Description of the problem

On a certain day, the customer feedback ICBC head office to a branch of the large customer business packet loss, PING packet packet length of 15,000 bytes, packet loss rate of more than 1%.

Network Topology

The head office site of ICBC uses Huawei OSN1500 equipment, and the other sites use OSN3500, version 18.50P01.

The OSN1500 at the head office site of ICBC is connected to the MAP router at the head office of ICBC through the EGS4 single board. The EGS4 single board is configured with EVPL to aggregate the services of each outlet, and the bandwidth of each outlet is 2*VC12, and the intermediate link is Huawei's SDH equipment, and the networking mode is MSP and SNCP, and the terminal link is connected to Huawei's equipment through the STM-1/4 fiber optic port by using the R-factory 3500E equipment, and the outlet is configured with box equipment such as R-factory 3107, etc., and connected to Huawei equipment through the FE electric port. Configuration R manufacturer 3107 and other box equipment through the FE electrical port and the outlet MaiPu router docking, the connection relationship is as follows:

ICBC head office Maipu router-(EGS4)ICBC OSN1500 (SL16)-Transmission L-Transmission C-Transmission E(SLQ1)-(STM-1)R manufacturer 3500E(STM- 1)-R Manufacturer 3107(FE)-Outlet Mapper Router

The ICBC OSN1500 equipment is located in the ICBC head office, the Transmission L is located in the local network subnet, the Transmission C and Transmission E are located in the MAN subnet, the R Manufacturer 3500E equipment is in the same room as the Transmission E equipment, and the R Manufacturer 3107 is located in the outlets.

There are no abnormal alarm messages.

Processing

1、Link error code leads to packet loss

2、Docking router port failure

3、Insufficient bandwidth

4, R manufacturers outlets equipment or aggregation EGS4 single board failure

5, Ethernet encapsulation docking problems

6、Other network configuration

Root cause

Huawei OSN1500 device clock configuration is not synchronized

Solution

1, through the network management query performance, did not find that the link exists in the error code, and the same route other services are normal, to rule out the error code causes packet loss.

2、Configure a route with the same, different Ethernet ports at both ends of the business for testing, packet loss rate is still above 1%, can be ruled out ICBC router failure.

3, increase the bandwidth to 4 * VC12, after testing the packet loss rate is still more than 1%, so we can rule out the bandwidth shortage caused by packet loss.

4, adjust the EGS4 single board Ethernet configuration, and other normal operation of the business alignment binding time slot and VC trunk, after testing still packet loss, using the same type of equipment to replace the R manufacturers of business outlets equipment, after testing still packet loss, the initial suspicion that the equipment docking mismatch problem.

5, in and R manufacturers docking transmission E equipment to increase EFS4 Ethernet single board, test EFS4 and R manufacturers equipment docking situation, respectively, configure 1 port to the head office of ICBC EPL services, 2 port to the R manufacturers equipment EPL services, using segmentation test, the test results for the two segments are not packet loss phenomenon, according to the surface phenomenon of the preliminary judgment to use the EFS4 and the R manufacturers docking is not a problem, the use of EGS4 docking there are packet loss. EGS4 docking there is a packet loss, but because the EGS4 single board has been reconciled before the VC trunk, and other business outlets are configured with the same equipment type, and can be reversed to prove that there is no problem with the equipment docking.

6, due to the customer's network has occurred before the R manufacturers of equipment clock module batch problems caused by the failure, the customer requested R manufacturers R & D on-site support for the analysis of the problem, and from the headquarters of the other models of equipment, the use of new equipment to replace the packet loss after the test, at this time it can be determined that the R manufacturers of new and old equipment to deal with the docking data there are differences in order to completely locate the root cause, the configuration of a line from the transmission of the L to the transmission of the E VC12 level of service, under the transmission of E equipment, an idle optical port, R manufacturers of R & D using the instrument to test the optical port, found that there is a pointer to adjust the test 10 minutes to adjust the count of 1000, R manufacturers of R & D replied that the value of the value has exceeded the R manufacturers of the old equipment processing capacity, so it will lead to the equipment to lose packets, in addition to the transmission of the E site query the 15-minute performance of this service, the pointer to adjust the count is around 1300 blocks, so the existence of pointer adjustment is confirmed. Since the cross-subnet service is relatively small, and other services are only penetrated from the MAN, it is not affected.

7. Query the Huawei equipment clock configuration, the transmission L equipment is located in the local network, and the clock source is the clock outside the BITS, and the transmission C and transmission E equipment are located in the MAN, and the clock source is the internal clock of transmission C. The clock source levels of the two subnets are inconsistent, change the clock configuration so that transmission C and transmission E track transmission L, test the end of pointer adjustment through the instrument, retest the original service, and test for six hours without finding No packet loss was found in 6 hours of testing, and the problem of packet loss was finally solved.

Recommendations and Summary

Based on different subnet management of the transmission network, generally only focus on the clock configuration of the same subnet, no ring, follow each other, there is protection can be, but the clock level between the subnets is often ignored, this case is different subnets with different clock levels caused by the loss of packets, it is recommended that each subnet is configured with a BITS clock source, which can be realized by tracking the clock source of the same level for each network element.

This chapter of related technical information and SDH equipment troubleshooting process is collected and organized by Shenzhen Optical Transmission Network Technology Company Limited ( www.opticaltrans.com), reproduced please retain! Our company specializes in the sale of Huawei SDH optical transmission equipment, SDH transmission equipment.