Cisco ONS 15454 SDH Troubleshooting Guide, Release 4.0
Chapter 1, General Troubleshooting
Downloads: This chapterpdf (PDF - 0.98MB) The complete bookPDF (PDF - 3.66MB) | Feedback

General Troubleshooting

Table Of Contents

General Troubleshooting

1.1 Network Troubleshooting Tests

1.2 Identify Points of Failure on a E3-12 Circuit Path

1.2.1 Perform a Facility Loopback on a Source E3-12 Port

Create the Facility Loopback on the Source E3-12 Port

Test the Facility Loopback Circuit

Test the E3-12 Cabling

Test the E3-12 Card

Test the FMEC

1.2.2 Perform a Hairpin on a Source Node Port

Create the Hairpin on the Source Node Port

Test the Hairpin Circuit

Test the Standby Cross-Connect Card

Retest the Original Cross-Connect Card

1.2.3 Perform a Terminal Loopback on a Destination E3-12 Port

Create the Terminal Loopback on a Destination E3-12 Port

Test the Terminal Loopback Circuit on the Destination E3-12 Port

Test the Destination E3-12 Card

1.2.4 Perform a Hairpin on a Destination Node

Create the Hairpin on the Destination Node

Test the Hairpin Circuit

Test the Standby Cross-Connect Card

Retest the Original Cross-Connect Card

1.2.5 Perform a Facility Loopback on a Destination E3-12 Port

Create a Facility Loopback Circuit on a Destination E3-12 Port

Test the Facility Loopback Circuit

Test the E3-12 Cabling

Test the E3-12 Card

Test the FMEC

1.3 Identify Points of Failure on an STM-N Circuit Path

1.3.1 Perform a Facility Loopback on a Source-Node STM-N Port

Create the Facility Loopback on the Source STM-N Port

Test the Facility Loopback Circuit

Test the STM-N Card

1.3.2 Perform a Terminal Loopback on a Source-Node STM-N Port

Create the Terminal Loopback on a Source Node STM-N Port

Test the Terminal Loopback Circuit

Test the STM-N card

1.3.3 Create the XC Loopback on the Source STM-N Port

Test the XC Loopback Circuit

Test the Standby Cross-Connect Card

Retest the Original Cross-Connect Card

1.3.4 Perform a Facility Loopback on an Intermediate-Node STM-N Port

Create the Facility Loopback on an Intermediate-Node STM-N Port

Test the Facility Loopback Circuit

Test the STM-N Card

1.3.5 Perform a Terminal Loopback on an Intermediate-Node STM-N Port

Create the Terminal Loopback on an Intermediate-Node STM-N Port

Test the Terminal Loopback Circuit

Test the STM-N card

1.3.6 Perform a Facility Loopback on a Destination-Node STM-N Port

Create the Facility Loopback on a Destination-Node STM-N Port

Test the Facility Loopback Circuit

Test the STM-N Card

1.3.7 Perform a Terminal Loopback on a Destination-Node STM-N Port

Create the Terminal Loopback on a Destination-Node STM-N Port

Test the Terminal Loopback Circuit

Test the STM-N Card

1.4 Restoring the Database and Default Settings

1.4.1 Restore the Node Database

Restore the Database

1.4.2 Restore the Node to Factory Configuration

Use the Reinitialization Tool to Clear the Database and Upload Software (Windows)

Use the Reinitialization Tool to Clear the Database and Upload Software (UNIX)

1.5 PC Connectivity Troubleshooting

1.5.1 Unable to Verify the IP Configuration of Your PC

Verify the IP Configuration of your PC

1.5.2 Browser Login Does Not Launch Java

Reconfigure the PC Operating System Java Plug-in Control Panel

Reconfigure the Browser

1.5.3 Unable to Verify the NIC Connection on Your PC

1.5.4 Verify PC Connection to the ONS 15454 SDH (ping)

Ping the ONS 15454 SDH

1.5.5 The IP Address of the Node is Unknown

Retrieve Unknown Node IP Address

1.6 CTC Operation Troubleshooting

1.6.1 Unable to Launch CTC Help After Removing Netscape

Reset Internet Explorer as the Default Browser for CTC

1.6.2 Unable to Change Node View to Network View

Reset the CTC_HEAP Environment Variable for Windows

Reset the CTC_HEAP Environment Variable for Solaris

1.6.3 Browser Stalls When Downloading CTC JAR Files From TCC2

Disable the VirusScan Download Scan

1.6.4 CTC Does Not Launch

Redirect the Netscape Cache to a Valid Directory

1.6.5 Sluggish CTC Operation or Login Problems

Delete the CTC Cache File Automatically

Delete the CTC Cache File Manually

1.6.6 Node Icon is Grey on CTC Network View

1.6.7 CTC Cannot Launch Due to Applet Security Restrictions

Manually Edit the java.policy File

1.6.8 Java Runtime Environment Incompatible

Launch CTC to Correct the Core Version Build

1.6.9 Different CTC Releases Do Not Recognize Each Other

Launch CTC to Correct the Core Version Build

1.6.10 Username or Password Do Not Match

Verify Correct Username and Password

1.6.11 No IP Connectivity Exists Between Nodes

1.6.12 DCC Connection Lost

1.6.13 "Path in Use" Error When Creating a Circuit

1.6.14 Calculate and Design IP Subnets

1.6.15 Ethernet Connections

Verify Ethernet Connections

1.6.16 VLAN Cannot Connect to Network Device from Untag Port

Change VLAN Port Tag and Untagged Settings

1.6.17 Cross-Connect Card Oscillator Fails

Resolve the XC Oscillator Failure When Slot 8 XC Card is Active

Resolve the XC Oscillator Failure When Slot 10 XC Card is Active

1.7 Circuits and Timing

1.7.1 Circuit Transitions to Partial State

View the State of Circuit Nodes

1.7.2 DS3 Card Does Not Report MS-AIS From External Equipment

1.7.3 STM-1 and DCC Limitations

1.7.4 ONS 15454 SDH Switches Timing Reference

1.7.5 Holdover Synchronization Alarm

1.7.6 Free-Running Synchronization Mode

1.7.7 Daisy-Chained BITS Not Functioning

1.7.8 Blinking STAT LED after Installing a Card

1.8 Fiber and Cabling

1.8.1 Bit Errors Appear for a Traffic Card

1.8.2 Faulty Fiber-Optic Connections

Verify Fiber-Optic Connections

Remove GBIC or SFP Connectors

Installing a GBIC with Clips

Installing a GBIC with a Handle

1.8.3 Optical Card Transmit and Receive Levels

1.9 Power and LED Tests

1.9.1 Power Supply Problems

Isolate the Cause of Power Supply Problems

1.9.2 Power Consumption for Node and Cards

1.9.3 Lamp Test for Card LEDs

Verify Card LED Operation


General Troubleshooting


This chapter provides procedures for troubleshooting the most common problems encountered when operating a Cisco ONS 15454 SDH. To troubleshoot specific ONS 15454 SDH alarms, see "Alarm Troubleshooting." If you cannot find what you are looking for, contact the Cisco Technical Assistance Center (Cisco TAC).

This chapter includes the following sections on network problems:

Network Troubleshooting Tests—Describes loopbacks and hairpin circuits, which you can use to test circuit paths through the network or logically isolate faults.


Note For network acceptance tests, refer to the Cisco ONS 15454 SDH Procedure Guide.


Identify Points of Failure on a E3-12 Circuit Path—Explains how to perform the tests described in the "Network Troubleshooting Tests" section on a E3-12 circuit.

Identify Points of Failure on an STM-N Circuit Path—Explains how to perform the tests described in the "Network Troubleshooting Tests" section on an STM-N circuit.

The remaining sections describe symptoms, problems, and solutions that are categorized according to the following topics:

Restoring the Database and Default Settings—Provides procedures for restoring software data and restoring the node to the default setup.

PC Connectivity Troubleshooting—Provides troubleshooting procedures for PC and network connectivity to the ONS 15454 SDH.

CTC Operation Troubleshooting—Provides troubleshooting procedures for CTC login or operation problems.

Circuits and Timing—Provides troubleshooting procedures for circuit creation and error reporting as well as timing reference errors and alarms.

Fiber and Cabling—Provides troubleshooting procedures for fiber and cabling connectivity errors.

Power and LED Tests—Provides troubleshooting procedures for power supply and LED indicator problems.

1.1 Network Troubleshooting Tests

Use loopbacks and hairpins to test newly created circuits before running live traffic or to logically locate the source of a network failure. All ONS 15454 SDH line (traffic) cards, except E-Series and ML-Series Ethernet cards, allow loopbacks and hairpins.


Caution On STM-N cards, a facility loopback applies to the entire card and not an individual circuit. Exercise caution when using loopbacks on an STM-N card carrying live traffic.

A facility loopback tests the line interface unit (LIU) of a card, the FMEC card, and related cabling. After applying a facility loopback on a port, use a test set to run traffic over the loopback. A successful facility loopback isolates the LIU, the FMEC card, or the cabling plant as the potential cause of a network problem. Figure 1-1 shows a facility loopback on a E3-12 card.

Figure 1-1 Facility Loopback Process on a E3-12 Card

To test the LIU on an STM-N card, connect an optical test set to the STM-N port and perform a facility loopback or use a loopback or hairpin on a card that is farther along the circuit path. Figure 1-2 shows a facility loopback on an STM-N card.


Caution Before performing a facility loopback on an STM-N card, make sure the card contains at least two data communications channel (DCC) paths to the node where the card is installed. A second DCC provides a nonlooped path to log into the node after the loopback is applied, thus enabling you to remove the facility loopback. Ensuring a second DCC is not necessary if you are directly connected to the ONS 15454 SDH containing the loopback STM-N card.

Figure 1-2 Facility Loopback Process on an STM-N Card

A terminal loopback tests a circuit path as it passes through the cross-connect card (XC10G or XCVXL) and loops back from the card with the loopback. Figure 1-3 shows a terminal loopback on an STM-N card. The test-set traffic comes in on the E3-12 card and goes through the cross-connect card to the STM-N card. The terminal loopback on the STM-N card turns the signal around before it reaches the LIU and sends it back through the cross-connect card to the E3-12 card. This test verifies that the cross-connect card and terminal circuit paths are valid, but does not test the LIU on the STM-N card.

Setting a terminal loopback on the G-Series card may not stop the Tx Packets counter or the Rx Packet counters on the CTC card-level view Performance > Statistics page from increasing. The counters can increment even though the loopbacked port has temporarily disabled the transmit laser and is dropping any received packets.

The Tx Packet statistic continues to increment because the statistic is not based on the packets transmitted by the Tx laser but on the Tx signal inside the G-Series card. In normal in-service port operation, the Tx signal being recorded does result in the Tx laser transmitting packets, but in a terminal loopback this signal is being looped back within the G-Series card and does not result in the Tx laser transmitting packets.

The Rx Packet counter may also continue to increment when the G-Series card is in terminal loopback. Rx packets from any connected device are dropped and not recorded, but the internally looped back packets follow the G-Series card's normal receive path and register on the Rx Packet counter.

Figure 1-3 Terminal Loopback Process on an STM-N Card

Figure 1-4 shows a terminal loopback on a E3-12 card. The test-set traffic comes in on the STM-N card and goes through the cross-connect card to the E3-12 card. The terminal loopback on the E3-12 card turns the signal around before it reaches the LIU and sends it back through the cross-connect card to the STM-N card. This test verifies that the cross-connect card and terminal circuit paths are valid, but does not test the LIU on the E3-12 card.

Figure 1-4 Terminal Loopback Process on a E3-12 Card

A hairpin circuit brings traffic in and out on a E3-12 port rather than sending the traffic onto the STM-N card. A hairpin loops back only the specific VC3 or VC4 circuit and does not cause an entire STM-N port to loop back, thus preventing a drop of all traffic on the STM-N port. The hairpin allows you to test a specific STS or VT circuit on nodes running live traffic. Figure 1-5 shows the hairpin circuit process on a E3-12 card.

Figure 1-5 Hairpin Circuit Process on a E3-12 Card

A cross-connect loopback tests a circuit path as it passes through the cross-connect card and loops back to the port being tested. Testing and verifying circuit integrity often involves taking down the whole line; however, a cross-connect loopback allows you to create a loopback on any embedded channel at supported payloads at the VC3 granularity and higher. For example, you can loop back a single STM-1, STM-4, STM-16, etc. on an optical facility without interrupting the other STM circuits.

You can create a cross-connect loopback on all working or protect optical ports unless the protect port is used in a 1+1 protection group and is in working mode. If a terminal or facility loopback exists on a port, you cannot use the cross-connect loopback.

Figure 1-6 shows a cross-connect loopback on an STM-N port.

Figure 1-6 Cross-Connect Loopback Process on an STM-N Card

1.2 Identify Points of Failure on a E3-12 Circuit Path

Facility loopbacks, terminal loopbacks, and hairpin circuits are often used to test a circuit path through the network or to logically isolate a fault. Performing a loopback test at each point along the circuit path systematically isolates possible points of failure.

The example in this section tests a E3-12 circuit on a two-node, multiplex section-shared protection ring (MS-SPRing). Using a series of facility loopbacks, terminal loopbacks, and hairpins, the path of the circuit is traced and the possible points of failure are tested and eliminated. A logical progression of five network test procedures apply to this example scenario:


Note The test sequence for your circuits will differ according to the type of circuit and network topology.


1. A facility loopback on the source-node E3-12

2. A hairpin on the source-node E3-12

3. A terminal loopback on the destination-node E3-12

4. A hairpin on the destination-node E3-12

5. A facility loopback on the destination E3-12


Note All loopback tests require on-site personnel.


1.2.1 Perform a Facility Loopback on a Source E3-12 Port

The facility loopback test is performed on the node source port in the network circuit, in this example, the E3-12 port in the source node. Completing a successful facility loopback on this port isolates the cabling, the E3-12 card, and the FMEC card as possible failure points. Figure 1-7 shows an example of a facility loopback on a source E3-12 port.

Figure 1-7 Facility Loopback on a Circuit Source E3-12 Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Facility Loopback on the Source E3-12 Port


Step 1 Connect an electrical test set to the port you are testing.

Use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the electrical test set to the FMEC card connectors or DSx panel for the port you are testing. The Tx and Rx terminals connect to the same port. Adjust the test set accordingly.

Step 2 Use CTC to create the facility loopback on the port being tested:

a. In node view, double-click the card where you will perform the loopback.

b. Click the Maintenance > Loopback tabs.

c. Choose OOS_MT from the State column for the port being tested. If this is a multiport card, select the appropriate row for the port being tested.

d. Choose Facility (Line) from the Loopback Type column for the port being tested. If this is a multiport card, select the appropriate row for the port being tested.

e. Click the Apply button.

f. Click the Yes button in the Confirmation Dialog box.


Note It is normal for a LPBKFACILITY condition to appear during loopback setup. The condition clears when you remove the loopback.


Step 3 Complete the "Test the Facility Loopback Circuit" procedure.


Procedure: Test the Facility Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the facility loopback.

a. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

b. Go to the "Perform a Hairpin on a Source Node Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty E3-12 card, faulty cabling from the E3-12 card to the DSx panel or the FMEC card, or a faulty FMEC card.

Step 5 Complete the "Test the E3-12 Cabling" procedure.


Procedure: Test the E3-12 Cabling


Step 1 Replace the suspect cabling (the cables from the test set to the DSx panel or the FMEC card ports) with a cable that is known to be good.

If a cable that is known to be good is not available, test the suspect cable with a test set. Remove the suspect cable from the DSx panel or the FMEC card and connect the cable to the transmit (Tx) and receive (Rx) terminals of the test set. Run traffic to determine whether the cable is good or defective.

Step 2 Resend test traffic on the loopback circuit with a cable that is known to be good installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective cable.

a. Replace the defective cable.

b. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

c. Go to the "Perform a Hairpin on a Source Node Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty card or a faulty FMEC card.

Step 5 Complete the "Test the E3-12 Card" procedure.


Procedure: Test the E3-12 Card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the faulty card. See the "Physically Replace a Card" procedure for details.

c. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Go to the "Perform a Hairpin on a Source Node Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty FMEC card.

Step 5 Complete the "Test the FMEC" procedure.


Procedure: Test the FMEC


Step 1 Remove and reinstall the FMEC card to ensure a proper seating:

a. Unscrew the screws on the FMEC cover and pull the cover forward.

b. Loosen the faceplate screws that hold the FMEC card in place.

c. Pull the FMEC card outward by the faceplate to unseat it from the shelf assembly.

d. Push the FMEC card back inward by the faceplate to reseat it in the shelf assembly.

Step 2 Resend test traffic on the loopback circuit with known-good cabling, a known-good card, and the reinstalled FMEC.

Step 3 If the test set indicates a good circuit, the problem is probably an improperly seated FMEC.

a. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

b. Proceed to Step 8.

Step 4 If the test set indicates a faulty circuit, the problem is probably a defective FMEC card:

a. Return the defective FMEC card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the faulty FMEC card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 5 Resend test traffic on the loopback circuit with known-good cabling, a known-good card, and the replacement FMEC card.

Step 6 If the test set indicates a faulty circuit, repeat all of the facility loopback procedures.

Step 7 If the test set indicates a good circuit, the problem is probably the defective FMEC card.

Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

Step 8 Go to the "Perform a Hairpin on a Source Node Port" section.


1.2.2 Perform a Hairpin on a Source Node Port

The hairpin test is performed on the cross-connect card in the network circuit. A hairpin circuit uses the same port for both source and destination. Completing a successful hairpin through the card isolates the possibility that the cross-connect card is the cause of the faulty circuit. Figure 1-8 shows an example of a hairpin loopback on a source node port.

Figure 1-8 Hairpin on a Source Node Port


Note The ONS 15454 SDH does not support simplex operation on the cross-connect card. Two cross-connect cards of the same type must be installed for each node.


Procedure: Create the Hairpin on the Source Node Port


Step 1 Connect an electrical test set to the port you are testing:

a. If you just completed the "Perform a Facility Loopback on a Source E3-12 Port" section, leave the electrical test set hooked up to the E3-12 port in the source node.

b. If you are starting the current procedure without the electrical test set hooked up to the E3-12 port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the electrical test set to the DSx panel or the FMEC card connectors for the port you are testing. The Tx and Rx terminals connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the hairpin on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as Hairpin1.

c. Set the Circuit Type and Size to the normal preferences.

d. Uncheck the Bidirectional check box and click the Next button.

e. In the Circuit Source dialog box, select the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, use the same Node, card Slot, Port, and Type used for the Circuit Source dialog box and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a one-way circuit.

Step 4 Go to the "Test the Hairpin Circuit" procedure.


Procedure: Test the Hairpin Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the hairpin circuit.

a. Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Confirm that the hairpin circuit is deleted form the Circuits tab list.

b. Go to the "Perform a Terminal Loopback on a Destination E3-12 Port" section.

Step 4 If the test set indicates a faulty circuit, there might be a problem with the cross-connect card.

Step 5 Complete the "Test the Standby Cross-Connect Card" procedure.


Procedure: Test the Standby Cross-Connect Card


Step 1 Perform a reset on the standby cross-connect card:

a. Determine the standby cross-connect card. On both the physical node and the Cisco Transport Controller (CTC) window, the ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. Position the cursor over the standby cross-connect card.

c. Right-click and choose RESET CARD.

Step 2 Initiate an external switch command (side switch) on the cross-connect cards before retesting the loopback circuit:


Caution Cross-connect side switches are service-affecting. Any live traffic on any card in the node endures a hit of up to 50 ms.

a. Determine the standby cross-connect card. The ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. In the node view, select the Maintenance > Cross Connect tabs.

c. In the Cross Connect Cards menu, click the Switch button.

d. Click the Yes button in the Confirm Switch dialog box.


Note After the active cross-connect goes into standby, the original standby slot becomes active. This causes the ACT/STBY LED to become green on the former standby card.


Step 3 Resend test traffic on the loopback circuit.

The test traffic now travels through the alternate cross-connect card.

Step 4 If the test set indicates a faulty circuit, assume the cross-connect card is not causing the problem.

a. Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Confirm that the hairpin circuit is deleted form the Circuits tab list.

b. Go to the "Perform a Terminal Loopback on a Destination E3-12 Port" section.

Step 5 If the test set indicates a good circuit, the problem might be a defective cross-connect card.

Step 6 To confirm a defective original cross-connect card, complete the "Retest the Original Cross-Connect Card" procedure.


Procedure: Retest the Original Cross-Connect Card


Step 1 Initiate an external switch command (side switch) on the cross-connect cards to make the original cross-connect card the active card:

a. Determine the standby cross-connect card. The ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. In node view, select the Maintenance > Cross Connect tabs.

c. From the Cross Connect Cards menu, choose Switch.

d. Click the Yes button in the Confirm Switch dialog box.

Step 2 Resend test traffic on the loopback circuit.

Step 3 If the test set indicates a faulty circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the defective cross-connect card. See "Replace Hardware" for details.

c. Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Confirm that the hairpin circuit is deleted form the Circuits tab list.

d. Go to Step 5.

Step 4 If the test set indicates a good circuit, the cross-connect card might have had a temporary problem that is cleared by the side switch.

Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Confirm that the hairpin circuit is deleted form the Circuits tab list.

Step 5 Go to the "Perform a Terminal Loopback on a Destination E3-12 Port" section.


1.2.3 Perform a Terminal Loopback on a Destination E3-12 Port

The terminal loopback test is performed on the node destination port in the circuit, in this example, the E3-12 port in the destination node. First, create a bidirectional circuit that starts on the source node E3-12 port and loops back on the destination node E3-12 port. Then, proceed with the terminal loopback test. Completing a successful terminal loopback to a destination node E3-12 port verifies that the circuit is good up to the destination E3-12. Figure 1-9 shows an example of a terminal loopback on a destination E3-12 port.

Figure 1-9 Terminal Loopback on a Destination E3-12 Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Terminal Loopback on a Destination E3-12 Port


Step 1 Connect an electrical test set to the port you are testing:

a. If you just completed the "Perform a Hairpin on a Source Node Port" section, leave the electrical test set hooked up to the E3-12 port in the source node.

b. If you are starting the current procedure without the electrical test set hooked up to the E3-12 port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the electrical test set to the DSx panel or the FMEC card connectors for the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the terminal loopback circuit on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as "DSNtoDSN".

c. Set Circuit Type and Size to the normal preferences.

d. Leave the Bidirectional check box checked and click the Next button.

e. In the Circuit Source dialog box, fill in the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, fill in the destination Node, card Slot, Port, and Type (the E3-12 port in the destination node) and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a 2-way circuit.


Note It is normal for a LPBKTERMINAL condition to appear during a loopback setup. The condition clears when you remove the loopback.


Step 4 Create the terminal loopback on the destination port being tested:

a. Go to the node view of the destination node:

Choose View > Go To Other Node from the menu bar.

Choose the node from the drop-down list in the Select Node dialog box and click the OK button.

b. In node view, double-click the card that requires the loopback, such as the E3-12 card in the destination node.

c. Click the Maintenance > Loopback tabs.

d. Select OOS_MT from the State column. If this is a multiport card, select the row appropriate for the desired port.

e. Select Terminal (Inward) from the Loopback Type column. If this is a multiport card, select the row appropriate for the desired port.

f. Click the Apply button.

g. Click the Yes button in the Confirmation Dialog box.

Step 5 Go to the "Test the Terminal Loopback Circuit on the Destination E3-12 Port" procedure.


Procedure: Test the Terminal Loopback Circuit on the Destination E3-12 Port


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic being received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary on the loopback circuit.

a. Clear the terminal loopback:

Double-click the E3-12 card in the destination node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

Clear the terminal loopback circuit before testing the next segment of the network circuit path:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

b. Go to the "Perform a Hairpin on a Destination Node" procedure.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty card.

Step 5 Go to the "Test the Destination E3-12 Card" procedure.


Procedure: Test the Destination E3-12 Card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC.

b. Replace the defective E3-12 card. See the "Physically Replace a Card" procedure for details.

c. Clear the terminal loopback:

Double-click the E3-12 card in the destination node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Clear the terminal loopback:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 4 Go to the "Perform a Hairpin on a Destination Node" section.


1.2.4 Perform a Hairpin on a Destination Node

The hairpin test is performed on the cross-connect card in the network circuit. A hairpin circuit uses the same port for both source and destination. Completing a successful hairpin through the card isolates the possibility that the cross-connect card is the cause of the faulty circuit. Figure 1-10 shows an example of a hairpin loopback on a destination node.

Figure 1-10 Hairpin on a Destination Node


Note The ONS 15454 SDH does not support simplex operation on the cross-connect card. Two cross-connect cards of the same type must be installed for each node.


Procedure: Create the Hairpin on the Destination Node


Step 1 Connect an electrical test set to the port you are testing.

Use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the electrical test set to the FMEC card connectors or DSx panel for the port you are testing. The Tx and Rx terminals connect to the same port. Adjust the test set accordingly.

Step 2 Use CTC to set up the hairpin on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as Hairpin1.

c. Set the Circuit Type and Size to the normal preferences.

d. Uncheck the Bidirectional check box and click the Next button.

e. In the Circuit Source dialog box, select the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, use the same Node, card Slot, Port, and Type used for the Circuit Source dialog box and click the Finish button.

Step 3 Confirm that the newly created circuit appears in the Circuits tab list as a one-way circuit.

Step 4 Complete the "Test the Hairpin Circuit" procedure.


Procedure: Test the Hairpin Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the hairpin circuit.

a. Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits box.

Confirm that the hairpin circuit is deleted form the Circuits tab list.

b. Complete the "Perform a Facility Loopback on a Destination E3-12 Port" procedure.

Step 4 If the test set indicates a faulty circuit, there might be a problem with the cross-connect card.

Step 5 Complete the "Test the Standby Cross-Connect Card" procedure.


Procedure: Test the Standby Cross-Connect Card


Step 1 Perform a reset on the standby cross-connect card:

a. Determine the standby cross-connect card. On both the physical node and the CTC window, the ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. Position the cursor over the standby cross-connect card.

c. Right-click and choose RESET CARD.

Step 2 Perform an external switch command (side switch) on the cross-connect cards before retesting the loopback circuit:


Caution Cross-connect side switches are service-affecting. Any live traffic on any card in the node endures a hit of up to 50 ms.

a. Determine the standby cross-connect card. The ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. In the node view, select the Maintenance > Cross Connect tabs.

c. In the Cross Connect Cards menu, click the Switch button.

d. Click the Yes button in the Confirm Switch dialog box.


Note After the active cross-connect goes into standby, the original standby slot becomes active. This causes the ACT/STBY LED to become green on the former standby card.


Step 3 Resend test traffic on the loopback circuit.

The test traffic now travels through the alternate cross-connect card.

Step 4 If the test set indicates a faulty circuit, assume the cross-connect card is not causing the problem.

a. Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Confirm that the hairpin circuit is deleted form the Circuits tab list.

b. Go to the "Perform a Facility Loopback on a Destination E3-12 Port" section.

Step 5 If the test set indicates a good circuit, the problem might be a defective cross-connect card.

Step 6 To confirm a defective original cross-connect card, complete the "Retest the Original Cross-Connect Card" procedure.


Procedure: Retest the Original Cross-Connect Card


Step 1 Perform an external switching command (side switch) on the cross-connect cards to make the original cross-connect card the active card:

a. Determine the standby cross-connect card. The ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. In node view, select the Maintenance > Cross Connect tabs.

c. In the Cross Connect Cards menu, click the Switch button.

d. Click the Yes button in the Confirm Switch dialog box.

Step 2 Resend test traffic on the loopback circuit.

Step 3 If the test set indicates a faulty circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the defective cross-connect card. See "Replace Hardware" for details.

c. Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

d. Go to Step 5.

Step 4 If the test set indicates a good circuit, the cross-connect card might have had a temporary problem that is cleared by the side switch.

Clear the hairpin circuit:

Click the Circuits tab.

Choose the hairpin circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 5 Go to the "Perform a Facility Loopback on a Destination E3-12 Port" section.


1.2.5 Perform a Facility Loopback on a Destination E3-12 Port

The facility loopback test is performed on the node source port in the circuit, in this example, the destination E3-12 port in the destination node. Completing a successful facility loopback on this port isolates the possibility that the destination node cabling, E3-12 card, LIU, or FMEC card is responsible for a faulty circuit. Figure 1-11 shows an example of a facility loopback on a destination E3-12 port.

Figure 1-11 Facility Loopback on a Destination E3-12 Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create a Facility Loopback Circuit on a Destination E3-12 Port


Step 1 Connect an electrical test set to the port you are testing:

a. If you just completed the "Create the Hairpin on the Destination Node" procedure, leave the electrical test set hooked up to the E3-12 port in the destination node.

b. If you are starting the current procedure without the electrical test set hooked up to the E3-12 port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the electrical test set to the DSx panel or the FMEC card connectors for the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to create the facility loopback on the port being tested:

a. In node view, double-click the card where the loopback will be performed.

b. Click the Maintenance > Loopback tabs.

c. Select Facility (Line) from the Loopback Type column for the port being tested. If this is a multiport card, select the row appropriate for the desired port.

d. Click the Apply button.

e. Click the Yes button in the Confirmation Dialog box.


Note It is normal for a LPBKFACILITY condition to appear during loopback setup. The condition clears when you remove the loopback.


Step 3 Complete the "Test the Facility Loopback Circuit" procedure.


Procedure: Test the Facility Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the loopback circuit.

Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

The entire E3-12 circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty E3-12 card, faulty cabling from the E3-12 card to the DSx panel or the FMEC card, or a faulty FMEC card.

Step 5 Complete the "Test the E3-12 Cabling" procedure.


Procedure: Test the E3-12 Cabling


Step 1 Replace the suspect cabling (the cables from the test set to the DSx panel or the FMEC card ports) with a cable that is known to be good.

If a cable that is known to be good is not available, test the suspect cable with a test set. Remove the suspect cable from the DSx panel or the FMEC card and connect the cable to the transmit (Tx) and receive (Rx) terminals of the test set. Run traffic to determine whether the cable is good or defective.

Step 2 Resend test traffic on the loopback circuit with a cable that is known to be good installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective cable.

a. Replace the defective cable.

b. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

The entire E3-12 circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty card or a faulty FMEC card.

Step 5 Complete the "Test the E3-12 Card" procedure.


Procedure: Test the E3-12 Card


Step 1 Replace the suspect card with a known-good card.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC.

b. Replace the faulty card. See the "Physically Replace a Card" procedure for details.

c. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

The entire E3-12 circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty FMEC card.

Step 5 Complete the "Test the FMEC" procedure.


Procedure: Test the FMEC


Step 1 Remove and reinstall the FMEC card to ensure a proper seating:

a. Unscrew the screws on the FMEC cover and pull the cover forward.

b. Loosen the faceplate screws that hold the FMEC card in place.

c. Pull the FMEC card outward by the faceplate to unseat it from the shelf assembly.

d. Push the FMEC card back inward by the faceplate to reseat it in the shelf assembly.

Step 2 Resend test traffic on the loopback circuit with known-good cabling, a known-good card, and the reinstalled FMEC card.

Step 3 If the test set indicates a good circuit, the problem is probably an improperly seated FMEC card.

Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

The entire E3-12 circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.

Step 4 If the test set indicates a faulty circuit, the problem is probably the defective FMEC card:

a. Return the defective FMEC card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the faulty FMEC card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 5 Resend test traffic on the loopback circuit with known-good cabling, a known-good card, and the replacement FMEC card.

Step 6 If the test set indicates a faulty circuit, repeat all of the facility loopback procedures.

If the faulty circuit persists, contact the Cisco Technical Assistance Center (Cisco TAC).

Step 7 If the test set indicates a good circuit, the problem is probably the defective FMEC card.

Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

The entire E3-12 circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.


1.3 Identify Points of Failure on an STM-N Circuit Path

Facility loopbacks, terminal loopbacks, and cross-connect loopback circuits are often used together to test the circuit path through the network or to logically isolate a fault. Performing a loopback test at each point along the circuit path systematically isolates possible points of failure.

The example in this section tests an STM-N circuit on a three-node, multiplex section-shared protection ring (MS-SPRing). Using a series of facility loopbacks and terminal loopbacks, the path of the circuit is traced and the possible points of failure are tested and eliminated. A logical progression of six network test procedures apply to this example scenario:


Note The test sequence for your circuits will differ according to the type of circuit and network topology.


1. A facility loopback on the source-node STM-N port

2. A terminal loopback on the source-node STM-N port

3. A cross-connect loopback on the source STM-N port

4. A facility loopback on the intermediate-node STM-N port

5. A terminal loopback on the intermediate-node STM-N port

6. A facility loopback on the destination-node STM-N port

7. A terminal loopback on the destination-node STM-N port


Note All loopback tests require on-site personnel.


1.3.1 Perform a Facility Loopback on a Source-Node STM-N Port

The facility loopback test is performed on the node source port in the network circuit, in this example, the source STM-N port in the source node. Completing a successful facility loopback on this port isolates the STM-N port as a possible failure point. Figure 1-7 shows an example of a facility loopback on a circuit source STM-N port.

Figure 1-12 A Facility Loopback on a Circuit Source STM-N Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Facility Loopback on the Source STM-N Port


Step 1 Connect an optical test set to the port you are testing.

Use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. The Tx and Rx terminals connect to the same port. Adjust the test set accordingly.

Step 2 Use CTC to create the facility loopback circuit on the port being tested:

a. In node view, double-click the card where you will perform the loopback.

b. Click the Maintenance > Loopback tabs.

c. Choose OOS_MT from the State column for the port being tested. If this is a multiport card, select the appropriate row for the desired port.

d. Choose Facility (Line) from the Loopback Type column for the port being tested. If this is a multiport card, select the appropriate row for the desired port.

e. Click the Apply button.

f. Click the Yes button in the Confirmation Dialog box.


Note It is normal for a LPBKFACILITY condition to appear during loopback setup. The condition clears when you remove the loopback.


Step 3 Complete the "Test the Facility Loopback Circuit" procedure.


Procedure: Test the Facility Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the facility loopback.

a. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

b. Go to the "Perform a Terminal Loopback on a Source-Node STM-N Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty STM-N card.

Step 5 Complete the "Test the STM-N Card" procedure.


Procedure: Test the STM-N Card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the faulty card. See the "Physically Replace a Card" procedure for details.

c. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

Step 4 Go to the "Perform a Terminal Loopback on a Source-Node STM-N Port" section.


1.3.2 Perform a Terminal Loopback on a Source-Node STM-N Port

The terminal loopback test is performed on the node destination port in the circuit, in this example, the destination STM-N port in the source node. First, create a bidirectional circuit that starts on the node source STM-N port and loops back on the node destination STM-N port. Then, proceed with the terminal loopback test. Completing a successful terminal loopback to a node destination STM-N port verifies that the circuit is good up to the destination STM-N. Figure 1-13 shows an example of a terminal loopback on a destination STM-N port.

Figure 1-13 Terminal Loopback on a Source-Node STM-N Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Terminal Loopback on a Source Node STM-N Port


Step 1 Connect an optical test set to the port you are testing:

a. If you just completed the "Perform a Facility Loopback on a Source-Node STM-N Port" section, leave the optical test set hooked up to the STM-N port in the source node.

b. If you are starting the current procedure without the optical test set hooked up to the STM-N port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the terminal loopback circuit on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as "OCN1toOCN2".

c. Set Circuit Type and Size to the normal preferences.

d. Leave the Bidirectional check box checked and click the Next button.

e. In the Circuit Source dialog box, fill in the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, fill in the destination Node, card Slot, Port, and Type (the STM-N port in the source node) and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a 2-way circuit.


Note It is normal for a LPBKTERMINAL condition to appear during a loopback setup. The condition clears when you remove the loopback.


Step 4 Create the terminal loopback on the destination port being tested:

a. In node view, double-click the card that requires the loopback, such as the destination STM-N card in the source node.

b. Click the Maintenance > Loopback tabs.

c. Select OOS_MT from the State column. If this is a multiport card, select the row appropriate for the desired port.

d. Select Terminal (Inward) from the Loopback Type column. If this is a multiport card, select the row appropriate for the desired port.

e. Click the Apply button.

f. Click the Yes button in the Confirmation Dialog box.

Step 5 Complete the "Test the Terminal Loopback Circuit" procedure.


Procedure: Test the Terminal Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic being received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary on the loopback circuit.

a. Clear the terminal loopback:

Double-click the STM-N card in the source node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

b. Clear the terminal loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

c. Go to the "Perform a Facility Loopback on an Intermediate-Node STM-N Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty card.

Step 5 Complete the "Test the STM-N card" procedure.


Procedure: Test the STM-N card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the defective STM-N card. See the "Physically Replace a Card" procedure for details.

c. Clear the terminal loopback:

Double-click the STM-N card in the source node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Clear the terminal loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 4 Go to the "Create the XC Loopback on the Source STM-N Port" section.


1.3.3 Create the XC Loopback on the Source STM-N Port

The XC loopback test is performed on the cross-connect card in a network circuit. An XC loopback circuit uses the same port for both source and destination. Completing a successful XC loopback through the cross-connect card isolates the possibility that the cross-connect card is the cause of the faulty circuit. Figure 1-14 shows an example of an XC loopback on a source STM-N port.

Figure 1-14 XC Loopback on a Source STM-N Port


Step 1 Connect an optical test set to the port you are testing:


Note Refer to the manufacturer's instructions for detailed information on connection and setup of the optical test set.


a. If you just completed the "Perform a Terminal Loopback on a Source-Node STM-N Port" section, leave the optical test set hooked up to the STM-N port in the source node.

b. If you are starting the current procedure without the optical test set hooked up to the STM-N port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. The Tx and Rx terminals connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to put the circuit being tested out of service:

a. In node view, double-click the card where the test set is connected. The card view appears.

b. In card view, click the Provisioning > Line tabs.

c. Choose OOS (out of service) or OOS_MT (out of service maintenance) from the Status column for the port being tested.

d. Click Apply.

e. Click Yes in the confirmation dialog.

Step 3 Use CTC to set up the XC loopback on the circuit being tested:

a. In card view, click the Provisioning > SONET STS tabs.

b. Click the check box in the XC Loopback column for the port being tested.

c. Click Apply.

d. Click Yes in the confirmation dialog.

Step 4 Go to "Test the XC Loopback Circuit" procedure.


Procedure: Test the XC Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the cross-connect.

a. Clear the XC loopback:

In card view, click the Provisioning > SONET STS tabs.

Uncheck the check box in the XC Loopback column for the circuit being tested.

Click Apply.

Click Yes in the confirmation dialog.

b. Go to the "Perform a Facility Loopback on an Intermediate-Node STM-N Port" section.

Step 4 If the test set indicates a faulty circuit, there might be a problem with the cross-connect card.

Step 5 Go to "Test the Standby Cross-Connect Card" procedure.


Procedure: Test the Standby Cross-Connect Card


Step 1 Perform a reset on the standby cross-connect card:

a. Determine the standby cross-connect card. On both the physical node and the CTC window, the ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. Position the cursor over the standby cross-connect card.

c. Right-click and choose RESET CARD.

Step 2 Perform an external switching command (side switch) on the cross-connect cards before retesting the loopback circuit:


Caution Cross-connect side switches are service-affecting. Any live traffic on any card in the node endures a hit of up to 50 ms.

a. Determine the standby cross-connect card. The ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. In the node view, select the Maintenance > Cross Connect tabs.

c. In the Cross Connect Cards menu, click the Switch button.

d. Click the Yes button in the Confirm Switch dialog box.


Note After the active cross-connect goes into standby, the original standby slot becomes active. This causes the ACT/STBY LED to become green on the former standby card.


Step 3 Resend test traffic on the loopback circuit.

The test traffic now travels through the alternate cross-connect card.

Step 4 If the test set indicates a faulty circuit, assume the cross-connect card is not causing the problem.

Clear the XC loopback circuit:

a. Click the Circuits tab.

b. Choose the XC loopback circuit being tested.

c. Click the Delete button.

d. Click the Yes button in the Delete Circuits dialog box.

e. Confirm that the XC loopback circuit is deleted form the Circuits tab list.

Step 5 If the test set indicates a good circuit, the problem might be a defective cross-connect card.

Step 6 To confirm a defective original cross-connect card, complete the "Retest the Original Cross-Connect Card" procedure.


Procedure: Retest the Original Cross-Connect Card


Step 1 Initiate an external switching command (side switch) on the cross-connect cards to make the original cross-connect card the active card:

a. Determine the standby cross-connect card. The ACT/STBY LED of the standby cross-connect card is amber and the ACT/STBY LED of the active cross-connect card is green.

b. In node view, select the Maintenance > Cross Connect tabs.

c. In the Cross Connect Cards menu, click the Switch button.

d. Click the Yes button in the Confirm Switch dialog box.

Step 2 Resend test traffic on the loopback circuit.

Step 3 If the test set indicates a faulty circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the defective cross-connect card. See "Replace Hardware" for details.

c. Clear the XC loopback circuit:

Click the Circuits tab.

Choose the XC loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 4 If the test set indicates a good circuit, the cross-connect card might have had a temporary problem that is cleared by the side switch.

Clear the XC loopback circuit:

Click the Circuits tab.

Choose the XC loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.


1.3.4 Perform a Facility Loopback on an Intermediate-Node STM-N Port

The facility loopback test is performed on the node source port in the network circuit, in this example, the source STM-N port in the intermediate node. Completing a successful facility loopback on this port isolates the STM-N port as a possible failure point. Figure 1-15 shows an example of a facility loopback on a intermediate node circuit source STM-N port.

Figure 1-15 Facility Loopback on an Intermediate-Node STM-N Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Facility Loopback on an Intermediate-Node STM-N Port


Step 1 Connect an optical test set to the port you are testing:

a. If you just completed the "Perform a Terminal Loopback on a Source-Node STM-N Port" section, leave the optical test set hooked up to the STM-N port in the source node.

b. If you are starting the current procedure without the optical test set hooked up to the STM-N port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the facility loopback circuit on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as "OCN1toOCN3".

c. Set Circuit Type and Size to the normal preferences.

d. Leave the Bidirectional check box checked and click the Next button.

e. In the Circuit Source dialog box, fill in the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, fill in the destination Node, card Slot, Port, and Type (the STM-N port in the intermediate node) and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a 2-way circuit.


Note It is normal for a LPBKFACILITY condition to appear during a loopback setup. The condition clears when you remove the loopback.


Step 4 Create the facility loopback on the destination port being tested:

a. Go to the node view of the intermediate node:

Choose View > Go To Other Node from the menu bar.

Choose the node from the drop-down list in the Select Node dialog box and click the OK button.

b. In node view, double-click the card that requires the loopback, such as the destination STM-N card in the intermediate node.

c. Click the Maintenance > Loopback tabs.

d. Select OOS_MT from the State column. If this is a multiport card, select the row appropriate for the desired port.

e. Select Terminal (Inward) from the Loopback Type column. If this is a multiport card, select the row appropriate for the desired port.

f. Click the Apply button.

g. Click the Yes button in the Confirmation Dialog box.


Note It is normal for a LPBKFACILITY condition to appear during loopback setup. The condition clears when you remove the loopback.


Step 5 Complete the "Test the Facility Loopback Circuit" procedure.


Procedure: Test the Facility Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the facility loopback.

a. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the confirmation dialog box.

b. Clear the facility loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

c. Go to the "Perform a Terminal Loopback on an Intermediate-Node STM-N Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty STM-N card.

Step 5 Complete the "Test the STM-N Card" procedure.


Procedure: Test the STM-N Card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently carries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONS 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the faulty card. See the "Physically Replace a Card" procedure for details.

c. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Clear the facility loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 4 Go to the "Perform a Terminal Loopback on an Intermediate-Node STM-N Port" section.


1.3.5 Perform a Terminal Loopback on an Intermediate-Node STM-N Port

The terminal loopback test is performed on the node destination port in the circuit, in this example, the destination STM-N port in the intermediate node. First, create a bidirectional circuit that starts on the node source STM-N port and loops back on the node destination STM-N port. Then proceed with the terminal loopback test. Completing a successful terminal loopback to a node destination STM-N port verifies that the circuit is good up to the destination STM-N. Figure 1-16 shows an example of a terminal loopback on an intermediate node destination STM-N port.

Figure 1-16 Terminal Loopback on an Intermediate-Node STM-N Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Terminal Loopback on an Intermediate-Node STM-N Port


Step 1 Connect an optical test set to the port you are testing:

a. If you just completed the "Perform a Facility Loopback on an Intermediate-Node STM-N Port" section, leave the optical test set hooked up to the STM-N port in the source node.

b. If you are starting the current procedure without the optical test set hooked up to the STM-N port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the terminal loopback circuit on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as "OCN1toOCN4".

c. Set Circuit Type and Size to the normal preferences.

d. Leave the Bidirectional check box checked and click the Next button.

e. In the Circuit Source dialog box, fill in the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, fill in the destination Node, card Slot, Port, and Type (the STM-N port in the intermediate node) and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a 2-way circuit.


Note It is normal for a LPBKTERMINAL condition to appear during a loopback setup. The condition clears when you remove the loopback.


Step 4 Create the terminal loopback on the destination port being tested:

a. Go to the node view of the intermediate node:

Choose View > Go To Other Node from the menu bar.

Choose the node from the drop-down list in the Select Node dialog box and click the OK button.

b. In node view, double-click the card that requires the loopback, such as the destination STM-N card in the intermediate node.

c. Click the Maintenance > Loopback tabs.

d. Select OOS_MT from the State column. If this is a multiport card, select the row appropriate for the desired port.

e. Select Terminal (Inward) from the Loopback Type column. If this is a multiport card, select the row appropriate for the desired port.

f. Click the Apply button.

g. Click the Yes button in the Confirmation Dialog box.

Step 5 Complete the "Test the Terminal Loopback Circuit" procedure.


Procedure: Test the Terminal Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic being received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary on the loopback circuit.

a. Clear the terminal loopback:

Double-click the STM-N card in the intermediate node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

b. Clear the terminal loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

c. Go to the "Perform a Facility Loopback on a Destination-Node STM-N Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty card.

Step 5 Complete the "Test the STM-N card" procedure.


Procedure: Test the STM-N card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently caries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONs 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the defective STM-N card. See the "Physically Replace a Card" procedure for details.

c. Clear the terminal loopback:

Double-click the STM-N card in the source node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Clear the terminal loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 4 Go to the "Perform a Facility Loopback on a Destination-Node STM-N Port" section.


1.3.6 Perform a Facility Loopback on a Destination-Node STM-N Port

The facility loopback test is performed on the node source port in the network circuit, in this example, the source STM-N port in the destination node. Completing a successful facility loopback on this port isolates the STM-N port as a possible failure point. Figure 1-17 shows an example of a facility loopback on a destination node circuit source STM-N port.

Figure 1-17 Facility Loopback on a Destination-Node STM-N Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Facility Loopback on a Destination-Node STM-N Port


Step 1 Connect an optical test set to the port you are testing:

a. If you just completed the "Perform a Terminal Loopback on an Intermediate-Node STM-N Port" section, leave the optical test set hooked up to the STM-N port in the source node.

b. If you are starting the current procedure without the optical test set hooked up to the STM-N port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the facility loopback circuit on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as "OCN1toOCN5".

c. Set Circuit Type and Size to the normal preferences.

d. Leave the Bidirectional check box checked and click the Next button.

e. In the Circuit Source dialog box, fill in the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, fill in the destination Node, card Slot, Port, and Type (the STM-N port in the destination node) and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a 2-way circuit.


Note It is normal for a LPBKFACILITY condition to appear during a loopback setup. The condition clears when you remove the loopback.


Step 4 Create the facility loopback on the destination port being tested:

a. Go to the node view of the destination node:

Choose View > Go To Other Node from the menu bar.

Choose the node from the drop-down list in the Select Node dialog box and click the OK button.

b. In node view, double-click the card that requires the loopback, such as the destination STM-N card in the destination node.

c. Click the Maintenance > Loopback tabs.

d. Select OOS_MT from the State column. If this is a multiport card, select the row appropriate for the desired port.

e. Select Terminal (Inward) from the Loopback Type column. If this is a multiport card, select the row appropriate for the desired port.

f. Click the Apply button.

g. Click the Yes button in the Confirmation Dialog box.


Note It is normal for a LPBKFACILITY condition to appear during loopback setup. The condition clears when you remove the loopback.


Step 5 Complete the "Test the Facility Loopback Circuit" procedure.


Procedure: Test the Facility Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the traffic received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary with the facility loopback.

a. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the confirmation dialog box.

b. Clear the facility loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

c. Go to the "Perform a Terminal Loopback on a Destination-Node STM-N Port" section.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty STM-N card.

Step 5 Complete the "Test the STM-N Card" procedure.


Procedure: Test the STM-N Card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently caries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONs 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card installed.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the faulty card. See the "Physically Replace a Card" procedure for details.

c. Clear the facility loopback:

Click the Maintenance > Loopback tabs.

Choose None from the Loopback Type column for the port being tested.

Choose the appropriate state (IS, OOS, OOS_AINS) from the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Clear the facility loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

Step 4 Go to the "Perform a Terminal Loopback on a Destination-Node STM-N Port" section.


1.3.7 Perform a Terminal Loopback on a Destination-Node STM-N Port

The terminal loopback test is performed on the node destination port in the circuit, in this example, the destination STM-N port in the destination node. First, create a bidirectional circuit that starts on the node source STM-N port and loops back on the node destination STM-N port. Then proceed with the terminal loopback test. Completing a successful terminal loopback to a node destination STM-N port verifies that the circuit is good up to the destination STM-N. Figure 1-18 shows an example of a terminal loopback on an intermediate node destination STM-N port.

Figure 1-18 Terminal Loopback on a Destination-Node STM-N Port


Caution Performing a loopback on an in-service circuit is service-affecting.

Procedure: Create the Terminal Loopback on a Destination-Node STM-N Port


Step 1 Connect an optical test set to the port you are testing:

a. If you just completed the "Perform a Facility Loopback on a Destination-Node STM-N Port" section, leave the optical test set hooked up to the STM-N port in the source node.

b. If you are starting the current procedure without the optical test set hooked up to the STM-N port, use appropriate cabling to attach the transmit (Tx) and receive (Rx) terminals of the optical test set to the port you are testing. Both Tx and Rx connect to the same port.

c. Adjust the test set accordingly.

Step 2 Use CTC to set up the terminal loopback circuit on the port being tested:

a. Click the Circuits tab and click the Create button.

b. Give the circuit an easily identifiable name, such as "OCN1toOCN6".

c. Set Circuit Type and Size to the normal preferences.

d. Leave the Bidirectional check box checked and click the Next button.

e. In the Circuit Source dialog box, fill in the same Node, card Slot, Port, and Type where the test set is connected and click the Next button.

f. In the Circuit Destination dialog box, fill in the destination Node, card Slot, Port, and Type (the STM-N port in the destination node) and click the Finish button.

Step 3 Confirm that the newly created circuit appears on the Circuits tab list as a 2-way circuit.


Note It is normal for a LPBKTERMINAL condition to appear during a loopback setup. The condition clears when you remove the loopback.


Step 4 Create the terminal loopback on the destination port being tested:

a. Go to the node view of the destination node:

Choose View > Go To Other Node from the menu bar.

Choose the node from the drop-down list in the Select Node dialog box and click the OK button.

b. In node view, double-click the card that requires the loopback, such as the destination STM-N card in the destination node.

c. Click the Maintenance > Loopback tabs.

d. Select OOS_MT from the State column. If this is a multiport card, select the row appropriate for the desired port.

e. Select Terminal (Inward) from the Loopback Type column. If this is a multiport card, select the row appropriate for the desired port.

f. Click the Apply button.

g. Click the Yes button in the Confirmation Dialog box.

Step 5 Complete the "Test the Terminal Loopback Circuit" procedure.


Procedure: Test the Terminal Loopback Circuit


Step 1 If the test set is not already sending traffic, send test traffic on the loopback circuit.

Step 2 Examine the test traffic being received by the test set. Look for errors or any other signal information that the test set is capable of indicating.

Step 3 If the test set indicates a good circuit, no further testing is necessary on the loopback circuit.

a. Clear the terminal loopback:

Double-click the STM-N card in the intermediate node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

b. Clear the terminal loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

The entire STM-N circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.

Step 4 If the test set indicates a faulty circuit, the problem might be a faulty card.

Step 5 Complete the "Test the STM-N Card" procedure.


Procedure: Test the STM-N Card


Step 1 Replace the suspect card with a known-good card. See the "Physically Replace a Card" procedure for details.


Caution Removing a card that currently caries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONs 15454 SDH Procedure Guide for information.

Step 2 Resend test traffic on the loopback circuit with a known-good card.

Step 3 If the test set indicates a good circuit, the problem is probably the defective card.

a. Return the defective card to Cisco through the returned materials authorization (RMA) process. Contact the Cisco Technical Assistance Center (Cisco TAC).

b. Replace the defective STM-N card. See the "Physically Replace a Card" procedure for details.

c. Clear the terminal loopback:

Double-click the STM-N card in the source node with the terminal loopback.

Click the Maintenance > Loopback tabs.

Select None from the Loopback Type column for the port being tested.

Select the appropriate state (IS, OOS, OOS_AINS) in the State column for the port being tested.

Click the Apply button.

Click the Yes button in the Confirmation Dialog box.

d. Clear the terminal loopback circuit:

Click the Circuits tab.

Choose the loopback circuit being tested.

Click the Delete button.

Click the Yes button in the Delete Circuits dialog box.

The entire STM-N circuit path has now passed its comprehensive series of loopback tests. This circuit qualifies to carry live traffic.


1.4 Restoring the Database and Default Settings

This section contains troubleshooting for node operation errors that require restoration of software data or the default node setup.

1.4.1 Restore the Node Database

Symptom: One or more node(s) are not functioning properly or have incorrect data.

Table 1-1 describes the potential cause(s) of the symptom and the solution(s).

Table 1-1 Restore the Node Database 

Possible Problem
Solution

Incorrect or corrupted node database.

Perform a Restore the Database procedure. Refer to the "Restore the Database" procedure.


Procedure: Restore the Database


Note The following parameters are not backed up and restored: node name, IP address, mask and gateway, and IIOP port. If you change the node name and then restore a backed up database with a different node name, the circuits will map to the new renamed node. Cisco recommends keeping a record of the old and new node names.



Caution E1000-2 cards lose traffic for approximately 90 seconds when an ONS 15454 SDH database is restored. Traffic is lost during the period of spanning-tree reconvergence. The CARLOSS alarm will appear and clear during this period.


Caution If you are restoring the database on multiple nodes, wait until the TCC2 reboot has completed on each node before proceeding to the next node.


Step 1 Log into the node where you will restore the database:

a. On the PC connected to the ONS 15454 SDH, start Netscape or Internet Explorer.

b. In the Netscape or Internet Explorer Web address (URL) field, enter the ONS 15454 SDH IP address.

A Java Console window appears the CTC file download status. The web browser displays information about your Java and system environments. If this is the first login, CTC caching messages appear while CTC files are downloaded to your computer. The first time you connect to an ONS 15454 SDH, this process can take several minutes. After the download, the CTC Login dialog box appears.

c. In the Login dialog box, type a user name and password (both are case sensitive) and click the Login button. The CTC node view window will appear.

Step 2 Ensure that there are no ring or span (four-fiber only) switch events; for example, ring-switch east or west, and span-switch east or west. In network view, click the Conditions tab and click Retrieve Conditions to view a list of conditions.

Step 3 If there are switch events that need to be cleared, in node (default) view, click the Maintenance > MS-SPRing tabs and view the West Switch and East Switch columns:

a. If there is a switch event (not caused by a line failure), clear the switch by choosing CLEAR from the drop-down menu and click Apply.

b. If there is a switch event caused by the Wait to Restore (WTR) condition, choose LOCKOUT SPAN from the drop-down menu and click Apply. When the LOCKOUT SPAN is applied, choose CLEAR from the drop-down menu and click Apply.

Step 4 In node view, click the Maintenance > Database tabs.

Step 5 Click Restore.

Step 6 Locate the database file stored on the workstation's hard drive or on network storage.

Step 7 Click the database file to highlight it.

Step 8 Click Open. The DB Restore dialog box appears.


Caution Opening a restore file from another node or from an earlier backup can affect traffic on the login node.

Step 9 Click Yes.

The Restore Database dialog box monitors the file transfer.

Step 10 Wait for the file to complete the transfer to the TCC2 card.

Step 11 Click OK when the "Lost connection to node, changing to Network View" dialog box appears. Wait for the node to reconnect.

Step 12 If you cleared a switch in Step 3, reapply the switch as needed.


1.4.2 Restore the Node to Factory Configuration

Symptom    A node has both TCC2 cards in standby state, and you are unable reset the TCC2 cards to make the node functional.

Table 1-2 describes the potential cause(s) of the symptom and the solution(s).

Table 1-2 Restore the Node to Factory Configuration 

Possible Problem
Solution

Failure of both TCC2 cards in the node.

This procedure describes how to restore the node to factory configuration using the RE-INIT.jar JAVA file, which is referred to as the reinitialization tool in this documentation. Use this tool to upload the software package and/or restore the database after it has been backed up. You need the CD containing the latest software, the node's NE defaults, and the recovery tool.

To restore the node to factory configuration, refer to the "Use the Reinitialization Tool to Clear the Database and Upload Software (Windows)" procedure or the "Use the Reinitialization Tool to Clear the Database and Upload Software (UNIX)" procedure.

Replacement of both TCC2 cards at the same time.



Caution If you are restoring the database on multiple nodes, wait until the TCC2 cards have rebooted on each node before proceeding to the next node.


Caution Cisco strongly recommends that you keep different node databases in separate folders. This is because the reinitialization tool will choose the first product-specific software package in the specified directory if you only use the Search Path field. You might accidentally copy an incorrect database if multiple databases are kept in the specified directory.


Note If the software package files and database backup files are located in different directories, complete the Package and Database fields (Figure 1-19).



Note The following parameters are not backed up and restored: node name, IP address, mask and gateway, and IIOP port. If you change the node name and then restore a backed up database with a different node name, the circuits map to the new renamed node. Cisco recommends keeping a record of the old and new node names.


Procedure: Use the Reinitialization Tool to Clear the Database and Upload Software (Windows)


Note The TCC2 cards reboot several times during this procedure. Wait until they are completely rebooted before continuing.



Step 1 Insert the system software CD containing the reinit tool (Figure 1-19) into the local craft interface PC drive. If the CTC Installation Wizard opens, click Cancel.

Step 2 To find the recovery tool file, go to Start > Run > Browse and select the CD drive.

Step 3 On the CD drive, go to the CISCO15454 folder and set the Files of Type drop-down menu to All Files.

Step 4 Select the RE-INIT.jar file and click Open to open the reinit tool (Figure 1-19).

Figure 1-19 Reinitialization Tool in Windows

Step 5 If the node you are reinitializing is an external network element (ENE) in a proxy server network, enter the IP address of the gateway network element (GNE) in the GNE IP field. If not, leave it blank.

Step 6 Enter the node name or IP address of the node you are reinitializing in the Node IP field (Figure 1-19).

Step 7 Verify that the Re-Init Database, Upload Package, and Confirm check boxes are checked. If one is not checked, click the check box.

Step 8 In the Search Path field, verify that the path to the CISCO15454 folder on the CD drive is listed.


Caution Cisco strongly recommends that you keep different node databases in separate folders. This is because the reinit tool chooses the first product-specific software package in the specified directory if you use the Search Path field instead of the Package and Database fields. You might accidentally copy an incorrect database if multiple databases are kept in the specified directory.


Caution Before you perform the next step, be sure you are uploading the correct database. You cannot reverse the upload process after you click Yes.

Step 9 Click Go.

Step 10 A confirmation dialog box opens (Figure 1-20). Click Yes.

Step 11 The status bar at the bottom of the window will appear Complete when the node has activated the software and uploaded the database.


Note The Complete message only indicates that the TCC2 successfully uploaded the database, not that the database restore was successful. The TCC2 then tries to restore the database after it reboots.


Step 12 If you are logged into CTC, close the browser window and disconnect the straight-through LAN cable from the RJ-45 (LAN) port on the TCC2 or on the hub or switch to which the ONS 15454 SDH is physically connected. Reconnect your straight-through LAN cable to the LAN port and log back into CTC. Refer to the Cisco ONS 15454 SDH Procedure Guide.

Step 13 Manually set the node name and network configuration to site-specific values. Refer to the Cisco ONS 15454 SDH Procedure Guide for information on setting the node name, IP address, mask and gateway, and IIOP port.

Figure 1-20 Confirm NE Restoration


Procedure: Use the Reinitialization Tool to Clear the Database and Upload Software (UNIX)


Note JRE 1.03_02 must also be installed on the computer you use to perform this procedure.



Note The TCC2 cards will reboot several times during this procedure. Wait until they are completely rebooted before continuing.



Step 1 Insert the system software CD containing the reinit tool, software, and defaults database into the local craft interface PC drive. If the CTC Installation Wizard opens, click Cancel.

Step 2 To find the recovery tool file, go to the CISCO15454 directory on the CD (usually /cdrom/cdrom0/CISCO15454).

Step 3 If you are using a file explorer, double click the RE-INIT.jar file to open the reinitialization tool (Figure 1-21). If you are working with a command line interface, run java -jar RE-INIT.jar.

Figure 1-21 Reinitialization Tool in UNIX

Step 4 If the node you are reinitializing is an external network element (ENE) in a proxy server network, enter the IP address of the gateway network element (GNE) in the GNE IP field. If not, leave it blank.

Step 5 Enter the node name or IP address of the node you are reinitializing in the Node IP field (Figure 1-21).

Step 6 Verify that the Re-Init Database, Upload Package, and Confirm check boxes are checked. If any are not checked, click that check box.

Step 7 In the Search Path field, verify that the path to the CISCO15454 folder on the CD drive is listed.


Caution Cisco strongly recommends that you keep different node databases in separate folders. This is because the reinit tool chooses the first product-specific software package in the specified directory if you use the Search Path field instead of the Package and Database fields. You might accidentally copy an incorrect database if multiple databases are kept in the specified directory.


Caution Before you perform the next step, be sure you are uploading the correct database. You cannot reverse the upload process after you click Yes.

Step 8 Click Go.

Step 9 A confirmation dialog box opens (Figure 1-20). Click Yes.

Step 10 The status bar at the bottom of the window displays Complete when the node has activated the software and uploaded the database.


Note The Complete message only indicates that the TCC2 successfully uploaded the database, not that the database restore was successful. The TCC2 then tries to restore the database after it reboots.


Step 11 If you are logged into CTC, close the browser window and disconnect the straight-through LAN cable from the RJ-45 (LAN) port on the TCC2 or on the hub or switch to which the ONS 15454 SDH is physically connected. Reconnect your straight-through LAN cable to the LAN port and log back into CTC. Refer to the Cisco ONS 15454 SDH Procedure Guide.

Step 12 Manually set the node name and network configuration to site-specific values. Refer to the Cisco ONS 15454 SDH Procedure Guide for information on setting the node name, IP address, mask and gateway, and IIOP port.


1.5 PC Connectivity Troubleshooting

This section contains troubleshooting procedures for PC and network connectivity to the ONS 15454 SDH.

1.5.1 Unable to Verify the IP Configuration of Your PC

Symptom    When connecting your PC to the ONS 15454 SDH, you are unable to successfully ping the IP address of your PC to verify the IP configuration.

Table 1-3 describes the potential cause(s) of the symptom and the solution(s).

Table 1-3 Unable to Verify the IP Configuration of your PC 

Possible Problem
Solution

The IP address was typed incorrectly.

Verify that the IP address used to ping the PC matches the IP address displayed in the Windows IP Configuration information retrieved from the system. See "Verify the IP Configuration of your PC" procedure.

The IP configuration of your PC is not properly set.

Verify the IP configuration of your PC. See the "Verify the IP Configuration of your PC" procedure. If this procedure is unsuccessful, contact your Network Administrator for instructions to correct the IP configuration of your PC.


Procedure: Verify the IP Configuration of your PC


Step 1 Open a DOS command window by selecting Start > Run from the Start menu.

Step 2 In the Open field, type command and then click the OK button. The DOS command window will appear.

Step 3 At the prompt in the DOS window, type one of the following appropriate commands:

For Windows 98, NT, and 2000, type ipconfig and press the Enter key.

For Windows 95, type winipcfg and press the Enter key.

The Windows IP configuration information appears, including the IP address, subnet mask, and the default gateway.

Step 4 At the prompt in the DOS window, type ping followed by the IP address shown in the Windows IP configuration information previously displayed.

Step 5 Press the Enter key to execute the command.

If the DOS window displays multiple (usually four) replies, the IP configuration is working properly.

If you do not receive a reply, your IP configuration might not be properly set. Contact your Network Administrator for instructions to correct the IP configuration of your PC.


1.5.2 Browser Login Does Not Launch Java

Symptom    The message "Loading Java Applet" does not appear and the JRE does not launch during the initial login.

Table 1-4 describes the potential cause(s) of the symptom and the solution(s).

Table 1-4 Browser Login Does Not Launch Java 

Possible Problem
Solution

The PC operating system and browser are not properly configured.

See the "Reconfigure the PC Operating System Java Plug-in Control Panel" procedure and the "Reconfigure the Browser" procedure.


Procedure: Reconfigure the PC Operating System Java Plug-in Control Panel


Step 1 From the Windows start menu, click Settings > Control Panel.

Step 2 If Java Plug-in Control Panel does not appear, the JRE might not be installed on your PC. Complete the following steps:

a. Run the Cisco ONS 15454 SDH software CD.

b. Open the [CD drive]:\Windows\JRE folder.

c. Double-click the j2re-1_3_1_02-win icon to run the JRE installation wizard.

d. Follow the JRE installation wizard steps.

Step 3 From the Windows start menu, click Settings > Control Panel.

Step 4 In the Java Plug-in Control Panel window, double-click the Java Plug-in 1.3.1_02 icon.

Step 5 Click the Advanced tab on the Java Plug-in Control Panel.

Step 6 From the Java Run Time Environment menu, select JRE 1.3 in C:\ProgramFiles\JavaSoft\JRE\1.3.1_02.

Step 7 Click the Apply button.

Step 8 Close the Java Plug-in Control Panel window.


Procedure: Reconfigure the Browser


Step 1 From the Start Menu, launch your browser application.

Step 2 If you are using Netscape Navigator:

a. On the Netscape Navigator menu bar, click the Edit > Preferences menus.

b. In the Preferences window, click the Advanced > Proxies categories.

c. In the Proxies window, click the Direct connection to the Internet check box and click the OK button.

d. On the Netscape Navigator menu bar, click the Edit > Preferences menus.

e. In the Preferences window, click the Advanced > Cache categories.

f. Confirm that the Disk Cache Folder field shows one of the following paths:

For Windows 95/98/ME, C:\ProgramFiles\Netscape\Communicator\cache

For Windows NT/2000, C:\ProgramFiles\Netscape\<username>\Communicator\cache

g. If the Disk Cache Folder field is not correct, click the Choose Folder button.

h. Navigate to the file listed in Step f, and click the OK button.

i. Click the OK button on the Preferences window and exit the browser.

Step 3 If you are using Internet Explorer:

a. On the Internet Explorer menu bar, click the Tools > Internet Options menus.

b. In the Internet Options window, click the Advanced tab.

c. In the Settings menu, scroll down to Java (Sun) and click the Use Java 2 v1.3.1_02 for <applet> (requires restart) check box.

d. Click the OK button in the Internet Options window and exit the browser.

Step 4 Temporarily disable any virus-scanning software on the computer. See the "Browser Stalls When Downloading CTC JAR Files From TCC2" section.

Step 5 Verify that the computer does not have two network interface cards (NICs) installed. If the computer does have two NICs, remove one.

Step 6 Restart the browser and log on to the ONS 15454 SDH.


1.5.3 Unable to Verify the NIC Connection on Your PC

Symptom    When connecting your PC to the ONS 15454 SDH, you are unable to verify the NIC connection is working properly because the link LED is not illuminated or flashing.

Table 1-5 describes the potential cause(s) of the symptom and the solution(s).

Table 1-5 Unable to Verify the NIC Connection on your PC 

Possible Problem
Solution

The CAT-5 cable is not plugged in properly.

Confirm both ends of the cable are properly inserted. If the cable is not fully inserted due to a broken locking clip, the cable should be replaced.

The CAT-5 cable is damaged.

Ensure that the cable is in good condition. If in doubt, use a cable that is known to be good. Often, cabling is damaged due to pulling or bending

Incorrect type of CAT-5 cable is being used.

If connecting an ONS 15454 SDH directly to your laptop/PC or a router, use a straight-through CAT-5 cable. When connecting the ONS 15454 SDH to a hub or a LAN switch, use a crossover CAT-5 cable.

For details on the types of CAT-5 cables, see the "Crimp Replacement LAN Cables" section.

The NIC is improperly inserted or installed.

If you are using a PCMCIA based NIC, remove and re-insert the NIC to make sure the NIC is fully inserted.

If the NIC is built into the laptop/PC, verify that the NIC is not faulty.

The NIC is faulty.

Confirm that the NIC is working properly. If you have no issues connecting to the network (or any other node), then the NIC should be working correctly.

If you have difficulty connecting to the network (or any other node), then the NIC might be faulty and needs to be replaced.


1.5.4 Verify PC Connection to the ONS 15454 SDH (ping)

Symptom    The TCP/IP connection was established and then lost, and a DISCONNECTED alarm appears on CTC.

Table 1-6 describes the potential cause(s) of the symptom and the solution(s).

Table 1-6 Verify PC connection to ONS 15454 SDH (ping) 

Possible Problem
Solution

A lost connection between the PC and the ONS 1554 SDH.

Use a standard ping command to verify the TCP/IP connection between the PC and the ONS 15454 SDH TCC2 card. A ping command will work if the PC connects directly to the TCC2 card or uses a LAN to access the TCC2 card.

See the "Ping the ONS 15454 SDH" procedure.


Procedure: Ping the ONS 15454 SDH


Step 1 Display the command prompt:

a. If you are using a Microsoft Windows operating system, from the Start Menu choose Run, type command prompt in the Open field of the Run dialog box, and click OK.

b. If you are using a Sun Solaris operating system, from the Common Desktop Environment (CDE) click the Personal Application tab and click Terminal.

Step 2 For both the Sun and Microsoft operating systems, at the prompt type:

ping ONS-15454-SDH-IP-address

For example:

ping 192.1.0.2.

Step 3 If the workstation has connectivity to the ONS 15454 SDH, the ping is successful and displays a reply from the IP address. If the workstation does not have connectivity, a "Request timed out" message appears.

Step 4 If the ping is successful, an active TCP/IP connection exists. Restart CTC.

Step 5 If the ping is not successful, and the workstation connects to the ONS 15454 SDH through a LAN, check that the workstation's IP address is on the same subnet as the ONS node.

Step 6 If the ping is not successful and the workstation connects directly to the ONS 15454 SDH, check that the link light on the workstation's NIC is illuminated.


1.5.5 The IP Address of the Node is Unknown

Symptom    The IP address of the node is unknown and you are enable login.

Table 1-7 describes the potential cause(s) of the symptom and the solution(s).

Table 1-7 Retrieve the unknown IP address of the node 

Possible Problem
Solution

The node is not set to the default IP address.

Leave one TCC2 card in the shelf. Connect a PC directly to the remaining TCC2 card and perform a hardware reset of the card. The TCC2 card will transmit the IP address after the reset to enable you to capture the IP address for login.

See the "Retrieve Unknown Node IP Address" procedure.


Procedure: Retrieve Unknown Node IP Address


Step 1 Connect your PC directly to the active TTC+TCC2 card Ethernet port on the faceplate.

Step 2 Start the Sniffer application on your PC.

Step 3 Perform a hardware reset by pulling and reseating the active TCC2 card.

Step 4 After the TCC2 card completes resetting, it will broadcast its IP address. The Sniffer software on your PC will capture the IP address being broadcast.


1.6 CTC Operation Troubleshooting

This section contains troubleshooting procedures for CTC login or operation problems.

1.6.1 Unable to Launch CTC Help After Removing Netscape

Symptom    After removing Netscape and running CTC using Internet Explorer, the user is unable to launch the CTC Help and receives an "MSIE is not the default browser" error message.

Table 1-9 describes the potential cause(s) of the symptom and the solution(s).

Table 1-8 Unable to Launch CTC Help After Removing Netscape 

Possible Problem
Solution

Loss of association between browser and Help files.

When the CTC software and Netscape are installed, the Help files are associated with Netscape by default. When you remove Netscape, the Help files are not automatically associated with Internet Explorer as the default browser.

Reset Internet Explorer as the default browser so that CTC will associate the Help files to the correct browser.

See the "Reset Internet Explorer as the Default Browser for CTC" procedure to associate the CTC Help files to the correct browser.


Procedure: Reset Internet Explorer as the Default Browser for CTC


Step 1 Open the Internet Explorer browser.

Step 2 From the menu bar, click Tools > Internet Options. The Internet Options window appears.

Step 3 In the Internet Options window, click the Programs tab.

Step 4 Click the Internet Explorer should check to see whether it is the default browser check box.

Step 5 Click the OK button.

Step 6 Exit any and all open and running CTC and Internet Explorer applications.

Step 7 Launch Internet Explorer and open a new CTC session. You should now be able to access the CTC Help.


1.6.2 Unable to Change Node View to Network View

Symptom    When activating a large, multi node MS-SPRing from Software Release 3.2 to Software Release 3.3, some of the nodes appear grayed out. Logging into the new CTC, the user is unable to change node view to network view on any and all nodes, from any workstation. This is accompanied by an "Exception occurred during event dispatching: java.lang.OutOfMemoryError" in the java window.

Table 1-9 describes the potential cause(s) of the symptom and the solution(s).

Table 1-9 Browser Stalls When Downloading Files From TCC2 

Possible Problem
Solution

The large, multinode MS-SPRing requires more memory for the graphical user interface (GUI) environment variables.

Reset the system or user CTC_HEAP environment variable to increase the memory limits.

See the "Reset the CTC_HEAP Environment Variable for Windows" procedure or the "Reset the CTC_HEAP Environment Variable for Solaris" procedure to enable the CTC_HEAP variable change.

Note This problem typically affects large networks where additional memory is required to manage large numbers of nodes and circuits.


Procedure: Reset the CTC_HEAP Environment Variable for Windows


Step 1 Exit any and all open and running CTC and Netscape applications.

Step 2 From the Windows Desktop, right-click on My Computer and choose Properties in the popup menu.

Step 3 In the System Properties window, click the Advanced tab.

Step 4 Click the Environment Variables button to open the Environment Variables window.

Step 5 Click the New button under the User variables field or the System variables field.

Step 6 Type CTC_HEAP in the Variable Name field.

Step 7 Type 256 in the Variable Value field, and then click OK to create the variable.

Step 8 Click OK in the Environment Variables window to accept the changes.

Step 9 Click OK in the System Properties window to accept the changes.

You may now restart the browser and CTC software.


Procedure: Reset the CTC_HEAP Environment Variable for Solaris


Step 1 From the user shell window, kill any CTC applications.

Step 2 Kill any Netscape applications.

Step 3 In the user shell window, set the environment variable to increase the heap size:

% setenv CTC_HEAP 256 

You may now restart the browser and CTC software in the same user shell window.


1.6.3 Browser Stalls When Downloading CTC JAR Files From TCC2

Symptom    The browser stalls or hangs when downloading a CTC JAR file from the TCC2 card.

Table 1-10 describes the potential cause(s) of the symptom and the solution(s).

Table 1-10 Browser Stalls When Downloading jar File From TCC2 

Possible Problem
Solution

McAfee VirusScan software might be interfering with the operation. The problem occurs when the VirusScan Download Scan is enabled on McAfee VirusScan 4.5 or later.

Disable the VirusScan Download Scan feature. See the "Disable the VirusScan Download Scan" procedure.


Procedure: Disable the VirusScan Download Scan


Step 1 From the Windows start menu, choose Programs > Network Associates > VirusScan Console.

Step 2 Double-click the VShield icon listed in the VirusScan Console dialog box.

Step 3 Click the Configure button on the lower part of the Task Properties window.

Step 4 Click the Download Scan icon on the left of the System Scan Properties dialog box.

Step 5 Uncheck the Enable Internet download scanning check box.

Step 6 Click Yes when the warning message appears.

Step 7 Click OK on the System Scan Properties dialog box.

Step 8 Click OK on the Task Properties window.

Step 9 Close the McAfee VirusScan window.


1.6.4 CTC Does Not Launch

Symptom    CTC does not launch, usually an error message appears before the login window appears.

Table 1-11 describes the potential cause(s) of the symptom and the solution(s).

Table 1-11 CTC Does Not Launch 

Possible Problem
Solution

The Netscape browser cache might point to an invalid directory.

Redirect the Netscape cache to a valid directory. See the "Redirect the Netscape Cache to a Valid Directory" procedure.


Procedure: Redirect the Netscape Cache to a Valid Directory


Step 1 Launch Netscape.

Step 2 Display the Edit menu.

Step 3 Choose Preferences.

Step 4 Under the Category column on the left side, expand the Advanced category and choose the Cache tab.

Step 5 Change your disk cache folder to point to the cache file location.

The cache file location is usually C:\ProgramFiles\Netscape\Users\yourname\cache. The yourname segment of the file location is often the same as the user name.


1.6.5 Sluggish CTC Operation or Login Problems

Symptom    You experience sluggish CTC operation or have problems logging into CTC.

Table 1-12 describes the potential cause(s) of the symptom and the solution(s).

Table 1-12 Sluggish CTC Operation or Login Problems 

Possible Problem
Solution

The CTC cache file might be corrupted or might need to be replaced.

Delete the CTC cache file. This operation forces the ONS 15454 SDH to download a new set of jar files to your computer hard drive. See the "Delete the CTC Cache File Automatically" procedure or the "Delete the CTC Cache File Manually" procedure.


Procedure: Delete the CTC Cache File Automatically


Caution All running sessions of CTC must be halted before deleting the CTC cache. Deleting CTC cache can cause any CTC running on this system to behave in an unexpected manner.


Step 1 Enter an ONS 15454 SDH IP address into the browser URL field. The initial browser window shows a Delete CTC Cache button.

Step 2 Close all open CTC sessions and browser windows. The PC operating system does not allow you to delete files that are in use.

Step 3 Click the Delete CTC Cache button on the initial browser window to clear the CTC cache. Figure 1-22 shows the Delete CTC Cache window.

Figure 1-22 Deleting the CTC Cache


Procedure: Delete the CTC Cache File Manually


Caution All running sessions of CTC must be halted before deleting the CTC cache. Deleting CTC cache can cause any CTC running on this system to behave in an unexpected manner.


Step 1 To delete the jar files manually, from the Windows Start menu choose Search > For Files or Folders.

Step 2 Enter *.jar in the Search for files or folders named field on the Search Results dialog box and click Search Now.

Step 3 Click the Modified column on the Search Results dialog box to find the jar files that match the date when you downloaded the files from the TCC2. These files can include CTC*.jar, CMS*.jar, and jar_cache*.tmp.

Step 4 Highlight the files and press the keyboard Delete key.

Step 5 Click Yes at the Confirm dialog box.


1.6.6 Node Icon is Grey on CTC Network View

Symptom    The CTC network view shows one or more node icons as grey in color and without a node name.

Table 1-13 describes the potential cause(s) of the symptom and the solution(s).

Table 1-13 Node Icon is Grey on CTC Network View 

Possible Problem
Solution

Different CTC releases not recognizing each other.

Correct the core version build as described in the "Different CTC Releases Do Not Recognize Each Other" section.

A username/password mismatch.

Correct the username and password as described in the "Username or Password Do Not Match" section.

No IP connectivity between nodes.

Usually accompanied by Ethernet-specific alarms. Verify the Ethernet connections as described in the "Ethernet Connections" section.

A lost DCC connection.

Usually accompanied by an embedded operations channel (EOC) alarm. Clear the EOC alarm and verify the DCC connection as described in the "EOC" section.


1.6.7 CTC Cannot Launch Due to Applet Security Restrictions

Symptom    The error message "Unable to launch CTC due to applet security restrictions" appears after you enter the IP address in the browser window.

Table 1-14 describes the potential cause(s) of the symptom and the solution(s).

Table 1-14 CTC Cannot Launch Due to Applet Security Restrictions 

Possible Problem
Solution

Did not execute the javapolicyinstall.bat file, or the java.policy file might be incomplete.

1. Verify that you have executed the javapolicyinstall.bat file on the ONS 15454 SDH software CD. This file is installed when you run the CTC Setup Wizard (refer to the CTC installation information in the Cisco ONS 15454 SDH Procedure Guide for instructions).

2. If you ran the javapolicyinstall.bat file but still receive the error message, you must manually edit the java.policy file on your computer. See the "Manually Edit the java.policy File" procedure.


Procedure: Manually Edit the java.policy File


Step 1 Search your computer for this file and open it with a text editor (Notepad or Wordpad).

Step 2 Verify that the end of this file has the following lines:

		// Insert this into the system-wide or a per-user java.policy file.
	// DO NOT OVERWRITE THE SYSTEM-WIDE POLICY FILE--ADD THESE LINES!

	grant codeBase "http://*/fs/LAUNCHER.jar" {
permission java.security.AllPermission;
	};

Step 3 If these five lines are not in the file, enter them manually.

Step 4 Save the file and restart Netscape.

CTC should now start correctly.

Step 5 If the error message is still reported, save the java.policy file as (.java.policy). On Win95/98/2000 PCs, save the file to the C:\Windows folder. On WinNT4.0 PCs, save the file to all of the user folders on that PC, for example, C:\Winnt\profiles\joeuser.


1.6.8 Java Runtime Environment Incompatible

Symptom    The CTC application does not run properly.

Table 1-15 describes the potential cause(s) of the symptom and the solution(s).

Table 1-15 Java Runtime Environment Incompatible 

Possible Problem
Solution

The compatible Java 2 JRE is not installed.

The JRE contains the Java virtual machine, runtime class libraries, and Java application launcher that are necessary to run programs written in the Java programming language.

The ONS 15454 SDH CTC is a Java application. A Java application, unlike an applet, cannot rely completely on a web browser for installation and runtime services. When you run an application written in the Java programming language, you need the correct JRE installed. The correct JRE for each CTC software release is included on the Cisco ONS 15454 SDH software CD and on the Cisco ONS 15454 SDH documentation CD. See the "Launch CTC to Correct the Core Version Build" procedure.

If you are running multiple CTC software releases on a network, the JRE installed on the computer must be compatible with the different software releases. Table 1-16 shows JRE compatibility with ONS 15454 SDH software releases.


Table 1-16 JRE Compatibility

ONS Software Release
JRE 1.2.2 Compatible
JRE 1.3 Compatible

ONS 15454 SDH Release 3.3

Yes

Yes

ONS 15454 SDH Release 3.4

No

Yes

ONS 15454 SDH Release 4.0

No

Yes

Note Software R4.0 will notify you if an older version JRE is running on your PC or UNIX workstation.


Procedure: Launch CTC to Correct the Core Version Build


Step 1 Exit the current CTC session and completely close the browser.

Step 2 Start the browser.

Step 3 Type the ONS 15454 SDH IP address of the node that reported the alarm. This can be the original IP address you logged on with or an IP address other than the original.

Step 4 Log into CTC. The browser downloads the jar file from CTC.


1.6.9 Different CTC Releases Do Not Recognize Each Other

Symptom    This situation is often accompanied by the INCOMPATIBLE-SW alarm.

Table 1-17 describes the potential cause(s) of the symptom and the solution(s).

Table 1-17 Different CTC Releases Do Not Recognize Each Other 

Possible Problem
Solution

The software loaded on the connecting workstation and the software on the TCC2 card are incompatible.

This occurs when the TCC2 software is upgraded but the PC has not yet upgraded the compatible CTC jar file. It also occurs on login nodes with compatible software that encounter other nodes in the network that have a newer software version.

Note Remember to always log into the ONS node with the latest CTC core version first. If you initially log into an ONS node running a CTC core version of 2.2 or lower and then attempt to log into another ONS node in the network running a higher CTC core version, the lower version node does not recognize the new node.

See the "Launch CTC to Correct the Core Version Build" procedure.


Procedure: Launch CTC to Correct the Core Version Build


Step 1 Exit the current CTC session and completely close the browser.

Step 2 Start the browser.

Step 3 Type the ONS 15454 SDH IP address of the node that reported the alarm. This can be the original IP address you logged on with or an IP address other than the original.

Step 4 Log into CTC. The browser will download the jar file from CTC.


1.6.10 Username or Password Do Not Match

Symptom    A mismatch often occurs concurrently with a NOT-AUTHENTICATED alarm.

Table 1-18 describes the potential cause(s) of the symptom and the solution(s).

Table 1-18 Username or Password Do Not Match 

Possible Problem
Solution

The username or password entered do not match the information stored in the TCC2.

All ONS nodes must have the same username and password created to display every ONS node in the network. You can also be locked out of certain ONS nodes on a network if your username and password were not created on those specific ONS nodes.

For initial logon to the ONS 15454 SDH, type the CISCO15 user name in capital letters and click Login (no password is required). If you are using a CTC Software Release 2.2.2 or earlier and CISCO15 does not work, type cerent454 for the user name.

See the "Verify Correct Username and Password" procedure.


Procedure: Verify Correct Username and Password


Step 1 Ensure that your keyboard Caps Lock key is not turned on and affecting the case-sensitive entry of the username and password.

Step 2 Contact your system administrator to verify the username and password.

Step 3 Call Cisco TAC to have them enter your system and create a new user name and password.


1.6.11 No IP Connectivity Exists Between Nodes

Symptom    The nodes have a grey icon and is usually accompanied by alarms.

Table 1-19 describes the potential cause(s) of the symptom and the solution(s).

Table 1-19 No IP Connectivity Exists Between Nodes 

Possible Problem
Solution

A lost Ethernet connection.

Usually is accompanied by Ethernet-specific alarms. Verify the Ethernet connections as described in the "Ethernet Connections" section.


1.6.12 DCC Connection Lost

Symptom    The node is usually accompanied by alarms and the nodes in the network view have a grey icon. This symptom is usually accompanied by an EOC alarm.

Table 1-20 describes the potential cause(s) of the symptom and the solution(s).

Table 1-20 DCC Connection Lost 

Possible Problem
Solution

A lost DCC connection.

Usually accompanied by an EOC alarm. Clear the EOC alarm and verify the DCC connection as described in the "EOC" section.


1.6.13 "Path in Use" Error When Creating a Circuit

Symptom    While creating a circuit, you get a "Path in Use" error that prevents you from completing the circuit creation.

Table 1-21 describes the potential cause(s) of the symptom and the solution(s).

Table 1-21 "Path in Use" error when creating a circuit 

Possible Problem
Solution

Another user has already selected the same source port to create another circuit.

CTC does not remove a card or port from the available list until a circuit is completely provisioned. If two users simultaneously select the same source port to create a circuit, the first user to complete circuit provisioning gets use of the port. The other user will get the "Path in Use" error.

Cancel the circuit creation and start over, or click the Back button until you return to the initial circuit creation window. The source port that was previously selected no longer appears in the available list because it is now part of a provisioned circuit. Select a different available port and begin the circuit creation process again.


1.6.14 Calculate and Design IP Subnets

Symptom    You cannot calculate or design IP subnets on the ONS 15454 SDH.

Table 1-22 describes the potential cause(s) of the symptom and the solution(s).

Table 1-22 Calculate and Design IP Subnets 

Possible Problem
Solution

The IP capabilities of the ONS 15454 SDH require specific calculations to properly design IP subnets.

Cisco provides a free online tool to calculate and design IP subnets. Go to http://www.cisco.com/techtools/ip_addr.html. For information about ONS 15454 SDH IP capability, refer to the Cisco ONS 15454 SDH Reference Manual.


1.6.15 Ethernet Connections

Symptom    Ethernet connections appear to be broken or are not working properly.

Table 1-23 describes the potential cause(s) of the symptom and the solution(s).

Table 1-23 Calculate and Design IP Subnets 

Possible Problem
Solution

Improperly seated connections.

You can fix most connectivity problems in an Ethernet network by following a few guidelines. See Figure 1-23 when consulting the steps in the "Verify Ethernet Connections" procedure.

Incorrect connections.


Figure 1-23 Ethernet Connectivity Reference

Procedure: Verify Ethernet Connections


Step 1 Verify that the alarm filter is turned OFF.

Step 2 Check for SDH alarms on the STS-N that carries the VLAN #1 Ethernet circuit. Clear any alarms by looking them up in "Alarm Troubleshooting."

Step 3 Check for Ethernet-specific alarms. Clear any raised alarms by looking up that alarm in "Alarm Troubleshooting."

Step 4 Verify that the ACT LED on the Ethernet card is green.

Step 5 Verify that Ports 1 and 3 on ONS 15454 SDH #1 and Ports 1 and 2 on ONS 15454 SDH #2 have green link-integrity LEDs illuminated.

Step 6 If no green link-integrity LED is illuminated for any of these ports:

a. Verify physical connectivity between the ONS 15454 SDH and the attached device.

b. Verify that the ports are enabled on the Ethernet cards.

c. Verify that you are using the proper Ethernet cable and that it is wired correctly, or replace the cable with a known-good Ethernet cable.

d. Check the status LED on the Ethernet card faceplate to ensure the card booted up properly. This LED should be steady green. If necessary, remove and reinsert the card and allow it to reboot.

e. It is possible that the Ethernet port is functioning properly but the link LED itself is broken. Run the procedure in the "Lamp Test for Card LEDs" section.

Step 7 Verify connectivity between device A and device C by pinging between these locally attached devices (see the "Verify PC Connection to the ONS 15454 SDH (ping)" section). If the ping is unsuccessful:

a. Verify that device A and device C are on the same IP subnet.

b. Display the Ethernet card in CTC card view and click the Provisioning > VLAN tabs to verify that both Port 1 and Port 3 on the card are assigned to the same VLAN.

c. If a port is not assigned to the correct VLAN, click that port column in the VLAN row and set the port to Tagged or Untag. Click Apply.

Step 8 Repeat Step 7 for devices B and D.

Step 9 Verify that the Ethernet circuit that carries VLAN #1 is provisioned and that ONS 15454 SDH #1 and ONS 15454 SDH #2 ports also use VLAN #1.


1.6.16 VLAN Cannot Connect to Network Device from Untag Port

Symptom    Networks that have a VLAN with one ONS 15454 SDH Ethernet card port set to Tagged and one ONS 15454 SDH Ethernet card set to Untag might have difficulty implementing Address Resolution Protocol (ARP) for a network device attached to the Untag port (Figure 1-24). They might also see a higher than normal runt packets count at the network device attached to the Untag port. This symptom/limitation also exists when ports within the same card or ports within the same chassis are put on the same VLAN, with a mix of tagged and untagged.

Figure 1-24 A VLAN with Ethernet ports at Tagged and Untag

Table 1-24 describes the potential cause(s) of the symptom and the solution(s).

Table 1-24 Verify VLAN Connection to Network Device from Untag Port 

Possible Problem
Solution

The Tagged ONS 15454 SDH adds the IEEE 802.1Q tag and the Untag ONS 15454 SDH removes the Q-tag without replacing the bytes. The NIC of the network device categorizes the packet as a runt and drops the packet.

The solution is to set both ports in the VLAN to Tagged to stop the stripping of the 4 bytes from the data packet and prevent the NIC card in the network access device from recognizing the packet as a runt and dropping it. Network devices with IEEE 802.1Q-compliant NIC cards accept the tagged packets. Network devices with non-IEEE 802.1Q compliant NIC cards still drop these tagged packets. The solution might require upgrading network devices with non-IEEE 802.1Q compliant NIC cards to IEEE 802.1Q-compliant NIC cards. You can also set both ports in the VLAN to Untag, but you lose IEEE 802.1Q compliance.

Dropped packets can also occur when ARP attempts to match the IP address of the network device attached to the Untag port with the physical MAC address required by the network access layer.


Procedure: Change VLAN Port Tag and Untagged Settings


Step 1 Display the CTC card view for the Ethernet card involved in the problem VLAN.

Step 2 Click the Provisioning > VLAN tabs (Figure 1-25).

Figure 1-25 Configuring VLAN Membership for Individual Ethernet Ports

Step 3 If the port is set to Tagged, continue to look at other cards and their ports in the VLAN until you find the port that is set to Untag.

Step 4 At the VLAN port set to Untag, click the port and choose Tagged.


Note The attached external devices must recognize IEEE 802.1Q VLANs.


Step 5 After each port is in the appropriate VLAN, click Apply.


1.6.17 Cross-Connect Card Oscillator Fails

Symptom: The XC-VXL or XC10G card can be affected by this problem. It is indicated by a CTNEQPT-PBPROT or CTNEQPT-PBWORK condition raised against all I/O cards in the node. The following conditions might also be raised on the node:

SWMTXMOD against one or both cross-connect cards

SDBER-EXCEED-LO against near-end or far-end line cards

MS-AIS against far-end line cards

LP-RFI against near-end line cards

Table 1-25 describes the potential cause(s) of the symptom and the solution(s).

Table 1-25 Cross-Connect Card Oscillator Fails 

Possible Problem
Solution

The XC-VXL or XC10G card has oscillator failure.

1. If the Slot 8 cross-connect card is active, see the "Resolve the XC Oscillator Failure When Slot 8 XC Card is Active" section.

2. If the Slot 10 cross-connect card is active, see the "Resolve the XC Oscillator Failure When Slot 10 XC Card is Active" section.


Procedure: Resolve the XC Oscillator Failure When Slot 8 XC Card is Active


Step 1 If the CTNEQPT-PBPROT condition is reported against all I/O cards in the node and the Slot 8 cross-connect card is active, right-click the Slot 10 cross-connect card.

Step 2 Choose Reset Card, then click OK. (Slot 8 remains active and Slot 10 remains standby.)

Step 3 If the alarm remains, reseat the Slot 10 card.

Step 4 If CTNEQPT-PBPROT does not clear, replace the Slot 10 cross-connect card with a spare card.

Step 5 If CTNEQPT-PBPROT does not clear, replace the spare card placed in Slot 10 with the original cross-connect card.

Step 6 Right-click the Slot 8 card and choose Reset Card.

Step 7 Click OK to activate the Slot 10 card and place the Slot 8 card in standby.

Step 8 If you then see the CTNEQPT-PBWORK condition raised against all I/O cards in the node, verify that CTNEQPT-PBPROT has cleared on all I/O cards. Seeing CTNEQPT-PBWORK on the cards indicates that Slot 8 card has a bad oscillator. If this is indicated, complete the following substeps. Otherwise, go toStep 9.

a. Replace the Slot 8 cross-connect card with a spare card. (Slot 8 remains standby.)

b. Reseat the Slot 10 cross-connect card to activate the Slot 8 card and make Slot 10 standby.

c. Verify that the CTNEQPT-PBWORK condition has cleared on all I/O cards.

Step 9 If you see CTNEQPT-PBPROT reported against all I/O cards in the node, this indicates that the Slot 10 card has a bad oscillator. If so, complete the following steps:

a. Replace the Slot 10 cross-connect card with a spare card. (The Slot 8 card is now active.)

b. Reseat the Slot 8 cross-connect card to make Slot 10 active.

c. Verify that the CTNEQPT-PBPROT condition has cleared on all I/O cards.


Procedure: Resolve the XC Oscillator Failure When Slot 10 XC Card is Active


Step 1 If the CTNEQPT-PBWORK condition is reported against all I/O cards in the node and the Slot 10 card is active, right-click the Slot 8 cross-connect card.

Step 2 Choose Reset Card and click OK. (Slot 10 remains active and Slot 8 remains standby.)

Step 3 If the CTNEQPT-PBWORK condition does not clear, reseat the Slot 8 cross-connect card.

Step 4 If the condition does not clear, replace the Slot 8 cross-connect card with an identical, spare card.

Step 5 If the condition does not clear, replace the spare card placed in Slot 8 with the original cross-connect card.

Step 6 Right-click the Slot 10 cross-connect card.

Step 7 Choose Reset Card and click OK. The Slot 8 cross-connect card becomes active and Slot 10 becomes standby.

Step 8 If you have switched the Slot 8 card to active and continue to see CTNEQPT-PBWORK reported against all I/O cards in the node, this indicates the Slot 8 card has a bad oscillator. If this is indicated, complete the following substeps. If not, go to Step 9.

a. Replace the Slot 8 cross-connect card with a spare card. (The Slot 10 card is made active.)

b. Reseat the Slot 10 cross-connect card to make Slot 8 active.

c. Verify that the CTNEQPT-PBWORK condition has cleared on all I/O cards.

Step 9 If you then see the CTNEQPT-PBPROT condition raised against all I/O cards, verify that CTNEQPT-PBWORK has cleared on the I/O cards. This indicates that Slot 10 has a bad oscillator. If so, complete the following substeps:

a. Replace the Slot 10 cross-connect card with a spare card. (Slot 10 remains standby.)

b. Reseat the Slot 8 cross-connect card to activate the Slot 10 card and make Slot 8 standby.

c. Verify that the CTNEQPT-PBPROT condition has cleared on all I/O cards.


1.7 Circuits and Timing

This section provides solutions to circuit creation and reporting errors, as well as common timing reference errors and alarms.

1.7.1 Circuit Transitions to Partial State

Symptom    An automatic or manual transition of a circuit from one state to another state results in one of the following partial state conditions:

OOS_PARTIAL  At least one of the connections in the circuit is in OOS state and at least one other connection in the circuit is in IS, OOS_MT, or OOS_AINS state.

OOS_MT_PARTIAL  At least one connection in the circuit is in OOS_MT state and at least one other connection in the circuit is in IS, OOS_MT, or OOS_AINS state.

OOS_AINS_PARTIAL  At least one connection in the circuit is in the OOS_AINS state and at least one other connection in the circuit is in IS or OOS_AINS state.

Table 1-26 describes the potential cause(s) of the symptom and the solution(s).

Table 1-26 Circuit in Partial State 

Possible Problem
Solution

During a Manual transition, CTC cannot communicate with one of the nodes or one of the nodes is on a version of software that does not support the new state model.

Repeat the Manual transition operation. If the partial state persists, determine which node in the circuit is not changing to the desired state. Refer to the "View the State of Circuit Nodes" procedure.

Log onto the circuit node that did not change to the desired state and determine the version of software. If the software on the node is Software R3.3 or earlier, upgrade the software. Refer to the Cisco ONS 15454 SDH Software Upgrade Guide for software upgrade procedures.

Note If the node software cannot be upgraded to R4.0, the partial state condition can be avoided by only using the circuit state(s) supported in the earlier software version.

During an automatic transition, some path-level defects and/or alarms were detected on the circuit.

Determine which node in the circuit is not changing to the desired state. Refer to the "View the State of Circuit Nodes" procedure.

Log onto the circuit node that did not change to the desired state and examine the circuit for path-level defects, improper circuit termination, or alarms. Refer to the Cisco ONS 15454 SDH Procedure Guide for procedures to clear alarms and change circuit configuration settings.

Resolve and clear the defects and/or alarms on the circuit node and verify that the circuit transitions to the desired state.

One end of the circuit is not properly terminated.


Procedure: View the State of Circuit Nodes


Step 1 Click the Circuits tab.

Step 2 From the Circuits tab list, select the circuit with the *_PARTIAL state condition.

Step 3 Click the Edit button. The Edit Circuit window appears.

Step 4 In the Edit Circuit window, click the State tab.

The State tab window lists the Node, CRS End A, CRS End B, and CRS State for each of the nodes in the circuit.


1.7.2 DS3 Card Does Not Report MS-AIS From External Equipment

Symptom    A DS3i-12 card does not report MS-AIS from the external equipment/line side.

Table 1-27 describes the potential cause(s) of the symptom and the solution(s).

Table 1-27 DS3 Card Does Not Report MS-AIS From External Equipment 

Possible Problem
Solution

The card is functioning as designed.

This card terminates the port signal at the backplane, so STS MS-AIS is not reported from the external equipment/line side.

DS3i-12 cards have DS3 header monitoring functionality, which allows you to view performance monitoring (PM) on the DS3 path. Nevertheless, you cannot view MS-AIS on the STS path. For more information on the PM capabilities of the DS3i-12 cards, refer to the Cisco ONS 15454 SDH Procedure Guide.


1.7.3 STM-1 and DCC Limitations

Symptom    Limitations to STM-1 and DCC usage.

Table 1-28 describes the potential cause(s) of the symptom and the solution(s).

Table 1-28 STM-1 and DCC Limitations 

Possible Problem
Solution

STM-1 and DCC have limitations for the ONS 15454 SDH.

For an explanation of STM-1 and DCC limitations, refer to the DCC Tunnels section of the Cisco ONS 15454 SDH Procedure Guide.


1.7.4 ONS 15454 SDH Switches Timing Reference

Symptom    Timing references switch when one or more problems occur.

Table 1-29 describes the potential cause(s) of the symptom and the solution(s).

Table 1-29 ONS 15454 SDH Switches Timing Reference 

Possible Problem
Solution

The optical or BITS input is receiving loss of signal (LOS), loss of frame (LOF), or AIS alarms from its timing source.

The ONS 15454 SDH internal clock operates at a Stratum 3E level of accuracy. This gives the ONS 15454 SDH a free-running synchronization accuracy of ± 4.6 ppm and a holdover stability of less than 255 slips in the first 24 hours or 3.7 x 10-7/day, including temperature.

ONS 15454 SDH free-running synchronization relies on the Stratum 3 internal clock. Over an extended time period, using a higher quality Stratum 1 or Stratum 2 timing source results in fewer timing slips than a lower quality Stratum 3 timing source.

The optical or BITS input is not functioning.

Sync Status Messaging (SSM) message is set to Don't Use for Sync (DUS).

SSM indicates a Stratum 3 or lower clock quality.

The input frequency is off by more than 15 ppm.

The input clock wanders and has more than three slips in 30 seconds.

A bad timing reference existed for at least two minutes.


1.7.5 Holdover Synchronization Alarm

Symptom    The clock is running at a different frequency than normal and the HLDOVRSYNC alarm appears.

Table 1-30 describes the potential cause(s) of the symptom and the solution(s).

Table 1-30 Holdover Synchronization Alarm 

Possible Problem
Solution

The last reference input has failed.

The clock is running at the frequency of the last known-good reference input. This alarm is raised when the last reference input fails. See the "HLDOVRSYNC" section for a detailed description of this alarm.

Note The ONS 15454 SDH supports holdover timing per Telcordia standard GR-4436 when provisioned for external (BITS) timing.


1.7.6 Free-Running Synchronization Mode

Symptom    The clock is running at a different frequency than normal and the FRNGSYNC alarm appears.

Table 1-31 describes the potential cause(s) of the symptom and the solution(s).

Table 1-31 Free-Running Synchronization Mode 

Possible Problem
Solution

No reliable reference input is available.

The clock is using the internal oscillator as its only frequency reference. This occurs when no reliable, prior timing reference is available. See the "FRNGSYNC" section for a detailed description of this alarm.


1.7.7 Daisy-Chained BITS Not Functioning

Symptom    You are unable to daisy-chain the BITS.

Table 1-32 describes the potential cause(s) of the symptom and the solution(s).

Table 1-32 Daisy-Chained BITS Not Functioning 

Possible Problem
Solution

Daisy-chaining BITS is not supported on the ONS 15454 SDH.

Daisy-chaining BITS causes additional wander buildup in the network and is therefore not supported. Instead, use a timing signal generator to create multiple copies of the BITS clock and separately link them to each ONS 15454 SDH.


1.7.8 Blinking STAT LED after Installing a Card

Symptom    After installing a card, the STAT LED blinks continuously for more than 60 seconds.

Table 1-33 describes the potential cause(s) of the symptom and the solution(s).

Table 1-33 Blinking STAT LED on Installed Card 

Possible Problem
Solution

The card cannot boot because it failed the Power On Shelf Test (POST) diagnostics.

The blinking STAT LED indicates that POST diagnostics are being performed. If the LED continues to blink more than 60 seconds, the card has failed the POST diagnostics test and has failed to boot.

If the card has truly failed, an EQPT alarm is raised against the slot number with an "Equipment Failure" description. Check the alarm tab for this alarm to appear for the slot where the card was installed.

To attempt recovery, remove and reinstall the card and observe the card boot process. If the card fails to boot, replace the card.


Caution Removing a card that currently caries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONs 15454 SDH Procedure Guide for information.

1.8 Fiber and Cabling

This section explains problems typically caused by cabling connectivity errors. It also includes instructions for crimping CAT-5 cable and lists the optical fiber connectivity levels.

1.8.1 Bit Errors Appear for a Traffic Card

Symptom    A traffic card has multiple bit errors.

Table 1-34 describes the potential cause(s) of the symptom and the solution(s).

Table 1-34 Bit Errors Appear for a Line Card 

Possible Problem
Solution

Faulty cabling or low optical-line levels.

Bit errors on line (traffic) cards usually originate from cabling problems or low optical-line levels. The errors can be caused by synchronization problems, especially if PJ (pointer justification) errors are reported. Moving cards into different error-free slots will isolate the cause. Use a test set whenever possible because the cause of the errors could be external cabling, fiber, or external equipment connecting to the ONS 15454 SDH. Troubleshoot cabling problems using the "Network Troubleshooting Tests" section. Troubleshoot low optical levels using the "Faulty Fiber-Optic Connections" section.


1.8.2 Faulty Fiber-Optic Connections

Symptom    A line card has multiple SONET alarms and/or signal errors.

Table 1-35 describes the potential cause(s) of the symptom and the solution(s).

Table 1-35 Faulty Fiber-Optic Connections 

Possible Problem
Solution

Faulty fiber-optic connections.

Faulty fiber-optic connections can be the source of SONET alarms and signal errors. See the "Verify Fiber-Optic Connections" procedure.

Faulty CAT-5 cables.

Faulty CAT-5 cables can be the source of SONET alarms and signal errors. See the "Crimp Replacement LAN Cables" section.

Faulty gigabit interface connectors.

Faulty gigabit interface converters can be the source of SONET alarms and signal errors. See the "Replace Faulty GBIC or SFP Connectors" section.



Warning Follow all directions and warning labels when working with optical fibers. To prevent eye damage, never look directly into a fiber or connector. Class IIIb laser. Danger, laser radiation when open. The STM-64 laser is off when the safety key is off (labeled 0). The laser is on when the card is booted and the safety key is in the on position (labeled 1). The port does not have to be in service for the laser to be on. Avoid direct exposure to the beam. Invisible radiation is emitted from the aperture at the end of the fiber optic cable when connected, but not terminated.


Procedure: Verify Fiber-Optic Connections


Step 1 Ensure that a single-mode fiber connects to the ONS 15454 SDH STM-N card.

SM or SM Fiber should be printed on the fiber span cable. ONS 15454 SDH STM-N cards do not use multimode fiber.

Step 2 Ensure that the connector keys on the SC fiber connector are properly aligned and locked.

Step 3 Check that the single-mode fiber power level is within the specified range:

a. Remove the receive (Rx) end of the suspect fiber.

b. Connect the receive end of the suspect fiber to a fiber-optic power meter, such as a GN Nettest LP-5000.

c. Determine the power level of fiber with the fiber-optic power meter.

d. Verify the power meter is set to the appropriate wavelength for the optical card being tested (either 1310 nm or 1550 nm depending on the specific card).

e. Verify that the power level falls within the range specified for the card; see the "Optical Card Transmit and Receive Levels" section.

Step 4 If the power level falls below the specified range:

a. Clean or replace the fiber patch cords. Clean the fiber according to site practice or, if none exists, follow the procedure in the Cisco ONS 15454 SDH Procedure Guide. If possible, do this for the STM-N card you are working on and the far-end card.

b. Clean the optical connectors on the card. Clean the connectors according to site practice or, if none exists, follow the procedure in the Cisco ONS 15454 SDH Procedure Guide. If possible, do this for the STM-N card you are working on and the far-end card.

c. Ensure that the far-end transmitting card is not an ONS intermediate-range (IR) card when an ONS long-range (LR) card is appropriate.

IR cards transmit a lower output power than LR cards.

d. Replace the far-end transmitting STM-N card to eliminate the possibility of a degrading transmitter on this STM-N card.


Caution Removing a card that currently caries traffic on one or more ports can cause a traffic hit. To avoid this, perform an external switch if a switch has not already occurred. Consult the Cisco ONs 15454 SDH Procedure Guide for information.

e. If the power level still falls below the specified range with the replacement fibers and replacement card, check for one of these three factors that attenuate the power level and affect link loss (LL):

Excessive fiber distance; single-mode fiber attenuates at approximately 0.5 dB/km.

Excessive number or fiber connectors; connectors take approximately 0.5 dB each.

Excessive number of fiber splices; splices take approximately 0.5 dB each.


Caution These are typical attenuation values. Refer to the specific product documentation for the actual values or use an optical time domain reflectometer (OTDR) to establish precise link loss and budget requirements.

Step 5 If no power level shows on the fiber, the fiber is bad or the transmitter on the optical card failed.

a. Check that the Tx and Rx fibers are not reversed. LOS and EOC alarms normally accompany reversed Tx and Rx fibers. Switching reversed Tx and Rx fibers clears the alarms and restores the signal.

b. Clean or replace the fiber patch cords. Clean the fiber according to site practice or, if none exists, follow the procedure in the Cisco ONS 15454 SDH Procedure Guide. If possible, do this for the STM-N card you are working on and the far-end card.

c. Retest the fiber power level.

d. If the replacement fiber still shows no power, replace the optical card.

Step 6 If the power level on the fiber is above the range specified for the card, ensure that an ONS LR card is not being used when an ONS IR card is appropriate.

LR cards transmit a higher output power than IR cards. When used with short runs of fiber, an LR transmitter is be too powerful for the receiver on the receiving STM-N card.

Receiver overloads occur when maximum receiver power is exceeded.


Tip To prevent overloading the receiver, use an attenuator on the fiber between the ONS STM-N card transmitter and the receiver. Place the attenuator on the receive transmitter of the ONS STM-N cards. Refer to the attenuator documentation for specific instructions.



Tip Most fiber has text printed on only one of the two fiber strands. Use this to identify which fiber is connected to Tx and which fiber is connected to Rx.



1.8.2.1 Crimp Replacement LAN Cables

You can crimp your own LAN cables for use with the ONS 15454 SDH. Use a cross-over cable when connecting an ONS 15454 SDH to a hub, LAN modem, or switch, and use a LAN cable when connecting an ONS 15454 SDH to a router or workstation. Use CAT-5 cable RJ-45 T-568B, Color Code (100 Mbps), and a crimping tool. Figure 1-26 shows the layout of an RJ-45 connector.

Figure 1-26 RJ-45 Pin Numbers

Figure 1-27 and Table 1-36 shows a LAN cable layout and pinouts.

Figure 1-27 LAN Cable Layout

Table 1-36 Lan Cable Pinout 

Pin
Color
Pair
Name
Pin

1

white/orange

2

Transmit Data +

1

2

orange

2

Transmit Data -

2

3

white/green

3

Receive Data +

3

4

blue

1

 

4

5

white/blue

1

 

5

6

green

3

Receive Data -

6

7

white/brown

4

 

7

8

brown

4

 

8


Figure 1-28 and Table 1-37 shows a cross-over cable layout and pinouts.

Figure 1-28 Cross-Over Cable Layout

Table 1-37 Cross-Over Cable Pinout 

Pin
Color
Pair
Name
Pin

1

white/orange

2

Transmit Data +

3

2

orange

2

Transmit Data -

6

3

white/green

3

Receive Data +

1

4

blue

1

 

4

5

white/blue

1

 

5

6

green

3

Receive Data -

2

7

white/brown

4

 

7

8

brown

4

 

8



Note Odd-numbered pins always connect to a white wire with a colored stripe.


1.8.2.2 Replace Faulty GBIC or SFP Connectors

GBICs and SFPs are hot-swappable and can be installed or removed while the card or shelf assembly is powered and running.


Warning GBICs are Class I laser products. These products have been tested and comply with Class I limits.



Warning Invisible laser radiation may be emitted from the aperture ports of the single-mode fiber optic modules when no cable is connected. Avoid exposure and do not stare into open apertures.


GBICs and SFPs are input/output devices that plug into a Gigabit Ethernet card to link the port with the fiber-optic network. The type of GBIC or SFP determines the maximum distance that the Ethernet traffic can travel from the card to the next network device. For a description of GBICs and SFPs and their capabilities, see Table 1-38, Table 1-39, and refer to the Cisco ONS 15454 SDH Reference Manual.


Note GBICs and SFPs must be matched on either end by type: SX to SX, LX to LX, or ZX to ZX.


GBICs are available in two different models. One GBIC model has two clips (one on each side of the GBIC) that secure the GBIC in the slot on the E1000-2-G, G1000-4, or G1K-4 card. The other model has a locking handle. Both models are shown in Figure 1-29.

Figure 1-29 Gigabit Interface Converters

Table 1-38 shows the available GBICs.


Note The GBICs are very similar in appearance. Check the GBIC label carefully before installing it.


Table 1-38 Available GBICs 

GBIC
Associated Cards
Application
Fiber
Product Number

1000BaseSX

E1000-2-G
G1000-4
G1K-4

Short reach

Multimode fiber up to 550 m long

15454E-GBIC-SX=

1000BaseLX

E1000-2-G
G1000-4
G1K-4

Long reach

Single-mode fiber up to 5 km long

15454E-GBIC-LX=

1000BaseZX

G1000-4
G1K-4

Extra long reach

Single-mode fiber up to 70 km long

15454E-GBIC-ZX=


Table 1-39 shows the available SFPs.

Table 1-39 Available SFPs 

SFP
Associated Cards
Application
Fiber
Product Number

1000BaseSX

ML1000-2

Short reach

Multimode fiber up to 550 m long

15454E-SFP -LC-SX=

1000BaseLX

ML1000-2

Long reach

Single-mode fiber up to 5 km long

15454E-SFP -LC-LX=


Procedure: Remove GBIC or SFP Connectors


Step 1 Disconnect the network fiber cable from the GBIC SC connector or SFP LC duplex connector.


Warning Invisible laser radiation may be emitted from disconnected fibers or connectors. Do not stare into beams or view directly with optical instruments.

Step 2 Release the GBIC or SFP from the slot by simultaneously squeezing the two plastic tabs on each side.

Step 3 Slide the GBIC or SFP out of the Gigabit Ethernet module slot. A flap closes over the GBIC or SFP slot to protect the connector on the Gigabit Ethernet card.

Step 4 To replace an SFP, see the "Replace the Small-Formfactor Pluggable Connector" section


Procedure: Installing a GBIC with Clips


Step 1 Remove the GBIC from its protective packaging.

Step 2 Check the label to verify that the GBIC is the correct type (SX, LX, or ZX) for your network.

Step 3 Verify that you are installing compatible GBICs; for example, SX to SX, LX to LX, or ZX to ZX.

Step 4 Grip the sides of the GBIC with your thumb and forefinger and insert the GBIC into the slot on the E1000-2, E1000-2-G, G1000-4, or G1K-4 card (shown in Figure 1-30).


Note GBICs are keyed to prevent incorrect installation.


Figure 1-30 GBIC Installation (with Clips)

Step 5 Slide the GBIC through the flap that covers the opening until you hear a click. The click indicates the GBIC is locked into the slot.

Step 6 When you are ready to attach the network fiber-optic cable, remove the protective plug from the GBIC and save the plug for future use.

Step 7 Return to your originating procedure (NTP).


Procedure: Installing a GBIC with a Handle


Step 1 Remove the GBIC from its protective packaging.

Step 2 Check the label to verify that the GBIC is the correct type (SX, LX, or ZX) for your network.

Step 3 Verify that you are installing compatible GBICs; for example, SX to SX, LX to LX, or ZX to ZX.

Step 4 Remove the protective plug from the SC-type connector.

Step 5 Grip the sides of the GBIC with your thumb and forefinger and insert the GBIC into the slot on the E1000-2-G, G1000-4, or G1K-4 card.


Note GBICs are keyed to prevent incorrect installation.


Step 6 Lock the GBIC into place by closing the handle down. The handle is in the correct closed position when it does not obstruct access to SC-type connector.

Step 7 Return to your originating procedure (NTP).


1.8.3 Optical Card Transmit and Receive Levels

Each STM-N card has a transmit and receive connector on its faceplate.

Table 1-40 Optical Card Transmit and Receive Levels 

Optical Card
Receive
Transmit

OC3 IR 4/STM1SH 1310

-28 to -8 dBm

-15 to -8 dBm

OC3 IR/STM 1SH 1310-8

-30 to -8 dBm

-15 to -8 dBm

OC12 IR/STM4 SH 1310

-28 to -8 dBm

-15 to -8 dBm

OC12 LR/STM4 LH 1310

-28 to -8 dBm

-3 to +2 dBm

OC12 LR/STM4 LH 1550

-28 to -8 dBm

-3 to +2 dBm

OC12 IR/STM4 SH 1310-4

-28 to -8 dBm

-3 to +2 dBm

OC48 IR/STM16 SH AS 1310

-18 to 0 dBm

-5 to 0 dBm

OC48 LR/STM16 LH AS 1550

-28 to -8 dBm

-2 to +3 dBm

OC48 ELR/STM16 EH 100GHz

-28 to -8 dBm

-2 to 0 dBm

OC192 SR/STM64 IO 1310

-11 to -1 dBm

-6 to -1 dBm

OC192 IR STM64 SH 1550

-14 to -1 dBm

-1 to +2 dBm

OC192 LR/STM64 LH 1550

-21 to -9 dBm

+7 to +10 dBm

OC192 LR/STM64 LH ITU 15xx.xx

-22 to -9 dBm

+3 to +6 dBm

TXP-MR-10G

Trunk side:
Client side:

-26 to -8 dBm
-14 to -1 dBm

-16 to +3 dBm
-6 to -1 dBm

MXP-2.5G-10G

Trunk side:
Client side:

-26 to -8 dBm
depends on SFP

-16 to +3 dBm
depends on SFP


1.9 Power and LED Tests

This section provides symptoms and solutions for power supply problems, power consumption, and LED indicators.

1.9.1 Power Supply Problems

Symptom    Loss of power or low voltage, resulting in a loss of traffic and causing the LCD clock to reset to the default date and time.

Table 1-41 describes the potential cause(s) of the symptom and the solution(s).

Table 1-41 Power Supply Problems

Possible Problem
Solution

Loss of power or low voltage.

The ONS 15454 SDH requires a constant source of DC power to properly function. Input power is -48 VDC. Power requirements range from -42 VDC to -57 VDC.

A newly installed ONS 15454 SDH that is not properly connected to its power supply does not operate. Power problems can be confined to a specific ONS 15454 SDH or affect several pieces of equipment on the site.

A loss of power or low voltage can result in a loss of traffic and causes the LCD clock on the ONS 15454 SDH to default to January 1, 1970, 00:04:15. To reset the clock, in node view click the Provisioning > General tabs and change the Date and Time fields.

See the "Isolate the Cause of Power Supply Problems" procedure.

Improperly connected power supply.



Warning When working with live power, always use proper tools and eye protection.



Caution Always use the supplied electrostatic discharge (ESD) wristband when working with a powered ONS 15454 SDH. Plug the wristband cable into the ESD jack located on the lower-right outside edge of the shelf assembly.


Caution Operations that interrupt power supply or short the power connections to the ONS 15454 SDH are service-affecting.

Procedure: Isolate the Cause of Power Supply Problems


Step 1 If a single ONS 15454 SDH show signs of fluctuating power or power loss:

a. Verify that the -48 VDC #8 power terminals are properly connected to a fuse panel. These power terminals are located on the lower section of the backplane FMEC card under the clear plastic cover.

b. Verify that the power cable is #12 or #14 AWG and in good condition.

c. Verify that the power cable connections are properly crimped. Stranded #12 or #14 AWG does not always crimp properly with Staycon type connectors.

d. Verify that 20 A fuses are used in the fuse panel.

e. Verify that the fuses are not blown.

f. Verify that a rack-ground cable attaches to the frame-ground terminal (FGND) on the ONS 15454 SDH FMEC. Connect this cable to the ground terminal according to local site practice.

g. Verify that the DC power source has enough capacity to carry the power load.

h. If the DC power source is battery-based:

Check that the output power is high enough. Power requirements range from -42 VDC to -57 VDC.

Check the age of the batteries. Battery performance decreases with age.

Check for opens and shorts in batteries, which might affect power output.

If brownouts occur, the power load and fuses might be too high for the battery plant.

Step 2 If multiple pieces of site equipment show signs of fluctuating power or power loss:

a. Check the uninterruptible power supply (UPS) or rectifiers that supply the equipment. Refer to the UPS manufacturer's documentation for specific instructions.

b. Check for excessive power drains caused by other equipment, such as generators.

c. Check for excessive power demand on backup power systems or batteries, when alternate power sources are used.


1.9.2 Power Consumption for Node and Cards

Symptom    You are unable to power up a node or the cards in a node.

Table 1-42 describes the potential cause(s) of the symptom and the solution(s).

Table 1-42 Power Consumption for Node and Cards 

Possible Problem
Solution

Improper power supply

Refer to power information in the Cisco ONS 15454 SDH Procedure Guide.


1.9.3 Lamp Test for Card LEDs

Symptom    Card LED does not light or you are unsure if LEDs are working properly.

Table 1-43 describes the potential cause(s) of the symptom and the solution(s).

Table 1-43 Lamp Test for Card LEDs 

Possible Problem
Solution

Faulty LED

A lamp test verifies that all the card LEDs work. Run this diagnostic test as part of the initial ONS 15454 SDH turn-up, a periodic maintenance routine, or any time you question whether an LED is in working order.

See the "Verify Card LED Operation" procedure.


Procedure: Verify Card LED Operation


Step 1 Click the Maintenance > Diagnostic tabs.

Step 2 Click Lamp Test.

Step 3 Watch to make sure all the LEDs on the cards illuminate for several seconds.

Step 4 Click OK on the Lamp Test Run dialog box.

If an LED does not light up, the LED is faulty. Call the Cisco TAC and fill out an RMA to return the card.