Table Of Contents
Release Notes for Cisco ONS 15454
Release 4.1.8.2Path Protection Over 1+1 Provisioning Prior to Upgrade
Maintenance and Administration
Path Protection Over 1+1 Operation
Changed Default Alarm Severities
CSCdx40462, CSCdx47176, CSCdw22170
ONS 15454 Conducted Emissions Kit
Upgrading to Use the G1000-4 Ethernet Card
CSCdv10824: Netscape Plugins Directory
CSCdy48478 Fan Tray Assembly Fan Fail Lamp Test Failure
CWDM and DWDM GBIC Compatibility with G1000-4 Cards and G1K-4 Cards
SONET and SDH Card Compatibility
Active Cross Connect (XC/XC10G/XCVT) or TCC+/TCC2 Card Removal
CSCds13769: Fujitsu FLM-150 and Nortel OC-3 Express
Active Cross Connect (XC/XC10G/XCVT) or TCC+/TCC2 Card Removal
Transponder (TXP_MR_10G) and Muxponder (MXP_2.5G_10G) Documentation
Resolved Software Caveats for Release 4.1.8.2
CSCec88426, CSCec88508, CSCed85088
CSCec59739, CSCed02439, CSCed22547
CSCec88402, CSCed31918, CSCed83309, CSCec85982
CSCea16455, CSCea37089, CSCea37185
This issue is resolved in Releases 4.1.8.2 5.08 6.2.2 7.02 7.2
Maintenance and Administration
Transmission Control Protocol Specification
New Features and Functionality
New Software Features and Functionality
DS3i-N-12 Card Support for SONET Platforms
ML-Series Resilient Packet Ring
Protect Threshold Crossing Alarms
Alarms, Conditions, and Errors Changed in Release 4.1.x
Cisco Optical Networking Product Documentation CD-ROM
Cisco Product Security Overview
Reporting Security Problems in Cisco Products
Obtaining Technical Assistance
Cisco Technical Support & Documentation Website
Definitions of Service Request Severity
Obtaining Additional Publications and Information
Release Notes for Cisco ONS 15454
Release 4.1.8.2
July 2008
Note
The terms "Unidirectional Path Switched Ring" and "UPSR" may appear in Cisco literature. These terms do not refer to using Cisco ONS 15xxx products in a unidirectional path switched ring configuration. Rather, these terms, as well as "Path Protected Mesh Network" and "PPMN," refer generally to Cisco's path protection feature, which may be used in any topological network configuration. Cisco does not recommend using its path protection feature in any particular topological network configuration.
Release notes address closed (maintenance) issues, caveats, and new features for the Cisco ONS 15454 SONET multiplexer. For detailed information regarding features, capabilities, hardware, and software introduced with this release, refer to the "Release 4.1.x and 4.5" version of the Cisco ONS 15454 Procedure Guide; Cisco ONS 15454 Reference Guide; Cisco ONS 15454 Troubleshooting Guide; and Cisco ONS 15454 and Cisco ONS 15327 TL1 Command Guide. For the most current version of the Release Notes for Cisco ONS 15454 Release 4.1.8.2, visit the following URL:
http://www.cisco.com/univercd/cc/td/doc/product/ong/15400/454relnt/index.htm
Cisco also provides Bug Toolkit, a web resource for tracking defects. To access Bug Toolkit, visit the following URL:
http://www.cisco.com/pcgi-bin/Support/Bugtool/launch_bugtool.pl
Contents
Resolved Software Caveats for Release 4.1.8.2
New Features and Functionality
Cisco Product Security Overview
Obtaining Technical Assistance
Obtaining Additional Publications and Information
Changes to the Release Notes
This section documents supplemental changes that have been added to the Release Notes for Cisco ONS 15454 Release 4.1.8.2 since the production of the Cisco ONS 15454 System Software CD for Release 4.1.8.2.
There have been no changes to the release notes for Release 4.1.8.2
Caveats
Review the notes listed below before deploying the ONS 15454. Caveats with tracking numbers are known system limitations that are scheduled to be addressed in a subsequent release. Caveats without tracking numbers are provided to point out procedural or situational considerations when deploying the product.
Software Upgrades
Path Protection Over 1+1 Provisioning Prior to Upgrade
The value of the NODE.circuits.upsr.AllowUpsrOverOnePlusOne NE default after a software upgrade is FALSE. To enable path protection over 1+1 you must set this default to TRUE in the Defaults editor window. To edit NE defaults refer to the "NTP-A164 Edit Network Element Defaults" procedure in the Cisco ONS 15454 Release 4.1 Network Element Defaults document.
Maintenance and Administration
Caution
Note
Path Protection Over 1+1 Operation
For path protection over 1+1 circuits double fault scenarios are not supported. If both the path protection and 1+1 are switched, traffic can be lost. To allow path protection functionality the 1+1 must be placed in "Forced to Working" state. The path protection selector can only be switched manually. To allow 1+1 functionality, the path protection selector must be in "Forced to working" state. In other words, at all times one of the selectors (path protection or 1+1) must be in the "Forced to Working" state to operate without traffic loss in this mode.
CSCeg35075
A protection switch of a path protection circuit over a 1+1 link does not update in the edit pane. To see the correct status, check the path protection switch status from the path protection selector window. This issue will not be resolved.
CSCeg40008
A false UNEQ-V alarm is raised when you create a one-way VT 1.5 path protection circuit over 1+1 SONET Protection. This issue will not be resolved.
CSCeg37198
Occasionally, an LOP-V alarm can fail to clear after a switch on a path protection over 1+1 VT 1.5 circuit. This can occur when you create a bidirectional VT 1.5 path protection circuit over 1+1 SONET protection, then apply a force to working to both selectors on one node. When you place the working trunk card of 1+1 Protection OOS, then place the port back in IS, LOP-V is raised but traffic is fine. Traffic comes back to the working 1+1 card after the WTR expires. But then when you release the force switch from the selector, the LOP-V fails to clear. This issue will not be resolved.
CSCeg37206
Creating a VT path protection over 1+1 protected circuit and then deleting the protection group results in AIS-V and a failure to switch to protect (FAILTOSW-PROT) alarm on the path protection protected circuit. This issue will not be resolved.
CSCea84427
CTC or Cisco Transport Manager (CTM) communication to a node might be lost while the node is still able to respond to a ping. The CORBA interface on the node can be locked by sending invalid (non-Corba) data to the TCC IIOP listener port. When the CORBA interface is locked, legitimate CORBA communications are lost. To prevent this from occurring, the ONS 15454 should be placed on a secure network or behind a firewall. In the event that this issue arises, CTC/CTM management can be regained by performing a Manual TCC reset using TL1. This issue is resolved in Release 4.6.
CSCec58326
When switching to the protect card in EC1 during 1:1 protection, IPPM goes blank. This is the result of incorrect handling of STS Path PMs (IPPM) in EC1 1:1 protection. This issue is resolved in Release 4.6.
CSCec48038
A PCA circuit might be displayed as INCOMPLETE after a span upgrade; for example, when a BLSR span that is carrying PCA traffic is upgraded from OC-48 to OC-192. Although the circuit is reported as INCOMPLETE in CTC, traffic should be unaffected in this instance. To correct the display, restart CTC. This issue is resolved in Release 4.6.
CSCec29230
The NE PWR-B alarm is not reported in CTC after an upgrade to Release 3.4.2. You can view the alarm through TL1. This issue is resolved in Release 5.0.
CSCeb45149
A 1+1 non-revertive system in which the protect card is reseated while the working card is present and active reverts back to working after you clear a Force to Protect or Manual to Protect. Affected cards are OC-3 8 port and OC-12 4 port. To work around this issue, soft-reset the working card with the protect card present and active. This issue will be resolved in Release 4.6.
Changed Default Alarm Severities
The following alarm severities have changed for Release 4.1.x.
Table 1 Changed Alarm Severities
VT-MON::AUTOSW-LOP
NSA-Minor
VT-MON::AUTOSW-UNEQ
NSA-Minor
VT-TERM::PLM-V
SA-Major
CSCeb20996
While using the orderwire capability of the AIC-I, you must not input a station number with less than 4 digits. If you enter, for example, 123, CTC will display 0123; however, you will not be able to ring the node by dialing either *123, or *0123. This issue will be resolved in Release 4.6.
CSCea93638
Path level alarms are displayed on the CTC conditions pane for deleted circuits. This issue may occur on any circuit deletion case. The conditions may be cleared by a TCC side switch. This issue is resolved in Release 4.6.
CSCea61887
Terminal loopback is provisionable even if the card is in transponder mode.
To see this, in the provisioning tab for a G1000 or G1K-4 card pick a pair of ports and set them to transpond with each other. The condition also holds true by picking one port and setting it to transpond with itself (one-port unidirectional). Once the transponder setting is provisioned, go to the Maintenance tab and attempt to provision terminal loopback on any of the ports that were previously provisioned for transponder functionality. CTC allows terminal loopback to be provisioned even though the setting has no effect due to the fact that the ports are performing transponder functions. If terminal loopback is truly intended, you should remove the transponder settings. A warning stating that terminal loopback has no effect if transponders are present will be displayed in Release 4.6.
CSCea74000
Rarely, a PLM-P alarm on an OC-12 card might remain after deleting the affected STS-1 circuit. This condition persists even after deleting the affected card. If this occurs, you can reset the TCC2s to remove the stuck alarm. This issue is resolved in Release 4.6.
CSCea71675
Rarely, when nodes appear gray in CTC and you view the BLSR tab at the network level, the tab may be blank. When this occurs, all tabs on the same level will also become blank. To avoid this issue, do not enter into the BLSR tab at the network view until all of the nodes are some color other than gray. This issue will be resolved in Release 4.6.
CSCea92969
The switch indicator does not clear on the Maintenance > Protection tabs when a revertive 1:1 protection group is provisioned to be non-revertive. If this occurs, from the Maintenance > Protection tabs, select the slot with the "Switch" indicator. Click the "Clear" button. This issue is resolved in Release 4.6.
CSCea78364
Simultaneous failure of working and protect cards in 1:N protection groups may not be alarmed as service affecting. This can occur when the working card of the protection group has been removed from the chassis, and the protect card of the protection group is subsequently issued a Manual Reset. Since the working and protect facilities are impaired, the Improper removal alarm should clear and be reissued as a Critical and service affecting condition. This issue will be resolved in Release 5.0.
CSCea81001
When a fault condition exists against a circuit or port that is in the OOS-MT or OOS-AINS state (or when you are using the "Suppress Alarms" check box on the CTC Alarm Behavior pane), the alarm condition is not assigned a reference number. If you were to place the circuit or port in service at this time, in the absence of the reference number, the CTC alarm pane would display the condition with a time stamp indicating an alleged, but incorrect, time that the autonomous notification was issued. Clicking the CTC alarm "Synchronize" button at this stage will correct the alarm time stamp. There is no way to remedy the lack of reference number. This issue is resolved in Release 6.0.
CSCdx35561
CTC is unable to communicate with an ONS 15454 that is connected via an Ethernet craft port. CTC does, however, communicate over an SDCC link with an ONS 15454 that is Ethernet connected, yielding a slow connection. This situation occurs when multiple ONS 15454s are on a single Ethernet segment and the nodes have different values for any of the following features:
•
•
•
When any of these features are enabled, the proxy ARP service on the node is also disabled. The ONS 15454 proxy ARP service assumes that all nodes are participating in the service.
This situation can also occur immediately after the aforementioned features are enabled. Other hosts on the Ethernet segment (for example, the subnet router) may retain incorrect ARP settings for the ONS 15454s.
To avoid this issue, all nodes on the same Ethernet segment must have the same values for Enable OSPF on the LAN, Enable Firewall, and Craft Access Only. If any of these values have changed recently, it may be necessary to allow connected hosts (such as the subnet router) to expire their ARP entries.
You can avoid waiting for the ARP entries to expire on their own by removing the SDCC links from the affected ONS 15454 nodes. This will disconnect them for the purposes of the proxy ARP service and the nodes should become directly accessible over the Ethernet. Network settings on the nodes can then be provisioned as desired, after which the SDCC can be restored.
This issue will not be resolved.
CSCdz84149
If a user is logged into CTC as a superuser (or other higher level security type), and then another superuser changes the first user's security level to "retrieve" (or another lower level security type) without first logging the user out, the lower level user is then still able to perform some actions authorized only for the original login security level. For example, a "provisioning" level user demoted to "retrieve" level in this manner can still provision and edit MS-SPRings (BLSRs) while logged into the current session, though the same user may no longer provision DCCs. To ensure that a user's level is changed completely, the superuser must log the user out prior to changing the security level. This issue is resolved in Release 4.6.
CSCdz90753
In the Maintenance > Cross Connect Resource Pane, the VT matrix port detail is inconsistent with the general VT matrix data. This can occur when a 1+1 protection scheme is in place. To avoid confusion, note that the VT matrix data counts the VTs for both the working and protect card, while the detail data counts the VTs only for the working card. This issue will be resolved in Release 4.6.
CSCdz35479
Rarely, CTC Network view can freeze following the deletion or addition of a node from or to a BLSR/MS-SPRing. This can result in the CTC Network view no longer updating correctly. If this occurs, restart CTC. This issue is resolved in Release 4.6.
CSCdy57891
An LOP-P alarm can be inadvertently cleared by an LOS that is raised and cleared. On OC48AS, OC192, and OC12-4 cards, when an LOP condition and an LOS condition are both present on the input, an LOS will be raised as per Telcordia GR 253 alarm hierarchy. However, upon clearing the LOS with the LOP still present, the LOP alarm, which should then be raised, is not. An AIS-P condition will be visible. This issue is resolved in Release 6.0.
CSCdy56693
Microsoft Windows XP uses more memory than previous Microsoft operating systems, and this may result in reduced CTC performance. To avoid reduced performance, you can:
•
•
•
•
•
CSCdy61275
Far end path FC-P is not counted on EC1 or OC3 cards. When a path defect is transmitted to the far end, it reports RDI-P. However, the condition is not examined and reported as a PM count. This issue is resolved in Release 5.0.
CSCdy10030
CVs are not positively adjusted after exiting a UAS state. When a transition has been made from counting UAS, at least 10 seconds of non-SES must be counted to exit UAS. When this event occurs, Telcordia GR-253 specifies that CVs that occurred during this time be counted, but they are not. This issue will not be resolved.
CSCdy55556
In a 1:N protection group, where a protect card is protecting a failed card and another working card, which is missing, has a lockon condition, upon removing the lockon condition from the missing working card, the protect card may switch from the card it had been protecting to carry the traffic of the missing working card that just had the lockon removed. To avoid this issue, replace the failed working card before removing the lockon. This issue is resolved in Release 6.0.
CSCdy11012
When the topology host is connected to multiple OSPF areas, but CTC is launched on a node that is connected to fewer areas, the topology host appears in CTC, and all nodes appear in the network view, but some nodes remain disconnected. This can occur when the CTC host does not have routing information to connect to the disconnected nodes. (This can happen, for example, if automatic host detection was used to connect the CTC workstation to the initial node.)
CTC will be able to contact the topology host to learn about all the nodes in all the OSPF areas, but will be unable to contact any nodes that are not in the OSPF areas used by the launch node. Therefore, some nodes will remain disconnected in the CTC network view.
To work around this issue, if no firewall enabled, then the network configuration of the CTC host can be changed to allow CTC to see all nodes in the network. The launch node must be on its own subnet to prevent network partitioning, and craft access must not be enabled. The CTC host must be provisioned with an address on the same subnet as the initial node (but this address must not conflict with any other node in the network), and with the default gateway of the initial node. CTC will now be able to contact all nodes in the network.
If a firewall is enabled on any node in the network, then CTC will be unable to contact nodes outside of the initial OSPF areas.
NE Defaults
The following caveats apply for NE defaults.
•
•
•
CSCdy35514
The terminology used for provisioning overhead circuits has changed as of Release 3.4 as follows.
Overhead Circuit Types
•
•
Overhead Channel Types
•
•
•
•
•
Note
CSCds88976
When a new circuit is created around a ring (path protection or BLSR), the SD BER or SF BER alarm can be raised depending on the order in which the spans are provisioned. The alarms will eventually clear by themselves. Traffic is not affected. This issue will not be resolved.
CSCdx40462, CSCdx47176, CSCdw22170
While upgrading nodes from releases prior to 3.2, CTC might lose connection to the far end nodes. When this occurs, you will not be able to ping the grayed-out nodes; however, if you continue the upgrade, this problem resolves itself. This issue is resolved in Release 3.2, but can still occur when upgrading from nodes with earlier software releases.
CSCdw66895
XCVTs (both active and standby) reboot continuously when the K3 byte is mapped to the E2 byte on one side of a WTR span. The rebooting occurs after the WTR timer expires. This has been seen on a two fiber BLSR with OC-48AS. To avoid this issue, if possible, change the K3 mapping on both ends of the span before creating the ring; or, alternatively, you can prevent the ring from reverting during the K3 mapping by setting the Ring Reversion time to "never." Once you have completed the mapping of the K3 byte to the E2 byte on both sides, return the Ring Reversion to its normal value. This issue is resolved in Release 4.6.
ONS 15454 Conducted Emissions Kit
If you are deploying the Cisco ONS 15454 within a European Union country that requires compliance with the EN300-386-TC requirements for Conducted Emissions, you must obtain and install the Cisco ONS 15454 Conducted Emissions kit (15454-EMEA-KIT) in order to comply with this standard.
Upgrading to Use the G1000-4 Ethernet Card
Before installing or seating the G1000-4 Ethernet card on node running Release 3.1 or prior, you must upgrade the software on that node to Release 3.2 or later. This is as designed.
CSCdv10824: Netscape Plugins Directory
If you use CTC, JRE, and the Netscape browser with a Microsoft Windows platform, you must ensure that any new installation of Netscape uses the same Netscape directory as the previous installation did, if such an installation existed. If you install Netscape using a different path for the plugins directory, you will need to reinstall JRE so that it can detect the new directory.
CSCdu82934
When you auto-route a VT circuit on an ONS 15454 node, a path is computed based on the availability of STSs on the nodes involved. This selection process, when combined with a lack of VT matrix (or STS-VT connections) on an auto-route selected node, can result in the VT circuit creation failing with the message "unable to create connection object at node." To correct this situation, manually route VT circuits in cases when auto-routing fails. The error message will indicate which node is at issue.
"Are you sure" Prompts
Whenever a proposed change occurs, the "Are you sure" dialog box appears to warn the user that the action can change existing provisioning states or can cause traffic disruptions.
Hardware
CSCdy48478 Fan Tray Assembly Fan Fail Lamp Test Failure
A user-initiated lamp test does not illuminate the fan fail LED on the fan tray assembly for the ONS 15454 or the ONS 15454 SDH. The lamp test successfully lights the LEDs for major, minor, and critical alarms on the fan tray, and for all cards in the chassis, but does not light the fan fail LED. An actual fan failure, however, will light the fan fail LED.
CWDM and DWDM GBIC Compatibility with G1000-4 Cards and G1K-4 Cards
G1000-4 cards support CWDM and DWDM GBICs. G1K-4 cards with the CLEI code of WM5IRWPCAA (manufactured after August 2003) support CWDM and DWDM GBICs. G1K-4 cards manufactured prior to August 2003 do not support CWDM or DWDM GBICs.
Note
CSCdw57215
In a configuration with OC-48 Any Slot cards and an STS-24c circuit, provisioned between G1000-4 cards with traffic going over the OC-48 span, extracting the G1000-4 card at one end of the STS-24c circuit before deleting the circuit will result in a traffic hit on all existing SONET circuits defined over that same span. This only occurs when the STS-24c is provisioned on timeslot 25.
In the Cisco ONS 15454 Procedure Guide, Release 4.1.x, refer to the "NTP-77 Delete Circuits" procedure to delete the 24c circuit before removing the card. Once you have deleted the circuit, refer to the "DLP-191 Delete a Card from CTC" task (also in the procedure guide) to delete the G1000-4 card. This issue is resolved in Release 6.0.
Line Cards
Note
CSCef43718
Under the conditions listed below, the DS1 card does not raise Loss of Frame for a Mismatched Frame Format. This can happen when you change from the correct frame format to an incorrect frame format, where the correct format is Unframed and the incorrect format is D4, and you switch from the incorrect, to correct, and back to the incorrect format. This issue is resolved in Release 5.0.
CSCef24504
When a circuit is deleted from a DS3XM while DS1 loopbacks exist on the ports associated with that circuit, the loopbacks become frozen on those ports and cannot be removed. If this occurs, reestablish a circuit that uses those ports, then remove the loopbacks prior to deleting the circuit. This issue is resolved in Release 5.0.
CSCef18088
A DS1 AINS circuit might change to IS state after two-fiber BLSR switch on a mixed-node network with one ONS 15327 and two ONS 15454s. This issue is resolved in Release 5.0.
CSCec39567
Deleting a DS3I 1:N protection group might leave the protect card LED in a standby state. This can occur in a DS3I 1:N protection group with a LOCKON applied to the working card (ONS 15454 ANSI chassis only). Upon deleting the protection group, the LED on the protect DS3I card and the CTC display are still in the standby state. Soft reset the protect card to update the LED on the card and in CTC. An alternative workaround is to remove the LOCKON before deleting the protection group. This issue will be resolved in a future release.
CSCeb45064
Terminal loopback on DS3XM-6 card may not work properly. AIS-P may be transmitted towards the system (away from the line side) direction. A VT circuit newly created to the DS3XM-6 with no valid input signal to the port can cause this. A terminal loopback will not work because AIS-P is being transmitted towards the system direction. To avoid this issue, apply a valid signal to the input port for the DS3XM card for at least 10 seconds with no terminal or facility loopback applied to that particular port. After receiving a valid signal for this time period, the terminal loopback will function properly even if the input signal is disconnected from the DS3XM port. This issue is resolved in Release 4.6.
CSCdz49928
When using KLM type fuses with specific types of fuse and alarm panels, the PWR-REDUN alarm may not be displayed once the fuse is blown. A KLM fuse does not have a blown fuse indicator build into it. As a result, the blown fuse detection circuitry on the FAP may continue to provide voltage on its output despite a blown fuse.
Ethernet Polarity Detection
The TCC2 does not support polarity detection & correction on the LAN Ethernet port. This is a change from prior common control cards, such as the TCC+. If your LAN Ethernet connection has the wrong polarity (due to incorrect wiring on the backplane wire-wrap pins), the connection will work when using a TCC+, but not with a TCC2. To avoid possible problems, ensure that your Ethernet cable has the correct mapping of the backplane wire-wrap pins. For Ethernet pin mappings, consult the "DLP-A 21 Install LAN Wires on the Backplane" procedure in the user documentation.
If you are using a TCC+, the Release 4.0 or 4.1.x software will report the polarity issue (previous releases do not), by raising the standing condition: LAN Connection Polarity Reverse Detected (COND-LAN-POL-REV). Also, notification will appear on the fan tray LCD, which will display "BP LAN POL. NEG." The issue will typically be reported during the software upgrade process, but can also be raised during a new installation when using TCC+ and Release 4.0 or 4.1.x.
If this is a new installation with a TCC2 and you have the Ethernet polarity reversed, the TCC2 will not communicate over the LAN Ethernet interface (no polarity correction will occur), and no condition will be reported, nor will the fan tray LCD indicate an issue.
SONET and SDH Card Compatibility
Tables 1, 2, and 3 list the cards that are compatible for the ONS 15454 SONET and ONS 15454 SDH platforms. All other cards are platform specific.
CSCdw27380
Performing cross connect card switches repeatedly might cause a signal degrade condition on the lines or paths that can trigger switching on these lines or paths. If you must perform repeated cross connect card switches, lock out the corresponding span (path protection, BLSR, or 1+1) first. This issue is resolved in Release 6.0.
LOS Behavior
When an OC-N card is seeing LOS and the problem is resolved (for example, the pulled fiber is reinserted) the LOS will normally clear quickly, and any other errors seen on the signal will be raised. However, in the special case where the restored signal is unframed, the LOS will remain raised (that is, the LOS will not be replaced by an LOF). This is standard SONET behavior per Telcordia GR-253 R6-57, Method 1, where to clear LOS the signal must also contain valid framing alignment patterns.
CSCdw66444
When an SDH signal is sent into an ONS 15454 OC-12/STM-4 (IR, 1310 LR and 1550 LR) or an OC-48/STM-16 high-speed (IR and LR) port which has been configured to support SDH, an SD-P (Signal Degrade) alarm will appear as soon as the circuit is created. This alarm will continue to exist until the circuit is deleted.
To avoid this problem, when provisioning an OC-12/STM-4 (IR, 1310 LR and 1550 LR) or an OC-48/STM-16 high-speed (IR and LR) port to support SDH, disable the signal degrade alarm at the path level (SD-P) on the port.
Also, PM data at the path level will not be reliable. You must set associated threshold values to 0 in order to avoid threshold crossing alerts (TCA) on that port. The path threshold values to set to zero are CV-P, ES-P, SES-P, and UAS-P.
These issues are the result of a hardware limitation, and there are no current plans to resolve them.
CSCdw09604
If you are using an XC10G with OC-48, you must use either OC-48AS or OC-48 cards with a revision number higher than 005D.
Jitter Performance with XC10G
During testing with the XC10G, jitter generation above 0.10 UI p-p related to temperature gradient testing has been observed. This effect is not expected to be seen under standard operating conditions. Changes are being investigated to improve jitter performance in a future release. Defect numbers related to this issue include CSCdv50357, CSCdv63567, CSCdv68418, CSCdv68441, CSCdv68389, CSCdv59621, and CSCdv73402.
Active Cross Connect (XC/XC10G/XCVT) or TCC+/TCC2 Card Removal
You must perform a lockout in BLSR, path protection, and 1+1 before physically removing an active cross connect (XC/XC10G/XCVT) or TCC+/TCC2 card. The following rules apply.
Active cross connect (XC/XC10G/XCVT) cards should not generally be physically removed. If the active cross connect or TCC+/TCC2 card must be removed, you can first perform an XC/XCVT/XC10G side switch or TCC+/TCC2 reset and then remove the card once it is in standby, or you can perform a lockout on all circuits that originate from the node whose active cross connect or active TCC+/TCC2 will be removed (performing a lockout on all spans will also accomplish the same goal). No lockout is necessary for switches initiated through CTC or through TL1.
CSCdv62565, CSCdv62573
In a 1:N protection group, traffic loss could occur if a DS-N card is preprovisioned and then added to the group while another working card in the group is removed from its slot. To avoid this, before adding slots to a protection group ensure that:
•
•
This issue is resolved in Release 5.0.
E Series and G Series Cards
Note
CSCdy37198
On Cisco ONS 15454s equipped with XC or XCVT cross-connect cards, neither the E100T-12 nor the E1000-2 cards raise an alarm or condition in CTC when Ethernet traffic is predictably lost due to the following circumstances:
Circuits exist between Ethernet cards (E100T-12 and/or E1000-2) built over Protection Channel Access (PCA) bandwidth on BLSR spans. When BLSR issues a switch, the PCA bandwidth is preempted. Since there is no longer a connection between the ends of the Ethernet circuit, traffic is lost.
Note
This issue will be resolved in a future release.
CSCdr94172
Multicast traffic can cause minimal packet loss on the E1000-2, E100-12, and E100-4 cards. Packet loss due to normal multicast control traffic should be less than 1%. This issue was resolved in Release 2.2.1 for broadcast, and in Release 2.2.2 for OSPF, and some multicast frames. As of Release 3.0.3, the ONS 15454 supports HSRP, CDP, IGMP, PVST, and EIGRP, along with the previously supported broadcast and OSPF.
Note
Note
Note
Multicast MAC addresses used by the following control protocols have been added to the static MAC address table to guarantee no loss of unicast traffic during normal usage of these MAC addresses:
E1000-2/E100T
Do not use the repair circuit option with provisioned stitched Ethernet circuits.This issue is under investigation.
Single-card EtherSwitch
Starting with Release 2.2.0, each E100/E1000 card can be configured as a single-card EtherSwitch configuration to allow STS-12c of bandwidth to be dropped at each card. The following scenarios for provisioning are available:
1.
2.
3.
4.
5.
6.
7.
When configuring scenario 3, the STS-6c must be provisioned before either of the STS-3c circuits.
Multicard EtherSwitch
When deleting and recreating Ethernet circuits that have different sizes, you must delete all STS circuits provisioned to the EtherSwitch before you create the new circuit scenario. (See the preceding "Single-card EtherSwitch" section for details on the proper order of circuit creation.) Enable front ports so that the VLANs for the ports are carried by the largest circuit first. A safe approach is to enable the front port before you create any circuits and then retain the front port VLAN assignment afterwards. If you break the rules when creating a circuit, or if you have to delete circuits and recreate them again, delete all circuits and start over with the largest first.
CSCds02031 E1000-2/E100
Whenever you drop two 3c multicard EtherSwitch circuits onto an Ethernet card and delete only the first circuit, you should not provision STS-1 circuits to the card without first deleting the remaining STS-3c circuit. If you attempt to create an STS-1 circuit after deleting the first STS-3c circuit, the STS-1 circuit will not work and no alarms will indicate this condition. To avoid a failed STS-1 circuit, delete the second STS-3c prior to creating any STS-1 circuit.
ML-Series
CSCeb56287
When an ML-series circuit's state is provisioned from In-Service (IS) to Out-of-Service (OOS), and then back to IS, data traffic does not recover. To avoid this issue, prior to changing the state from IS, set the POS port to shut down on the CLI. After the state is changed back to IS from OOS, set the POS port to "no shutdown." This issue is resolved in Release 4.6.
CSCeb24757
Disconnecting a transmit fiber on an ML1000 port causes only the neighboring port to take the link down. Ideally, both ports should identify that the link went down so upper layer protocols can reroute the traffic to a different port. To work around this situation, issue "shutdown" and "no shutdown" to the port that has the disconnected or faulty transmit fiber. This issue is resolved in Release 4.6.
CSCea46580
SPR input counters do not increment on a BVI with an SPR interface. This issue will not be resolved.
CSCea35971
A monitor command may disappear from the configuration after a TCC reboots. To avoid this issue, use the exec command, "terminal monitor," instead (a minor drawback is that this command applies to all VTYs), or, alternatively, reapply the monitor command after connection is lost. This is as designed.
CSCdy31775
Packets discarded due to output queue congestion are not included in any discard count. This occurs under either of the following conditions:
•
•
This issue is resolved in Release 4.6.
CSCdz49700
The ML-series cards always forward Dynamic Trunking Protocol (DTP) packets between connected devices. If DTP is enabled on connected devices (which might be the default), DTP might negotiate parameters, such as ISL, that are not supported by the ML-series cards. All packets on a link negotiated to use ISL are always counted as multicast packets by the ML-series card, and STP and CDP packets are bridged between connected devices using ISL without being processed. To avoid this issue, disable DTP and ISL on connected devices. This functionality is as designed.
CSCdz68649
Under certain conditions, the flow-control status may indicate that flow control is functioning, when it is not. Flow-control on the ML-series cards only functions when a port-level policer is configured. A port-level policer is a policer on the default and only class of an input policy-map. Flow-control also only functions to limit the source rate to the configured policer discard rate, it does not prevent packet discards due to output queue congestion.
Therefore, if a port-level policer is not configured, or if output queue congestion is occurring, policing does not function. However, it might still mistakenly display as enabled under these conditions. To avoid this issue, configure a port-level policer and prevent output queue congestion. This issue will not be resolved.
CSCdz69700
Issuing a shutdown/no shutdown command sequence on an ML1000 port clears the counters. This is a normal part of the startup process and there are no plans to change this functionality.
CSCea11742
When a circuit between two ML POS ports is provisioned OOS, one of the ports might erroneously report TPTFAIL. This issue exists for both ML100T-12 and ML1000-2 cards. If this occurs, open a console window to each ML card and configure the POS port to shutdown. This issue is resolved in Release 4.6.
CSCea20962
No warning is displayed when applying OOS to ML drop ports on the circuit provisioning window. This issue is resolved in Release 5.0.
CSCin29274
When configuring the same static route over two or more interfaces, use the following command:
ip route a-prefix a-networkmask a.b.c.d
Where a.b.c.d is the address of the outgoing gateway, or, similarly, use the command:
ip route vrf vrf-name
Do not try to configure this type of static route using only the interface instead of the address of the outgoing gateway.
CSCin32057
If no BGP session comes up when VRF is configured and all interfaces have VRF enabled ensure that at least one IP interface (without VRF) is configured and add an IP loopback interface on each node. This issue will not be resolved.
CSCin35960
POS ingress classification based on IP precedence does not match the packets when inbound policy map classifying based on IP precedence is applied to the POS interface, which is configured for HDLC or PPP encapsulation. To avoid this issue, use LEX encapsulation (default) or, at the Ethernet ingress point, mark the COS based on an IP precedence classification, then classify based on the COS during POS ingress. This issue is resolved in Release 4.6.
CSCdy47284
ML-100 FastEthernet MTU is not enforced. However, frames larger than 9050 bytes may be discarded and cause Rx and Tx errors.
CSCdz74432
Issuing a "clear IP route *" command can result in high CPU utilization, causing other processes to be delayed in their execution. To avoid this issue do not clear a large number of route table entries at once, or, if you must use the "clear IP route *" command, do not install more than 5000 EIGRP network routes.
Interoperability
CSCds13769: Fujitsu FLM-150 and Nortel OC-3 Express
You cannot provision the FLM-150 and OC-3 Express in 1+1 revertive switching mode. The problem occurs when the ONS 15454 issues a user request in revertive mode to the protect channel. When the user request is cleared, the ONS 15454 issues a No Request. However, the FLM-150 and OC-3 Express issues a Do Not Revert, which causes traffic to remain on the protection channel. Based on GR-253, section 5.3.5.5, the FLM-150 and the OC-3 Express should respond with a No Request.
BLSR Functionality
CSCec74273
If a four-fiber BLSR execute MS-R or WTR coexists with SF-P on the same span and FS-R is issued on a different span in the ring, the FS-R preempts the MS-R or WTR, causing a traffic outage. The ring switch is removed, but one side remains in the ring switch state. To recover from this situation, issue a lockout of the protect span on the span that raised the ring switch event. This issue is resolved in Release 4.6.
CSCeb24331 and CSCeb40119
If you create a four-fiber BLSR with a VT circuit on it, then delete the circuit and the ring, then created a two-fiber BLSR on the same ports, you may see an unexpected AIS-V on the path, even before any additional circuit is created. A soft switch of the TCC will clear the AIS-V condition. This issue is resolved in Release 4.6.
CSCeb09217
Circuit states are not updated after a span update. If you update a four node OC-12 two-fiber BLSR to a four node OC-192 two-fiber BLSR, the previous PCA circuits should be shown as two-fiber BLSR protected, but they are shown as "UNKNOWN" protected. If you relaunch CTC this situation is corrected. This issue is resolved in Release 5.0.
CSCea59342
DS3 PCA traffic may take up to 20 seconds to recover after a BLSR switch is cleared. This can occur with DS3 PCA traffic on two-Fiber or four-Fiber BLSR configuration with XCVT cards in the same nodes as the DS3 cards. This issue will not be resolved.
CSCea81000
In a two-fiber or four-fiber BLSR, MS-RFI is not reported for an LOS or LOF with a ring lockout in place on a different span. This issue is resolved in Release 5.0.
CSCdy68207
Failing the working and protect spans on a four-fiber BLSR while an extra traffic circuit runs over the span and a lockout is on the span can cause the extra traffic to permanently fail, with no AIS.
The failure scenario is only reproducible by failing and restoring fibers in the following sequence.
Step 1
Step 2
Step 3
Step 4
If this issue occurs, clear the lockout span. All extra traffic is immediately restored. You may then reissue the lockout span. This issue is resolved in Release 4.6.
CSCdy56668
Ethernet circuits may appear in the CTC circuit table with an INCOMPLETE status after a BLSR/MSSP span is upgraded. The circuits, when this occurs, are not truly incomplete. They are unaffected and continue to carry traffic. To see the circuit status correctly, restart CTC. This issue is resolved in Release 4.6.
CSCdy45902
Traffic that should be dropped remains unaffected when a BLSR Protection Channel Access (PCA) VT tunnel is placed OOS. You must place all circuits in the tunnel OOS before the traffic will be dropped. This issue is resolved in Release 5.0.
CSCdw32540
The two protect OC48AS cards at the ends of a four fiber BLSR span must both be configured as either K3 or Z2 (not a mixture). If both ends are not the same, the BLSR may fail to switch correctly. In Release 3.4 the BLSR wizard ensures that both ends are configured correctly; however, you must still avoid manually changing the value on one side only (and hence, causing a mismatch) at the card level. If you do mismatch bytes at the card level, you can discover this by going to the BLSR edit map tied in with the BLSR wizard. The mismatched span will be red, and right-clicking on the span will allow you to correct the problem.
CSCdw58950
You must lock out protection BLSR, 1+1, and path protection traffic to avoid long, or double traffic hits before removing an active XC, XCVT, or XC10G card. You should also make the active cross connect card standby before removing it.
CSCdv70175
When configuring a node with one 4 Fiber BLSR and one 2 Fiber BLSR, or with two 2 fiber BLSRs, an issue exists related to the version of XC deployed. Revision 004H and earlier revisions of the XC do not support these configurations. All later revisions of the XC and all versions of the XCVT and XC10G cross connects support all permutations of two BLSRs per node.
CSCdv53427
In a two ring, two fiber BLSR configuration (or a two ring BLSR configuration with one two fiber and one four fiber ring) it is possible to provision a circuit that begins on one ring, crosses to a second ring, and returns to the original ring. Such a circuit can have protection vulnerabilities if one of the common nodes is isolated, or if a ring is segmented in such a way that two non-contiguous segments of the circuit on the same ring are each broken.
CSCct03919
VT1.5 and VC3/VC12 squelching is not supported in BLSR/MSSPR.
Database Restore on a BLSR
When restoring the database on a BLSR, follow these steps:
Step 1
Step 2
Step 3
Path Protection Functionality
CSCeb37707
With a VT path protection circuit, if you inject signals with a thru-mode test set into one path of the circuit in a particular order, you may not see the appropriate alarms. This can occur when you first inject LOP-P, then clear, then inject LOP-V. This issue will not be resolved.
CSCdv42151
When a path protection circuit is created end-to-end, CTC might not create the cross-connection on all the nodes along the path at the same time. This might cause an SD-P condition along the path. When the circuit is fully provisioned on all nodes, the SD-P will clear automatically. Other conditions that can be expected while the circuit is being created are LOP-P and UNEQ-P. To reduce the risk of unexpected transient conditions, circuits should be created in the OOS_AINS state.
Active Cross Connect (XC/XC10G/XCVT) or TCC+/TCC2 Card Removal
As in BLSR and 1+1, you must perform a lockout on path protection before removing an active cross connect or TCC+/TCC2 card. The following rules apply to path protection.
Active cross connect (XC/XC10G/XCVT) cards should not generally be physically removed. If the active cross connect or TCC+/TCC2 card must be removed, you can first perform an XC/XCVT/XC10G side switch or TCC+/TCC2 reset and then remove the card once it is in standby, or you can perform a lockout on all circuits that originate from the node whose active cross connect or active TCC+/TCC2 will be removed (performing a lockout on all spans will also accomplish the same goal). No lockout is necessary for switches initiated through CTC or through TL1.
Performance Monitoring
CSCeb41916
If you create a 1+1 protection group, create a circuit on the working line, and then try to retrieve the path PMs on the protect side using TL1, the request is denied. To work around this issue, use CTC to retrieve the Path PMs on the protect line. This issue is resolved in Release 5.0.
CSCea38791
In the CTC Performance > Statistics tab of the G1000-4 or G1000-2, there are no entries for Rx/Tx Multicast and Broadcast packets. This issue is resolved in Release 4.6.
Documentation
The following two notes on page 5-30 of the Cisco ONS 15454 Reference Manual, R4.1.x and R4.5 should be replaced.
Note
Note
The replacement information is contained in CWDM and DWDM GBIC Compatibility with G1000-4 Cards and G1K-4 Cards.
Transponder (TXP_MR_10G) and Muxponder (MXP_2.5G_10G) Documentation
The documentation set for the Cisco ONS 15454 contains references to new transponder (TXP_MR_10G) and muxponder (MXP_2.5G_10G) cards. These portions of the documentation are meant to refer to cards that will become available with a future release. The nXP cards are not available with Release 4.1.x. Cisco apologizes for any confusion this may cause.
TL1
Note
CSCsh41324
When running release 4.1.4, if a circuit is created within CTC and if that circuit is retrieved via TL1, all looks as expected. However, after the software is upgraded to release 6 and later, the circuit retrieve does not show the same value as before. For example FAC-4-1 changes to FAC-4-0. Workaround is to manually reset the active TCC after the upgrade.
CSCeh64248
The TL1 ED-USER-SECU command fails to modify an empty password. This issue will be resolved in a future release.
CSCed08144
Rarely, TL1 autonomous messages might not be displayed in a session after several days of PM-related provisioning changes. This issue will be resolved in Release 5.0.
CSCeb33033
An exception is raised when retrieving PM stats via TL1 for the protect card of a 1:1 protection group when the working card is active. To avoid this issue, retrieve stats from the working card instead of the protect card. This issue is resolved in Release 4.6.
CSCdz86121
In one rare case, the ONS 15454/15327 times out a user session without communicating the timeout to TL1. If this happens, the TL1 user remains logged in, although the session is actually timed out. This can occur when you log into the node with a timeout of X minutes. If the user session sits idle for all but 5 seconds of the X minutes, then you have only 5 seconds to type in a command to notify the node that the session is active. If you try this, you will likely miss the five second window, in which case the node will respond as though the session is inactive and deny access. However, because you have typed a key, irrespective of the five second window, TL1 responds as though the session is active and does not log you out (time out). You will not have access to the node and will receive a "DENY" response to TL1 commands. The error message may vary depending on commands issued. To recover from this situation, log out and log back in to TL1. This issue is resolved in Release 4.5.
CSCdz26071
The TL1 COPY-RFILE command, used for SW download, database backup, and database restore, currently does not allow a user-selected port parameter to make connections to the host. The command expects the default parameter of Port 21 and will only allow that number. This issue is resolved in Release 5.0
Resolved Software Caveats for Release 4.1.8.2
The following items are resolved in Release 4.1.8.2
CSCdu53509
When a TL1 session to a remote node (ENE) is established via a gateway node (GNE) and you have changed the node name of the ENE via either TL1, CTC or SNMP, then you must wait for about 30 seconds to issue a TL1 command via the GNE. This delay is to permit the updates to propagate to all nodes in the network. During this transition, neither the old node name nor the new node name can be used in the TL1 session to access the ENE. This is resolved now for all releases.
CSCse60231
Networks containing mixed Releases of 3.x and 6.x nodes are not supported, such a mix might occur briefly as nodes are upgraded. OSPF-HELLO alarms will be raised on DCCs when upgrading nodes to Release 6.x or later if the network contains a node running any release prior to Release 4.0. OSPF-HELLO alarms indicate that there is no management communication along the affected DCC. If the DCC was used for management communication prior to the upgrade, some nodes might become isolated. To avoid this issue ensure that all nodes in the network are first upgraded to Release 4.0 or later before upgrading any node to Release 6.0 or later. This issue is resolved in following releases 6.2.2, 7.02, 7.22., 8.0, 4.182
CSCsd47710
LAN-connected NEs become unreachable for about 15 minutes every 99 days if it has been running without a shelf controller reset for 99 days. The Ethernet port will fail without any alarms or warnings for 15 minutes. If the NE is LAN-connected during this failure, it might be unreachable by management hosts (CTC and Tl1). This occurs for all models of shelf controller, on the following platforms: ONS 15454, ONS 15327, ONS 15600, ONS 15310-CL, and ONS 15310-MA. However the management communication over the DCC will not be affected. Thus, if a node is managed over the DCC, it will not experience problems. On the other hand, if the NE is a GNE, communication to ENEs behind the GNE will be affected. This is caused because the NE has an internal clock that rolls over to zero every 99 days (approximately) the last 15 minutes prior to rollover the Ethernet interface incorrectly calculates the age of ARP entries and assumes the ARP entries have expire regardless of the true age and stops transmitting IP packets This issue is resolved in Releases 4.182 5.08 6.22 7.02 8.0.
CSCse11933
RAI alarm on DS3E was card getting raised even as LOF/LOS was present. LOS/LOF should have masked RAI. FEAC (Far End Alarm and Control) alarms fluctuate in DS3E card, when a LOF/LOS/AIS conditions are fluctuating. The LOF/LOS/AIS defect will clear FEAC alarms when the latter don't need to soak. As a result, the FEAC alarms toggle on and off rapidly, this can cause alarm storm.The workaround is to use the Alarm Profile to change FEAC alarm severity to avoid an alarm storm. This issue is resolved in Release 4.182 7.02 8.0
CSCse50795
ML Card System returns to ROMMON due to a bus error. Show tech/Show Stack after the crash will give a message similar to the following: System was restarted by bus error at PC 0x3ADC0C, address 0x3ADC0C. The software has been modified to prevent the reload when it hits these conditions.
CSCsd06997
Display of Daylight Savings Time is incorrect in CTC, TL1. As per the change in energy policy in the United States of America, the DST start date is now 2nd Sunday of March and DST end date is 1st Sunday of November. Between 2nd Sunday of March 2007 and 1st Sunday of April 2007 and last Sunday of October 2007 to 1st Sunday in November of 2007, CTC, TL1 would follow the old DST Rules. For example 11:00 03/15/2007 CDT would be reported as 11:00 03/15/2007 CST.This issue is resolved in Releases 4.1.8.2, 5.0.8, 7.0.2, and 8.0.
CSCse08560
On a DS3E card with a port provisioned for C-BIT, you might see multiple RAI alarms and alarm clears within a second. This issue is resolved in 4.18.2 and 7.0.2
CSCse42563
CTC merges VLANs from different rings when nodes belonging to different rings are added to same CTC session
•
•
•
•
•
•
This causes E100T instability when many VLANs on different rings exists (in Service Provider's networks) and can cause E100T resets and unpredictable behavior. CTM does not have this issue. Software Version 5.X and newer do not have this issue. This issue is resolved in 4.1.8.2
CSCse88396
Static routes stop working, if for some reason LAN connection on a 15454 node flaps, static routes configured with next hop as the node itself stop working. Static routes with next hop as some other node are not affected. Workaround: There are two workarounds:
1. Provision the same static route on some other node in the network.
2. Edit or re-enter host route in affected 15454. Resolution:
The issue is resolved in 4.1.8.2
Hardware
CSCdv05723 and CSCdw37046
DS-3 traffic hits can occur during synchronization changes (frequency offsets applied) on the node's timing.
A specific scenario under which this has been seen involves configurations with multiple nodes line-timed off each other in series. If a network configuration has a DS-3 circuit routed over a chain of nodes that are line-timed off each other in sequence (more than 1 line-timed "hop"), the DS-3 traffic might exhibit errors on timing disturbances applied on the source node.
The other scenario involves an abrupt change in reference frequency between two nodes. This can result in test set errors.
This issue is resolved in Release 4.1.
Line Cards
CSCeh19956
When DS1-14 cards are used and ports are provisioned for B8ZS/ESF, random DS1-14 ports raise LOF and Tx-AIS, though terminating equipment is configured and connected properly. DS1 traffic goes down and terminating equipment raises AIS. Setting the DS1-14 ports on the ONS 15454 to unframed should avoid this issue. Connected DS1 equipment can remain provisioned for B8ZS/ESF. This issue is resolved in Releases 4.1.8, 5.0.2, and 6.0.
CSCei37835
Rarely, a MFGMEM alarm is raised incorrectly against the backplane (BP) on an ONS 15454 node using the 15454-SA-HD (high density) or 15454-SA-ANSI chassis with TCC2 cards. The alarm is raised due to the inability of the TCC2 card to correctly read the EEPROM. Less than 0.02% of nodes in this hardware configuration are known to be affected.
In Release 5.0.2, when using the TCC2 in combination with the 15454-SA-HD, and the DS3/EC1-48 card, failure of the TCC2 to correctly read the EEPROM can cause a loss of traffic. The failure scenario can occur because, while using the ONS 15454 high density (HD) shelf, the inability to read the EEPROM correctly leads the TCC2 to incorrectly identify the shelf as an ONS 15454 ANSI shelf, and, in Release 5.0.2, when using DS3/EC1-48 cards, this can result in the DS3/EC1-48 cards being deleted from the database. This aspect of the issue only occurs in Release 5.0.2. The alarm itself is not service affecting in releases prior to Release 5.0. This issue is attributed to a failure to check and verify the correct content in the EEPROM.
This issue is most likely to occur during turn-up of a new node. The issue can also occur if a new TCC2 is installed in the shelf, or during a first switchover operation. Cisco recommends upgrading to Release 5.0.4 if you deploy the DS3/EC1-48 to avoid this issue. To recover from this issue, side switch the affected TCC2 back to the protect card. If the issue persists, call the Cisco TAC. This issue is resolved in Releases 4.1.8.1, 4.6.6, 5.0.4, 5.0.5, and 6.0.
CSCeg52788
When a DS1 port on the XTC card (ONS 15327) or the on the DS1 card (ONS 15454) is configured as ESF frame format, and is connected to external DS1
Guest
