When a user maintains OceanStor 18500 V100R001C00SPC200 at a site, the user cannot log in to DeviceManager using SVP and an error is reported indicating that the system is busy. When the user tries to log in by using CLI, a message is displayed, stating, "The communication is abnormal or the system is busy."
Restart the MMT process of the storage array to temporarily resolve this problem. Specifically, perform the following steps on each controller:
a. Run the change user_mode current_mode user_mode=developer command to enter the developer mode. The default password is debug@storage. In developer mode, run the minisystem command to enter the minisystem mode. The command output is as follows:
2. Run the ps -ef command. In the command output, check the process ID of the MMT process /OSM/bin/mmt (the value in the second column, which varies with site conditions and is 6844 in this example). The following figure provides an example:
c. Run kill -9 6844 (6844 indicates the process ID in the preceding figure and is used as an example). After command execution, the process is restarted automatically. See the following figure.
d. Run ps -ef to query information about the MMT process. The process ID becomes different, indicating that the process is successfully restarted. See the following figure.
e. Run exit and type y to exit the minisystem mode. (You must exit from the minisystem mode in this manner.)
f. Perform steps a to e on other controllers and then log in using the CLI or DeviceManager. If the communication fault persists, restart the SVP.
When a network management client is used to log in to the storage array, the client needs to set up a connection with the MMT of the OM. The MMT uses a thread pool when it concurrently connects to multiple network management clients. For the connection between the OM and DeviceManager or the CLI of SVP, SSL is used. OpenSSL 0.9.8y has a known vulnerability in concurrent (multithreaded) scenarios, and this vulnerability exactly causes this problem. The external client fails to set up a connection with the server, so the user fails to log in to the storage array.
1. To fix this issue, upgrade the storage array to V100R001 C00SPC300 or later.
2. If the current network is not suitable for array upgrade, use the method provided above to work around this issue.