Veritas-bu

[Veritas-bu] [Veritas-bu]Problem restoring an Oracle Database under OFS

2001-12-11 13:28:11
Subject: [Veritas-bu] [Veritas-bu]Problem restoring an Oracle Database under OFS
From: dcaddell AT intersil DOT com (Caddell, Scott)
Date: Tue, 11 Dec 2001 13:28:11 -0500
Hi Everyone,

We've recently encountered a problem restoring an Oracle Database.  This is the
first time we have tried to restore a database that is setup under Oracle Fail
Safe and we are not sure if this is the problem. Below is the detail.  Please
advise if you have any suggestions.

Thanks,

Scott Caddell
Intersil Corporation
Palm Bay, FL


### What version and patchset level of the database are you running? ###
Oracle8i Enterprise Edition Release 8.1.7.1.1 
PL/SQL Release 8.1.7.1.0 
CORE 8.1.7.0.0 
TNS for 32-bit Windows: Version 8.1.7.1.0 
NLSRTL Version 3.4.1.0.0 
In addition, the databases are setup in Oracle Fail Safe Version 3.1.2.


### Please describe your problem: ###
We are using Veritas media software and RMAN to do backups on a database 
created through the Oracle Applications 11i Rapid Install. The database is 
setup in OFS on a clustered environment. We've had no problems with the 
backups, but when I attempted to restore to a point-in-time from the RMAN 
backup the Veritas log file indicated that the operation completed successfully
(in 11 minutes) and the NT batch job was still sitting at RMAN-03022: compiling
command: IRESTORE. Only the 3 control files were restored. There were 3 tapes
needed for the restore (1 full and 2 incrementals). We verified that the correct
tapes were in the ATL and checked the log files from the RMAN backups and all of
the files were successfully backed up. We attempted to do the point-in-time
recovery with various dates/times and always had the same results. We chose
times that would have used just the first tape, the first and second tapes, and
all three. No errors are returned but the job just sits on the RMAN command
above. The database was taken off-line in OFS before the restore was attempted,
and the database was started in nomount mode. The database is ~20Gg.
We are using Veritas Netbackup Datacenter 3.4. We've applied patches J0850645
and 3411. The Oracle backups are done on DLT7000 with DLT Type IV tapes.

### If you are receiving errors, please list exact error messages and text: ###
A screen dump of the RMAN restore is below (no error is returned - it just 
hangs: (I replaced pwd's with xxx)
C:\nonsms\veritas_scripts>rman target oem_admin/[email protected] catalog rman2/[email protected] 
cmdfile='point_in_time_recovery_apq1.rcv'
Recovery Manager: Release 8.1.7.1.1 - Production
RMAN-06006: connected to target database: APQ1 (not mounted)
RMAN-06008: connected to recovery catalog database
RMAN> run {
2> # Allocate two channels, one for tape and one for disk.
3> allocate channel dev1 type 'sbt_tape';
4> allocate channel dev2 type disk;
5> # Restore the controlfile from backup. Its name/path would be found in the
init file.
6> restore controlfile to 'e:\oradata\apq1data\cntrl01.dbf';
7> # Replicate other control files from the one just restored.
8> replicate controlfile from 'e:\oradata\apq1data\cntrl01.dbf';
9> # Mount the database.
10> sql "alter database mount";
11> # Restore the database files.
12> restore database;
13> # recover database;
14> # Recover the database to a point in time. The preceding line, which is
commented out, # would be used to restore up to the time of failure.
15> recover database until time 
"TO_DATE('20011204:03:07:00','yyyymmdd:hh24:mi:s
s')";
16> # Open the database and reset the logs.
17> sql "alter database open resetlogs";
18> # Release the channels.
19> release channel dev1;
20> release channel dev2;
21> }
22>
RMAN-03022: compiling command: allocate
RMAN-03023: executing command: allocate
RMAN-08030: allocated channel: dev1
RMAN-08500: channel dev1: sid=11 devtype=SBT_TAPE
RMAN-08526: channel dev1: VERITAS NetBackup for Oracle8 - Release 3.4.1GA 
(030800)
RMAN-03022: compiling command: allocate
RMAN-03023: executing command: allocate
RMAN-08030: allocated channel: dev2
RMAN-08500: channel dev2: sid=12 devtype=DISK
RMAN-03022: compiling command: restore
RMAN-03022: compiling command: IRESTORE
RMAN-03023: executing command: IRESTORE
RMAN-08016: channel dev1: starting datafile backupset restore
RMAN-08502: set_count=14 set_stamp=447704341 creation_time=05-DEC-01
RMAN-08021: channel dev1: restoring controlfile
RMAN-08505: output filename=E:\ORADATA\APQ1DATA\CNTRL01.DBF
RMAN-08023: channel dev1: restored backup piece 1
RMAN-08511: piece handle=APQ1_bk_14_1_447704341 tag=HOT_DB_BK_LEVEL1 
params=NULL
RMAN-08024: channel dev1: restore complete
RMAN-03022: compiling command: replicate
RMAN-03023: executing command: replicate
RMAN-08058: replicating controlfile
RMAN-08506: input filename=E:\ORADATA\APQ1DATA\CNTRL01.DBF
RMAN-08505: output filename=E:\ORADATA\APQ1DATA\CNTRL02.DBF
RMAN-08505: output filename=E:\ORADATA\APQ1DATA\CNTRL03.DBF
RMAN-03022: compiling command: sql
RMAN-06162: sql statement: alter database mount
RMAN-03023: executing command: sql
RMAN-03022: compiling command: restore
RMAN-03025: performing implicit partial resync of recovery catalog
RMAN-03023: executing command: partial resync
RMAN-08003: starting partial resync of recovery catalog
RMAN-08005: partial resync complete
RMAN-03022: compiling command: IRESTORE

Here's the Veritas Log File:
VERITAS LOG from PBAPDQ1A
Restore started Fri Dec 07 13:55:13 2001
13:59:08 (110709.xxx) Restore job id 110709 will require 1 image.
13:59:08 (110709.xxx) Media id DRH636 is needed for the restore.
13:59:20 (110709.001) Restoring from image created 12/05/01 18:32:23
13:59:23 (110709.001) INF - Waiting for mount of media id DRH636 on server
semibkup1.
13:59:28 (110709.001) INF - Data socket = SEMIBKUP1.4057
13:59:28 (110709.001) INF - Name socket = SEMIBKUP1.3488
13:59:28 (110709.001) INF - New data socket = SEMIBKUP1.3833
13:59:28 (110709.001) INF - Backup time = 1007595143
13:59:29 (110709.001) INF - Encrypt = 0
13:59:29 (110709.001) INF - Use shared memory = 0
13:59:29 (110709.001) INF - Restore id = 110709.001
13:59:29 (110709.001) INF - Client read timeout = 300
13:59:29 (110709.001) INF - Media mount timeout = 0
13:59:30 (110709.001) INF - client = pbapdq1a
13:59:30 (110709.001) INF - requesting_client = pbapdq1a
13:59:30 (110709.001) INF - browse_client = pbapdq1a
14:01:04 (110709.001) INF - Waiting for positioning of media id DRH636 on server
semibkup1.
14:10:31 (110709.001) INF - Beginning restore from server semibkup1 to client
pbapdq1a.
14:11:11 (110709.001) Status of restore from image created 12/05/01 18:32:23 =
the requested operation was successfully completed
14:11:17 INF - Server status = 0
14:11:17 (110709.xxx) INF - Status = the requested operation was successfully
completed.

<Prev in Thread] Current Thread [Next in Thread>
  • [Veritas-bu] [Veritas-bu]Problem restoring an Oracle Database under OFS, Caddell, Scott <=