THanks a lot Crvelin.

Sorry since I was on leave for last few days so I could not respond at proper time.

But I have one doubt that if through RMAN they copied the database then how the new host will come to know from where to restore?

Is RMAN itself will be able to detect it automatically?

However as per my understanding I should follow following steps (Please correct me if I am wrong):

1. Install NMO on the new host.
2. Configure nsrnmo script on new host
(Do I need to configure RMAN client for new host in backup server)
3. Database team will copy database to new host.
4. Database team will start the restore and recovery operations.

Please suggest me how to proceed.

1) is correct. Install it and relink library.
2) not required for restore
3) This can be done by RMAN script with duplicate command. Inside you point NSR_CLIENT to source (where original backup happened).
4) Correct.

Thanks Crvelin.

I will perform the test and in case I get problem I will update.

Hi Crvelin,

I have installed NMO in new host and linked the libraries.

Database team has created new instance in the new host with different name. Necessary tns entry has been placed for connectivity. We are able to connect to the target (original) database and catalog from new host. In the nw host we are able to connect to the new instance as sysdba.

We can allocate ordinary channel of disk type as well as SBT_TAPE type. But could not allocate auxiliary channel. Following is the command outputs:

With Auxialiary Channel

RMAN> run
2> {
3> ALLOCATE AUXILIARY CHANNEL ch1 DEVICE TYPE 'SBT_TAPE' parms 'ENV=(NSR_DATA_VOLUME_POOL=DailyNI)';
4> }
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of allocate command on ch1 channel at 09/18/2009 12:57:35
ORA-19554: error allocating device, device type: SBT_TAPE, device name:
ORA-27211: Failed to load Media Management Library
Additional information: 2

With Normal Channel

RMAN> run
2> {
3> ALLOCATE CHANNEL ch1 DEVICE TYPE 'SBT_TAPE' parms 'ENV=(NSR_DATA_VOLUME_POOL=DailyNI)';
4> }
allocated channel: ch1
channel ch1: sid=122 devtype=SBT_TAPE
channel ch1: NMO v5.0.0.0
released channel: ch1


Please suggest.

Was this script created with NMO wizard? I think in the past with Oracle 9 or 10 we used to allocate first stb?tape channel and then auxiliary one. Not sure if things have changed since - especially with NMO 5. The reason I asked for wizard was that in previous version there something causing it to fail.

By looking strictly to error message, I see that it failed due to issue with loading module library. Can you double check relinking was ok? Oh well, dummy me, otherwise with normal channel it would not fail. Can you check sbtio.log to see if it contains more information?

Hi Crvelin,

We are not using NMO wizard. We are using manual scripts.

I checked sbtio.log in original database but it does not contain any messages. It is totally blank. In the new host I have not found sbtio.log itself.

We are using following script for recovery:
#tst is the new database created on new host.
#rmantest is the username/password for test database with which we have taken #backup.
#test is the original database in original host;
#testcat is the user/pass for recovery catalog rcatqc from which we have taken #backup.
#db2 is the original host
#data is the backup server


export ORACLE_SID=tst;
connect target rmantest/rmantest@test;
connect catalog testcat/testcat@rcatqc;
connect auxiliary /;
run
{
allocate channel d0 device type disk;
allocate channel t0 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=data.go2uti.com, NSR_CLIENT=db2.qc.go2uti.com)';
allocate auxiliary channel a0 type 'SBT_TAPE' parms 'ENV=(NSR_SERVER=data.go2uti.com, NSR_CLIENT=db2.qc.go2uti.com)';
duplicate target database to tst;
release channel d0;
release channel t0;
release channel a0;
}

Please suggets.

Hi Crvelin,

I have deleted and recreate library linking. Now I am able to allocate auxiliary channel but getting error on recovery. The script goes fine upto the reading catalog but it is not contacting the backup server. I have checked the volume and these save sets are visible in the volume save sets. The error that I get is:

ORA-19870: error reading backup piece /FULL_TEST_02kpftob_1_1_20090917/
ORA-19507: failed to retrieve sequential file, handle="/FULL_TEST_02kpftob_1_1_2
ORA-27029: skgfrtrv: sbtrestore returned error
ORA-19511: Error received from media manager layer, error text:
lnm_index_cfx_connect retry failed. See privious error messages. (3:3:11)
channel a0: starting datafile backupset restore
channel a0: specifying datafile(s) to restore from backup set
restoring datafile 00001 to +DATA
channel a0: reading from backup piece /FULL_TEST_01kpftoa_1_1_20090917/
ORA-19870: error reading backup piece /FULL_TEST_01kpftoa_1_1_20090917/
ORA-19507: failed to retrieve sequential file, handle="/FULL_TEST_01kpftoa_1_1_2
ORA-27029: skgfrtrv: sbtrestore returned error
ORA-19511: Error received from media manager layer, error text:
lnm_index_cfx_connect retry failed. See privious error messages. (3:3:11)
failover to previous backup

released channel: d0
released channel: t0
released channel: a0
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of Duplicate Db command at 09/18/2009 18:11:29
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 5 found to restore
RMAN-06023: no backup or copy of datafile 4 found to restore
RMAN-06023: no backup or copy of datafile 3 found to restore
RMAN-06023: no backup or copy of datafile 2 found to restore
RMAN-06023: no backup or copy of datafile 1 found to restore

Please suggest.

Does your user have right to read index as oracle user? I assume you have oradba on both systems so in admin list put oradba@hotst.domain (list all domains if not sure via which one contact is going) and root@host.domain on admin list. On source client, in remote access list oradba@target.machine.domain.

If it fails after that, check what it says on server side in daemon.raw and sbtio.log on client.

Hi Crvelin,

On the script we have changed the NSR_CLIENT to include only db2 and now I am able to go ahead one more step. Now in backup server it is showing one browse session for new host and loads the required volumes. After that it again fails with following error:

ORA-19870: error reading backup piece /FULL_TEST_03kpftob_1_1_20090917/
ORA-19507: failed to retrieve sequential file, handle="/FULL_TEST_03kpftob_1_1_20090917/", 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: Remote system error (1:5:239)
channel a0: starting datafile backupset restore
channel a0: specifying datafile(s) to restore from backup set
restoring datafile 00003 to +DATA
restoring datafile 00004 to +DATA
channel a0: reading from backup piece /FULL_TEST_02kpftob_1_1_20090917/
ORA-19870: error reading backup piece /FULL_TEST_02kpftob_1_1_20090917/
ORA-19507: failed to retrieve sequential file, handle="/FULL_TEST_02kpftob_1_1_20090917/", parms=""
ORA-27029: skgfrtrv: sbtrestore returned error
ORA-19511: Error received from media manager layer, error text:
There is already a NW session opened by this thread. (0:3:11)
channel a0: starting datafile backupset restore
channel a0: specifying datafile(s) to restore from backup set
restoring datafile 00001 to +DATA
channel a0: reading from backup piece /FULL_TEST_01kpftoa_1_1_20090917/
ORA-19870: error reading backup piece /FULL_TEST_01kpftoa_1_1_20090917/
ORA-19507: failed to retrieve sequential file, handle="/FULL_TEST_01kpftoa_1_1_20090917/", 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: Remote system error (1:5:239)
failover to previous backup

released channel: d0
released channel: t0
released channel: a0
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of Duplicate Db command at 09/18/2009 20:12:54
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 5 found to restore
RMAN-06023: no backup or copy of datafile 4 found to restore
RMAN-06023: no backup or copy of datafile 3 found to restore
RMAN-06023: no backup or copy of datafile 2 found to restore
RMAN-06023: no backup or copy of datafile 1 found to restore


Do I still need to put the suggested user in admin list.

Please suggest.

Yes you do. The error now, first one nsr_bind_mm() I have no idea what it is, but I suspect it tries to hook on nsrmmd which is reading data. I only guess that due to this error you get second one which claimed there is already session opened by the same thread.

When we did this we used 4.2. Since we updated to 4.5, but we didn't play with duplicate any more. I do plan to upgrade our GeoRAC systems to 5.0 build 306 (it is newer than one found on PowerLink) due to some issues with nwora locking mechanism we observed in 4.5 we use, but I do not have list of issues fixed by that build so I can't rule possible bug.

When this error occurs, do you see any message in daemon.raw on backup server?

Hi Crvelin,

I checked the admin list and found that *@* is already there, so it there should not be any issue related to permissions. However if you suggest I can put specifically for this.

In daemon there is no error messages for this.

No, then admin list can be left as it is (even *@* is asking for troubles). Nevertheless, you should still update source's remote access list as I have see that even when you have everyone in admin list with remote access right it still asks for that to be filled additionally.

Then I'm out of idea. Try to allocate just one channel and see what happens then.

Hi Crvelin,

The source and destination both client resources have *@* in the Remote Access atribute already.