How to backup clustered Windows clients

[OpsAnalyst]

I'm on Netbackup 6.5 and I need to set up a new backup policy for two clustered nodes.  Easy enough to set it up if I want the passive node's backup to fail every night, but how do I set it to where both the active and passive nodes will complete successfully, but only the active node is actually backing up the data? Of course it also needs to be able to tell which node is active and such. Thanks!

[VirtualED]

The best option for backing up two cluster base clients is to:

1. Create 1 Policy to backup the local data and system state with both active and passive nodes as the client name.

2. Create another policy to backup the shared data using the virtual client name.  So when it fails over it will always follow the shared resources.

[J_H_Is_gone]

install NB on the two physical servers, with the client name being the physical servers.

I setup a policy to backup the physical servers: (I call this one physical_clusters)
C:\
System State/Shadow copy (depending on the OS version)

I then setup a policy(s) to get the other stuff.
Logging into one of physical servers I go to Cluster Administrator.
Under Groups I look at the virtual names and see what resources each of them owns.
I then setup a policy to get the resources.

CLSTR_Vserver1
M:\
N:\

CLSTR_Vserver2
P:\
R:\

Now I know I get the items from the physical that I need
and I get the resources no matter what physical server they are currently on.
(NOTE:  if you log into a Vserver and look at the BAR you will NOT see the Vserver name - that is fine.  It works just fine so long as NB is installed on the physical servers)
 

[OpsAnalyst]

I appreciate the help so far, but I don't think you guys realize just how new I am to all of this. lol I've been thrust into the backup admin position recently and I am pretty much just picking it up on the fly with no training or anything.  So you'll have to forgive me for not understanding the advice so far.  I'll explain a little more in depth what I have and what I need.  I got a request from a systems analyst to set up a backup for this cluster, let's call it LVBGH654a and LVBGH654b.  The netbackup client is already installed on the servers so I just needed to set up a policy.  Right now I can RDP to LVBGH654 and it takes me to the active node.  In the netbackup client settings on the server the only client listed is LVBGH654a.  Now in my Netbackup policy I added clients LVBGH654a and LVBGH654b.  The directories being request include to shared drives N & M on whichever node is active and then some Program Files on each node's C drive.  As of now the active client backup succeeds which is the most important thing, but the passive node fails after sitting there until the backup window closes.  I know there's a way to get that to not happen, but I'm not sure if it's a setting in the GUI or if it's a setting on the client. I hope this makes it a little more clear. Thanks a ton for the advice so far and any to come in the future!!!!

[J_H_Is_gone]

you have physical server 654a and 654b - and both have NB installed.

in NB create a policy

CLSTR_physical

add 654a and 654b as clients in this policy doing windows backups

for your backup selection

C:\

Shadow Copy Components:\

################################

now you have to figure out your virtual server names and what they have.

From what you say N and M are the shared drives - meaning they could be running off of either physical server - so you DO NOT want to back them up in the physical policy.

so talk to the people who told you to backup N and M and ask them what is the  virtual server name that OWNS N and M

now crate a policy

CLSTR_<virtualservername>

again this is a window backup

the client will ONLY be the virtualserver name that owns N and M

your backup selection will be

N:\

M:\

################################

so what you have now is the C drive and SCC of each physical server will be backed up

N and M will be backed up by the virtual name - which means it does not matter where N or M is at the time - it will always find it via the virtual name.

Now if you get more clusters - just add the physical servers to the CLSTR_physical server policy

and then do the steps to backup the virtual server and its drives in a new policy.

#############################

once you get a good backup from both policies - verify with the server admin via the Backup Archive and Restore GUI - that everything they want backed IS getting backed up.

[OpsAnalyst]

Great advice! One problem though. Now when I run the virtual policy it is failing and the error says "client did not start".  Would that be an issue in the policy or on the server?

[Mark_Solutions]

What is the status code of that failure - is it a 49?

Could you post the text in the detailed status of the job please so that we can see exactly what happens here.

We know the node names so could you also confirm the virtual name and that now you have that in the policy you can connect to it via Client Host Properties section and when connected its client name is that of the active node in the cluster.

Everything so far has talked about drive letters so i was assuming that this was a Windows system - but a 49 error seems strange unless it is a loopback issue.

Check the hosts files on each node to make sure you dont have anything that looks like this which will casue a loopback issue with NetBackup:

127.0.0.1      LVBGH654a      LVBGH654a.domain.com

Another thing that can cause this, especially on 6.5 is VSP so add each node and the virtual name to the Client Attributes section of the Master Servers host properties, select each and on the Open Files tab ensure that VSS is selected and not VSP

Hope this helps

[Marianne]

"Would that be an issue in the policy or on the server?"

In addition to the text in the job details as per Mark's request, please also post the policy config.
On master, run the following from cmd:
cd <install-path>\veritas\netbackup\bin\admincmd
bppllist <policy-name> -U
(it MUST be big U)

[J_H_Is_gone · Accepted solution]

you are on 6.5 and I would bet that this cluster is a 2008 server.

If so you need to go into the properties of the master server.

Go to Client Attributes

Add all 3 server names to this list

then for each of these go to the Windows Open File Backup tab

and select the radio button for VSS.

[OpsAnalyst]

You guys and gals are amazing!  The policy settings above were very helpful, but then I hit the snag with the 49 errors.  Saw the last three posts today and, being that the servers in question are 2008, I added the VSS option under the Client Attributes\Open File tab on the master server and they all kicked in a ran like champs!  I really appreciate you all being willing to share your knowledge. Thanks a lot!

[J_H_Is_gone]

Would you mind marking the post that most helped you with your solution.