cancel
Showing results for 
Search instead for 
Did you mean: 

Server physically down I need to restore form other windows or unix server

Netbackup_fan
Level 5

Hi experts.

I have a Windows 2003 server physcially down.

I need to restore the backups in another windows server or unix server.

What would be the detailed steps?

 

The backup id are telpri-eng02_1438739159, and telpri-eng02_1438574302

1 ACCEPTED SOLUTION

Accepted Solutions

sdo
Moderator
Moderator
Partner    VIP    Certified

I think you've found the problem...

...the master server has a hosts file entry of "10.0.112.74 vz-ca-wfm01", but you are expecting the restore to go to a server with an IP of 192.168.109.30

I think you need to consider, and carefully double check, whether the intended target client is in fact the target that you expect it to be... i.e. whilst an E: drive does appear to exist on server with IP of 192.168.109.30, it would seem that server with an IP address of 10.0.112.74 does not have an E: drive.  In short, your restore attempt appears to not have been going to the server that you expected/wanted.

View solution in original post

25 REPLIES 25

sdo
Moderator
Moderator
Partner    VIP    Certified
One doesn't normally attempt to restore from Windows backups to Unix clients, it is unsupported and therefore results would be unpredictable. To push restore to a different Windows client, use a NetBackup Server, and specify source and target client names as you wood for any other push restore.

sdo
Moderator
Moderator
Partner    VIP    Certified

Do you need a hand filling in the fields of the restore GUI?

Nicolai
Moderator
Moderator
Partner    VIP   

What type of data are we talking about - database ?

 

Marianne
Level 6
Partner    VIP    Accredited Certified

If Filesystem backups - See the steps that I have shared over here: 

https://www-secure.symantec.com/connect/forums/restoration-test#comment-11155851 

If database backups, let us know which db type - we can point you to the relevant agent guide.

Netbackup_fan
Level 5

I have experience restoring Netbackup for Oracle using Rman and also without the Rman using Netbackup client.

The problem is that the restore is failing from windows to another window server and from windows to unix.

sdo
Moderator
Moderator
Partner    VIP    Certified

Does the E: drive actually exist on server vz-ca-wfm01 ?

If E: exists on the restore target server, then is the E: drive a local volume, or SAN FC/iSCSI attached volume, or a USB drive, or a CIFS/SMB mounted share?

Marianne
Level 6
Partner    VIP    Accredited Certified

If the backup type was Oracle using RMAN scripts, you need to follow steps in NBU for Oracle Guide under this topic: Redirecting a restore to a different client 

You can only initiate restore from another Windows machine where Oracle is installed using RMAN. 
That is after you have authorized redirected restore on the master by creating altnames file as per the instructions in above section of the manual.

Netbackup_fan
Level 5

Thank you.

Was not using Rman.

Marianne
Level 6
Partner    VIP    Accredited Certified

So, why the reference to rman?

What was the policy type used for backups?

Have a look at sdo's post - does e:\ exist on the destination client as local drive (not network mount)?

Netbackup_fan
Level 5

Because Nicolai asked;  What type of data are we talking about - database ?

The policy is MS-Windows.

 

e:\ does not exists on the destination client as a local drive.

 

sdo
Moderator
Moderator
Partner    VIP    Certified

Does it exist at all?

If so, what form doth the beast take?

Netbackup_fan
Level 5

Sory sdo. 

I don't understand.

sdo
Moderator
Moderator
Partner    VIP    Certified

You said E: does not exist as a local drive.  This implies that it does exist, but clearly not as a local disk.  You didn't say whether it doesn't exist at all, nor whether it does exist as a SAN disk, nor whether it exists as a USB disk, nor whether it exists as a mapped drive?

Does it even exist at all?  If it does exist, then what is the connectivity/config/construct/device-type for E:?

I'm only asking because the restore job log shows that you attempted to restore to a device named E: on the restore target client.  So I wanted to find out what form it took.

Netbackup_fan
Level 5

I have uploaded printscreen of the destination server.

 

Netbackup_fan
Level 5

Disk e properties printscreen.

sdo
Moderator
Moderator
Partner    VIP    Certified

Ok - can we do some basic fact finding first...

1) On the master server what do these show:

hostname

cat /usr/openv/netbackup/version

/usr/openv/netbackup/bin/admincmd/bpgetconfig -g vz-ca-wfm01

/usr/openv/netbackup/bin/admincmd/bpgetconfig -M vz-ca-wfm01 client_name

nslookup vz-ca-wfm01

nslookup vz-ca-wfm01

(yes do the nslookup command twice)

2) On the client what do these show:

ipconfig

hostname

wmic logicaldisk list brief

reg query "HKLM\Software\Veritas\NetBackup\CurrentVersion\Config" | findstr /i "client server"

cacls E:\

...post the output, and we'll see what we need to ask next.

Thanks.

Marianne
Level 6
Partner    VIP    Accredited Certified

I'm battling to understand.

First you say :

e:\ does not exists on the destination client as a local drive.

Then you send us evidence of drive letter E:\ on the client.

So, it DOES exist?

Can you verify that the logon account used for NBU Client Service on destination client has write access to the E-drive?

Create bpcd and tar log folders under ...\netbackup\logs on the destination client, then try a test restore of a small file or folder to E-drive on this client.

Check the logs for errors.

Before you do this test restore - ensure bprd log folder exists on the master server.
If not, create it and restart NBU.

If restore fails again, copy all of the logs mentioned above to .txt files (e.g. bprd.txt) and upload as File attachments.

Netbackup_fan
Level 5

Sory Marianne for the confusion.

E:  does exists - Type Local Disk File system: Ntfs

 

Working on the other instructions.

sdo
Moderator
Moderator
Partner    VIP    Certified

OK - names and IPs and disk and MFD permissions look ok, but the restore job failed with a 'path not found'.

I think we're going to need a verbose=5 "tar" log from the client.  Do you need help configuring verbose logging and creating the log folder for "tar" on the client, and then re-try the restore, and then post up as a text file attachment the tar log file.

.

And can we just check hosts file too please, on master:

grep -i "vz-ca-wfm01" /etc/hosts

...and just check who the client side NetBackup Client services are running as, on the target client:

for /f "tokens=1,*" %a in ('net start') do (if "%a"=="NetBackup" (sc qc "%a %b">>a.txt))

findstr /i "^$ service_name service_start" a.txt

Thanks.