Release Notes for the
Cisco 6100 Series System Release 2.4.0

System controller (SC) module software

9651-001-31

Network interface (NI) module software

9601-001-26

Subtending host module (STM) software

9601-005-09

CAP ATU-C module software

9101-001-20

DMT-2 ATU-C module software

9101-002-01

RDF File

9601-003-11

2.4.0

Yes

Yes

No

No

Yes

Yes

Yes

Yes

2.3.5

No

No

Yes

Yes

Yes

Yes

Yes

Yes

2.3.0

No

No

Yes³

Yes⁴

Yes

Yes

Yes

2.2.1

No

Yes

Yes⁵

Yes⁵

Yes⁵

Yes

Yes

Yes

2.2.0

No

Yes

Yes⁵

Yes⁵

Yes⁵

Yes

Yes⁵

Yes

2.1.0

No

Yes

Yes⁶

Yes⁶

Yes⁶

Yes⁶

Yes⁶

Yes

2.4.0

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

2.3.5

No

No

Yes

Yes

Yes

Yes

Yes

Yes

2.3.0

No

No

Yes³

Yes³

Yes⁴

Yes

Yes

Yes

2.2.0

No

No

Yes⁵

Yes⁵

Yes⁵

Yes

Yes

Yes

Cisco 6130⁴

Yes

Yes⁵

Yes

No

No

Cisco 6100⁶

Yes⁷

No

No

Yes¹

Yes¹

Dual-port CAP ATU-C

Yes

Yes²

No

No

No

Dual-port DMT-2 ATU-C³

Yes⁴

No

Yes⁴

No

Yes⁴

CAP ATU-C and DMT-2 ATU-C

Circuit ID field activation

Provides more line and location detail for a particular subscriber. You can configure and display a circuit ID for each subscriber. The circuit field accepts up to 20 characters. Valid characters include
A to Z, a to z, 0 to 9, and{.,-=;:'#!$%&*()<>_+/\}.

DMT-2 ATU-C

ViewRunner support

Transmission rates

• Upstream bit range from 864 kbps to 32 kbps
  
  
• Downstream bit range from 8032 kbps to 32 kbps
  
  

Bit rates are determined by the number of carriers and the number of bits per carrier

CAP ATU-C
DMT-2 ATU-C

Per-subscriber power setting transmittal

You can refine signal attenuation by setting the power transmittal in dBm/Hz per-subscriber.

Configure the transmit power attribute using the Subscriber Properties dialog box in the PSD Transmit Power field. Default values are

• Upstream: -38
  
  
• Downstream: -40
  
  

DMT-2 ATU-C

Interleaved path support¹

Default is Interleaved and unavailable for change (dimmed).

Although you can configure interleaved values on the DMT-2 Subscriber Properties dialog box, the interleaved option is the read-only default per subscriber on the Service Provisioning dialog box. Delay options include: 0, 500, 1000, 2000, 4000, 8000, 16000, 32000, 64000. Delay settings 16000 to 64000 result in a maximum setting.

DMT-2 ATU-C

Training mode

The read-only default is set to Standard Train.

CAP ATU-C
DMT-2 ATU-C

Priority queue support

You can configure the priority queue for each subscriber PVC through the Service Provisioning dialog box. The default Priority Queue setting is 3. You can also configure the priority queue for each transit subscriber PVC through the Transit Subscriber dialog box. A new header field labeled Priority displays priority in a
list box.

DMT-2 ATU-C

Performance monitoring: parameter and value

You can access the Parameter field in the DMT-2 Port Status dialog box to view performance counters such as HecVioll and FarCvflfe. Performance counters measure the DMT-2 port performance according to the following factors:

• Attenuation equations
  
  
• Signal-to-noise ratios
  
  
• Number of anomalies per interval
  
  
• Types of defects such as SEF or LPR
  
  
• Number of cells used
  
  

The Value field contains the value of the performance counter.

DMT-2 ATU-C

Forward error correction (FEC)

You can configure the number of redundancy bytes in both upstream and downstream. The default value is 16. The unit is bytes. Redundancy byte amount options range in multiples of two from 2 to 16.

DMT-2 ATU-C

Overhead framing structure

The default overhead framing structure value is 3. You can configure overhead framing according to the structure definitions and values below.

• Full asynchronous—Full overhead framing with asynchronous bit-to-modem timing (enabled synchronization control mechanism). Value = 0.
  
  
• Full synchronous—Full overhead framing with synchronous bit-to-modem timing (disabled synchronization control mechanism). Value = 1.
  
  
• Reduced separate fast—Reduced overhead framing with separate fast and sync bytes in fast and interleaved latency buffer respectively (64 kbits/s framing overhead). Value = 2.
  
  
• Reduced merged fast—Reduced overhead framing with merged fast and dsync byte, using either the fast or interleaved latency buffer (32kbit/s framing overhead). Value = 3.
  
  

CAP ATU-C
DMT-2 ATU-C

NI module information

You can determine the physical layer and the number of ports from the NI module Port Configuration dialog box.

DMT-2 ATU-C

ATM virtual circuit support

Support for the ATM PVC is similar to the existing CAP ATU-C module. Subscriber provisioning is similar to the CAP ATU-C module, except for training rates.

CSCdm36984

CLI command show stat icp client should display clearer info

CSCdm33193

Subtend port loses cell at full ds3 rate

CSCdm31593

CAP RADSL Training Problems

CSCdm31114

NI DS3 does not throttle in the upstream direction

CSCdm29561

Postcard message generation from SCs UI does not send correct data

CSCdm29330

lrNumTrapReceivers does not get updated when trap rcvr added

CSCdm29272

Agent reports bad value error when adding LIM chassis

CSCdm25847

NI constantly resets in direct connect system if modems are training

CSCdm25734

Chassis Alarms returns bad values

CSCdm25731

Cannot pre provision a lim chassis

CSCdm25708

Walk of lrSlotTable for lim chassis returns invalid lrSlotEntries

CSCdm25689

System reset causes bad NI state

CSCdm25489

Chassis list (from cr) resets the sc when lim chassis present

CSCdm13350

MCLI I/O allocates memory of long-term storage region

CSCdm13099

Y2K—Event occuring on 2/29/2000 is logged on 3/1/2000

CSCdm04581

MCLI commands/responses echoed on all mcli session terminals

CSCdm04577

Password for NI hidden cmds menu visible in CT session terminal

CSCdm02229

NI hidden debug menu is left exposed when CLI session closed

CSCdm00653

6100 DS3 Interface Reports FIFO FULL Locks Up

CSCdk90876

NI command pd all or dump causes the NI to hang

CSCdk72226

SHM and NI images wont distribute if certain size

CSCdk71731

User input not received by NI correctly when done via cut-thru

CSCdk67493

DS3 NI not sending PLCP yellow during PLCP OOF

CSCdk62129

NI CT:When Reset command is given , CT window hangs

CSCdk60677

NIM debug i/f "dr" and "mr" cmds not range checking device number argument.

CSCdk49649

In CLI for SendByte and RevcvByte the error codes are incorrect.

CSCdk47034

Concurrent NI debug sessions may give incorrect output values.

CSCne02408

SC and LC reset after modifying dip switches ids to 14 and 15 from 1 and 2

CSCne02132

In alarms test, clearing AIS alarm gives PLCP OOF rather than RAI alarm

CSCdm87985

ADI/Aware DSP Upload Can Corrupt ADI/Aware Flashs

Impact: If an ADI/Aware DSP upgrade is terminated prematurely (normally by pulling the card out of the chassis during the upgrade) there is the potential that the ADI/Aware flash will be corrupted. Once the flash is corrupted it can not be programmed. The board will need to be returned to Cisco and the flash popped off and re-burned. This problem can occur whenever the upgrade process is prematurely interrupted. This could include removing the line card from the chassis or failing to lock the subscriber port so that a connection request is received during the upgrade.

Workaround: Ensure that the DSP upgrade process is not interrupted prematurely. To do so, be certain that the subscriber port is locked and that the card is not removed prior to completion of the upgrade. TAC will also have an external Sun Solaris DSP upgrade program that will be available on an an-needed basis for customers. Improvements to the DSP upgrade process are expected in a follow-on project.

CSCdm87044

ATUC: generates IPCtoSMB Unable to allocate buffer!

Impact: After training up 64 modems at 864/8032 and requesting "status x" from all 64 ATUC channels (via serial ports), the following error occurs for all ATUCs:

CSCdm81817

DMT2 various mode, upstream rate combos cause high ES, CB.

Impact: Excessively high corrected/uncorrected block ratios and errored second (ES) counts are seen when the following combinations are provisioned for a subscriber -- overhead framing mode-2, 64 Kbps upstream; or overhead framing mode-3, 96 Kbps upstream

The high error rates only occur with the overhead framing mode-2, 64K upstream, or overhead framing mode-3, 96K upstream combinations.

Workaround: Do not provision a subscriber with either of these combinations of overhead framing mode and upstream rate. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm77282

DMT2 margins cannot be set lower than 6 dB.

Impact: Setting margin for DMT-2 lower than 6 does not change the actual setting, which will still be 6 dB. This occurs with any DMT-2 margin setting less than 6 dB.

Workaround: There is no workaround. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm76074

DMT 16000, 32000 and 64000 interleave all equal max

Impact: Setting interleave for DMT-2 line cards to 16000, 32000 or 64000 all results in the maximum interleave setting being used. This will show up on the CPE output as interleave = 64. This occurs whenever any of these interleave settings are utilized.

Workaround: There is no workaround. This is the maximum interleave that can currently be obtained with DMT-2 in the 6130. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm69068

DMT with trellis enable,CB/UB go down, but intermittent high ES

Impact: Trellis coding enabled for a subscriber causes the connection to show up with a high ratio of corrected/uncorrected blocks and excessive errored second (ES) counts. This occurs when trellis coding is enabled at both the CO and CPE so that trellis is then active for the connection.

Workaround: Do not enable trellis for the subscriber. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm69047

DMT Setting PSD -43=-52;-46=-40;-49=-40.

Impact: Setting the PSD to -46 dB or -49 dB in ViewRunner results in the downstream PSD still being the default of -40 dB. This occurs only with DMS-2.

Workaround: There is no workaround. The inability to set PSD lower than -43 dB is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem. It is strongly recommended the default PSD of -40 dB be used.

CSCdm59472

DMT cannot train at 128K increment for upstream

Impact: When a subscriber is configured for an upstream rate that is an increment of 128K such as 128K, 256K, etc., the trained rate will always be at least 32K below the configured rate. This only occurs with upstream increments of 128K.

Workaround: There is no workaround to obtain an increment of 128K. If a higher rate is desired, setting the next higher rate will allow this. For example, to get greater than 128K, provision for 160K, which will result in a 160K trained rate.

CSCdm52542

DMT Downstream file transfer perf overly impacted by upstream rt.

Impact: When the downstream rate is set much higher than the upstream rate -- for example, 1544/96 -- the data transfer rates will likely not be as high as would be expected for downstream data transfers. This occurs when upstream rates are set very low relative to downstream rates, and particularly for file transfers that require acknowledgements, such as FTP.

Workaround: Do not set very low upstream rates, such as below 256 K, when using high downstream rates.

CSCdm46110

DS3 subtending bandwidth is less than 40.7 Mbps.

Impact: When a rate greater than 40.2 Mbps of traffic is sent through the DS3 subtend ports, traffic will be sent no faster than 40.2 Mbps. This occurs only when attempting to send traffic at a rate greater than 40.2 Mbps through the subtend ports.

Workaround: None required. This change was need to keep cells from being lost due to the inability of the subtend ports to handle traffic greater than 40.2 Mbps.

CSCdm45054

Clearing of fan shelf alarm causes system reset.

Impact: When a fan tray alarm is cleared, the entire system reset. This occurs with any kind of fan tray alarm that indicates a fan tray is required - as with DMT-2 - but is not currently present.

Workaround: There is no workaround.

CSCdm43638

DMT-2 fails to meet T1.413 loop midCSA6-cannot set margin of 3.

Impact: When running T1.413 MidCSA 6 loop, the required rate cannot be met. MidCSA 6 requires a margin setting of 3; however, there is no way to set the margin below 6. Although you can set the value below 6 in ViewRunner, the margin will remain at 6.

Workaround: There is no workaround. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm41964

DMT-2 line card FE Corrected Blocks dont match CPE.

Impact: When corrected/uncorrected blocks are checked at the CO side then compared with similar statistics on the CPE side, the values may not necessarily match. Other statistics may also be out of sync. This can occur with any kind of connection.

Workaround: There is no workaround. There are slight differences in the manner in which the firmware reports statistics to the CO and the CPE side. The discrepancies are not severe enough to impact accurate reporting of general functionality. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm40771

Could not login to NI debug mode after creating 1600 PVCs and 1.

Impact: Can not login to NI debug mode after creating 1600 PVCs and 1500 transit subscribers. Occurs after creating greater than 1600 subscribers and 1100 transit subscribers.

Workaround: There is none. Due to memory limitations, you cannot enter NI debug mode after exceeding this number of subscribers and transit subscribers. The fix is scheduled to be incorporated in the next major release.

CSCdm36644

9000-9350 ft/ -34 dBM/Hz, CPE cannot train.

Impact: Setting a subscriber to a PSD setting other than the default value of -40 dB may cause unpredictable results. In particular, a setting of -34 dB will cause the subscriber to continually retrain.

Workaround: Leave the subscriber PSD setting at the default of -40 dB. This problem is due to a limitation with third-party hardware. There is currently no scheduled date from the vendor for resolution of the problem.

CSCdm21026

Reseating OC-3 NI generate buffer overflow msg in VR4W.

Impact: When an NI is reseated in the chassis, an error message is generated in the ViewRunner log indicating the following:

CSCdm20798

DMT2 does not accurately report trained/not trained.

Impact: When checking in ViewRunner on line card status or when viewing Active Connections, a connection may be incorrectly reported as not trained although it actually is. With Active Connections, the problem manifests itself by not listing all actual active connections.

Workaround: Refresh status. Due to limitations in how quickly information can be reported, there will be periodic incorrect reporting regarding line card connection status.

CSCdk59039

If a user is sending debug messages and enters an invalid instance type in the message, the SC can crash. This problem only occurs in a lab debug environment.

Impact: There is no impact. This only occurs in a lab debug environment.

Workaround: Do not enter a debug message with invalid insttype.

CSCdk57824

Downstream RS errors when adjacent modem channel hangs up.

Impact: If both channels of an ATU-C central office modem card are trained to customer premise equipment and one channel hangs up, the other channel on the card produces a burst of downstream reed-solomon errors.

Workaround: These errors are corrected by the CPE equipment.

CSCdk57362

If the system is in Direct Connect mode, the CPE TIMER alarm is not cleared after lock/unlock of the line port.

Impact: This only impacts systems in Direct Connect Mode. If a CPE TIMER alarm is raised against a port in Direct Connect Mode, the alarm will not clear until the hour timer has expired.

Workaround: The user must wait the whole hour before the alarm is cleared against the line port.

CSCdk55957

ATU-C does not give enough margin when trained to a c660.

Impact: Lowers performance when using a c660.

Workaround: If you want to use this release against a c660, the system administrator should set the downstream margin to 12 to avoid problems, or if this is unacceptable, the system administrator should set the margin setting and analyze it on a case-by-case basis (By analyzing RS Error counters at the CPE.) Upstream margin of 6 should prove sufficient to prevent problems, but the real margin may vary depending on the rate selected.

CSCdk55611

If you specify a bad TFTP server IP address from the Boot Rom Menu Screen, the SC fails the download and does not run.

Impact: The SC does not load with the proper image if you specify a bad TFTP server IP address from the Boot Rom Menu Screen.

Workaround: The user must specify the proper TFTP server IP address during downloads.

CSCdk53848

Life Line not preserved when POTS splitter card is removed.

Impact: When the POTS card is removed, phone service is lost.

Workaround: Do not remove the POTS card.

CSCdk53830

The counter "Failed Trains due to non-timer enabled CPE" not visible while the system is in Direct Connect Mode.

Impact: If the system is in Direct Connect Mode, you not know how many times a line port failed to train because the CPE gear is not timer-enabled.

Workaround: There is no workaround.

CSCdk53806

Command Line Interface does not show "CPE NOT TIMER ENABLED" alarm text.

Impact: When showing the alarms through the Command Line Interface, an alarms appears that shows no textual information about the alarm. The "CPE not Timer Enabled" Alarm is not visible from the Command Line Interface.

Workaround: Show the alarms with ViewRunner.

CSCdk52588

Interrupted upgrade causes system controller (SC) to be orphaned.

Impact: If an upgrade is interrupted after the old SC image is deleted and before the new SC image has been successfully tftp'd, the Cisco 6100 will be orphaned.

Workaround: The user must download the new SC image by using the Craft interface.

CSCdk51475

NI debug command DS3 status (ss all) shows an unexpected m23 format.

Impact: No impact.

Workaround: There is no workaround.

CSCdk46493
CSCdk49143

When four VCs are sending data simultaneously and are configured on one port, the bandwidth is not distributed fairly.

Impact: If four VCs are configured for one line port and sending data simultaneously, the bandwidth for some of the VCs are not be distributed fairly amongst each of the VCs. Users on some of those VCs may not get their provisioned bandwidth.

Workaround: There is no workaround. The NI does not perform per VC queuing.

CSCdk43651

LIM controller SMB failure over Temperature and Humidity.

Impact: After a temperature transition from +50 C at 10% humidity to +70 C at 50% humidity, the LIM controller stops responding to the System Monitor on the SC over the external SMB bus.

Workaround: There is no workaround for this problem.

CSCdk43208

Disconnecting tip/ring momentarily causes loss of upstream data

Impact: If the tip/ring cable is momentarily removed for under 1 second, the upstream data path does not operate until the modem power is recycled.

Workaround: Cycle the power on the CPE equipment.

CSCdk37403

If you are using a snmp command line tool or mib browser, the set fails. This is not a problem when using ViewRunner. If the Cisco 6100 is sent a snmp "row create" of a subscriber with the line port of the subscriber included in the set (lrSubscrRowStatus.1=createAndGo; lrSubscrLinePort.1=lrLpLPoolId.2.1.1), the set fails.

Impact: This is only an impact if not using ViewRunner.

Workaround: Send in the set to separate SNMP PDUs. First send the "row create," and then the Line port for the subscriber.

CSCdk34684

CAP ATU-C margin can be lower than specified for 1024 k and 896 k.

Impact: After training is completed, the margin may not be as high as specified.

Workaround:

1. Provision the line for different rates (such as 1280 for 1024, 960 for 896)

2. Provision the line with an additional 3 dB of downstream margin.

CSCne02362
CSCne02112

System control IP information becomes corrupted after save or after BOOTP is completed.

Impact: Only an installation impact when you use the BOOTP capability.

Workaround: Re-enter the SC IP address in the boot menu.

CSCne02176

When locked, Cisco 6100 modules still respond with alarms when pulled from chassis.

Impact: No system impact, unimplemented feature.

Workaround: Not required.

CSCne02002

On the SHM module, the RAI and OCD alarms do not clear when the alarm condition is corrected.

Impact: Incorrect alarms.

Workaround: No workaround.

CSCne01970
CSCne02364

Fabric Control does not configure transit VPCs.

Impact: No system impact, unimplemented feature.

Workaround: Use VCCs only.

CSCne01912
CSCne01913

DS3 Subtending port does not block data flow upon port or module lock. Unimplemented feature at this time.

Impact: Cannot block data by unlocking Subtend port.

Workaround: To block the data, pull out the DS3 cable.

CSCne01901

Infrequently, in a single LIM chassis system, if you replace a LIM controller, this causes ViewRunner to display two LIM chassis.

Impact: Displays a LIM chassis, which does not exist.

Workaround: Delete second LIM chassis.

CSCne01854

Following an NI reset, the following non-fatal events are seen in the event window:

FC_SUBTEND_PORT_BUFFER_OVERFLOW INFO

FC_SUBTEND_PORT_UTOPIA_ERROR INFO

FC_SUBTEND_PORT_INGRESS_ERROR INFO

FC_SUBTEND_PORT_INGRESS_2_ERROR INFO

FC_SUBTEND_PORT_EGRESS_PARITY INFO

Impact: No impact, events are generated at start up and is based on the startup sequence of NI and SHM.

Workaround: Not required.

CSCne01497

Identical fault message text is provided when NI DS3 C-bit parity detection is asserted and cleared.

Impact: The message "Unexpected frame format" is provided in both the asserting and clearing cases.

Workaround: Not required.

CSCne01131

During an NI reset, active OC3 cell flow may cause NI alarm and NI shutdown on power-up.

Impact: If the fiber is currently plugged in and active, during install or reset of the system, then the NI may shutdown. An alarm in ViewRunner is provided to alert the user of this problem:
"ATM SWITCH POLLING RATE IS INADEQUATE, CELLS DROPPED"

Workaround: Disconnect fiber and restart NI.