Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

4547

February 22nd, 2007 03:00

Oracle Recoveries

I have worked on Oracle module, its linking and backup & recovery scripts but one thing I have not worked on is Recovering an Oracle database on a different server.

Does anyone have a procedure known for this? I am using NW 7.2, NMO 4.2, Oracle 9.2 on Linux.

I am using the recovery script -

connect target sys/sys@onmdb;
connect catalog rmannew/rman786@revcat;
run {
allocate channel t0 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t1 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t2 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t3 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t4 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t5 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t6 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
allocate channel t7 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=ebsbkp,NSR_CLIENT=standbyonmamr,NSR_DATA_VOLUME_POOL=Database Sunday)';
restore database;
release channel t0;
release channel t1;
release channel t2;
release channel t3;
release channel t4;
release channel t5;
release channel t6;
release channel t7;
}

In the script, standbyonmamr is the client whose database needs to be recovered. I don't want to recover control files if it is possible to avoid that.

It goes to the tape, picks up the correct tape and ends up with the error -

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of restore command at 02/21/2007 11:31:12
ORA-19507: failed to retrieve sequential file, handle="0tiadac2_1_1", parms=""
ORA-27029: skgfrtrv: sbtrestore returned error
ORA-19511: Error received from media manager layer, error text:
The function nsr_bind_recov_mm() failed with the error: Unknown host (1:5:0)

Please suggest

2 Intern

 • 

14.3K Posts

March 6th, 2007 07:00

No idea, but I guess yes. For PiT you simply set until time based on Oracle data or SCN number and that's it.

2 Intern

 • 

14.3K Posts

February 22nd, 2007 03:00

That "Unknown host" might suggest what proble is (DNS, name mismatch, not being defined in NW, remote access among other things...). Check NW log on both client and server - if you get unknown host there then you know what it is.

2 Intern

 • 

14.3K Posts

February 28th, 2007 06:00

What about names you used in RMAN script?

2K Posts

February 28th, 2007 06:00

We have checked those issues.

We are able to recover a File System backup using nwrecover on the same server.

2K Posts

February 28th, 2007 07:00

onmdb is the name of the database.

standbyonmamr is the client whose database is to be restored on this client named abc.

recvcat is the new recovery catalog with is a copy of the original recovery catalog. We do not want to use the original catalog as while restoring that would get updated.

2 Intern

 • 

14.3K Posts

February 28th, 2007 11:00

Why would be using old catalog be a problem?

2K Posts

March 4th, 2007 00:00

I am not good at Oracle and the customer's Oracle DBA claims that if they recover using the existing catalog, it will mark some entries in the catalog that would not permit recovering the database again.

Actually, we are testing recovering the online production database on a test server.

2 Intern

 • 

14.3K Posts

March 4th, 2007 10:00

Ask him to test duplicate with RMAN... it will use backup of original db and restore it to test system (of course, you can rename it and RMAN takes care of that).

2K Posts

March 6th, 2007 07:00

We have got the DNS issue sorted out and recovered the database today.

Moving ahead with point in time recovery, I just wanted to find out if the time when NW updates is index is same as the time at which RMAN updates its catalog.

2K Posts

March 6th, 2007 07:00

I am doing the same but using the index date & time of NW assuming it to be same as RMAN's timestamp but it still picks up the older media for recoveries.

17 Posts

June 22nd, 2007 03:00

Hi there,
we have the same problem .
To solve the problem is to recover RMAN from a different catalogue or to search for a network problem (DNS incorrect configuration, switch etc)
Thanks for any suggestions .
Andrea

2K Posts

June 22nd, 2007 22:00

We finally resolved it by checking DNS & Network issues.

17 Posts

June 25th, 2007 00:00

Anju,
thanks very much for your answer.
Please, do you know what means the error : "Unknown host (1:5:0)"
Thanks again
Regards,
Andrea

2 Intern

 • 

14.3K Posts

June 25th, 2007 00:00

It means one of the hosts used during backup operation is unknown (can't find its IP).

30 Posts

April 1st, 2020 06:00

How do you select the correct timestamp for a Point in Time recovery, such that it only uses a particular saveset.

For instance, the datafile Level 0 backup shows its save time as 9:00pm. But when I select that as the timestamp for recovery, it looks for savesets of earlier dates.

When I select a timestamp after the backup has completed, e.g 10pm , it'll restore but then look for an archive log in the follwoing day's backup saveset.

I want to be able to pick a timestamp in which the recovery only uses that particular saveset, without depending on backups done on other days.

Which timestamp should I use?

 

No Events found!

Top