Friday, July 8, 2016

Mismatching Between MA Groups and Associated VCTRUNKs, Hence Causing the Binding Failure

IMA groups fail to be bound with associated VCTRUNKs. You need to configure IMA groups to match with associated VCTRUNKs.

Fault Type

Configuration_Problem

Symptom

The principle of binding VCTRUNKs is met, but VCTRUNKS fail to be bound.

Cause Analysis

Mutual check relation exists between IMA groups and associated VCTRUNKs.

Procedure

  1. If IMA groups are symmetrically configured, make sure that the mapping VCTRUNKs are symmetrical.
  2. If VCTRUNKs are nonsymmetrical, make sure that the mapping IMA groups are nonsymmetrical.

Services Are Interrupted After the Protection Hold-Off Time Is Set to 10s During Configuration of Client-Side 1+1 Protection on the OptiX OSN 6800, The System Reports The R_LOS Alarm

During configuration of client-side 1+1 protection on the OptiX OSN 6800, Working Channel Hold-Off Time (100ms) is set to 100. In this case, the protection hold-off time is too long and services are interrupted, the system reports the R_LOS alarm.

Fault Type

Protection
R_LOS
CLIENT_PORT_PS

Symptom

After the fiber cable on the main optical path of the OptiX OSN 6800 in a network is broken, certain services in the IP bearer network are interrupted for 40 seconds.
Query the alarms of the OptiX OSN 6800 on the T2000. According to the query result, an R_LOS alarm is reported and a protection switching alarm is reported 10s after the R_LOS is reported.

Cause Analysis

The possible causes of the preceding problem are as follows:
  • Protection switching on the DCP board is abnormal.
  • There are excessive bit errors and therefore the performance of the optical path is degraded. In addition, the SD bit errors are not enabled as a trigger condition of switching.
  • The protection hold-off time is inappropriate.

Procedure

  1. Query alarms on the NE by using the T2000. It is found that the R_LOS and CLIENT_PORT_PS alarms are reported.
  2. Query the attributes of client-side 1+1 protection on the NE. It is found that the protection hold-off time is set at the transmit and receive optical interfaces in the east direction between the NEs where the services are interrupted. Protection switching at the transmit and receive optical interfaces in the west direction is normal.
  3. Query the value of Working Channel Hold-Off Time (100ms) on the T2000. The value is 100. It is confirmed that the unit for setting the protection hold-off time is 100 ms. In this case, the value 100 indicates that the actual protection hold-off time is 10s. That is, protection switching is performed 10s after a switching-triggering alarm, such as the R_LOS alarm, is received. As a result, services are interrupted.
  4. Set Working Channel Hold-Off Time (100ms) to 1. Then the protection switching is normal.

Result

The problem is resolved.

Thursday, July 7, 2016

Service Interruption After Enabling the Shaping Function of the N1EMS4 Board

After the shaping function of the N1EMS4 board is enabled for egress port queues, the services are interrupted.

Fault Type

  • RPR
  • Ethernet fault
Configuration problem

Symptom

Figure 1 shows the VCTRUNK-shared EVPL service between NE2 and NE3.
Figure 1 Configuration of the VCTRUNK-shared EVPL service
Two pairs of users (D and D', and E and E') share the 300 Mbit/s bandwidth of VCTRUNK1.
  1. Create a flow of the Port type for PORT1 and PORT2 separately.
  2. Create a CoS of the simpletype. Groom the port flow of PORT1 to queue 1 of VCTRUNK1, and groom the port flow of PORT2 to queue 3 of VCTRUNK1.
  3. Enable the traffic shaping function for queue 3 of VCTRUNK1. Set the CIR to 300 Mbit/s.
  4. Enable the traffic shaping function for queue 1 of VCTRUNK1. Use the default shaping parameter values.
Now if the traffic that enters PORT2 reaches a rate of 300 Mbit/s, the service of PORT1 is interrupted. Hence, although certain remaining bandwidth is available, it cannot allocated to queue 1.

Cause Analysis

This problem has the following possible causes.
  1. In this fault case, the priority of queue 3 is higher than the priority of queue 1. As a result, the packets of the higher priority queue 3 occupy the entire 300 Mbit/s bandwidth, and there is no remaining bandwidth for the lower priority queue 1. Consequently, the service between D and D' is interrupted.
  2. The shaping function is enabled for queue 1, but the shaping bandwidth is not set. As the default shaping bandwidth is 0, the service is interrupted.

Procedure

  1. Cause 1: The packets in queue 3 of higher priority occupy all the available bandwidth.
    1. Decrease the bandwidth of the higher priority queue 3.
      1. On the Main Topology, right-click the NE icon and choose NE Explorer from the shortcut menu.
      2. Select the relevant Ethernet board in the Object Tree. Choose Configuration > QoS Management > Port Shaping Management from the Function Tree.
      3. Set the committed information rate (CIR) and burst size (less than 300 Mbit/s).
  2. Cause 2: The shaping function is enabled for queue 1, but the committed bandwidth (CIR) of the shaping function is not set.
    1. Set the shaping bandwidth for queue 1.
      1. On the Main Topology, right-click the NE icon and choose NE Explorer from the shortcut menu.
      2. Select the relevant Ethernet board in the Object Tree. Choose Configuration > QoS Management > Port Shaping Management from the Function Tree.
      3. Select queue 1, and then set the relevant parameters for the queue.

Similar Problems

After a flow is bound with a CAR, the service is interrupted. The possible cause is that, although the CAR is enabled, the CIR and PIR (which are both 0 by default) are not set for the CAR.



How to Change the Link Type of an Interface?

Failure in Restricting the Bandwidth After Binding the CAR to the Flows

A flow is created after the relevant Ethernet service is created. After the flow is bound with a CAR, it is found that the CAR does not take effect. (On the instrument, the transmit bandwidth is equal to the receive bandwidth.)

Fault Type

  • QoS
  • Ethernet fault

Symptom

As shown in Figure 1, a unidirectional point-to-point EPL service is configured from PORT1 to PORT2 on the same board.
Figure 1 Unidirectional point-to-point EPL service from PORT1 to PORT2
Create the flow at PORT1. Set the flow type to port+VLAN+priority.
Configure the CAR. Set both the CIR and PIR to 10 Mbit/s. Bind the CAR to the flow. Use the SmartBits to transmit packets at a rate higher than 10 Mbit/s. The receive bandwidth of the instrument, however, is found to be equal to the transmit bandwidth of the instrument.

Cause Analysis

This problem has the following possible causes.
  1. The CAR is not enabled. After the CAR is bound to the flow, the CAR is effective for the flow only if the CAR is enabled.
  2. The flow type does not match the service type. In this fault case, a port flow should be configured for the point-to-point EPL service.
  3. Certain boards do not support the flow of the Port type, because the transmitted service packets probably do not match the specified VLAN+priority. Hence, the CAR does not take effect on the flow.

Procedure

  1. Cause 1: The CAR is not enabled.
    1. Enable the CAR.
      1. On the Main Topology, right-click the NE icon and choose NE Explorer from the shortcut menu.
      2. Select the relevant Ethernet board in the Object Tree. Choose Configuration > QoS Management > Flow Management from the Function Tree.
      3. Click the CAR Configuration tab.
      4. Set Enable/Disable to Enable.
  2. Cause 2: The flow type does not match the service type.
    Re-create a flow of the type (that is, port flow) that matches the service type. Then bind the flow with a CAR.
    1. Re-create a flow.
      1. On the Main Topology, right-click the NE icon and choose NE Explorer from the shortcut menu.
      2. Select the relevant Ethernet board in the Object Tree. Choose Configuration > QoS Management > Flow Management from the Function Tree. Then click .
      3. Click the Flow Configuration tab.
      4. Click New. Set the flow attributes in the New Flow dialog box, according to the completed flow planning and the description of flow parameters.
      5. Click OK. The Operation Result dialog box is displayed, indicating that the flow is successfully created.
    2. Bind the flow with a CAR.
      1. On the Main Topology, right-click the NE icon and choose NE Explorer from the shortcut menu.
      2. Select the relevant Ethernet board in the Object Tree. Choose Configuration > QoS Management > Flow Management from the Function Tree. Then click .
      3. Click the Flow Configuration tab.
      4. Select a flow for which you want to configure the binding. In the Bound Shaping, Bound CoS and Bound CAR tabs, select the Shaping ID, CoS ID and CAR ID respectively.
      5. Click Apply. The Operation Result dialog box is displayed, indicating that the binding is successful.
  3. Cause 3: The transmitted service packets do not match the specified VLAN+priority.
    1. Select the relevant board from the NE Explorer. Choose Configuration > Ethernet Service > Ethernet LAN Service from the Function Tree.
    2. Check whether each VLAN priority matches the service packets. If not, set the VLAN priority to the other value.




The N2SLQ16 Board Interconnected with the TQS Board Reports an R_LOS Alarm Because The TQS Board Emits White Lights When the Input Optical Power Is Normal

When the N2SLQ16 board on an OptiX OSN 7500 NE is interconnected with the TQS board on an OptiX OSN 6800 NE, the N2SLQ16 board reports an R_LOS alarm because the TQS board emits white lights when the input optical power is normal.

Fault Type

Equipment Interconnection
Client Equipment
R_LOS
R_LOF

Symptom

When the N2SLQ16 board on an OptiX OSN 7500 NE is interconnected with the TQS board on an OptiX OSN 6800 NE at an office and no service is deployed on the TQS board, the N2SLQ16 board reports an R_LOS alarm when the optical power is normal. After a software outloop is performed on the TQS board, the alarm reported on the N2SLQ16 board is changed to R_LOF.

Cause Analysis

No service is deployed on the TQS board. Therefore, an R_LOF not an R_LOS alarm should be reported when the TQS board is interconnected with the N2SLQ16 board and when the optical power is normal. Since the N2SLQ16 is not faulty, the possible cause of the R_LOS alarm is that signals are lost rather than that the optical power is abnormal. That is, the N2SLQ16 board may receive white lights.
Under the condition that signals are scrambled but not framed on the TQS board, the initial values of the output signals on the WDM side of the TQS board are all zero. At this time, the optical module on the TQS board outputs white lights. When a loopback is performed on the TQS board, that is, signals on the optical module are framed and scrambled for initialization, the TQS no longer outputs white lights but random signal lights. As a result, the downstream N2SLQ16 board does not report the R_LOS alarm.

Procedure

  1. Query the alarm mode of the OptiX OSN 6800 NE on the T2000. The result shows that the alarm mode is set to automatic reversion for the NE but the automatic reversion function is not enabled at the optical interface.
  2. Replace the N2SLQ16 board or the optical module on the board but the problem persists. Perform a hardware loopback at the optical interfaces on the N2SLQ16 board by using an optical fiber. The R_LOS alarm disappears. The possibility that the N2SLQ16 board or the optical module on the board is faulty is excluded.
  3. Analyze that the TQS board interconnected with the N2SLQ16 board may be faulty. That is, the lights transmitted by the TQS board may be white, and therefore signals on the N2SLQ16 board are lost. On the T2000, set the client-side optical interfaces on the TQS board to Outloop. At this time, the alarm on the N2SLQ16 board is changed from R_LOS to R_LOF. The fault is rectified.

Result

The problem is resolved.

Wednesday, July 6, 2016

How Signals Are Selected When the DCP Board Is Used

This section describes how signals are selected when the DCP board is used.

Fault Type

Optical Protection Unit

Symptom

None.

Cause Analysis

When the DCP board is not used for protection but physically connect to an OTU board through the RI11 and RI12 optical interfaces to receive the client-side signals on the OTU board, the difference between the receive optical power values at the two optical interfaces is small and the receive optical power values are in the normal range. In this case, you need to observe the following points.
  • When the DCP board is running:
    • Because protection is not configured, the DCP board does not perform protection switching regardless of the channel status. That is, the system receives signals from the RI11 or RI12 optical interface even when no signal is available at the RI11 or RI12 optical interface.
  • When the DCP board is powered off:
    • The system always receives signals from the RI12 optical interface no matter whether signals are available at the RI11 or RI12 optical interface.
When the DCP board is used for configuring Inter-subrack client 1+1 protection or Intra-subrack client 1+1 protection and connects to an OTU board through the RI11 and RI12 optical interfaces to receive the client-side signals on the OTU board, the difference between the receive optical power values at the two optical interfaces is small and the receive optical power values are in the normal range. In this case, you need to observe the following points.
  • When the DCP board is running:
    • With protection configured, the system selectively receives signals on the working channel by default, that is, the system receives signals from the RI11 optical interface.
    • When no signal is available at the RI11 optical interface, the system selectively receives signals from the RI12 optical interface (protection channel).
    • When no signal is available at the RI12 optical interface, the system selectively receives signals from the RI11 optical interface (working channel).
  • When the DCP board is powered off:
    • The system selectively receives signals from the RI12 optical interface no matter whether signals are available at the RI11 or RI12 optical interface.

Procedure

  1. None.

Result

The problem is resolved.

40G Wavelengths Are Unavailable Due to Inappropriate PMD

40G wavelengths are unavailable due to inappropriate polarization mode dispersion (PMD).

Product

OptiX BWS 1600G, OSN6800,OSN8800

Fault Type

PMD Abnormity

Symptom

An OptiX BWS 1600G network transmits both 10G and 40G wavelengths. The network diagram is shown in the following figure. 40G wavelengths are transmitted from station A to station C. They pass through station B. 10G wavelengths are added and dropped at station B. 40G wavelengths in two directions between station A and station C are unavailable. Performance of the 10G wavelengths between station A and station C and those between station A and station B fluctuates and the BER before correction exceeds two orders of magnitude; however, performance of the wavelengths between station B and station C is normal.

Cause Analysis

The possible causes of the performance fluctuation of the wavelengths are as follows:
  • The PMD is excessively high.
  • The optical amplifier boards work unstably.
The receive optical power on the OTU board at station C is stable. Therefore, the possibility that the optical amplifier boards work unstably is excluded. The measured PMD on the optical fibers during network design is low. In addition, the optical fibers are not rerouted. Therefore, it is preliminarily determined that the PMD is inappropriate.

Procedure

  1. Perform self-loops on the 40G OTU boards at stations A and C. The result is normal. This indicates that the 40G OTU boards work normally. The BERs of the 10G OTU boards fluctuate but there is only a slight change in the receive optical power. It is concluded that both the 10G and 40G OTU boards work normally.
  2. Check the total optical power and the optical power of each wavelength on each optical amplifier board along the links. The result is normal and the OSNR at the receive end reaches the specified value. This indicates that the links in the network are normal.
  3. Add a 40G board at station B. That is, add a 40G wavelength from station A to station B and from station B to station C each. The 40G wavelength from station A to station B is still unavailable but that from station B to station C is normal. It is concluded that the link from station A to station B is abnormal.
  4. Check the DCM modules at each station on the link from station A to station B. All DCM modules are properly connected. Therefore, it is determined that the PMD may be inappropriate.
  5. Measure the PMD span by span from station A to station B. It is found that the PMD of the optical fibers in the transmit and receive directions in a span is much higher than the design value. After the optical fibers in the span are replaced, the bidirectional 40G wavelengths between station A and station C are available.

Result

The problem is resolved.