Hi there!
This post will be sharing a case about the SCC board fails to be reset due to large DCN.
Problem Description
One day, the OSN7500 (gateway NE) of a site was remotely upgraded. The activation failed. The system control board was reset repeatedly in the BIOS state. The hard insertion and removal of the board were invalid. After the network cable of the 7500 is removed, the system control board is reset successfully. Device version: V100R8C02SPC200 The NMS displays a message indicating that the activation fails. When you log in to the NMS through the CLI, a message is displayed, indicating that the NE software is in the BIOS state. In this case, loading the software through the CLI fails.
Process
1. The SCC board successfully loads the software and fails to be activated. The possible cause is that the SCC board is faulty.
2. During the activation of the system control board, other boards go online and offline. As a result, the system control board fails to be reset.
3. The system control board restarts repeatedly in the BIOS state. The possible cause is that the loaded software data is abnormal.
4. When the system control board is restarted in the BIOS state, it may be attacked by DCN packets.
5. When the DCN is too large, a large amount of ECC information is filled with the message queue of the main control board.
Root cause
The DCN scale is too large, and the mailbox communication message queue overflows.
solution
1. When the activation fails, the indicator shows that the main control board restarts repeatedly to determine whether the board is faulty. After the board is replaced and then activated again, the system control board is still in the BIOS state. Insert the board into another 7500 devices. It is found that the board is upgraded successfully. Therefore, the fault is not caused by the board.
2. The new system control board is inserted into the slot and is not upgraded. Only a cold reset is performed on the board. It is found that the system control board still cannot start normally. It is determined that DCN packets may be attacked. After the network cable of the 7500 is removed, the board is reset successfully. After multiple tests, it is found that the system control board cannot be reset after the network cable is connected. Therefore, it is determined that the fault is related to the DCN.
3. The length of the message queue for initializing the mailbox communication in the BIOS is 128. Assume that the BIOS does not process any message. If the preceding function returns a failure, that is, the message queue is full of 128 messages, a new message enters the system and the board is reset.
4. This problem occurs when the ECC is too large on the live network. In R10C03, this problem is solved.
Suggestions and Summary
If a GNE is connected to a large number of non-GNEs and the ECC information is excessive, it is recommended that you replan the DCN or add GNEs. Generally, the number of DCN NEs does not exceed 64.
That's all, welcome everyone to leave a message and exchange in the comment area!
Thank you！
