Save Set: All gives error on Sun cluster nodes

[jlh16]

Good Morning,

We are running Networker 7.1.1 on a Solaris 8 server with an attached storage node running Solaris 9. When trying to back up a Sun cluster node using the All save set we get the following error:

* cluster4node1:All 1 retry attempted
* cluster4node1:All nsrexec: authtype nsrexec
* cluster4node1:All savefs: nothing to save

The error does not occur when we specify the filesystems in the save set field.

Any ideas?

Thanks

 

[605]

May it be that you try to backup the physical node instead of the virtual node?

 

[jlh16]

Yes, we are trying to back up the system files on the physical node. Which works with the save -c backup command when we specify what filesystems to back up. Once we change the save set to All we get the above error.

 

[605]

This is not what you should do in a cluster. Please follow the instructions in the Installation Guide and the Admin Guide.

 

[jlh16]

I have looked in the admin guide and it states:

Backing up data from a physical client within a cluster is is similar to backing up data on a computer that is not part of a cluster. For detailed instructions on how to back up data, see "Chapter 3:Server Management.

 

[605]

OK. So your actual intention is to backup the physical node(s).

Actually, the NW client software would look up in /etc/vfstab for mounted filesystems to be backed up. It could be that NW can not read this configuration for whatever reason.

 

[ag100]

Hey - I haven't seen this pop up on UNIX systems before, but it's pretty common on Windows boxes... Check the DNS from both the server and the client, and ensure that the aliases are properly entered.

Good luck!

-ag100

 

[ag100]

One other thing... You mention you used the All save set... Have you tried specifying "/" to see if it would work? Does the issue only occurr when using All?

Also - As 605 mentioned, ensure the disks are not manually mounted and that they are listed in vfstab.

-ag100

 

[pure8888]

You should setup cluster as "NW aware nodes" so that you
can backup logical hosts. LGTO.clnt resouce needed for every
logical host.

For physical node backup, use every local file systems instead of ALL.
Sometimes, server or client's directive must be setup to skip files
and dirs.