08-24-2020 09:42 AM - edited 08-24-2020 09:44 AM
It will be great help if anyone can help on below issue.
I have one old clustered master server which is on 7.7.3 version which are going to decommissioned in 1 month.First we have started decommissioning media server. It has only one media server which is RCH now. Before proceeding on decommissioing RCH, i am trying to move RCH to master server. This is a clustered master server and it has 2 nodes active and passive. Zoning is done and i can see tape drive and robot on master server node. While doing robot/tape drive configuration using wizard, i am not getting active node name listed in the list where i need to select RCH, however virtual name is listed there. I tried configuring by selecting virtual name, it has given me error "Invalid host(88) : host is not defind in EMM". Please help on it. Details are given below.
vmoprcmd -devmon hs command output (it listed only one node which is passive node and one media server which we will decomm first)
Host Name Version Host Status
masternode2 773000 ACTIVE-DISK
nbumedia 773000 ACTIVE-DISK
emmcmd -listhost output:
08-24-2020 09:48 PM
on a clustered master server, RCH will be tagged to virtual name and wont show active node name. However, it might be necessary for you to run device configuration wizard when you failover to the other node.
Hope it helps.
08-24-2020 11:57 PM
Please double-check that you have correct licenses added for all nodes.
You need to have Library-based tape drive licenses added.
This status looks like this license might be missing: ACTIVE-DISK
08-25-2020 12:24 AM
One more thing -
Please be sure to go through 'Device configuration guidelines' and 'Configuring devices ' in Veritas NetBackup™ Clustered Master Server Administrator's Guide.
This section says that you need to configure devices on the active node, then fail over and repeat device config, because device paths will most likely be different on the 2 nodes.
08-25-2020 12:55 AM
08-25-2020 02:26 AM
The HOST STATUS from nbemmcmd -listhosts -verbose is supposed to reflect licensed features, not configuration details.
I have been once on a particular site with a clustered master server where I had issues with host status and device config.
This worked for me:
nbemmcmd -updatehost -machinename <media_server> -machinestateop set_tape_active -machinetype media -masterserver <master_server>
(media_server will be the active nodename and master_server the virtual hostname)
Next, run device config wizard from cmd:
vmoprcmd -h <media-server> -timeout 3600 -autoconfig -a
Restart MM processes (stoplid ; ltid -v)
You may need to freeze the cluster to prevent failover or restart action by the cluster.