Description of the problem

BUS_ERR, HSC_UNAVAIL alarms are reported on OSN7500 device, 10-UXCSA primary, and standby 9-UXCSA veneer;

Business is normal.

Board version 9-UXCSA: 8.13; 10-UXCSA: 8.13;

Mainframe version: 5.21.18.50

9-UXCSA single board alarm parameters:

1. HSC_UNAVAIL 0X04 0X01 0X09 0XFF 0XFF

2.BUS_ERR 0X04 0X03 0X13 0X02 0XFF

3.BUS_ERR 0X04 0X04 0X13 0X02 0XFF

Processing

1, detailed query BUS_ERR alarm parameters: para1 indicates the logical board position, in this case para1 value of 0x04, cross board monitoring 4-SLO1 board position to their own business bus there are problems; para2 indicates that the board position in the bus serial number in this case, the value of 0x03/04; para3 value of different bit for the existence of different alarms state, the return value will be converted to hexadecimal number and then query against; if para4 = 0x01, it means that different alarm exists, the return value of 0x01 is 0x02 0XFF; if para4 = 0x01, it means that the alarm exists. The return value will be converted to hexadecimal numbers and then query against; if para4 = 0x01, it means that the cross board to detect the business board over the bus, if para4 = 0x02, it means that the two cross board handshake detected.

2, according to the above, confirmed for the 4-SLO1 signal sent to the cross-board, the two cross-board handshake information has a problem.

Root cause

The main crossover board is not plugged in properly.

Solution

1. Replace the 9-UXCSA, the problem remains;

2. Replace 4-SLO1, the problem remains;

3, suspect that the main crossover has a problem, but due to HSC_UNAVAIL, the crossover can not be network management inversion, hard switching is a risk;

4, after consulting the company's R & D, for SLO1 single board, only 2 buses, that is, if it is really SLO1 or standby cross board failure, bus_err alarm parameter of the second bit should be 0X01 or 0X02, this time for the 0X023/OX04, for the bus is not in use, it should be a cross false alarms, you can force the switching, does not affect the business;

5, after forcibly plugging 10-UXCSA, cross inverted to 9-UXCSA, the alarm disappeared, observed half a day disappeared not reproduced; will be manually switched back to the veneer 10-UXCSA, the problem does not reappear.

Suggestions and Summary

1, BUS_ERR is a common alarm for NG SDH, the problem may be a cross board to receive the bus, may also be a business slot to send the bus failure, the actual positioning can not be accurately located to the specific board, it is best to bring all the boards to solve the problem at once;

2, this failure is relatively rare, the main backup cross board handshake signal failure, and the main failure, replace the backup cross can not solve the problem. However, due to the existence of alarms on the spare, it is not possible to softly invert the crossover to try to locate the problem, and it is necessary to hard-plug the main crossover in order to trigger the inversion and locate the problem.