Restoring NSR backup to target server in homogeneous copy

[Former Member]

Hi

I am trying to restore/recover the source database onto the target system as part of the setup of a standby database. The target database is backed up using Networker (NSR) to tape.

Am I able to copy the files containing the backup history from the source to the target system so that I can perform the recovery using the external (NSR) backup??

I have tried to create a full FILE backup but this is extremely slow (NSR takes 1.5 hours and the FILE backup to a local drive is taking >10 hours). Is there a way to speed up type of backup??

Thanks in advance

Doug

[Former Member]

This issue was Windows resources when reading the files. Increasing the number of backup threads (and thereby decreasing the size of each backup file produced) we were able to recover the system.

[Former Member]

Thanks Lars.

I have opened a support call and I will update this message with the resolution.

Thanks again.

Doug

[Former Member]

Thanks Lars.

I'm still struggling with restoring data from a remote instance to create a standby database.

We are using Networker (Legato) for the tape backups and it hangs when I try to recover this to the standby system (no I/O activity on the Networker recover processes). Please note that the backup & restore works perfectly on the local instance.

In the interim I have been trying the FILE backup and, with this forums help, I have been able to save to a remote share. The problem I am having with restoring the standby db from the file backup is that the recover_start is terminating with the return code of -8020 and an Error of end of file.

Luckly we have several different installations with varying database sizes. I have been able to restore to the standby database using a small database (21GB). However, when I have tried the 40GB and 120GB database I get the error below.

dbmcli on SID>recover_start DR_SRC DATA
OK
Returncode              -8020
Date                    20081218
Time                    00102401
Server                  sourcedb.domain.au
Database                SRC
Kernel Version          Kernel    7.6.04   Build 016-123-191-583
Pages Transferred       1793984
Pages Left              3217930
Volumes
Medianame               DR_SRC
Location                \\sharehost\db_backup\FULL_DATA
Errortext               end of file
Label                   DAT_000000081
Is Consistent           true
First LOG Page          6655740
Last LOG Page
DB Stamp 1 Date         20081218
DB Stamp 1 Time         00090547
DB Stamp 2 Date
DB Stamp 2 Time
Page Count              5011914
Devices Used            1
Database ID             sourcedb.domain.au:SRC_20080907_211746
Max Used Data Page      0
Converter Page Count    2697

---
dbmcli on SID>db_restartinfo
ERR
-24988,ERR_SQL: SQL error
-104,DBM command impossible at this time

---

Is this a known issue and is there a workaround???

Thanks

Doug

[lbreddemann]

> In the interim I have been trying the FILE backup and, with this forums help, I have been able to save to a remote share. The problem I am having with restoring the standby db from the file backup is that the recover_start is terminating with the return code of -8020 and an Error of end of file.

> dbmcli on SID>db_restartinfo

> ERR

> -24988,ERR_SQL: SQL error

> -104,DBM command impossible at this time

> Is this a known issue and is there a workaround???

Actually this is not an issue but it works as designed.

The -8020 message (it's not an error) just means "I reached the end of one backup medium but until now there are still pages to be read from other mediums".

So either the backup had been taken to parallel mediums (which is not the case here since the message shows "Devices Used 1" or the backup was done to multiple mediums one after the other).

So the next best thing I propose you do is to check whether taking the backup was completely successfull and if there is possibly another backup file.

If so, use recover_replace in the same backup session you used for the backup_start to recover the second backup file as well.

Once you're sure that all backup data had been restored, you may end the backup process with recover_ignore (be aware that this will also start the db, so no further backups e.g. log backups can be applied afterwards).

The db_restartinfo is not supposed to be run within a open backup session while a backup is active.

Hope that helps,

Lars

[Former Member]

Hi Natalia

Thanks for your response.

In regards to my questions:

1) I noticed that the DBM Gui will allow you to recover from another database on a different server. However, when I tried this I was getting an error on the destination database. From the documentation, it would appear that this may be caused by the incorrect user at the service level (Windows).

I have updated the Xserver service to run as the sidadm user but a simple test failed. Do I need to change the user for the MAXDB: ... services as well?

2) The database version is 7.6.04.16 and the DB is approx. 170GB. I

Thanks for your assistance.

Doug

[lbreddemann]

> 1) I noticed that the DBM Gui will allow you to recover from another database on a different server. However, when I tried this I was getting an error on the destination database. From the documentation, it would appear that this may be caused by the incorrect user at the service level (Windows).

>I have updated the Xserver service to run as the sidadm user but a simple test failed. Do I need to change the user for the MAXDB: ... services as well?

Basically all you've to do is to make sure that the user that runs the MAXDB-Services is allowed to read the backup files.

So whether you change the users for the services or change the permissions of the backup files - it's up to you.

Since it's only the MaxDB Kernel (which is the executable 'behind' the MAXDB: Services) needs to read the backup, there is no need to change the user for the XServer service.

> 2) The database version is 7.6.04.16 and the DB is approx. 170GB.

Nice to know but irrelevant for this context

regards,

Lars

[former_member229109]

Hello Gerd,

1) "Am I able to copy the files containing the backup history from the source to the target system so that I can perform the recovery using the external (NSR) backup??"

-> Yes, you can.

2) "I have tried to create a full FILE backup but this is extremely slow (NSR takes 1.5 hours and the FILE backup to a local drive is taking >10 hours). Is there a way to speed up type of backup??"

What do you mean?

You could check if you are using the fast disks for the databackup, for example.

Are you using a separate disk for the databackup?

Please give more details about the size of the data in your database, version of the database, configuration of the backup mediums.

Link :: http://maxdb.sap.com/doc/7_6/default.htm -> Glossary -> Backup -> Backing Up and Restoring

->Backing Up Data

&& Also see "Notes on the Different Types of Data Carriers" at

http://maxdb.sap.com/doc/7_6/43/d0c45de3f63ac3e10000000a422035/content.htm

Thank you and best regards, Natalia Khlopina