stack to stay joined) to timeout. This can be rectified by increasing the
stack timeout value. The default is 16 seconds and the maximum is 65535 seconds. It's also
important to ensure that the latest Catalyst 3900 revision is being run by checking CCO.
Q. Can the Catalyst 3900/3920 and Catalyst 5000/5500 detect soft errors on a per port basis
and isolate the offending station?
A. The Catalyst 3900/3920 Switch Software Release 4.1(1) and later performs
error detection and isolation by monitoring the Report Soft Error MAC frames generated by stations
on each port. Soft errors occur during normal ring operation and do not typically disrupt traffic
on the ring. However, soft errors can occur at a rate that could potentially degrade the performance
of the ring. Using the Catalyst 3900 or Catalyst 3920, you can configure soft error thresholds
and sampling intervals for a port. During the interval you define, the Catalyst 3900 monitors
the stations on the port. If the threshold is exceeded, the switch can be configured to generate
a trap indicating the port number and station on which the threshold was exceeded. If
necessary, you can issue a Remove Ring Station MAC frame to remove the station from
the ring. For more information, refer to the Catalyst 3900 Configuration Guide Configuring Soft
Error Monitoring .
The Catalyst 5000/5500 Token Ring Blade Software Release 3.1(1) and later also performs
error detection and isolation by monitoring the Report Soft Error MAC frames generated by
stations on each port. Refer to Configuring Soft Error Monitoring in the Configuring Token
Ring Switching for Catalyst 5000 Guide for more information.
Q. Is there a way to reduce Explorers on the Catalyst 5000 and 3900 switches?
A. In release 3.2(3) of the Token Ring Blade in the Catalyst 5000 family,
the number of explorer frames forwarded by a Token Ring port can be configured. This is important in
networks that have a significant amount of explorer frames that are impacting the network.
Use the set tokenring explorer-throttle command to control the number of incoming
explorer frames per second allowed on a Token Ring module port. For the Catalyst 3900 and
3920 there is an option called Max Explorer rate under port configuration. Refer to
Configuring Port Parameters in the Catalyst 3900 Configuration Guide.
Q. How do I design a redundant switched backbone?
A. The simplest and most effective way to do this is to provide two
(and only two) paths from
each ring to every other ring. The following is a picture of four switches in a parallel
backbone. In this diagram, you will notice that there are two completely parallel paths. The
two sides of the network do not have to be interconnected via ISL or ATM. Doing so only
adds complexity with no gain in availability. Each side of the network will be a different
TrBRF VLAN. For each ring, there will be a TrCRF on each side of the network with the
same ring number, but not the same VLAN id (you could make the VLAN id the same if you
make sure that the two halves of the network are not interconnected via VTP). To connect
servers directly to the switch, use a separate switch that is then connected to both halves of
the backbone via ISL. To connect to the router via ISL, two links can be used as shown
below:
Q. How does RI-RO support work on the Catalyst 3900 and 5000 switches?
A. Ring In/Ring Out (RI/RO) support is provided on the fiber Token Ring ports on
both the Catalyst 3900 and 5000. In addition, ports 19 and 20 of the Catalyst 3900 support RI/RO.
These ports can be used to connect to IBM 8230 compatible RI/RO ports on hubs. The
following diagram shows how this works. If the switch is connected to both the RI and RO
port of a hub, it will provide a backup capability in case a link between hubs is broken. In the
normal case, there will be a single ring connected to two switch ports and spanning tree will
block one of the ports. If the ring breaks, there will now be two rings that are bridged together
via SRS ( the two ports must be defined in the same CRF). Therefore, the ring will be healed
and will now have twice the bandwidth. It is not necessary to connect both the RI and RO,
unless redundancy is desired. Note also, that the switch will signal the hub to wrap the RI or
RO port if only the transmit fiber from the hub to switch is cut. This ensures that the hub will
maintain the integrity of the ring.
Q. Why can't we configure the typical parallel bridge scenario below?
A. The above is a valid scenario. If done in a single switch, two TrBRF VLANs (with
different bridge numbers) are defined with two TrCRF VLANs each (ring 2 and ring 3). The
ports are then connected to the MAUs. This simply appears as two parallel bridges between
the rings. Spanning Tree will prevent loops in the above configuration.
Note: There is no direct connection within the switch between TrBRF 1 and TrBRF 2.
Q. How do you configure redundancy for ISL in a Token Ring network?
A. ISL backup of another ISL connection is configured by default. All ISL ports are
configured by default to trunk all VLANs. Spanning Tree will block redundant paths between
multiple ISL ports. Spanning Tree allows the automatic configuration of backup ISL paths.
Q. Why does the entire TokenChannel go down if a single link fails in the Catalyst 3900?
A. This limitation was removed in release 4.1.1., when fault-tolerant Channels was added
(TokenChannel and ISL Channel). The fault-tolerant feature enables TokenChannel and ISL
Channel configurations to function as long as there is at least one port active in the channel.
This capability ensures that large portions of a network are not disrupted in the event a port or
cable fails within the channel by transferring the traffic to one or more of the remaining ports
in the channel.
Q. What type of errors cause the Catalyst 3900 to switch to store-and-forward?
A. The decision to transition between cut-thru and store-and-forward when a port is
configured for "auto" is based on the percentage of errored frames to all frames seen on that
port during the sampling interval .Errored frames are:
Frames with CRC errors
Aborted frames
Frames that are too short
At the end of a sampling interval, a calculation is done to determine the percentage of these
errored frames to all frames seen on the port. If the resulting percentage is greater than the
"error high threshold", the port will go to store-and-forward mode. If the percentage is lower
than the "error low threshold", the port will go to cut-thru mode.
The sampling interval, the error high threshold, and the error low threshold are all
configurable on the "Port Configuration" panel. The defaults are:
sampling interval: 10 minutes
error high threshold: 10%
error low threshold: 1%
Q. Can Ethernet be tunneled over the Catalyst 3900 ISL links?
A. No. Pass-thru of Ethernet traffic from one ISL port to another on the Catalyst
3900 is not supported. In addition, the Catalyst desktop Ethernet switches that support ISL do NOT
support Token Ring ISL passthrough.
Q. What Catalyst 5000 Fast Ethernet and Gigabit modules support TRISL?
A. Catalyst 5000 Modules That Support TRISL
* NOTES:
1. If using Catalyst 3900 ISL up-links, 3900 Main Image Release 3.0(3), or higher, is recommended.
Also, review Release Notes at:
Fiber versions of this module supports full size 17,800 byte frames.
Q. What is the minimum IOS release to support routing of Token Ring VLANs on an RSM as well as
MLS for Fast Ethernet on the same RSM/Catalyst 5000?
A. The minimum IOS release would be 12.0(3)T.
Q. What Routing Modules of the Cat5000/6000 family support Token Ring VLANS?
A.The Route Switch Module (RSM) is the ONLY Cat5000 routing module that supports
Token Ring VLANs. The Cat5000's Route Switch Feature Card (RSFC) does NOT support
Token Ring VLANs. There is NO support for Token Ring VLANs in any Cat6000 product.
Support for the Token Ring RSM feature was first introduced in the Cisco IOS 11.3(5)T. The
Token Ring RSM feature is supported on all RSM IOS 12.0T software release images.
Q. Is there a command on the Catalyst switches that can display a listing of the order of stations by
MAC address on the monitored rings?
A. Yes, on the Catalyst 5000 and 5500 in Token Ring blade release 3.2(5), you can use the
command show station ordertable to display a listing of the order of stations on a per port
basis, here is a sample display:
Pteradactyl-Sup> (enable) sh station ordertable 7/4
