In Linux:

Virtual Storage Platform One SDS Block System Administrator Guide
Version
1.16.x
Audience
anonymous
Part Number
MK-24VSP1SDS001-03

If connection is established by a means other than nvme-cli, contact customer support for the corresponding connection method.

  1. Verify the connection information on the compute node that has a duplicated initiator name (host NQN). 
    # nvme list-subsys

    Note down Subsystem NQN and tcp traddr out of the displayed information.

    After the command is run, Subsystem NQN and tcp traddr are indicated as shown in the following example display.

  2. Disconnect from the initiator that has a duplicated initiator name (host NQN). For <Subsystem NQN>, specify the one that you noted down in step 1. 
    # nvme disconnect -n <Subsystem NQN>
  3. Edit the /etc/nvme/hostnqn file using an editor such as vi commands. 
    # vi /etc/nvme/hostnqn
  4. In the file, change the duplicated initiator name (host NQN) to a unique initiator name (host NQN).
  5. Reconnect the initiator. For <Subsystem NQN> and <tcp traddr>, specify the ones that you noted down in step 1. 
    # nvme connect -t tcp -n <Subsystem NQN> -a <tcp traddr> -s 4420
  6. After reconnection, verify that volumes are visible from the compute node.

    If volumes are still not visible after reconnection, check for other failures by following If a failure status is detected in the storage cluster.