Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

Master server migration to new hardware

Created: 21 Feb 2014 • Updated: 26 Feb 2014 | 3 comments
This issue has been solved. See solution.

We are planning and testing our NBU master server (7.5.0.6 on HP-UX 11.31 IA 64, EMM and media server as well) to a new hardware with same OS.

I got a few questions somebody may help. thanks in advance.

1. we have around 600GB catalog. we are planning to run full backup in previous day, then on migration day, we will suspend backup jobs, run a incr backup. then reocver catalog from incr backup. But somehow the incr recovery failed with following. we have exactly same installation path on old and new master. same hostname but different IP.

Catalog recovery is in progress. Please wait...

ERR - Failed to find last backup image record for client mmdcux17, client_type 7, with policy Hot_Catalog_Backup and sched_type 4 (227)

Failed to recover catalog (227)

any idea?

2. during new master installation, i found under /usr/openv, there are a few files .psummary .pseq and .phistory. what are these files for? can we delete them?

3. on old and new master, we have different FS mount points. we try to origanize the FSs better. Wish these changes will not cost any issue.

old master: /usr/openv,   /usr/openv/netbackup/db

new master: /usr/opevn, /usr/openv/netbackup/db/images, /usr/openv/logs

4. we have two tape libraries configured on old master/media, one is decidated to master/media, another is shared with another media server with SSO option. afte migration, i assume we have to reconfgiure them on new master after we migrated FC zones. right?

Operating Systems:

Comments 3 CommentsJump to latest comment

Will Restore's picture

Failed to recover catalog (227)

You must install NetBackup to the exact same path and location as the old server

http://www.symantec.com/business/support/index?pag...

Will Restore -- where there is a Will there is a way

gurminder singh's picture

It is possible for the parent job of a hot catalog backup to fail with a Status 227 even though the child jobs complete with a Status 0. The failure will occur when the bpdbm daemon attempts cannot find the catdirinfo file which contains the DR information from the policy.It will be necessary to check the Disaster Recovery tab of the hot catalog backup policy to see if a DR path has been specified.

Go tot the Policies section and  Click on the Disaster Recovery tab to display the DR settings.

Deb Wilmot's picture

In reply to the other questions:
 

2. during new master installation, i found under /usr/openv, there are a few files .psummary .pseq and .phistory. what are these files for? can we delete them?

A.  Those files are used to track patches.  You can delete them if you like, but they are so small, I can't imagine that they would be problematic.  The roll back files will still be in the /usr/openv/pack directory.  

3. on old and new master, we have different FS mount points. we try to origanize the FSs better. Wish these changes will not cost any issue.

old master: /usr/openv,   /usr/openv/netbackup/db

new master: /usr/opevn, /usr/openv/netbackup/db/images, /usr/openv/logs

A.  As stated above in the post from wr, a catalog restore requires the same mount points so the addition of /usr/openv/netbackup/db/images would cause the catalog restore to fail.  

4. we have two tape libraries configured on old master/media, one is decidated to master/media, another is shared with another media server with SSO option. afte migration, i assume we have to reconfgiure them on new master after we migrated FC zones. right?

A.  Yes, you will have to reconfigure devices as the paths will most likely change.

Now - a comment relating to #1.  If the processes are shutdown, you can simply tar/gzip up the /usr/openv/ directory and copy it to the new server.  

You would need to install netbackup on the server, with the same patch level before you do so (or the links will not work well).

SOLUTION