Product and version information:
On the clustered node named ClusterNode1, the gabconfig -a command is executed and the command output shows that only ports a, b, and n are successfully started.
ClusterNode1:~ # gabconfig -a
GAB Port Memberships
Port a gen 181a08 membership 01
Port b gen 181a0b membership 01
Port h gen 181a0a membership 01
Step 1 Use the Console IP address to log in to the N8000 as user support.
Attention：In the following steps, the file system unmounting and disk group exporting operations will interrupt ongoing services. Before performing the operations, inform your customer of the interruption and gain the customer's approval.
Step 2 Run cfsumount all to unmount all file systems.
Step 3 Run vxdg deport sfsdg to export disk group.
Step 4 Run vxdg import sfsdg to import the disk group.
Step 5 Run vxvol -g sfsdg startall to start disk group.
Step 6 Run hagrp -online cvm -sys ClusterNode1 to start the cvm service group.
Step 7 Mounting the File System.
1. Run cfsmntadm display to check the mounting status of file systems.
If a file system's STATUSis NOT MOUNTED, the file system is not mounted.
SHdatasyncSr_01:~ # cfsmntadm display
Cluster Configuration for Node: ClusterNode1
MOUNT POINT TYPE SHARED VOLUME DISK GROUP STATUS
/vx/__LogFS__ Regular __LogFS__ sfsdg NOT MOUNTED
/var/lib/nfs/sm Regular _nlm_ sfsdg NOT MOUNTED
/vx/SourceFs01 Regular SourceFs01 sfsdg NOT MOUNTED
2. If a file system is not mounted, run cfsmount FSpath to mount the file system. Replace FSpath with the actual directory where you want to mount the directory.
Step 8 Use the faulty node's IP address to log in to the faulty node as user support.
Step 9 Run hagrp -online cvm -sys ClusterNode1 to check whether the cvm service group is successfully started.
If yes, no further actions are required.
If no, contact R&D engineers.
Port status reflects the status of service groups and disk groups. Perform the following operations to find the fault:
1. On the faulty node, run hagrp -online cvm -sys ClusterNode1 to manually start the “cvm” group. The start fails and a message is displayed indicating an error in adding the node to the cluster.
2. On the faulty node, run vxdisk -o alldgs list. The command output shows that none of the disk group “sfsdg” is imported.
3. Run the same command on the active node. The command output shows that all disks are in “online dgdisabled share” state.
Therefore, it can be concluded that the active node's setting all disks to “online dgdisabled share” has caused the disk group importing to fail. As a result, the cvm service group fails to start and the port anomalies are displayed after the gabconfig -a command is executed.