Description

This alarm is generated when the subordinate network port in a bonded port becomes abnormal.

Attribute

Parameters

Impact on the System

Network bonding in active/standby or load balancing mode cannot take effect.

Possible Causes

The target host contains only one subordinate network port, or a subordinate network port is invalid.

Procedure

1. Perform the required operation based on the alarm information.
   • If only one subordinate network port is available and it is not in the unknown state, go to 2.
   • If more than one subordinate network port is available and one network port is in the unknown state, go to 3.
   • If more than one subordinate network port is available and one network port is in the unknown state, go to 7.

2. Add subordinate ports to the bonded port as follows:
   • If no service is running on the host, log in to the FusionSphere OpenStack web client, locate the group to which the target host belongs, and delete the bonded port. Then create a bonded port with the same name as the deleted one and add more than one subordinate port to the bonded port.
   • If the host has services running, log in to the FusionSphere OpenStack system and run the following command using the command-line interface (CLI) to update the subordinate network ports in the bonded port:

     cps hostcfg-item-update --item bond --bond-name trunk_name --slaves nic1,nic2,... --type network group_name

     cps commit

     In this command, the group_name value can be obtained using the cps hostcfg-list --type network command. bond-name and trunk_name values can be obtained using the cps hostcfg-show --type network group_name command.

     For example, run the following commands:

     cps hostcfg-item-update --item bond --bond-name trunk1 --slaves nic3,nic4 --type network group1

     cps commit

     Then, go to 10.

3. Perform the required operation based on the network port type.

   Log in to the FusionSphere OpenStack web client and locate the group to which the target host belongs.

   Log in to the FusionSphere OpenStack system and run the following command using the command-line interface (CLI) to check whether the option attributes of the subordinate network ports are AUTO_PXE.

   cps hostcfg-show --type network group_name

   • If the unknown subordinate port is of the AUTO_PXE type, go to 4.
   • If the unknown subordinate port is not of the AUTO_PXE type, go to 5.

4. Perform the required operation to replace the unknown subordinate network port with a planned physical network port on the host.
   • If the planned physical NIC on the host is not registered with the host group, go to 6.
   • If the planned physical NIC has been registered with the host group, log in to the FusionSphere OpenStack system and run the following command on the CLI to deregister the NIC from the host group:

     cps hostcfg-item-delete --item nic --nic-name nic_name --type network group_name

     cps commit

     The host group name can be queried using the cps hostcfg-list --type network command, and the NIC name can be queried using the cat /usr/bin/ports_info |python -mjson.tool|grep eth command.

     For example, run the following commands:

     cps hostcfg-item-delete --item nic --nic-name nic3 --type network group1

     cps commit

     After this step is complete, go to 6.

5. Replace the unknown subordinate network port in the bonded port as follows:
   • If no service is running on the host, log in to the FusionSphere OpenStack web client, locate the group to which the target host belongs, and delete the bonded port. Then create a bonded port with the same name as the deleted one and use normal network ports in the bonded port.
   • If the host has services running, log in to the FusionSphere OpenStack system and run the following command using the CLI to update the subordinate network ports in the bonded port:

     cps hostcfg-item-update --item bond --bond-name trunk_name --slaves nic1,nic2,... --type network group_name

     cps commit

     In this command, the group_name value can be obtained using the cps hostcfg-list --type network command. bond-name and trunk_name values can be obtained using the cps hostcfg-show --type network group_name command.

     For example, run the following commands:

     cps hostcfg-item-update --item bond --bond-name trunk1 --slaves nic3,nic4 --type network group1

     cps commit

     Then, go to 10.

6. Configure the switch to enable the planned physical NIC to communicate with the preboot execution environment (PXE) network port at layer 2.

   Then, go to 10.

7. View the additional information in the alarm.
   • If "status is down" is displayed, go to 8.
   • If "aggregator id mismatch" is displayed, go to 9.

8. Check abnormal and normal NICs in the additional information and rectify the fault one if any by referring to ALM-6021 Host Network Port Fault.

   Then, go to 10.

9. Log in to the peer host or switch of the aggregation network port and check whether the trunk configuration is correct.
   • If the configuration is incorrect, correct the configuration and go to 10.
   • If the configuration is correct, go to 11.

10. Wait for 5 minutes and check whether the alarm is cleared.
    • If yes, no further action is required.
    • If no, go to 11.

11. Contact technical support for assistance.

Related Information

None