Networker restore

[Roshan]

Hi Erman,

I am performing a restore on a test instance. The backup is on Networker. When I am launching the restore, I get the error below:

Starting restore at 20-DEC-16

released channel: CH1
released channel: CH2
released channel: CH3
released channel: CH4
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of Duplicate Db command at 12/20/2016 13:46:04
RMAN-03015: error occurred in stored script Memory Script
RMAN-06026: some targets not found - aborting restore
RMAN-06023: no backup or copy of datafile 78 found to restore

RMAN> **end-of-file**


I did RMAN> delete expired backup. Still I am getting same error.

Do  I need to to perform  fresh backup?

Regards,
ROshan

[Roshan]

I did list backup of datafile 78, no output was returned

[ErmanArslansOracleBlog]

Send me the whole script and the whole output..

[Roshan]

rmanbil_log.txt

pmtbil.sql

Please find attached.

Rgds,
ROshan

[ErmanArslansOracleBlog]

Register Tape backups to your target database (actually the actual database) controlfile. (considering you don't have rman catalog)

Here is the way for doing that:

Note 550082.1 How to Catalog Tape Backup Pieces

[Roshan]

How will I catalogue the backup piece? I do not know the backup piece name

[ErmanArslansOracleBlog]

Ask the Backup admin.. (The admin of networker to learn the piece names)

[Roshan]

I have checked with backup admin. He is telling me I have to catalog the backup instead of reading the control file. You can check this

https://community.emc.com/thread/224763?tstart=0

Partitioning occurs on production database instance. It backups a datafile before deleting it. So I think we configure catalog on our production database.

[ErmanArslansOracleBlog]

If you have a catalog db, then your catalog db would record the backup info.
 But I guess you don't have a catalog db.
At least, you didnt have one , when you were taking backup.
So in this situation, the controlfile of your database(backed up) will record the backup info.
However, as I see you don't have that info as well.
Why don't you have that info in your controlfile? (you recreated it maybe?)
That's why I say catalog them...

[ErmanArslansOracleBlog]

"Partitioning occurs on production database instance. It backups a datafile before deleting it. So I think we configure catalog on our production database."

--> so you can view all the backups from rman prompty (using list backup) except the dropped datafile?

[ErmanArslansOracleBlog]

Are you hitting this ? I m asking it becausae I couldn't exactly understand your situation .

Note that -> There is a limitation in 11.2.0.3 and below - you cannot perform a backup based duplicate to point in time prior to the backup of datafile.

Bug 12799744 is fixed in 12.1

Solutions:

Take backup of datafile immediately following datafile addition then duplicate to point in time after this backup

OR

Perform RMAN restore and recovery rather than duplicate

OR

If added datafile is not part of a SYSTEM, UNDO or SYSAUX tablespace then you can consider skipping the tablespace from the duplicate process