cancel
Showing results for 
Search instead for 
Did you mean: 

Installing Netbackup on 2nd node of cluster

Manish_Bhatia
Level 4
Partner

Hi,

In my environment, we have a single node cluster running with vcs 6.0.3 and netbackup 7.5.0.6 on RHEL 6.4.

 

Now we need to add a 2nd node to the cluster, and upgrade the vcs to 6.1. Plan shared by Linux team is:

Install OS on 2nd node

Install VCS 6.1 on 2nd node

Zone the san disk which has NBU vol and database

Install NBU on 2nd node and manuall failover the netbackup on the 2nd node, by making 1st node down.

if successful, then we will join the 2nd node in cluster for automatic failover.

 

Is this approach correct? if so, please let me know, steps to install NBU on 2nd node and how it will judge that it is a part of cluster, how about adding to resource groups. As the failover will be tested first and then the node will be added to cluster (as told by linux team, i dont knowm how it is possible)

 

Thanks

Manish

 

 

1 ACCEPTED SOLUTION

Accepted Solutions

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

I believe you need to install same SF/HA version (6.0.3) on node 2, join the existing cluster as per VCS documentation, install NBU on node 2 as per NBU documentation, then use steps in this article to perform rolling upgrade of SF/HA: 
https://www-secure.symantec.com/connect/forums/sfha-solutions-61-performing-rolling-upgrade-storage-...

View solution in original post

4 REPLIES 4

Deb_Wilmot
Level 6
Employee Accredited Certified

Also note the instructions in the install guide for clusters:

Special installation guidelines for UNIX clustered environments
Use the following guidelines when you install NetBackup in clustered systems:
■ Ensure that each node in the NetBackup cluster can run the rsh command or
its equivalent (on HP-UX, the command is remsh). Starting with NetBackup
7.5, you can also use the ssh command. The root user must be able to perform
a remote login to each node in the cluster without entering a password. This
remote login is necessary for installation and configuration of the NetBackup
server and any NetBackup options. After installation and configuration are
completed, it is no longer required.
■ You must install, configure, and start the cluster framework before you install
NetBackup. For additional installation prerequisites and installation notes,
see the NetBackup Clustered Master Server Administrator’s Guide.
■ You must have defined a virtual name using DNS, NIS, or /etc/hosts. The IP
address is defined at the same time. (The virtual name is a label for the IP
address.) Use this virtual name and IP address only for the NetBackup resource.

Install Guide is available here:

http://kbdownload.symantec.com/resources/sites/BUSINESS/content/live/DOCUMENTATION/5000/DOC5154/en_US/NetBackup7.5_Install_UNIX.pdf?__gda__=1403793455_37e1e33aa32d1e2e4ef80ca5d2185fb8

 

After you install the second node, check the NBU_RSP files on both nodes (in /usr/openv/netbackup/bin/cluster) to make sure both nodes are listed as "NODES".

 

Otherwise that 'should' work.

 

You may also want to check the high availability guide to make sure I didn't miss anything. :)

http://www.symantec.com/business/support/index?page=content&id=DOC5183

 

 

Deb

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

I don't believe the steps will work.

The 2nd node will need to join the existing cluster at VCS level before NBU is installed.
Not sure if 6.1 node can join 6.0.3 node - you will have to check VCS documentation.

When both nodes are in the same cluster, NBU installation on node 2 will detect existing cluster config when virtual hostname is specified as master server and will proceed to update NBU/EMM on both nodes with cluster info as well as modify NBU service group with additional node.
rsh comms is needed during this phase along with all requirements documented in NBU Clustered Master Server Guide.

Manish_Bhatia
Level 4
Partner

Thanks Marianne.  

Even i am in doubt about the same, so want to get it clear. Also, the current situation is that exixting node has 6.0.3 vcs version, and on 2nd node team has already upgraded vcs to 6.1. so in this case, we need to downgrade vcs version on 2nd node, and install 6.0.3 and join cluster, install nbu, as cluster node, do failover to test, and then move forward for vcs upgrade.

 

Please confirm if this is ok.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

I believe you need to install same SF/HA version (6.0.3) on node 2, join the existing cluster as per VCS documentation, install NBU on node 2 as per NBU documentation, then use steps in this article to perform rolling upgrade of SF/HA: 
https://www-secure.symantec.com/connect/forums/sfha-solutions-61-performing-rolling-upgrade-storage-...