on 04-03-2008 8:36 AM
Dear Gurus,
I am planning to for the Migration of my SAP system.
My current configuration is
*Operating System :- Windows 2000 Server SP4*
*Hardware: - 64Bit Intel Xeon Processor 3.4 GHz*
*SAP :- R/3 4.7 Enterprise Ext.200 SR1, Kernel 6.40*
*Database :- INFORMIX.*]
New System Configuration :-
*Operating System :- Windows 2003 Server Standard edition 64 Bit R2*
*Hardware: - 64Bit Intel Xeon Processor 3.4 GHz*
*SAP :- R/3 4.7 Enterprise Ext.200 SR1, Kernel 6.40*
*Database :- ORACLE 10g*
I am planning to do the Heterogeneous system copy for migration.
I want more details about Migration/Heterogeneous System Copy.
Please guide me on this.
Thanks & Regards,
Sachin Sable
hi sachin,
SAP NetWeaver 04 SR1
Installation Guide
Homogeneous and
Heterogeneous
System Cop
Heterogeneous
System Copy for SAP Systems Based
on SAP Web AS 6.40 SR1
Purpose
When a homogeneous system copy is performed, the target SAP system is installed on the
same operating system and the same database system as the source SAP system. The
contents of the database are copied from the source to the target system.
During a heterogeneous system copy, the operating system or the database system is
changed. Migration is a synonym for heterogeneous system copy. Familiarize yourself with
the migration procedure. For detailed information, see:
SAP System Copy & Migration page on SAP Service Marketplace at
service.sap.com/systemcopy.
SAP OS/DB Migration page on SAP Service Marketplace at
service.sap.com/osdbmigration.
Additionally to the information contained on this page, check the SAP OS/DB Migration
Planning Guide that is available in the Media Library.
SAP Note 82478
A homogeneous system copy should only be done by a person with
experience in copying systems and with knowledge of the operating system,
the database, and the ABAP Dictionary.
A heterogeneous system copy must be done by a certified system support
consultant or a certified SAP Technical Consultant.
Create an SAPNet - R/3 Frontend message in the application area BC-INSUNX
(UNIX), BC-INS-NT (Windows), or BC-INS-AS4 (IBM eServer iSeries) if
you experience problems during the homogeneous system copy.
In the case of a heterogeneous system copy (OS/DB migration), create an
SAPNet - R/3 Frontend message in the application area BC-INS-MIG if you
experience problems.
Terminology
Homogeneous System Copy
The copy of the system is installed on the same operating system and database platform
as the original system.
Heterogeneous System Copy
During the copy, either the operating system or the database system is switched, or both.
Heterogeneous system copy is a synonym for migration.
Source System and Target System
The SAP system containing the original database is called the source system and the
system to which the database copy is to be imported is called the target system. Their
SAP system names are abbreviated to SOURCE_SAPSID and TARGET_SAPSID (IBM
1 Planning
9
eServer iSeries: source_<SID> and target_<SID>). The terms source database and
target database are also used in this description.
System Copy
Duplication of an SAP system. Certain SAP parameters may change in a copy. When a
system copy is made, all the instances are newly installed, but the database is set up
using a copy of the source system database.
Database Copy
Database-dependent part of the system copy.
Placeholders
Placeholders such as <SAPSID> are used in commands. They are used in the same way
as in the SAP system installation documentation, and must be replaced with the values
valid for your site.
The following additional placeholders are used:
Placeholder Meaning How to find out
<S_HOST> System name of the source host Command hostname
<T_HOST> System name of the target host Command hostname
<S_SAPSID> SAP system ID of the source system <SAPSID> of the original system
<T_SAPSID> SAP system ID of the target system <SAPSID> of the target system
<S_DBSID> Database ID of the source system <DBSID> of the original system
<T_DBSID> Database ID of the target system <DBSID> of the target system
Constraints
A system copy should only be done by a person with experience in copying systems
and with knowledge of the operating system, the database, and the ABAP Dictionary.
Client transport is not supported as a system copy method. Transporting production
clients is not supported at all. You can use client transport for the initial set up of a SAP
system infrastructure. The client copy procedure is not handled in this documentation.
Export and import of a database with the installation tools for reorganization purposes
is not described in this documentation, and is not supported by SAP. Use the
appropriate tools for database reorganization.
If you have made modifications in your development system, and want to copy your
quality assurance or production system onto the development system, see SAP Note
130906.
Copying data from SAP R/2 Systems to an SAP system based on SAP Web
Application Server ABAP is not described in this documentation. Copying data from
non-SAP systems to SAP systems is not described in this documentation either.
If you want to convert a non-Unicode systems to a Unicode systems or perform system
copy of a Unicode systems, see SAP Note 548016.
1 Planning
10
1 Planning
Purpose
Before you begin with the practical system copy tasks, it is essential to have a planning
phase in which you make a number of fundamental decisions that influence the subsequent
system copy procedure. Careful planning is a prerequisite for the successful system copy of
the system.
SAP recommends that you make a system copy in order to set up a test, demo, training or
standby system (Oracle and Informix: standby systems cannot be created with a system
copy). You should perform upgrades in a test system first. This way you can identify
customer-specific problems which might result from modifications.
The SAP system infrastructure (development, quality assurance and production system) can
be set up without making a system copy as follows:
Install all SAP systems (begin with the development system). Customize the
development system as described in the implementation documentation.
Transport the client-dependent and client-independent data to the quality assurance
and production systems.
However, if you do not follow this concept, you can also install a system, customize it and
then perform a system copy.
When copying a system which contains production data it is important to choose the right
moment for the copy. This could be a month-end or year-end closing. Make sure that there is
a well-defined starting point for the data in the new system.
Prerequisites
Read the following SAP Note before beginning the system copy. This note
contains the most recent information regarding the system copy.
SAP Note 784931 System Copy for SAP Systems Based on SAP Web AS
6.40 SR1
Make sure that you have the most recent version of the note. SAP Notes are
located on SAP Service Marketplace at service.sap.com/notes.
A consistent system copy can only be ensured if you perform all steps described in this
documentation.
The system copy is applicable for:
Setting up system landscapes (where the SAP systems have different SAPSIDs).
Creating systems for testing, demonstration, training and standby. Depending on the
purpose of the system, it may be advisable to use the same SAP system name, even
though the system cannot be included in a system group for transports in this case.
The SAP system release of the source and target systems must be the same.
Process Flow
...
1. Obtain the latest versions of the SAP Notes.
2. In the case of a major change in hardware configuration (for example, new machine
type, new hard disk configuration, new file system type), consult your SAP-authorized
hardware partner.
1 Planning
11
3. Create a plan for performing the system copy.
Take into account the downtime of the source system (for preparations and copying)
when planning the system copy.
4. Decide which homogeneous system copy method you want to use:
�� With database-specific tools (tools provided by the database vendor):
Some database vendors offer specific tools for copying a database. These tools allow
you to:
�� Restore a backup of one database (source DB) in another one (target
DB) (backup method)
�� Unload the source DB and load the data into the target DB.
�� With SAP tools (R3load procedure):
The system copy is carried out with SAP tools. This method should be used if
database-specific methods are either not available or not suitable.
Copying an SAP system with these tools is described in section R3load Procedures.
For a heterogeneous system copy, only the R3load procedure is available.
Neither of these methods is supported for all database systems. See the following table
to check which copy methods are available for your database system:
Database OS Platform Available Methods
UNIX Use one of the following:
The R3load method (see section R3load
Procedure on UNIX).
The MaxDB backup/restore method [Page 35]
MySQL MaxDB
In this
documentati
on we refer
to the
MySQL
MaxDB
database as
MaxDB from
now on.
Windows Use one of the following:
The R3load method (see section R3load
Procedure on Windows).
The MaxDB backup/restore method [Page 35]
IBM DB2
Universal
Database for
eServer iSeries
eServer iSeries Use the R3load method (see section R3load
Procedure on eServer iSeries) or the Save/Restore
Library method (see SAP Note 585277).
IBM DB2
Universal
Database for
UNIX and
Windows
UNIX or
Windows
Use one of the following:
The R3load method (see the corresponding
section R3load Procedure on UNIX or R3load
Procedure on Windows)
The backup method of IBM DB2 Universal
Database for UNIX and Windows is supported
for SAP systems based on SAP Web AS 6.40
SR1. For more information, see IBM DB2 UDB
for UNIX and Windows Specific Procedure
[Page 30].
1 Planning
12
IBM DB2
Universal
Database for z/OS
eServer zSeries Use one of the following:
The R3load method (see section R3load
Procedure on UNIX or R3load Procedure on
Windows).
The IBM DB2 Universal Database for z/OS
specific procedure as described in IBM DB2
UDB for z/OS Specific Procedure [Page 32].
UNIX Use one of the following:
The R3load method (see section R3load
Procedure on UNIX).
The Informix backup/restore method [Page 34].
Informix
Windows Use one of the following:
The R3load method (see section R3load
Procedure on Windows).
The Informix backup/restore method [Page 34].
UNIX Use one of the following:
The R3load method (see section R3load
Procedure on UNIX)
The R3load method with Export/Import
Monitors (see section R3load Procedure with
Export/Import Monitors)
The Oracle backup/restore method (see
section Database-Specific Procedures →
Oracle-Specific Procedure)
Oracle
Windows Use one of the following:
The R3load method (see section R3load
Procedure on Windows)
The Oracle backup/restore method (see
section Database-Specific Procedures →
Oracle-Specific Procedure, and SAP Note
676468).
The R3load method with Export/Import
Monitors (see section R3load Procedure with
Export/Import Monitors).
MS SQL Server Windows Use one of the following:
The R3load method (see section R3load
Procedure on Windows)
The Backup/Restore or Detach/Attach Method
(see section Database-Specific Procedures →
MS SQL Server-Specific Procedure).
5. Order the right version of the installation kit before starting the system copy.
6. Choose an SAP system name. The new SAP system name <TARGET_SAPSID> can
be chosen freely (during a new installation), but to meet the requirements of the
1 Planning
13
Workbench Organizer you must choose different SAP system names for different
systems.
7. Make sure that the versions of the SAP system and the installation tools are the same
on the target and source systems (exceptions are only allowed if they are described in
an SAP Note).
Several SAP systems can be operated on a single host without encountering
any problems. Nevertheless, SAP recommends that you use a separate host
for each system because an SAP system upgrade may depend on an OS
upgrade. If the SAP systems are on separate hosts, it is possible to upgrade
them at different times.
The source system must be in a consistent state before it can be copied.
8. Heterogeneous System Copy: Get Migration Key
If you change the operating system or the database system during the copy
(heterogeneous system copy), you need a migration key.
You can generate the migration key via SAP Service Marketplace at
service.sap.com/migrationkey. Enter the installation number of your
source system when your are requesting the key.
2 Preparations
14
2 Preparations
Before you start the system copy, you must perform steps to prepare the procedure.
Organizational Preparations
Required DVDs
Make sure that all required CDs / DVDs for the system copy are available. You do no
longer require a special migration CD. The tools are located on the SAP Installation
MasterDVD.
In case you want to perform a copy of a non-R/3 SAP system as a pilot project, see SAP
Note 543715 (APO, BW), SAP Note 693168 (IMIG) and SAP Note 611232 (released
CRM/EBP products).
Tool Versions
Check that you have the appropriate tool versions for your SAP Kernel.
SAP License
Once installation is complete and the SAP system copy has been imported, you will
require a new license key for the target system. The license key of the source system is
not valid for this system. You can order a new license key for the target system as
follows:
�� SAP Service Marketplace at service.sap.com.
�� Remote Connection to SAP support to request help with these tasks. For more
information see service.sap.com/remoteconnection.
�� Telefax
For more information, see SAP Note 94998.
Archive Files
You must make data archived in the source system (data that does not reside in the
database but was moved to a different storage location using SAP Archive Management)
accessible in the target system. Adapt the file residence information in the target system.
Refer to the SAP Online Documentation (SAP Library → Cross-Application Components
→ Archiving Application Data) for help.
Access to archive files is platform-independent.
Configuration Analysis / Hardware Analysis
The following factors have to be determined:
�� The number of application servers
�� The expected size of the database
�� Additional disks or other hardware required
�� Required memory
See Hardware and Software Requirements Check in the SAP system
installation documentation to determine the system requirements.
Test Run / Schedule
2 Preparations
15
�� Perform a test run of the system copy. The time taken by the test run is used to
calculate the system downtime:
�� If your target system will replace your source system, try to perform a
complete test run, meaning that the entire database is exported from the
source system, transferred to the target system and imported there.
Approximately system downtime will be equal to the total test time (that
is, time for export, transport and import).
�� If you do not want to replace your source system, a partial test run
(export of entire database or parts of it) may suffice to calculate the
system downtime. The source system will only be down for the time of
the export.
Calculating the system downtime is particularly important for very large
databases (VLDB) or when tapes are being used. The test run is also
done to determine the amount of export data. You should choose the
best data transfer method (for example, FTP or tape). We recommend to
perform read/write actions only on local file systems. Do not use NFSmounted
file systems, as:
�� Reading from NFS-mounted file systems may fail.
�� Writing to NFS-mounted file systems may cause corrupted dumps.
�� Define a schedule for the test migration and the final migration.
Technical Preparations
In order to make a consistent copy of the database, it is necessary to prepare the source
system and to carry out some subsequent actions on the target system. This is not necessary
when performing a test run.
The following list describes important preparatory actions. For further information on SAP
system administration, see the SAP Online Documentation.
Preparing the Source System
�� Before starting a system copy, check the minimal kernel patch level which is
required by the support package level of the source system. It may be
necessary to replace the SAP kernel which is delivered with the kernel CD of
the installation kit and installed during the installation of the target system by a
newer kernel patch level before starting the target system. If you have to replace
the delivered SAP kernel, you can do that after the installation of the central
instance.
�� No canceled or pending update requests should be in the system. Check this
via: Tools → Administration → Monitor → Update (SM13). If canceled or
pending updates exist, these must be updated again or deleted from all clients.
You can see whether canceled or pending updates exist by checking if table
VBDATA contains any entries. Proceed as follows in order to find the canceled or
open updates:
i. Call transaction SM13.
ii. Delete the default values for the client, user, and time.
iii. Choose all update requests.
If canceled or pending records exist, then these must be updated again
or deleted. Check whether this action was successful using transaction
SE16 for table VBDATA.
2 Preparations
16
�� Cancel all released jobs: Tools → CCMS → Jobs → Maintenance. This also
applies to the jobs which must run periodically (see SAP Note 16083). Select all
jobs (include start after event): Job → Schedule Job → Cancel.
�� Adapt the operation mode timetable to make sure that no switching of operating
modes takes place while a system is being copied: Administration → CCMS →
Configuration → Operation mode calendar.
�� Note the logical system names of all clients:
...
i. If you plan to overwrite an existing system with a system copy (i.e. the
source and target systems will both exist after the system copy), make
sure you note the logical system names of all clients in the system that
will be overwritten (transaction SCC4). As the logical system names will
be overwritten, in case of differences, they must be changed back to their
original names (as they existed in the system that is overwritten) in the
follow-on actions after the system copy.
ii. BW customers: When planning an SAP BW system copy, read SAP
Note 325525.
iii. If you create a new system with a system copy (i.e., create an upgrade
test system), the logical naming strategy for this new system should be
consistent with your existing logical system naming convention. See
OSS note 184447 if you are still planning your BW system landscape.
iv. If your system copy is used to replace hardware for the DB server,
migrate to a different database system or operating system (i.e., source
system for the copy is the same as the copy target), no changes to
logical system names are required.
�� Before the export, delete QCM tables from your system. Proceed as follows:
...
i. Before deleting you must always check
�� that the tables are consistent (no restart log or conversion procedure
termination must be displayed)
�� that the data of the original table is legible
�� If application programs do not run correctly which use the affected
original table, do not delete the QCM table yet.
ii. Start transaction SE14.
iii. Choose Extras → Invalid temp. table
All QCM tables that can be deleted are displayed.
iv. Mark the tables and delete them.
�� FI customers: You can perform an additional consistency check by running the
job SAPF190 before the system copy in the source system, as well as after the
copy in the target system, and then comparing the results. No customer data
may be changed in the meantime. (Accounting → Financial Accounting →
General ledger → Periodic Processing → Closing → Check/count →
Comparison)
�� FI customers: You can further check consistency by running the jobs RFUMSV00
(tax on sales/purchases), RAGITT01 (asset history sheet), RAZUGA01 (asset
acquisitions), RAABGA01 (fixed asset retirements) before the system copy in the
source system, as well as after the copy in the target system, and then
comparing the results. No customer data may be changed in the meantime.
2 Preparations
17
�� CO customers: You can perform an additional consistency check by running the
report group 1SIP before the system copy in the source system, as well as after
the copy in the target system, and then comparing the results. No customer data
may be changed in the meantime.
Prerequisites for an export:
Before performing an export, make sure that:
- No upgrade-prepare is performed.
- No incremental conversion is in progress.
To test if an incremental conversion is in progress, start the transaction ICNV.
If there are any table entries in the TICNV, an incremental conversion is in
progress. In this case, you have the following options:
1. Defer the migration until the incremental conversion has finished.
2. Try to finish the incremental conversion by performing the following steps:
If the tables are in state For conversion or in state Done, delete
the entries by choosing Control �� Delete Entry.
If the tables are in any other state, you have to finish the incremental
conversion. Choose the button Assistant and proceed according to
the Online Documentation.
Only Heterogeneous System Copy:
Before you start the export of your source system, make sure that the tables
TATGPC and TATGPCA are empty. To do so, use your database utility and
delete the contents of these tables with the following statements:
DELETE from TATGPC
DELETE from TATGPCA
Normally both tables are empty. If you do not delete the contents of these
tables you will encounter problems while importing the data to your target
system because of non NULL capable fields in these tables.
3 Database-Specific Procedures
18
3 Database-Specific Procedures
The following sections describe database-specific procedures for the homogeneous system
copy. Database-specific methods are not available for all database systems. For information
on the availability, see the section Planning [Page 10] in this documentation and check the
SAP Service Marketplace for SAP Notes describing the homogeneous system copy for your
database system.
3 Database-Specific Procedures
19
3.1 Oracle-Specific Procedure
Purpose
In an SAP system environment, you can create a homogeneous copy of an Oracle database
by copying database files. This method is suitable for creating an exact copy of an existing
database. The source of the copy can be an offline backup or the file system of your source
host.
SAPinst is used for the installation on the target system host as described in the installation
documentation for your SAP component. Only the SAPinst steps for setting up and loading
the database steps are modified.
As of SAP Web AS 6.40 SR 1, the new OraBRCopy Java tool replaces the former R3COPY
shell script. It works on Unix and Windows as well. The tool generates a file CONTROL.SQL
into the current working directory, which can be used without further adaptions on the target
system..
The tool is located on the SAP Installation Master DVD:
Unix:
(<DVD-DIR>:/IM_<xx>/SAPINST/UNIX/COMMON/INSTALL/ORA/ORABRCOPY.SAR)
Windows:
(<DVD-DIR>:\IM_<xx>\SAPINST\NT\COMMON\INSTALL\ORA\ORABRCOPY.SAR)
For a detailed description of the OraBRCopy tool refer to the delivered documentation
ORABRCopy.pdf which is part of the OraBRCOPY.SAR archive.
Advantages
Existing offline backups can be used (provided redo logs were cleaned up with forced
log switches).
Procedure is faster than the R3load method.
Disadvantages
Offline backup/copy of database files in a heterogeneous environment is not possible as
hardware of the source and target systems must be binary compatible.
We recommend that you use different host for the source system and the target system.
SAP system and database must be shut down during offline backup/copy of database
files.
You cannot change the database schema and the tablespace names.
Prerequisites
Make sure that the following requirements are met:
You use the same Oracle release and patch level for your database in the source and
target system.
If you want to upgrade your database from 32-bit to 64-bit, add the following lines at the
bottom of the control.sql file:
shutdown immediate;
startup restrict
3 Database-Specific Procedures
20
spool utlirp.log
@?/rdbms/admin/utlirp.sql
spool off
alter system disable restricted session;
You must have installed JRE version 1.4.1 or higher on your system
The JAVA_HOME environment variable must point to the JRE directory.
The classes12.jar must exist in the <ORACLE_HOME>/jdbc/lib directory
(installed using a standard Oracle installation).
The backup must be done offline.
The source and target systems must run on different hosts for security reasons.
The source and target systems must be binary compatible.
If you use Windows, note that the source or target system must be 32 or 64-
bit systems (I386 or IA64). You can copy from 32-bit systems to 64-bit
systems and vice versa.
UNIX only: The hard disk type and the disk management system of the source and
target systems must be the same.
Make sure that all redo log groups are archived.
If your source system uses the US7ASCII character set, you must choose this character set
when installing the target system. SAPinst prompts for the character set during the installation
(key: Database Character Set). The installation default is WE8DEC or UTF8 for Unicode
systems. To find out the character set used by the source system, connect to the source
database as user sap<schemaid> or sapr3 with sqlplus and enter: SELECT * FROM
V$NLS_PARAMETERS;
Process Flow
...
1. Generate the control file structure for the target database on the source system.
2. Prepare the target system.
3. If required, create an offline backup of the source database.
4. Restore the data files, log files, trace file and structure information file on the target
system by using the offline backup or by copying the listed files directly from the source
host.
5. Restart SAPinst.
Result
You finished this part of the system copy. To complete the system copy, you have to perform
the steps in section Final Activities [Page 81].
3 Database-Specific Procedures
21
3.1.1 Generating the Control File Structure
Prerequisites
We recommend that you shut down the SAP system before you perform the
following steps. The database must still be running.
Procedure
...
1. Create an installation directory <INSTDIR> (UNIX: with permissions 777) on the source
system.
2. Copy the ORABRCOPY.SAR archive from the SAP Installation Master DVD to the
installation directory and extract it using SAPCAR:
You find the archive in the following directory:
�� For Unix:
(<DVDDIR>:/
IM_<xx>/SAPINST/<UNIX>/COMMON/INSTALL/ORA/ORABRCOPY.SAR)
�� Windows:
(<DVDDIR>:\
IM_<xx>\SAPINST\NT\COMMON\INSTALL\ORA\ORABRCOPY.SAR)
3. Start the OraBRCopy tool as an OS user with Oracle DBA privileges (user ora<dbsid>
on UNIX, user <sapsid>adm on Windows):
�� On UNIX, enter the following commands:
./ora_br_copy.sh targetSid <TARGET_DBSID> -password <systems
password> -listenerPort <listener port>
�� On Windows, enter the following commands:
ora_br_copy.bat targetSid <TARGET_DBSID> -password <systems
password> -listenerPort <listener port>
The tool will create the files CONTROL.SQL, CONTROL.TRC and
init<targetSID>.ora in your installation directory, shutdown and restart the
database and perform the required log switches.
If an error occurs, check the log file:
<INSTDIR>/ora.brcopy.log
4. Verify and, if necessary, update the CONTROL.SQL control file using the CONTROL.TRC
trace file as follows:
3 Database-Specific Procedures
22
Example for Windows
In the following example for Windows, entries of CONTROL.SQL written in bold
should be compared to the trace file:
REM
==========================================================
==========
REM CONTROL.SQL
REM
REM SAP AG Walldorf
REM Systeme, Anwendungen und Produkte in der
Datenverarbeitung
REM
REM (C) Copyright SAP AG 2004
REM
==========================================================
==========
REM Generated at:
REM Fri Sep 17 08:33:25 CEST 2004
REM for target system NEW
REM on
REM Windows 2000 5.0 x86
CONNECT / AS SYSDBA
STARTUP NOMOUNT
CREATE CONTROLFILE REUSE
SET DATABASE "NEW"
RESETLOGS
ARCHIVELOG
MAXLOGFILES 255
MAXLOGMEMBERS 3
MAXDATAFILES 1022
MAXINSTANCES 50
MAXLOGHISTORY 1134
LOGFILE
GROUP 1 (
'D:\ORACLE\NEW\ORIGLOGA\LOG_G11M1.DBF',
'D:\ORACLE\NEW\MIRRLOGA\LOG_G11M2.DBF'
) SIZE 50M,
GROUP 2 (
'D:\ORACLE\NEW\ORIGLOGB\LOG_G12M1.DBF',
'D:\ORACLE\NEW\MIRRLOGB\LOG_G12M2.DBF'
3 Database-Specific Procedures
23
) SIZE 50M,
GROUP 3 (
'D:\ORACLE\NEW\ORIGLOGA\LOG_G13M1.DBF',
'D:\ORACLE\NEW\MIRRLOGA\LOG_G13M2.DBF'
) SIZE 50M,
GROUP 4 (
'D:\ORACLE\NEW\ORIGLOGB\LOG_G14M1.DBF',
'D:\ORACLE\NEW\MIRRLOGB\LOG_G14M2.DBF'
) SIZE 50M
DATAFILE
'D:\ORACLE\NEW\SAPDATA1\SYSTEM_1\SYSTEM.DATA1',
'D:\ORACLE\NEW\SAPDATA3\IMS_1\IMS.DATA1',
'D:\ORACLE\NEW\SAPDATA3\IMS_2\IMS.DATA2',
'D:\ORACLE\NEW\SAPDATA3\IMS_3\IMS.DATA3',
'D:\ORACLE\NEW\SAPDATA3\IMS_4\IMS.DATA4',
'D:\ORACLE\NEW\SAPDATA4\IMS_5\IMS.DATA5',
'D:\ORACLE\NEW\SAPDATA4\IMS_6\IMS.DATA6',
'D:\ORACLE\NEW\SAPDATA4\IMS_7\IMS.DATA7',
'D:\ORACLE\NEW\SAPDATA4\IMS_8\IMS.DATA8',
'D:\ORACLE\NEW\SAPDATA4\IMS_9\IMS.DATA9',
'D:\ORACLE\NEW\SAPDATA1\IMS640_1\IMS640.DATA1',
'D:\ORACLE\NEW\SAPDATA1\IMS640_2\IMS640.DATA2',
'D:\ORACLE\NEW\SAPDATA1\IMS640_3\IMS640.DATA3',
'D:\ORACLE\NEW\SAPDATA1\IMS640_4\IMS640.DATA4',
'D:\ORACLE\NEW\SAPDATA2\IMS640_5\IMS640.DATA5',
'D:\ORACLE\NEW\SAPDATA2\IMS640_6\IMS640.DATA6',
'D:\ORACLE\NEW\SAPDATA2\IMS640_7\IMS640.DATA7',
'D:\ORACLE\NEW\SAPDATA2\IMS640_8\IMS640.DATA8',
'D:\ORACLE\NEW\SAPDATA2\IMS640_9\IMS640.DATA9',
'D:\ORACLE\NEW\SAPDATA3\IMS640_10\IMS640.DATA10',
'D:\ORACLE\NEW\SAPDATA4\IMS640_11\IMS640.DATA11',
'D:\ORACLE\NEW\SAPDATA1\IMSUSR_1\IMSUSR.DATA1',
'D:\ORACLE\NEW\SAPDATA2\ROLL_1\ROLL.DATA1'
;
ALTER DATABASE OPEN RESETLOGS;
ALTER TABLESPACE PSAPTEMP ADD TEMPFILE
'D:\ORACLE\NEW\SAPDATA3\TEMP_1\TEMP.DATA1'
SIZE 350M REUSE AUTOEXTEND OFF;
3 Database-Specific Procedures
24
This is an example for Windows, but the following changes to be made are
valid for UNIX, too.
Changes to be made:
a. MAXLOGFILES 255
The numbers must be greater or equal to the corresponding numbers in the trace file.
b. GROUP 1 (
'D:\ORACLE\NEW\ORIGLOGA\LOG_G11M1.DBF',
'D:\ORACLE\NEW\MIRRLOGA\LOG_G11M2.DBF'
) SIZE 50M,
Group 2 (
The sizes of the respective groups must be equal to the sizes of the corresponding
groups in the trace file.
c. 'D:\ORACLE\NEW\SAPDATA1\SYSTEM_1\SYSTEM.DATA1',
'D:\ORACLE\NEW\SAPDATA3\IMS_1\IMS.DATA1',
'D:\ORACLE\NEW\SAPDATA1\IMS640_1\IMS640.DATA1'
The count of the data files must be equal to the count of the corresponding data files in
the trace file.
d. ALTER TABLESPACE PSAPTEMP ADD TEMPFILE
'D:\ORACLE\NEW\SAPDATA3\TEMP_1\TEMP.DATA1'
SIZE 350M REUSE AUTOEXTEND OFF;
The size must be equal to the corresponding size in the trace file.
e. The number of the rows with ALTER TABLESPACE must be equal to the number of the
corresponding rows in the trace file.
3.1.2 Preparing the Target System
Prerequisites
Make sure that the sapdata<n> file systems on the target system host are large enough.
Procedure
...
1. Install the target SAP system with SAPinst as described in the installation
documentation for your SAP component.
When you install the central instance and the database instance, make sure that
you cannot change the database schema names. The schema names of the
source and target system must be identical.
...
�� When SAPinst prompts for the installation type, choose:
3 Database-Specific Procedures
25
System Copy / Oracle Backup/Restore.
�� Install until SAPinst stops to restore the database files on the target system.
The following message is displayed:
SAPinst now stops the installation. Please proceed as
follows:....
2. UNIX: Install the Oracle database software on your target system:
a. If necessary, extract the Oracle stage archives manually.
b. Install the Oracle Software as described in the installation documentation for
your SAP component.
3. Create the following directories on the target system, if they do not exist:
�� UNIX:
�� /oracle/<TARGET_DBSID>/mirrlog<x>
�� /oracle/<TARGET_DBSID>/origlog<x>
�� /oracle/<TARGET_DBSID>/sapdata<x>
�� /oracle/<TARGET_DBSID>/sapreorg
�� /oracle/<TARGET_DBSID>/saparch
�� /oracle/<TARGET_DBSID>/oraarch
�� /oracle/<TARGET_DBSID>/saptrace
�� /oracle/<TARGET_DBSID>/saptrace/background
�� /oracle/<TARGET_DBSID>/saptrace/usertrace
�� /oracle/<TARGET_DBSID>/origlogA/cntrl
�� /oracle/<TARGET_DBSID>/sapdata1/cntrl
�� /oracle/<TARGET_DBSID>/saparch/cntrl
�� /oracle/<Target_DBSID>/sapcheck
�� Windows:
�� <drive>:\oracle\<TARGET_DBSID>\mirrlog<x>
�� <drive>:\oracle\<TARGET_DBSID>\origlog<x>
�� <drive>:\oracle\<TARGET_DBSID>\sapdata<x>
�� <drive>:\oracle\<TARGET_DBSID>\sapreorg
�� <drive>:\oracle\<TARGET_DBSID>\saparch
�� <drive>:\oracle\<TARGET_DBSID>\oraarch
�� <drive>:\oracle\<TARGET_DBSID>\saptrace
�� <drive>:\oracle\<TARGET_DBSID>\saptrace\background
�� <drive>:\oracle\<TARGET_DBSID>\saptrace\usertrace
�� <drive>:\oracle\<TARGET_DBSID>\origlogA\cntrl
3 Database-Specific Procedures
26
�� <drive>:\oracle\<TARGET_DBSID>\sapdata1\cntrl
�� <drive>:\oracle\<TARGET_DBSID>\saparch\cntrl
�� <drive>:\oracle\<Target_DBSID>\sapcheck
Note that for Windows, one of the cntrl files can also be located in the
directory
<drive>:\oracle\<TARGET_DBSID>\sapdata1\system_1\cntrl
4. Make sure that the following directories are empty (except the subdirectory
saparch/cntrl).
�� UNIX: /oracle/<TARGET_DBSID>/saparch and
/oracle/<TARGET_DBSID>/oraarch
�� Windows: <drive>:\oracle\<TARGET_DBSID>\saparch and
<drive>:\oracle\<TARGET_DBSID>\oraarch
5. UNIX: All directories must be owned by the user ora<target_dbsid>.
To achieve this enter the following command:
chown ora<target_dbsid>:dba<directory>
6. Windows: For all Oracle directories set the security settings for the built-in accounts
and groups SYSTEM, Administrators, SAP_<SAPSID>_GlobalAdmin (domain
installation), and SAP_<SAPSID>_LocalAdmin (local installation) as follows:
a. In the Windows Explorer, right-click the Oracle root directory and choose
Properties.
b. Under Security, choose Advanced.
c. Uncheck Allow inheritable permissions from the parent . (Windows Server
2003), or Inherit from parent the permission entries that apply to child objects
(Windows 2000).
d. In the upcoming dialog, choose Copy, to copy the permission entries that were
previously applied from the parent to this object.
e. Choose OK.
f. Set the permissions for the above-mentioned accounts SYSTEM,
Administrators, SAP_<DBSID>_GlobalAdmin, or
SAP_<DBSID>_LocalAdmin to Full Control.
g. Delete all other accounts.
3 Database-Specific Procedures
27
3.1.3 Creating an Offline Backup
Create an offline backup if required. There are different possibilities to prepare the actual
transfer of the database files:
If you have an existing offline backup which is up-to-date, you can use it (provided redo
logs were cleaned up with forced log switches).
If you want to transport the database file (for example, on tape) or if you have to
perform the database shut-down at a certain time, stop the database (normal shutdown)
and perform a complete offline backup. You can use the trace file
CONTROL.TRC created by OraBrCOPY to determine the file system trees that have to
be saved.
Otherwise, stop the database (normal shutdown) and copy the database files when the
actual transfer to the target system takes place. You do not have to perform any
preparations for the actual transfer now. Proceed with the next step.
3.1.4 Restoring the Database Files on the Target
System
If you do not use an offline backup but copy the database files directly from
the source to the target system host, make sure that you shut down the
database on the source system before you copy the listed files from the
source to the target directories. ...
1. Copy the following files from the source to the target system host either by using an
offline backup or by copying the listed files from the source directories to the target
directories:
The table shows the directories on Unix.
For Windows you have the corresponding directories, for example
<drive>:\oracle\<DBSID>\sapdata<x>:
Source and Target Directory Files
/oracle/<DBSID>/sapdata<x> All files
/oracle/<DBSID>/origlog<x> All files
/oracle/<DBSID>/mirrlog<x> All files
source: <INSTDIR>
target: <INSTDIR>
CONTROL.SQL
source: <INSTDIR>
target:
/oracle/<DBSID>/<DB_VERSION>_<BIT>/dbs
init<TARGET_DBSID>.ora
If you use an existing offline backup, the source data files and log files
are not located in the directories shown in the table.
On Windows, the installation directory of the target system is normally
located in the directory:
3 Database-Specific Procedures
28
%programfiles%\sapinst_instdir\NW04SR1\WebAS_ABAP_O
RA_NUC\DB
2. Windows: After you have copied the database files, make sure that the files on the
source and target system are not located in different directories or drives. If required,
make the corresponding changes in the control.sql and the init<DBSID>.ora
file.
3. UNIX: Verify that the created directories and copied files have the owner
ora<target_dbsid>, belong to the group dba, and have the permissions 740.
4. Make sure that the control files are not restored. If necessary, remove them.
The file names are specified by the parameter control_files of the
init<TARGET_DBSID>.ora file.
3.1.5 Restarting SAPinst
Procedure
...
1. Restart SAPinst. Continue as described in the installation documentation for your SAP
component.
2. Request a new license key from SAP. For more information, see the installation
documentation for your SAP component, section Post-Installation Activities - Installing
and Using the SAP License and SAP Note 94998.
3 Database-Specific Procedures
29
3.2 MS SQL Server-Specific Procedure
Purpose
In an SAP system environment, you can create a homogeneous system copy of an MS SQL
Server database by using the backup/restore method or the detach/attach method. For more
information, see SAP Notes 193816 and 151603. The SAPinst installation tool supports both
methods.
The backup/restore method and the detach/attach method have the following advantages
compared to the R3load method:
You can use an existing backup.
These methods are much faster than the R3load method.
Process Flow
3. You run SAPinst to install the central instance.
4. You back up or detach the MS SQL Server database that you want to copy.
5. You restore or attach the database on the target database server instance.
6. You download the SAP Tools for MS SQL Server from SAP Service Marketplace at
service.sap.com/msplatforms → Microsoft → SQL Server.
For more information, see SAP Note 683447.
7. You run SAP Tools for MS SQL Server, choose Migration setup in the root menu, and
follow the dialogs.
Alternatively, you can perform the Postprocessing step of SAP Note 151603.
However, this involves many manual steps. It is much easier to make a system
copy with SAP Tools for MS SQL Server.
3 Database-Specific Procedures
30
3.3 IBM DB2 UDB for UNIX and Windows
Specific Procedure
Purpose
In an SAP system environment, you can create a homogeneous system copy of a DB2
database using the backup method or by relocating your database. The relocation of the
database is usually used in conjunction with split mirror (for more information, see SAP Note
594301 and the DB2 documentation). This section provides information on the backup
method.
SAPinst is used for the installation on the target system host as described in the installation
documentation for your SAP component. Only the SAPinst steps for setting up and loading
the database steps are replaced by a database restore.
Advantages of the Backup Method
You can use existing offline backups.
Using the backup method is faster than the R3load method.
Disadvantages of the Backup Method
An offline backup or copy of database files in a heterogeneous environment is not possible as
the hardware of the source and target systems must be binary compatible.
With DB2 UDB for UNIX and Windows, it is possible to use backup images
cross platform for AIX, Solaris and HP-UX.
Prerequisites
For security reasons, the source and target systems should run on different hosts.
If your source and target system reside on the same host, you have to use the
redirected restore or relocate your database.
The source and target systems should be binary compatible.
If errors occur when restoring the backup on the target system, the complete restore
must be repeated.
Process Flow
...
1. You perform an offline backup or restore an existing backup copy. For the restore of
your database, you can choose between one of the following options:
�� Simple database restore
To perform a database restore, use the DB2 restore command. For more information
see the IBM DB2 documentation DB2 Command Reference.
�� Redirected restore
To perform a redirected restore, use tool brdb6brt that retrieves a database backup
and creates a CLP script for the restore of this backup image. For more detailed
information on how to use tool brdb6brt, see the documentation Database
Administration Guide: SAP on IBM DB2 Universal Database for UNIX and Windows
section Usage of Tool brdb6brt and the IBM DB2 documentation DB2 Command
Reference.
3 Database-Specific Procedures
31
Be aware of the following constraints, when using the backup method for a
homogeneous system copy:
You cannot change the connect user. During the dialog phase you have
to make sure, that you enter exactly the name of the connect user like you
did on your source system.
The tablespace names remain the same during the database restore.
However, you may change them after the installation.
If you want to change the container names on the target system, you have
to adapt the container names in the redirected restore script and then
perform a redirected restore. For more information, see the
documentation Database Administration Guide: SAP on IBM DB2
Universal Database for UNIX and Windows, section Usage of Tool
brdb6brt.
2. You run SAPinst to install the central instance.
3. You run SAPinst to install the database instance.
During the installation phase you will be prompted to perform the database restore. For
more information, see SAP Note 784931.
4. Perform the database restore and continue with the installation.
5. If required, you can modify the tablespace names after the installation using the
following command:
db2 rename tablespace <old name> to <new name>
db2 rename tablespace <SAPSID_SOURCE>#STABD to
<SAPSID_TARGET>#STABD
In addition, you have to update the tablespace names in tables TSDB6, IADB6 and
TADB6 using the following commands:
�� For table TSDB6, enter the following SQL command:
update <connect_user_name>.tsdb6 set tabspace =
'<SAPSID_TARGET>#'||substr(tabspace,5,length(tabspace)-4),
indspace='<SAPSID_TARGET>#'||substr(indspace,5,length(indspace)
-4)
�� For table IADB6, enter the following SQL command:
update <connect_user_name>.iadb6 set tabspace =
'<SAPSID_TARGET>#'||substr(tabspace,5,length(tabspace)-4)
�� For table TADB6, enter the following SQL command:
update <connect_user_name>.tadb6 set tabspace =
'<SAPSID_TARGET>#'||substr(tabspace,5,length(tabspace)-4)
See also:
Database Administration Guide: SAP on IBM DB2 Universal Database for UNIX and
Windows
IBM DB2 documentation → DB2 Command Reference
3 Database-Specific Procedures
32
3.4 IBM DB2 UDB for z/OS Specific
Procedure
Purpose
In an SAP system environment, you can create a homogeneous system copy of a DB2
database using the offline system copy method.
Advantage of the Offline System Copy Method
This method is faster than the R3load method.
Restriction of the Offline System Copy Method
At the moment, you cannot copy an individual MCOD component to another system. You can
only copy the complete system.
The offline system copy must be performed by an experienced database
administrator.
For more information, see SAP Note 680746.
3 Database-Specific Procedures
33
3.5 IBM DB2 UDB for iSeries Specific Procedure
Purpose
In an SAP system environment, you can create a homogeneous system copy of a DB2
database using the SAV/RSTLIB system copy method.
Advantage of the Offline System Copy Method
This method is faster than the R3load method.
For more information, see SAP Note 585277.
3 Database-Specific Procedures
34
3.6 Informix-Specific Procedure
Purpose
In an SAP system environment, you can create a homogeneous system copy of an Informix
database using the backup and restore method. This is possible for databases running on
UNIX or Windows operating systems. The SAP installation tool, SAPinst, supports the system
copy.
Advantages of the Backup and Restore Method
You can use an existing backup.
This method is faster than the R3load method.
Disadvantages of the Backup and Restore Method
Source system host and target system host should be different.
Prerequisites
You can find the documentation SAP Database Guide: Informix referred to below at:
help.sap.com → Documentation → SAPNetWeaver → English → SAP NetWeaver →
Application Platform (SAP Web Application Server) → Databases → IBM Informix
Process Flow
...
1. You back up the Informix database that you want to copy.
For more information on database backup, see the Informix documentation or the SAP
documentation SAP Database Guide: Informix.
2. You run SAPinst to install the central instance.
3. You run SAPinst to install the database instance.
During the installation phase SAPinst prompts you to select the database instance
installation method in screen Selecting the Database Instance installation Method.
4. You choose System Copy / Restore of the database using a database specific method.
SAPinst displays the following message and waits for you to perform the restore.
Please restore your Informix database NOW. After you have completed this and the
database is online again, press OK.
5. You restore your Informix database.
For more information on database restore, see the Informix documentation or the SAP
documentation SAP Database Guide: Informix.
6. If required, you rename your target database system ID, T_DBSID.
7. You drop the table SAPUSER in the target database system.
8. You choose OK in SAPinst to continue.
3 Database-Specific Procedures
35
3.7 MaxDB-Specific Procedure
Purpose
In an SAP system environment, you can create a homogeneous copy of a MaxDB database
by using the backup and restore method. This method is suitable for creating an exact copy
of an existing database. The source of the copy is a complete data backup of your source
database.
The SAPinst tool is used for the installation on the target system host as described in the
installation documentation for your SAP component. In SAPinst you select the backup and
restore method as the database installation method. SAPinst stops before the database
instance initialization and asks you to perform the recovery on the target database. After you
have performed recovery and post-recovery activities you can continue the installation in
SAPinst.
This description is not valid for the liveCache system copy.
Prerequisites
CPU architecture
You can use the backup and restore method to copy systems to the same CPU
architecture. That is, you can copy a system based on a swap byte architecture to
another system based on swap byte architecture. You can also copy a RISC-based
system to another RISC-based system. For example, the following combinations are
possible:
�� HP-UX <-> Sun Solaris
�� Sun Solaris <-> AIX
�� Windows <-> HP Tru64
�� Windows <-> Linux
�� Windows 32-bit <-> Windows 64-bit
However, you cannot copy between the following systems:
�� HP-UX <-> Windows NT
�� HP Tru64 <-> Sun Solaris
You can copy backups from 32-bit systems to 64-bit systems.
Data backup
You perform the complete data backup of your source database.
Recovery tool
You are using the MaxDB Database Manager (DBMGUI) version 7.5.0 Build 12 or above.
You can find more information on DBMGUI at either of the following:
�� dev.mysql.com/doc → MaxDB by MySQL → MaxDB Online Library →
Tools
�� help.sap.com → Documentation → SAPNetWeaver → English → SAP
NetWeaver → Application Platform (SAP Web Application Server) →
Databases → MySQL MaxDB → Tools → Database Manager GUI
Database software
3 Database-Specific Procedures
36
The database software on the target host must have the same version as the software on
the source host. The build number of the software version on the target host must be
greater than or equal to the version on the source host.
Size of the data on the target system
The size of the target system must be greater than the used space on the source system.
You can find the size of the used pages on the source system as follows:
dbmcli d <database_name> -u <dbm_user>,<password> -n
<database_server> -u SQL sap<sid>,<password> sql_execute 'SELECT
USEDPERM FROM SERVERDBSTATISTICS'
The result of this query is the amount of used space, expressed as the number of 8 KB
pages. By dividing this value by 128 you can get the used space expressed in MB. When
SAPinst prompts you, configure the database data volumes according to this value.
Process Flow
...
1. You do the following on the source system:
a. You create a complete data backup using the DBMGUI tool:
DBMGUI → Backup → Backup Wizard → Complete
b. You make the backup medium available on the target host.
2. You do the following on the target system:
a. You start SAPinst to configure the database instance parameters.
SAPinst stops before database initialization and asks you to perform the data
recovery.
b. You start the data recovery wizard from DBMGUI:
i. You register your database instance in the DBMGUI
ii. You check the database instance is in the admin state.
iii. You choose Recovery → Recovery with Initialization
iv. In type of recovery you select Restore a medium.
v. You specify the backup medium.
vi. You start the restore procedure.
The recovery wizard does not start the recovery immediately. It initializes the
database instance first. It takes some time for the database server to format
the database volumes.
c. After the restore, you check the state of the target database instance. Change
the database state to online if it is not already in online state.
d. You delete the entries from the following tables to make sure that information
about the backup history for update statistics in the Computing Center
Management System (CCMS) from the old system does not appear in the new
system:
CNHIST, CNREPRT, CNMEDIA, DBSTATHADA, DBSTAIHADA, DBSTATIADA,
DBSTATTADA, SDBAADAUPD
e. Reload the system tables in the target system as follows:
dbmcli -d <NEW_DBSID> -u control,control load_systab -u
superdba,admin -ud domain
3 Database-Specific Procedures
37
f. You change the database user SAP<OLD_SID> to SAP<NEW_SID>. The name
of the database user must be exactly SAP<SID>, where <SID> is the system ID
of the SAP system to be installed. The password of the user must be sap. This
is necessary for the post-installation steps executed by SAPinst.
You change the database user as follows:
dbmcli -d <NEW_DBSID> -u control,control u SQL superdba,admin
sql_execute rename user SAP<OLD_SID> to SAP<NEW_SID>
You change the password of the SAP<SID> database user as follows:
dbmcli d <NEW_DBSID> -u control,control u SQL superdba,admin
sql_execute alter password SAP<NEW_SID> sap
g. You continue with SAPinst or restart it if you stopped it during the recovery.
h. After installation is completed you maintain the database connection for CCMS.
For more information, see SAP Note 588515.
4 R3load Procedures
38
4 R3load Procedures
Purpose
With the SAP installation tool SAPinst, you can export and import your database in a
database-independent format. The procedure generates a database export of all SAP objects
that are defined in the ABAP Dictionary.
Prerequisites
R3load Restrictions
Keep the following restrictions of R3load procedures in mind:
SAPinst generates a database dump of all SAP objects that are defined in the ABAP
Dictionary. Other objects are not exported by SAPinst.
Changes to database objects that cannot be maintained in the ABAP Dictionary
(transaction SE14), such as the distribution of tables over several
tablespaces/dbspaces, are lost after the system copy.
No indexes longer than 18 characters are allowed on the database to be exported.
For a consistent database export, no transactions on export-relevant database objects
are allowed during the export. Otherwise, the export has to be restarted. Therefore, it is
recommended to shutdown the SAP system for the export. The database must still be
running.
Considerations concerning the System Copy Tools
Every installation service (dialog instance installation, for example) must have its own
separate installation directory whenever you start SAPinst.
If the target system is already existing and if you do NOT plan to perform a MCOD
installation, delete the database on the target system before the import according to
the corresponding description in section Additional Information of the installation
documentation for your SAP component.
If the database configuration of your database is stored in the file system, it is advisable
to backup these configuration files before deleting the database.
Splitting STR Files
During the standard system copy process, all tables of the SAP system are grouped
into packages, whereby all tables with the same data class belong to the same
package. The processing unit for one unload/load process is a package. The packages
usually differ in number and size of contained tables, resulting in varying unload/load
runtimes. The overall runtime can be reduced by creating packages of the same size,
that is creating packages with a similar processing time. Splitting the default packages
(one package per data class) into more and smaller pieces will help to reach this aim.
There are several options of how to split packages. For a detailed description of the
options, see the F1 help about the parameters prompted on the screen Split STR Files
while running SAPinst to export the database. The options can be used separately or -
when using the new Java based splitting tool - combined.
'Splitting of STR Files' is part of the 'Advanced Export Parameters' and is disabled by
default. If you select the splitting option and unless you did not already perform some
tests, using the splitting tool parameters selected by SAPinst is a good starting point.
4 R3load Procedures
39
If you want to split STR file, it is mandatory that EXT files for the target
database system are created before. You will find the EXT files in your export
dump directory, subdirectory DB/<DBTYPE>, for example DB/ORA.
Process Flow
On UNIX, see R3load on UNIX [Page 40].
On Windows, see R3load on Windows [Page 47].
On IBM eServer iSeries, see R3load on IBM eServer iSeries [Page 52]
4 R3load Procedures
40
4.1 R3load Procedure on UNIX
Purpose
This section describes the R3load system copy procedure for Oracle, Informix, IBM DB2 UDB
for UNIX and Windows, IBM DB2 UDB for z/OS, and SAP DB on UNIX platforms.
Process Flow
The R3load procedure consists of the following steps:
...
1. Heterogeneous system copy: Generate the migration key via SAP Service
Marketplace.
You need a migration key for a heterogeneous system copy.
You can generate the migration key required for the heterogeneous system
copy via SAP Service Marketplace at service.sap.com/migrationkey.
2. Export the source database:
a. Make sure that the QCM tables are deleted from your system [Page 14].
b. IBM DB2 UDB for UNIX and Windows, Informix and Oracle only:
Set the library path environment variable [Page 40].
c. Generate DDL statements [Page 55].
d. Run SAPinst to export the database [Page 42].
3. Set up the target system [Page 60].
Result
You finished this part of the system copy. To complete the system copy, you have to perform
the steps in section Final Activities [Page 81].
4.1.1 Setting the Library Path Environment
Variable
Use
This section is only valid for IBM DB2 UDB for UNIX and Windows, Informix and Oracle.
You need to set the library path environment variable of user root before starting SAPinst.
Procedure
As user root, set the library path environment variable on your installation host according to
the following tables:
<ORACLE_HOME> is the home directory that you set up for the oracle instance
<DBSID>: /oracle/<DBSID>/920_32 if your operating system is of 32 bit,
or /oracle/<DBSID>/920_64 if your operating system is of 64 bit.
Value of Library Path Environment Variable
4 R3load Procedures
41
DB
Version
Operating
System
Variable Value
AIX /usr/sap/<SAPSID>/SYS/exe/run:.
Linux (32
Bit); Linux
IA64 (64
Bit);
Solaris
/usr/sap/<SAPSID>/SYS/exe/run:.
IBM
DB2
UDB for
UNIX
and
Window
s
HP-UX: /usr/sap/<SAPSID>/SYS/exe/run:/opt/IBM/db2/V8.1/lib64:.
Informix All UNIX
operating
systems
/informix/<SAPSID>/lib:/ \
informix/<SAPSID>/lib/esql
HP Tru64
UNIX
<sapmnt>/<SAPSID>/exe: \
/<ORACLE_HOME>/lib:/oracle/client/ \
92x_64/lib
Linux <sapmnt>/<SAPSID>/exe: \
/<ORACLE_HOME>/lib
Oracle
8.1.7
Oracle
9.2.0
All other
UNIX
operating
systems
<sapmnt>/<SAPSID>/exe
Name of Library Path Environment Variable
Operating System Variable Name
AIX LIBPATH
All other UNIX operating
systems
LD_LIBRARY_PATH
If you are using the Bourne shell (sh):
<Variable Name>=<Variable Value> \
export <Variable Name>
That is, if your operating system is Linux and your database is Oracle:
LD_LIBRARY_PATH=/sapmnt/<SAPSID>/exe: \
<ORACLE_HOME/lib
If you restart SAPinst at a later time, make sure the variable is still set.
4 R3load Procedures
42
4.1.2 Generation of DDL statements
Use
To migrate non-standard database objects, you need to generate DDL statements using the
ABAP report SMIGR_CREATE_DDL.
You need to follow this procedure before starting SAPinst.
Procedure
4. Log on to the system as system administrator in the productive BW-client.
5. Call transaction SE38 and run the program SMIGR_CREATE_DDL.
6. Select the target database. Depending on the database manufacturer, you might need
to select the database version. Value help supports you in the selection of database
version. In general, you should not enter a database version that is not available in the
value help.
7. You are able to select Unicode Migration if you also wish to perform Unicode system
copy (from UC to UC) or a Unicode conversion (from non-UC to UC).
8. Specify an empty working directory to which the files generated by the report will be
written.
9. Optional: You can restrict the generation of DDL statements to specific table types or
individual tables.
10. Execute the program. The DDL statements are generated and are written to the
specified directory.
If no DB specific objects exists in the database, then no SQL files will be
generated. As long as the report terminates with status successfully, this is
not an error!
See also:
SAP Note 771209 for additional database specific information.
4 R3load Procedures
43
4.1.3 Running SAPinst to Export the Database
We recommend that you shut down the SAP system before the export. The
database must still be running. Otherwise, the target system may be
inconsistent.
Use
This procedure tells you how to run SAPinst to export the database of your SAP system.
The following sections describe a local export, that is, SAPinst and SAPinst GUI run on the
same host.
However, you can also perform a remote export using a standalone SAPinst GUI on a
separate Windows or UNIX host. This enables you to perform the installation on a remote
host while monitoring it with SAPinst GUI from a local host. If you want to perform a remote
installation, also see section Performing a Remote Installation with SAPinst [Page 87].
Prerequisites
...
Make sure that your operating system does not delete the temporary directory TEMP,
TMP, TMPDIR or /tmp and its subdirectories when the system is rebooted.
SAPinst normally creates the installation directory sapinst_instdir
directly below the temporary
directory. SAPinst finds the temporary directory by checking the value of the
environment variables TEMP, TMP, or TMPDIR. If no value is set for these
variables, SAPinst uses /tmp as default installation directory.
The SAPinst Self-Extractor extracts the SAPinst executables to the temporary
directory, TEMP, TMP, TMPDIR or /tmp. These executables are deleted again
after SAPinst has stopped running.
If SAPinst cannot find a temporary directory, the installation terminates with
the error FCO-00058.
Make sure that you have at least 50 MB of free space in the installation directory for
each ABAP installation service. In addition, you need 60 ‒ 200 MB free space for the
SAPinst executables. If you cannot provide 200MB free space in the temporary
directory, you can set one of the environment variables TEMP,TMP, or TMPDIR to
another directory with 200 MB free space for the SAPinst executables.
Each SAP instance requires a separate installation directory.
Make sure that umask is set to 022 for user root. As user root, enter the following
command: umask 022
If you want to perform a standard, that is, a local installation with SAPinst, the
DISPLAY environment variable to must be set to <host_name>:0.0, where
<host_name> is the host on which the SAPinst GUI will be displayed.
If required, you can terminate SAPinst and the SAPinst Self-Extractor by pressing
Ctrl+C.
4 R3load Procedures
44
If required, delete directories with the name sapinst_exe.xxxxxx.xxxx after
SAPinst has finished. Sometimes these remain in the temporary directory.
If there are errors with SAPinst Self-Extractor, you can find the Self-Extractor
log file dev_selfex.out in the temporary directory.
We recommend that you keep all installation directories until you are fully
satisfied that the system is completely and correctly installed.
Oracle only:
Make sure that the password for the database user SAP<SCHEMAID> or SAPR3 is SAP
and that the password for the database user system is manager:
a. Log on as user ora<dbsid>.
b. If the password of user system is not manager, enter:
brconnect -u system/<passwd> -c f chpass o system p
manager
c. Enter:
brconnect c f chpass -o SAP<SAPSCHEMAID> -p SAP
IBM DB2 Universal Database for Unix and Windows only:
Before you start the export of existing SAP System, you have to download the current
version of R3szchk from SAP Service Marketplace at the Internet address
service.sap.com/patches and copy it into directory
/usr/sap/<SAPSID>/SYS/exe/run/.
Procedure
...
1. Log on to your host as user root.
2. Mount the SAP Installation Master DVD.
For more information on mounting DVDs, see section Mounting a CD / DVD for <your
OS> in the documentation SAP Web Application Server 6.40 SR 1 ABAP on <your OS>:
<your database> Part II - Installation and Post-Installation.
Mount the DVDs locally. We do not recommend using Network File System
(NFS) as reading from NFS-mounted DVDs may fail.
3. Start SAPinst from the SAP Installation Master DVD in one of the following ways:
�� Using the default installation directory (recommended)
Enter the following commands:
cd <SAP_Installation_Master_DVD>/IM<x>_<OS>/SAPINST/UNIX/<OS>
./sapinst
SAPinst creates a directory called sapinst_instdir, which is the current working
directory for your installation, below the temporary directory of your operating system.
�� Using an alternative installation directory
If you want to use an alternative installation directory, set the environment
variable TEMP, TMP, or TMPDIR.
4 R3load Procedures
45
SAPinst uses the port 21212 during the installation for communication with
SAPinst GUI. If this port is already used by another service you must add the
parameter SAPINST_DIALOG_PORT=<free_port_number> to the relevant
sapinst command above. For example:
./sapinst SAPINST_DIALOG_PORT=<free_port_number>
SAPinst GUI normally starts automatically by displaying the Welcome screen.
However, if there is only one component to install, SAPinst
directly displays the first input dialog without the Welcome
screen.
4. In the Welcome screen, select ABAP System → <your database> →
<Unicode or non-Unicode> → ABAP Database Content Export.
5. Choose Next.
6. If you generated SQL files with DDL statements (see Generation of DLL Statements
[Page 55]), then copy now the generated files into the SAPinst installation directory.
7. Follow the instructions in the SAPinst input dialogs and enter the required parameters.
To find more information about each parameter, use the F1 key in SAPinst. If
you need information about input parameters, position the cursor on the field
of the respective parameter and choose F1.
After you have entered all required input parameters, SAPinst starts the export and displays
the progress during the processing phase.
Troubleshooting
If an export process aborts due to a hardware failure (e.g. filesystem full), you
have to repeat the export of the complete package. Remove the dump files
<package>.<nnn>, the TOC file <package>.TOC, the log file
<package>.log and make sure that all tables in the TSK file
<package>.TSK have the status flag 'xeq' or 'err' set.
If there is not enough disk space in the export directory, the R3load database export
will fail. You will then find error messages in the log files SAP*.log.
You can subsequently move the dump files that have been created from the file system in
which the export directory is located to a different file system during the export. Currently
there is no possibility to automatically distribute the export over different file systems.
If an error occurs during the dialog phase, SAPinst:
�� Stops the export.
�� Displays a dialog that informs you about the error.
You can now directly view the log file by choosing View Logs.
Finally you must abort the export with OK and try to solve the problem.
If an error occurs during the processing phase, SAPinst:
�� Stops the export.
�� Displays a dialog that informs you about the error.
You can now:
4 R3load Procedures
46
�� Directly view the log file by choosing View Logs.
�� Try to solve the problem.
�� Continue the export by choosing Retry.
�� Abort the export by choosing OK.
See also: Continuing an Interrupted Installation [Page 64].
4 R3load Procedures
47
4.2 R3load Procedure on Windows
Purpose
This section describes the R3load system copy procedure for Oracle, Informix, IBM DB2
Universal Database for UNIX and Windows, IBM DB2 UDB for z/OS, MS SQL Server and
SAP DB on Windows platforms.
Process Flow
The R3load procedure consists of the following steps:
...
1. For a heterogeneous system copy, generate the migration key in the SAP Service
Marketplace.
You need a migration key for a heterogeneous system copy.
You can generate the migration key required for the heterogeneous system
copy in SAP Service Marketplace at service.sap.com/migrationkey.
2. Export the source database:
...
a. Before the export, delete QCM tables from your system. Proceed as follows:
i. Before deleting you must check that
�� the tables are consistent (no restart log or conversion procedure
termination must be displayed)
�� the data of the original table is legible
If application programs do not run correctly which use the affected
original table, do not delete the QCM table yet.
ii. Start transaction SE14.
iii. Choose Extras → Invalid temp. table
All QCM tables that can be deleted are displayed.
iv. Mark the tables and delete them.
b. Generate DLL-statements [Page 55].
c. Run SAPinst to export the database [Page 48].
3. Set up the target system [Page 60].
Result
You finished this part of the system copy. To complete the system copy, you have to perform
the steps in section Final Activities [Page 81].
4 R3load Procedures
48
4.2.1 Generation of DDL statements
Use
To migrate non-standard database objects, you need to generate DDL statements using the
ABAP report SMIGR_CREATE_DDL.
You need to follow this procedure before starting SAPinst.
Procedure
4. Log on to the system as system administrator in the productive BW-client.
5. Call transaction SE38 and run the program SMIGR_CREATE_DDL.
6. Select the target database. Depending on the database manufacturer, you might need
to select the database version. Value help supports you in the selection of database
version. In general, you should not enter a database version that is not available in the
value help.
7. You are able to select Unicode Migration if you also wish to perform Unicode system
copy (from UC to UC) or a Unicode conversion (from non-UC to UC).
8. Specify an empty working directory to which the files generated by the report will be
written.
9. Optional: You can restrict the generation of DDL statements to specific table types or
individual tables.
10. Execute the program. The DDL statements are generated and are written to the
specified directory.
If no DB specific objects exists in the database, then no SQL files will be
generated. As long as the report terminates with status successfully, this is
not an error!
See also:
SAP Note 771209 for additional database specific information.
4.2.2 Running SAPinst to Export the Database
It is recommended to shutdown the SAP system before the export. The
database must still be running. Otherwise, the target system may be
inconsistent.
Use
...
You use this procedure to run SAPinst to export the database of your SAP system. The
following describes a standard export, that is, SAPinst and SAPinst GUI run on the same
host.
However, you can also perform a remote export using a standalone SAPinst GUI on a
separate Windows or UNIX host. This enables you to perform the installation on a remote
4 R3load Procedures
49
host while monitoring it with SAPinst GUI from a local host. If you want to perform a remote
installation, also see section Performing a Remote Installation with SAPinst [Page 87].
Prerequisites
...
SAPinst normally creates the installation directory sapinst_instdir directly below the
Program Files directory. If SAPinst is not able to create sapinst_instdir directly below
the Program Files directory, SAPinst tries to create sapinst_instdir in the directory, to
which the environment variable TEMP is set.
Each SAP instance requires a separate installation directory.
The SAPinst Self-Extractor extracts the executables to a temporary directory (TEMP,
TMP, TMPDIR, or SystemRoot). These executables are deleted after SAPinst has
stopped running. Directories with the name sapinst_exe.xxxxxx.xxxx sometimes
remain in the temporary directory. You can safely delete them.
In the temporary directory you can also find the SAPinst Self-Extractor log file
dev_selfex.out, which might be useful if an error occurs.
If you want to terminate SAPinst and the SAPinst Self-Extractor, do one of the
following:
�� Right-click the icon for the SAPinst output window located in the Windows tray
and choose Exit.
�� Click the icon for the SAPinst output window located in the Windows tray and
chooseFile → Exit.
You need at least 50 MB of free space in the installation directory for each ABAP
installation service. In addition, you need 60-200 MB free space for the SAPinst
executables.
We recommend that you keep all installation directories until the system is
completely and correctly installed.
If SAPinst cannot find a temporary directory, the installation terminates with
the error FCO-00058.
Oracle, MS SQL Server, IBM DB2 UDB for UNIX and Windows, IBM DB2 UDB for
z/OS:
Before you start the export of an existing SAP system, check if the current version of
R3szchk and R3ldcdl (IBM DB2 UDB for z/OS) is available on your kernel DVD.
If not, download the current file from SAP Service Marketplace at:
service.sap.com/patches and copy it to directory
\usr\sap\<SAPSID>\SYS\exe\run\.
Procedure
...
1. Log on to your host as a user who is a member of the local administration group.
2. Insert the SAP Installation Master DVD in your DVD drive.
3. Double-click sapinst.exe from the following path:
4 R3load Procedures
50
<DVD drive>:\IM<x>_<OS>\SAPinst\NT\<OS>
SAPinst uses the ports 21212 and 21213 during the installation for
communication with SAPinst GUI. You get an error message if one of these
ports is already in use. In this case, you must do the following:
Open a command prompt.
Change to <DVD drive>:\IM<x>_<OS>\SAPINST\NT\<OS> and run
.\sapinst.exe SAPINST_DIALOG_PORT=<port>
where <port> is an unused port on your host.
4. Select ABAP System → <your database> → <Unicode or Non-Unicode> → ABAP
Database Content Export and choose Next.
5. If you generated SQL files with DDL statements (see Generation of DLL Statements
[Page 55]), then copy now the generated files into the SAPinst installation directory.
6. Follow the instructions in the SAPinst input dialogs and enter the required parameters.
If you need more information about input parameters, position the cursor on
the field of the respective parameter and press F1.
If SAPinst prompts you to log off, log off and log on again.
The installation restarts automatically.
If you have entered all required information during the dialog phase, SAPinst starts the
export and displays the export progress during the processing phase.
Troubleshooting
If an export process aborts due to a hardware failure (e.g. filesystem full), you
have to repeat the export of the complete package. Remove the dump files
<package>.<nnn>, the TOC file <package>.TOC, the log file
<package>.log and make sure that all tables in the TSK file
<package>.TSK have the status flag 'xeq' or 'err' set.
If there is not enough disk space in the export directory, the R3load database export
will fail. You will then find error messages in the log files SAP*.log.
You can subsequently move the dump files that have been created from the file system in
which the export directory is located to a different file system during the export. Currently
there is no possibility to automatically distribute the export over different file systems.
If an error occurs during the input phase, SAPinst:
�� Stops the installation.
�� Displays a dialog that informs you about the error.
You can now directly view the log file by choosing View Logs.
You must abort the installation with OK and try to solve the problem.
If an error occurs during the processing phase, SAPinst:
�� Stops the installation.
�� Displays a dialog that informs you about the error.
You can now:
4 R3load Procedures
51
�� Directly view the log file by choosing View Logs.
�� Try to solve the problem
�� Retry the installation by choosing Retry.
�� Abort the installation by choosing OK.
For more information, see Continuing an Interrupted Installation with SAPinst [Page 64].
4 R3load Procedures
52
4.3 R3load Procedure on IBM eServer iSeries
Purpose
This section describes the R3load system copy procedure for IBM DB2 UDB on iSeries.
Process Flow
The R3load procedure consists of the following steps:
...
1. Heterogeneous system copy: Generate the migration key via SAP Service
Marketplace.
You need a migration key for a heterogeneous system copy.
You can generate the migration key required for the heterogeneous system
copy via SAP Service Marketplace at service.sap.com/migrationkey.
2. Preparing the Windows Host for the SAP System Installation [Page 52].
3. Preparing a Windows User Account and iSeries User Profile [Page 53].
4. Installing TMKSVR and Creating an Installation Share [Page 54].
5. Generate DDL statements [Page 55].
6. Running SAPinst to Export the Database [Page 56].
7. Setting up the Target System [Page 60].
If you have installed your SAP system on SAP R/3 release prior to 3.0D, you have to delete QCM tables from your system as described in SAP Note 9385 before the export.
Target database IBM DB2 UDB for OS/390 and z/OS only:
Result
You finished this part of the system copy. To complete the system copy, you have to perform
the steps in the section Final Activities [Page 81].
4.3.1 Preparing the Windows Host for the SAP
System Installation
Use
The Java-based SAPinst graphical user interface (GUI) called SAPinst GUI requires a Java
Development Kit (Java 2 SDK, Standard Edition) with graphical capabilities (AWT, Swing).
Since IBM eServer iSeries does not provide a graphical user interface, you must install the
JDK on a Windows host to perform the installation with SAPinst.
Prerequisites
To prepare the system for SAPinst and SAPinst GUI you need to do the following:
Necessary operating system versions: Windows NT/2000/2003/XP
Check your Java Runtime Environment (JRE) on the host where SAPinst GUI runs,
because the JRE cannot be integrated into the SAPinst GUI executable for all
platforms due to licensing issues.
Set the system path if you install on Windows.
4 R3load Procedures
53
Procedure
The J2EE Engine requires a Java Development Kit (Java 2 SDK, Standard Edition).
Therefore, make sure a valid JDK version is installed on every host on which you want to
install an SAP instance including the J2EE Engine.
For more information on the JDK versions that are released for the SAP Web Application
Server, SAP components based on SAP Web AS and the J2EE Engine, see SAP Service
Marketplace at service.sap.com/platforms → Product Availability Matrix → SAP
NetWeaver → SAP NetWeaver 04 → JSE Platforms.
JDK is not part of the SAP shipment. If necessary, you need to download and
install it.
To check the version of an already installed JDK, enter: java -version
If you have more than one Java Virtual Machine (JVM) installed on your
system
(for example, you have two JDKs with different versions installed), make sure
that the JAVA_HOME environment variable is set to the valid <JAVA_HOME>
directory. Make sure that %JAVA_HOME%\bin is included in your system path.
4.3.2 Preparing a Windows User Account and
iSeries User Profile
Use
For the installation you need to create a user account on your Windows installation host and
a user profile on the iSeries you want to install.
The following requirements apply:
The iSeries user profile and the Windows user account must have the same name and
password.
The iSeries user profile must have user class *SECOFR and all special authorities that
belong to user QSECOFR.
The Windows user account must have administrator rights on the Windows installation
host.
Procedure
The user name SAPINST and the password SAP are used in the procedures
as examples.
Windows :
...
Create a local user.
i. Create a local user.
ii. In the field User name, enter your installation user name, for example,
SAPINST.
iii. In the fields Password and Confirm password, enter the password SAP.
4 R3load Procedures
54
iv. Deselect User must change password at next logon.
v. Assign the new user SAPINST to group Administrators.
iSeries:
Execute the following command:
CRTUSRPRF USRPRF(SAPINST) PASSWORD(SAP) USRCLS(*SECOFR) TEXT('Test
User for SAP Installation') SPCAUT(*USRCLS)
4.3.3 Installing TMKSVR and Creating an
Installation Share
Use
The TMKSVR is the interface between iSeries and Windows for the installation with SAPinst.
SAPinst is running on Windows, but has to install the product on iSeries. This means that all
actions required for iSeries are initiated remotely on Windows but executed locally using the
TMKSVR. The communication is done using TCP/IP.
In addition, an installation share on the iSeries host needs to be created and mapped to the
Windows installation host, which is done automatically by the TMKSVR.
The TMKSVR has to be installed and an installation share has to be created on all iSeries
hosts where instances of an SAP system should be installed.
Prerequisites
An FTP server must be running on iSeries.
You must prepare a user. For more information on how users are prepared, see
Preparing a Windows User Account and iSeries User Profile [Page 53].
Copy the SAP Installation Master DVD from the DVD drive to the iSeries.
Procedure
...
1. Log on to your Windows host as the installation user. For more information, see
Preparing a Windows User Account and iSeries User Profile [Page 53].
2. Run SETUP.EXE from the directory IM(x)_OS400_64\SAPINST\AS400\TMKSVR on
the DVD containing the installation package. You can start the setup program by
double-clicking it in the Windows Explorer.
To find the SAPinst executable in your platform-specific IM<x> directory,
look in the README.TXT file on the SAP Installation Master DVD.
The following dialog box appears:
4 R3load Procedures
55
3. Enter the following values:
�� iSeries Hostname
Enter the name of the iSeries host where you want to install TMKSVR.
�� iSeries Administrator (QSECOFR or similar)
Enter iSeries user. For more information, see Preparing a Windows User
Account and iSeries User Profile [Page 53].
�� Update existing TMKSVR instances
Do not select this option.
�� Yes, create TMKSVR instance
Select this option.
�� TMKSVR instance number
Leave the value at 0.
�� TMKSVR Instance Port (also referred to as the Dispatcher Port)
Leave the value at 59975, if possible. Only change this port number if you
encounter problems during installation because the port is in use.
Result
The installation uses FTP to install and start the TMKSVR on iSeries. During installation, the
TMKSVR library is created on iSeries. If you want a TMKSVR instance to be created, a library
named TMKSVR<nn> is also created, where <nn> is the instance number (for example,
TMKSVR00).
A NetServer share named ROOTBIN will be created on the iSeries host. You can map it now
to your Windows PC or let SAPinst do it during the installation.
For more information, see the documentation install.pdf on the DVD in directory
IM<x>_OS400_64\SAPINST\AS400\TMKSVR.
4 R3load Procedures
56
4.3.4 Generation of DDL statements
Use
To migrate non-standard database objects, you need to generate DDL statements using the
ABAP report SMIGR_CREATE_DDL.
You need to follow this procedure before starting SAPinst.
Procedure
4. Log on to the system as system administrator in the productive BW-client.
5. Call transaction SE38 and run the program SMIGR_CREATE_DDL.
6. Select the target database. Depending on the database manufacturer, you might need
to select the database version. Value help supports you in the selection of database
version. In general, you should not enter a database version that is not available in the
value help.
7. You are able to select Unicode Migration if you also wish to perform Unicode system
copy (from UC to UC) or a Unicode conversion (from non-UC to UC).
8. Specify an empty working directory to which the files generated by the report will be
written.
9. Optional: You can restrict the generation of DDL statements to specific table types or
individual tables.
10. Execute the program. The DDL statements are generated and are written to the
specified directory.
If no DB specific objects exists in the database, then no SQL files will be
generated. As long as the report terminates with status successfully, this is
not an error!
See also:
SAP Note 771209 for additional database specific information.
4.3.5 Running SAPinst to Export the Database
This section refers to installation of an instance. This can be viewed as a
synonym for export an SAP system.
Use
This procedure tells you how to run SAPinst to install one or more SAP instances. It describes
an installation where SAPinst GUI and SAPinst server are running on the same Windows
host.
SAPinst creates the installation directory \usr\sap\sapinst on iSeries.
4 R3load Procedures
57
Prerequisites
TMKSVR is up and running: WRKACTJOB SBS (TMKSVR00) (there must be a DISPATCH
job). For more information, see Installing TMKSVR and Creating an Installation Share
[Page 54].
The Windows host is set up. For more information, see Preparing the Windows Host for
the SAP System Installation [Page 52].
The users required for the installation are prepared. For more information, see
Preparing a Windows User Account and iSeries User Profile [Page 53].
Make sure that the JAVA_HOME environment variable is set correctly on your Windows
host.
Procedure
...
1. Log on to the Windows host as the installation user. For more information, see
Preparing a Windows User Account and iSeries User Profile [Page 53]:
2. Start SAPinst from the SAP Installation Master DVD in one of the following ways:
Using the default installation directory
sapinst.exe
in the path <Mapped_Drive>:\<Copied SAP Installation Master
DVD>\IM<x>_OS400_64\SAPINST\OS400\AS400.
SAPinst uses the port 21212 and 21213 during the installation for communication with the
SAPinst GUI. If this port is already used by another service you must add the parameter
SAPINST_DIALOG_PORT=<free_port_number> to the relevant sapinst command
above.
For example:
sapinst.exe SAPINST_DIALOG_PORT=<free_port_number>
Using an alternative installation directory
Change to your installation directory.
Enter the following command to start SAPinst from the SAP Installation Master
DVD: <Mapped_Drive>:\<Copied SAP Installation Master
DVD>\IM<x>_OS400_64\SAPINST\OS400\AS400\sapinst.exe
3. The SAPinst/TMKSVR Session Parameters dialog box appears and prompts you for
the target iSeries parameters. Enter your values.
4 R3load Procedures
58
The SAPinst GUI now starts automatically by displaying the Welcome screen.
4. In the Welcome screen, select ABAP System �� <your database> �� <Unicode or Non-
Unicode> �� ABAP Database Content Export and then choose Next.
5. If you generated SQL files with DDL statements (see Generation of DLL Statements
[Page 55]), then copy now the generated files into the SAPinst installation directory.
6. Follow the instructions in the SAPinst input dialogs and enter the required parameters.
To find more information on each parameter, use the F1 key in SAPinst. If you
need information about input parameters, position the cursor on the field of
the respective parameter and press F1.
After you have entered all required input parameters, SAPinst starts the installation
and displays the progress of the installation.
When the installation has successfully completed, the screen Finished installation is
displayed.
Troubleshooting
If an export process aborts due to a hardware failure (e.g. filesystem full), you
have to repeat the export of the complete package. Remove the dump files
<package>.<nnn>, the TOC file <package>.TOC, the log file
<package>.log and make sure that all tables in the TSK file
<package>.TSK have the status flag 'xeq' or 'err' set.
If an error occurs during the dialog phase, SAPinst:
�� Stops the installation.
�� Displays a dialog that informs you about the error.
�� You can now directly view the log file by choosing View Logs.
�� Finally you must abort the installation with OK and try to solve the problem.
If an error occurs during the processing phase, SAPinst:
4 R3load Procedures
59
�� Stops the installation.
�� Displays a dialog that informs you about the error.
You can now directly view the log file by choosing View Logs.
�� Try to solve the problem.
�� Retry the installation by choosing Retry.
�� Abort the installation by choosing OK.
�� For more information, see Interrupted Installation with SAPinst [Page 64].
4 R3load Procedures
60
4.4 Setting up the Target System
Purpose
With the SAP installation tool SAPinst you can install the target system and import the
database files that you have exported from the source system.
Process Flow
Perform the following actions:
...
1. Transfer the export files to the target host. [Page 60]
2. Install the target system and use the database export to load the database during the
installation process. [Page 60]
3. If you want to load data from a system into the existing database of another system,
perform the RELOAD Procedure [Page 62]
4.4.1 Transferring the Export Files to the Target
Host
Procedure
...
1. On the target host, create a directory <EXPDIR> with sufficient space for the database
export files available.
2. Copy all files and directories (recursively) that are located on the source host in the
migration export directory from the source host to the target host.
If you transfer the files with ftp, make sure that you use binary mode for
transferring the files <EXPDIR>/DATA/*.00<n> and use ASCII mode for
transferring all other files.
3. Check the permissions of the transferred files on the target host. All files have to be
accessible for user <sapsid>adm of the target system.
4 R3load Procedures
61
4.4.2 Installing the Target System
Make sure there is enough free space on the target system for the database
load. To find out the size of the export and the sizes of the tablespaces or
dbspaces that will be created, look at the file DBSIZE.XML located in the
directory <DRIVE>:\<EXPDIR>\DB\<DATABASE> (Windows) or
<EXPDIR>/DB/<DATABASE> (UNIX).
Procedure
...
1. You start SAPinst as described in the installation documentation for your SAP
component.
2. To install the target system follow the instructions in the SAPinst input dialogs and
enter the required parameters up to the window Selecting the Database Instance
Installation Method. On this screen, you choose System Copy/Migration.
If you need more information about input parameters, position the cursor on
the field of the respective parameter and press F1.
3. When SAPinst displays the CD Browser-Window and asks for the CD Export Migration,
enter the path to the export directory <EXPDIR>.
4. When SAPinst displays the window General Load Parameters, specify the following
settings:
�� Migration Key: If you perform a heterogeneous system copy, enter the migration
key.
�� General Settings:
�� Specify the order in which the packages are loaded (alphabetical order,
according to the size or custom order)
If you choose Load packages in custom order the additional window Data
Load Options is displayed (see below).
�� DB code page: You normally do not have to change this value.
�� Number of parallel jobs: Specify the number of parallel R3load
processes.
�� Advanced Configuration of Packages
�� If you choose Individual configuration for task file generation the window
Task File Generation Options is displayed.
�� If you choose Individual configuration for data load the window Data
Load Options is displayed.
Advanced package configuration should only be performed by certified
database administrators. We recommend that you use the default settings if
possible.
5. Complete the installation as described in the installation documentation for your SAP
component.
4 R3load Procedures
62
If you have to restart the import after an error, just restart SAPinst. The import
is continued with the table that was not imported successfully.
4.4.3 RELOAD Procedure
Use
You want to load data from a system (for example, your productive system) into the existing
Oracle database of another system (for example, your test system).
The RELOAD service reinstalls the target database while keeping the SAP settings. Data
stored in the target database is deleted. Then, data from the source database is reloaded.
RELOAD is not intended to add additional data (that is, merge data from two databases by
keeping old data and adding new data). After the RELOAD procedure, only data from the
source database is left.
The RELOAD service is not available for MCOD systems. During the
RELOAD procedure the database is created again and therefore all data (and
all database schemas) is lost.
At the moment, this RELOAD service is available for Oracle databases only.
The RELOAD service cannot be used to refresh any database schema in an
MCOD system.
Procedure
...
1. Dialog input:
On the screen ABAP System > Database Instance Installation Method, choose System
Copy / Migration by Reload: Refresh the database content of a non-MCOD system
(R3load).
2. Additional input for the procedure:
Before you start the installation of the target system,
�� To save any changes made to this file, make a backup of the file
init<SID>.ora located in <ORACLE_HOME>/dbs.
�� Delete all entries of the directories saparch and oraarch in your target
system.
�� From your old instance, delete all data files located in your sapdata directories
and all control files.
3. To save your database configuration, proceed as follows after the installation of the
new system has been finished successfully:
Make sure that the parameters of your old init<SID>.ora file (that you backed up in
step 2) are also contained in the new init<SID>.ora file that was created by
SAPinst.
5 R3load Procedures using the Migration Monitor
63
5 R3load Procedures using the
Migration Monitor
Purpose
The Migration Monitor is a tool which helps you to perform and control the unload and load
process during the system copy procedure.
From SAP Web AS 6.40 on the Migration Monitor is integrated into the SAPinst system copy
tool, but it is also possible to use the monitor for copying older releases by starting it
manually.
The Migration Monitor will
create R3load command files
create R3load task files if required
start the R3load processes to unload the data
transfer packages from the source to the target host if required
start the R3load processes to load the data as soon as a package is available
inform the person performing the system copy in case of errors.
The Migration Monitor has to be started on the source database host (=> Export Monitor) and
on the target database host (=> Import Monitor).
The advantage of using the Migration Monitor is, that R3load Procedures run faster.
The Migration Monitor is currently available for Oracle only.
Prerequisites
You have to stick to the same restrictions and considerations that apply to the R3load
procedures without using Migration Monitor (see Prerequisites of R3load Procedures [Page
38]).
In addition, you have to keep the following restrictions in mind:
The version of your Java Runtime Environment (JRE) on both the export and the
import host must be 1.4.1 or higher.
The JAVA_HOME environment variable on both the export and the import host must
point to the relative JRE directory.
The correct directory structure for R3load dump files must exist on both the source and
target host.
Process Flow
On UNIX, see R3load with Migration Monitor on UNIX [Page 64].
On Windows, see R3load with Migration Monitor on Windows [Page 65].
5 R3load Procedures using the Migration Monitor
64
5.1 R3load Procedure with Migration Monitor
on UNIX
Purpose
This section describes the R3load system copy procedure with Migration Monitor for Oracle
on UNIX platforms.
Process Flow
The R3load procedure with Migration Monitor consists of the following steps:
...
1. Unpack the MIGMON.SAR SAPCAR archive:
The file is located in <SAP Installation Master DVD> /UNIX/COMMON/INSTALL.
2. Configure the export properties file [Page 66] export_monitor_cmd.properties.
3. Heterogeneous system copy: Generate the migration key on SAP Service Marketplace.
You need a migration key for a heterogeneous system copy.
You can generate the migration key required for the heterogeneous system
copy on SAP Service Marketplace at service.sap.com/migrationkey.
4. Set up the target system [Page 79].
5. Export the source database:
a. Make sure that the QCM tables are deleted from your system [Page 14].
b. Set the library path environment variable [Page 40].
c. Generate DDL statements [Page 55].
d. Run SAPinst to export the database [Page 42].
e. Start the Migration Monitor [Page 75].
f. Check the output files [Page 78] of the Migration Monitor.
Result
You have finished this part of the system copy. To complete the system copy, you have to
perform the steps in the Final Activities [Page 81] section.
5 R3load Procedures using the Migration Monitor
65
5.2 R3load Procedures with Migration
Monitor on Windows
Purpose
This section describes the R3load system copy procedure with Migration Monitor for Oracle
on Windows.
Process Flow
The R3load procedure with Migration Monitor consists of the following steps:
...
1. Unpack the MIGMON.SAR SAPCAR archive:
2. The file is located in <SAP Master CD> \NT\COMMON\INSTALL.
3. Configure the export properties file [Page 66] export_monitor_cmd.properties.
4. Heterogeneous system copy: Generate the migration key on SAP Service Marketplace.
You need a migration key for a heterogeneous system copy.
You can generate the migration key required for the heterogeneous system
copy on SAP Service Marketplace at service.sap.com/migrationkey.
5. Build up the target system [Page 60].
6. Export the source database:
a. Make sure that the QCM tables are deleted from your system [Page 14].
b. Generate DDL statements [Page 55].
c. Run SAPinst to export the database [Page 48].
d. Start the Migration Monitor [Page 75].
e. Check the output files [Page 78] for the Migration Monitor.
Result
You have finished this part of the system copy. To complete the system copy, you have to
perform the steps in the Final Activities [Page 81] section.
5 R3load Procedures using the Migration Monitor
66
5.3 Configuring the Import and Export
Properties Files
Use
You have to configure both the export properties file export_monitor_cmd.properties
and the import properties file import_monitor_cmd.properties as specified in the
tables below.
Options that refer to both scripts:
Common options
E-mail options
Trace options
Options, which concern only the export property file export_monitor_cmd.properties :
Export options
Network exchange options
FTP exchange options
Export socket options
FTP copy options
Some of these options are mandatory for the Export Monitor.
Options, which concern only the import property file import_monitor_cmd.properties :
Import options
Import exchange options
Import socket options
Some of these options are mandatory for the Import Monitor.
Features
Help Options
Name Description Comments
help Migration Monitor help
?
The Migration Monitor tool
will display the help on the
available parameters, if you
call it with either help or -
?
Common Options
Name Description Comments
monitorTimeout Migration Monitor timeout in
seconds
During a timeout, the
Migration Monitor thread
sleeps and checks every x
seconds whether there is
5 R3load Procedures using the Migration Monitor
67
work to do.
The normal sleep duration is
30 120 seconds.
Trace Options
Name Description Comments
Trace Trace level Possible values:
All, off, 1 (error), 2
(warning), 3 (info), 4 (config,
default), 5, 6, 7 (trace)
Email Options
Name Description Comments
mailServer SMTP server Server name or IP address of
the company SMTP server
mailFrom From email address
mailto To email address Can contain an address list
separated by ; or spaces.
Export Options
Option Description Comments
exportDirs List of export directories Separator on Windows is ;,
on UNIX :.
The exportDirs parameter
points to the directory where
the R3load dump files will be
written to. In the exportDirs
directory, the subdirectories
DATA, DB and
DB/<target_DBTYPE> (e.g.
DB/ORA) have to exist.
installDir SAPinst start directory The directory where the
installation tool (SAPinst,
R3SETUP) is started; if you
run the Migration Monitor
without using the installation
tools, then the installation
directory is the directory,
where the R3load TSK and log
files will be written.
omit R3load omit value (makes
sense for import only)
5 R3load Procedures using the Migration Monitor
68
All options below are for the server mode. The Import Monitor always runs in the
server mode. If you want to run the Export Monitor in the server mode, specify the
parameter Server in the Export Monitors properties file.
server Server operating mode Running in server mode
means, the Monitor will
creates R3load TSK files (if
necessary), R3load cmd files
and start the R3load
processes.
orderBy Package order Can be the 'name' or path of
the file that contains package
names.
r3loadExe Path of the R3load executable Optional; the default is R3load.
If only the name of the R3load
executable is available, then
JVM looks for the R3load
executable using OS-specific
process search rules.
tskFiles 'yes' to create task files; 'no'
to skip
Before version 4.6 must be set
to no; starting from version
4.7 yes.
If the R3load task files
*.TSK already exist then
the monitor will not overwrite
them.
extFiles 'yes' to include EXT files;
'no' to skip
Add EXT file entries to cmd
files;
If the EXT files cannot be
found in
/expDirs/DB/<target_DB
TYPE> the package
processing is aborted.
Always use extFiles=no
for the export!
dataCodepage Code page for data files See SAP Note 552464.
Possible values: 4102,
4103, 1100
taskArgs Additional R3load arguments
for the TASK phase
Appended to the R3load
command line.
Options already set by the
monitor:
-ctf; -l; -o (if the omit
argument is specified).
loadArgs Additional R3load arguments
for the LOAD phase
Appended to the R3load
command line.
Options already set by the
5 R3load Procedures using the Migration Monitor
69
monitor:
-e; -datacodepage; -l; -p;
-r; -socket (if the socket
option is specified); -o (if the
omit argument is specified and
task files are not used, that is,
the value of the tskFiles option
is no).
expJobNum Number of parallel export jobs;
the default is 1.
Any positive number; 0 for an
unlimited number of jobs.
Network exchange options
Option Description Comments
net Network operating mode Exported dump files must be
visible on the import host to
use this mode.
netExchangeDir Network exchange directory Used for communication
between the export and Import
Monitors.
Must be writable for Export
Monitor and readable for
Import Monitor. The Export
Monitor will write a file
<package>.SGN to the
netExchangeDir as a signal
for the Import Monitor, that the
package is exported
successfully and the import
could be started.
FTP exchange options
Option Description Comments
ftp FTP operating mode Exported dump files will be
transferred automatically from
the source host (directory
exportDirs) to the target
host (directory importDirs)
using FTP.
ftpHost Remote FTP host Name or IP address of the
import server
ftpUser Name of the remote FTP user The ftpUser specified here
should be the <sapsid>adm
to make sure, that the
package files can be read by
during the import (which is
started as <sapsid>adm).
5 R3load Procedures using the Migration Monitor
70
ftpPassword Password of the remote FTP
user
Security risk
ftpExportDirs List of remote FTP directories
for export dump
Both ; or : separators are
valid.
This is the directory on the
target host to which the dump
will be transfered. The value
will be the same as for
importDirs in the Import
Monitors property file.
ftpExchangeDir Remote FTP exchange
directory
Used for communication
between the export and Import
Monitors.
Must be writable for the Export
Monitor and readable for the
Import Monitor. The Export
Monitor will write a file
<package>.SGN to the
ftpExchangeDir as a signal for
the Import Monitor, that the
package is exported
successfully and the import
could be started.
ftpJobNum Number of parallel FTP jobs;
the default is 1.
Any positive number; 0 for an
unlimited number of jobs.
Export socket options
Option Description Comments
socket Socket operating mode R3load will not write dump
files to the file system but the
export and import work
through the socket connection.
host Remote import host Name or IP address of the
import host.
port Host port number Must be the same as the port
number on the import host.
Any free port on the import
host from 1024 to 65535.
FTP copy options
Option Description Comments
ftpCopy FTP copy operating mode Used as a separate program
call for migration with sockets.
All files produced by R3lctl
and R3szchk will be
transferred from the source to
5 R3load Procedures using the Migration Monitor
71
the target host using FTP.
exportDirs List of export directories Separator on Windows: ;
Separator on UNIX: :
In the exportDirs directory,
the subdirectories DATA, DB
and DB/<target_DBTYPE>
(for example DB/ORA) have to
exist. The R3load STR files
have to exist in the
subdirectory DATA, the
DDL*.TPL files in the
subdirectory DB, and the
R3load EXT files (if required)
in the subdirectory
DB/<target_DBTYPE>.
ftpHost Remote FTP host Name or IP address of the
import server.
ftpUser Name of the remote FTP user The ftpUser specified here
should be the <sapsid>adm
to make sure, that the
package files can be read by
during the import (which is
started as <sapsid>adm).
ftpPassword Password of the remote FTP
user
Security risk
ftpExportDirs List of remote FTP directories
for export dump
Both ; or : separators are
valid.
This is the directory on the
target host to which the dump
will be transfered. The value
will be the same as for
importDirs in the Import
Monitors property file.
Mandatory options for the Export Monitor
Mode Options
Client mode: installDir, exportDirs,
one of the options ftp, nfs, socket (and their
related parameters)
Server mode: installDir, exportDirs, tskFiles,
extFiles,
one of the options ftp, nfs, socket (and their
related parameters)
FTP copy exportDirs, ftpHost, ftpUser,
ftpPassword, ftpExportDirs,
ftpExchangeDir
5 R3load Procedures using the Migration Monitor
72
The value of the dbType option is determined automatically in the shell
script/batch files from the dbms_type environment variable.
Import Options
Option Description Comments
importDirs List of import directories Separator on Windows: ;
Separator on UNIX: :
The importDirs parameter
points to the directory with the
R3load dump files. In the
importDirs directory, the
subdirectories DATA, DB and
DB/<target_DBTYPE> (e.g.
DB/ORA) have to exist.
installDir Installation directory Directory where the installation
tool (SAPinst, R3SETUP) is
started; if you run the
Migration Monitor without
using the installation tools,
then the installation directory
is the directory, where the
R3load TSK and log files will
be written.
orderBy Package order This option is used only if the
Import Monitor works without
the Export Monitor in standalone
mode, that is, all export
dump files are available on the
import host before the Import
Monitor is started.
Values can be:
name : load packages in
alphabetical order,
size : load packages starting
with the largest one,
or a path of the file that
contains package names.
r3loadExe Path of the R3load executable Optional; the default is
R3load.
If only the name of the R3load
executable is available then
JVM looks for the R3load
executable using OS-specific
process search rules.
tskFiles 'yes' to create task files; 'no' Before version 4.6, must be
set to no; starting from
5 R3load Procedures using the Migration Monitor
73
to skip version 4.7 yes.
If task files already exist then
the monitor does not overwrite
them.
extFiles 'yes' to include EXT files; 'no'
to skip them
Add EXT file entries to cmd
files;
If the EXT files cannot be
found in
/importDirs/DB/<target
_DBTYPE> the package
processing is aborted.
dbCodepage Database code page for the
target database
See SAP Note 552464.
Possible values: 4102, 4103,
1100
migrationKey Migration key
omit R3load omit value Can contain only 'DTIPV'
letters.
-o D : omit data; do not load
data
-o T: omit tables; do not
create tables
-o I: omit indexes; do not
create indexes
-o P: omit primary keys; do
not create primary keys
-o V: omit views; do not
create views
If you want to combine several
omit options, list these options
without blank (for example
-o TV).
taskArgs Additional R3load arguments
for the TASK phase
Appended to the R3load
command line.
Options already used by the
monitor:
-ctf; -l; -o (when the omit
argument is specified).
loadArgs Additional R3load arguments
for the LOAD phase
Appended to the R3load
command line.
Options already used by the
monitor:
-i; -dbcodepage; -l; -p;
-k; -r; -socket (if the
socket option is specified); -
o (if the omit argument is
5 R3load Procedures using the Migration Monitor
74
specified and task files are not
used, that is, the value of
tskFiles option is no).
impJobNum Number of parallel import jobs;
the default is 1.
Any positive number; 0 for an
unlimited number of jobs
Import exchange options
Option Description Comments
exchangeDir Exchange directory If this option is not set, then
the monitor runs in standalone
mode, that is without
the Export Monitor.
All the export dump files or
the SAP export CDs from the
installation kit must be
available on the import host
and be specified with the
parameter importDirs (for
example, in the properties
file).
Import socket options
Option Description Comments
socket Socket operating mode
port Server port number Any free port from 1024 to
65535.
Mandatory options for the Import Monitor
Mode Options
Server mode (default) installDir, importDirs, tskFiles,
extFiles,
one of the options exchangeDir or socket
(and its related parameters)
Stand-alone mode installDir, importDirs, tskFiles,
extFiles
The value of the dbType option is determined automatically in the shell
script/batch files from the dbms_type environment variable.
5 R3load Procedures using the Migration Monitor
75
5.4 Starting the Migration Monitor
Use
The Migration Monitor can be started using:
UNIX shell scripts export_monitor.sh / import_monitor.sh
Windows batch files export_monitor.bat / import_monitor.bat
The JRE java tool (this way is preferable when the command line is generated in an
external application such as SAPinst).
The application allows you to specify options in the command line or in the export or import
property files [Page 66]. The names of the property files are export_monitor_cmd.properties
and import_monitor_cmd.properties.
Templates for these files are included in the application archive and must be located in the
current users working directory.
The options specified in the command line take precedence over the corresponding options in
the application property file. Options are case sensitive; any options that are not recognized
are ignored.
To specify an option in the command line, enter -optionName optionValue; in the
application property file, insert a new line optionName=optionValue.
Example of a command line for a UNIX terminal:
./export_monitor.sh ftp
./export_monitor.sh ftpCopy
./export_monitor.sh socket host import_server port
5000
Example of a command line for Windows cmd.exe:
export_monitor.bat net
export_monitor.bat -socket
Procedure
For the export with Migration Monitor, proceed as follows:
...
1. Run the export of the ABAP system.
2. Select the SAPinst option Export using Migration Monitor
3. To specify the correct monitor options, create or edit the
export_monitor_cmd.properties file.
4. If you use FTP access, verify that the required directories exist on the import server
before the Migration Monitor is started. The file import_dirs.sh or
import_dirs.bat can be used to create a correct directory structure.
5. To specify the correct monitor options, create or edit the file
export_monitor_cmd.properties.
6. If FTP access is used, verify that the required directories exist on the import server
before the export files are copied to the import server. The script import_dirs.sh /
the batch file import_dirs.bat can be used to create the correct directory structure.
5 R3load Procedures using the Migration Monitor
76
7. Sockets only: If FTP access is used, transfer all export files to the import server using
the ftpCopy monitor option
8. Sockets only: Start the Import Monitor and wait until it starts listening at the specified
port
9. Start the Migration Monitor as user <sapsid>adm after the export process has been
started
10. If any export errors occur, restart the Migration Monitor as soon as the problem is fixed.
11. After all packages have been loaded successfully, continue/restart the installation tool
and finish the installation.
Using the Export Monitor in parallel with SAPinst:
The Export Monitor can be started as an additional tool for the SAPinst
standard export process. In this case, the monitor transfers any completed
export packages to the target host.
The Export Monitor can be started in parallel with SAPinst, but only after the
dump directories are created on both the export and import server.
For the import with Import Monitor, proceed as follows:
...
1. Install the Central Instance.
2. Run the installation of the Database Instance.
If you want to start the installation of the target system before the export of the
source system has been started, make sure that at least the files
<importDir>/LABEL.ASC and
<importDir>/DB/<your database>/DBSIZE.{TPL|XML}
<importDir>/DB/DDL<your database>.TPL
exist and contain the correct data
3. Select the SAPinst option Installation using Migration Monitor
4. Create or edit the import_monitor_cmd.properties file to specify the correct
monitor options.
5. After the exit step, stop SAPinst
6. Sockets only: Make sure that all files on the export server generated by SAPinst (with
R3ldctl/R3szchk) are copied to or available on the import server (including all STR,
EXT package-specific files).
7. Start the Migration Monitor as user <sapsid>adm.
8. If any import errors occur, restart the Migration Monitor as soon as the problem is fixed.
9. After all packages have been loaded successfully, restart or continue with the
installation tool and finish the installation.
For more information, see the documentation Migration Monitor Users
Guide on the SAP Installation Master DVD
5 R3load Procedures using the Migration Monitor
77
5.5 Restarting R3load Processes
Use
The state file allows package states to be manually updated to restart failed R3load
processes.
For example, if package processing failed and the package state has the
value , the state can be set to 0 and processing of the package will be
started again.
Activities
To restart package processing, set the package state from to 0.
To skip package processing, set the package state from 0 or to +.
(This is not recommended, because it can cause inconsistent data files or database
content.)
If the package is currently being processed (the package state is ?), then any manual
modifications of the package state are ignored.
Sockets only: You cannot restart processes.
5 R3load Procedures using the Migration Monitor
78
5.6 Output Files of the Migration Monitor
Use
During the system copy procedure, the Migration Monitor writes to the following log files:
Export
export_monitor.log
export_state.properties
Import
import_monitor.log
import_state.properties
Features
The export state file contains package state lines such as:
SAPUSER=++
The format of the line is <PACKAGE>=<STATE>. Possible values for <STATE> are:
0 Package export/import has not started.
? Package export/import is in progress.
- Package export/import has finished with errors.
+ Package export/import has finished
successfully.
If any ftp/net exchange options are used, then the export state file may contain a second
<STATE> column, which refers to the state of the package transfer.
Then the export state file contains package state lines such as the following:
SAPUSER=++
The format of the line is <PACKAGE>=<STATE>. Possible values for <STATE> are:
0 Package export has not yet started.
? Package export is in progress
- Package export has finished with errors.
+0 Package export has finished successfully;
package transfer has not yet started.
+? Package transfer is in progress.
+- Package transfer has finished with errors.
++ Package transfer has finished successfully.
5 R3load Procedures using the Migration Monitor
79
5.7 Setting up the Target System using the
Migration Monitor
Purpose
With the SAP installation tool SAPinst you can install the target system and import the
database files that you have exported from the source system.
In addition you have to configure the import scripts [Page 66] and to invoke the Migration
Monitor [Page 75].
Process Flow
Perform the following actions:
...
1. Configure the import properties file [Page 66].
2. Install the target system [Page 79] and use the database export to load the database
during the installation process.
3. Invoke the Migration Monitor [Page 75].
4. Check the output files [Page 78] of the Migration Monitor.
5.7.1 Installing the Target System Using the
Migration Monitor
Prerequisites
Make sure there is enough free space in the target system for the database
load. To find out the size of the export and the sizes of the tablespaces or
dbspaces that are created, look at the file DBSIZE.XML located in the
directory <DRIVE>:\<EXPDIR>\DB\<DATABASE> (Windows) or
<EXPDIR>/DB/<DATABASE> (UNIX).
Procedure
...
1. You start SAPinst as described in the installation documentation for your SAP
component.
2. To install the target system follow the instructions in the SAPinst input dialogs and
enter the required parameters up to the window Selecting the Database Instance
Installation Method. On this screen, you choose System Copy/Migration using
Migration Monitor (R3load).
If you need more information about input parameters, position the cursor on
the field of the respective parameter and press F1.
3. When SAPinst displays the CD browser-window and asks for the Export Migration CD,
enter the path to the export directory <EXPDIR>.
5 R3load Procedures using the Migration Monitor
80
4. Continue as described in the installation documentation for your SAP component until a
dialog box appears that states:
If the export has been started on the source system and the
Export Monitor is running, you can now start the data load by
starting the Import Monitor.
5. Check that the prerequisites in the dialog box are fulfilled by your system. If so, you can
start the Migration Monitor.
6. Complete the installation as described in the installation documentation for your SAP
component.
If you have to restart the import after an error, just restart SAPinst. The import
is continues with the table that was not imported successfully.
6 Final Activities
81
6 Final Activities
To finish the system copy of your SAP system:
...
1. Perform follow-on actions in the source system [Page 82].
2. Perform follow-on actions in the target system [Page 83].
6 Final Activities
82
6.1 Performing Follow-On Actions in the
Source System
...
1. Reschedule your canceled jobs:
Tools → CCMS → Jobs → Maintenance (SM37).
2. Using CCMS, adapt your operation mode timetable to the original status:
Tools → CCMS → Configuration → Operation mode calendar (SM63).
6 Final Activities
83
6.2 Performing Follow-On Actions in the
Target System
Procedure
Actions on Operating System Level
...
1. Adapt the configuration files on operating system level to meet network and SAP
requirements.
2. Adapt additional SAP software components (for example, RFC, CPIC, SAP
ArchiveLink) if required.
3. Adapt additional non-SAP software components (for example, archiving systems,
monitoring tools, job schedulers) if required.
4. Adapt backup programs (for example BRBACKUP, BRARCHIVE, BACKINT) if
required.
5. Adapt non-SAP directories, file systems, NFS mounts, etc. if required.
6. Check the SAP parameters of the default and instance profiles.
7. Check your UNIX shell files for special entries.
8. Check crontab or AT jobs.
9. Check operating system files (for example, .netrc, .rhosts).
10. Check operating system printers.
11. Oracle: Adapt the database profiles init<SAPSID>.ora, init<SAPSID>.dba and
init<SAPSID>.sap.
Actions on Database Level
...
1. Before starting the SAP system, make sure that the logging mechanism of the
database is active.
2. Check the parameters in the database profiles.
3. Oracle: Delete all entries from the following tables: DBSTATHORA, DBSTAIHORA,
DBSTATIORA, DBSTATTORA.
4. Oracle: Delete the user OPS$<SOURCE_SAPSID>ADM.
5. Oracle: If you changed the <DBSID> during the system copy, it is recommended to
adapt the global_name parameter with the following SQL command:
alter database rename global_name to <NEW_DBSID>;
If the parameter is not existing on your system, ignore this step.
Actions on SAP System Level
...
1. Run installation check: Administration → System administration → Administration →
Installation Check (transaction SM28).
2. Configure the Workbench Organizer (SE06) with the option Database Copy. This
releases all transport, repair, and customizing requests that have not been released in
the source system.
3. Adapt the transport parameters and transport routes in the Transport Management
System (TMS):
6 Final Activities
84
a. Choose transaction STMS → Overview → Systems.
b. Select the system and select the tab Transporttool.
To adapt the transport routes:
Choose transaction STMS → Overview → Transport routes.
4. Delete all entries from the following tables: ALCONSEG, ALSYSTEMS, DBSNP, MONI,
OSMON, PAHI, SDBAD, SDBAH, SDBAP, SDBAR.
5. Delete canceled and finished jobs.
Execute ABAP program RSBTCDEL, marking the field delete with forced mode: Tools →
ABAP Workbench → ABAP Editor (SE38).
6. Adapt all jobs needed in the target system:
a. Copy the old jobs.
b. Modify the new jobs.
c. Delete the old jobs.
7. Check the consistency of the Temporary Sequential Objects (TemSe) by searching for
files of TemSe objects for which no TemSe objects exist: Administration → CCMS →
Spool → TemSe administration (SP12). For more information, see SAP Note 16875.
8. Adapt the definition of the printers to meet the new system requirements:
�� Device types and character set definitions
�� Spool servers
�� Output management systems (OMS)
9. Delete entries in table DDLOG for buffer synchronization. Synchronize the buffers as
described in SAP Note 36283. Adapt the client information for the logical system.
10. Adapt the RFC destination: Tools → Administration → Administration → Network →
RFC destinations (SM59). Clean the transactional RFC: Tools → Administration →
Monitor → Transactional RFC (SM58). See the relevant description in the SAP Online
Documentation.
11. Create new operation modes and remove old ones:
...
a. Create new operation modes and instance definitions.
b. Maintain the timetable using the new operation modes.
c. Delete the old operation modes and old instance definitions.
12. Adapt the operation mode time tables (CCMS): Administration → CCMS →
Configuration → Operation mode calendar (SM63).
13. Adapt the instances and profiles (CCMS): Administration → CCMS → Configuration →
OP modes/instances (RZ04).
14. Define or remove the SAP system users: Tools → Administration → User maintenance
→ Users (SU01). Furthermore, revise the authorizations of the system users.
15. Delete all entries from tables TPFET and TPFHT. These contain information about
changes made to the profile of your source system: Administration → CCMS →
Configuration → Profile Maintenance (RZ10).
IBM DB2 UDB for iSeries: Use the commands CLRPFM R3<SID>DATA/TPFET and
CLRPFM R3<SID>DATA/TPFHT.
6 Final Activities
85
16. Adapt other CCMS settings (for example, alert thresholds, reorganization parameters of
CCMS table MONI) if required.
17. Delete all entries from table TLOCK which holds the repair requests from your source
system.
18. Make data archived in the source system (data that does not reside in the database but
was moved to a different storage location using SAP Archive Management) accessible
in the target system. Adapt the file residence information in the target system. Refer to
the SAP Online Documentation (SAP Web Application Server → System
Administration → Application Data Archiving and Reorganization) for help.
19. Redefine database actions (backup, update statistics, etc.) if you have used the DBA
calendar in the source system (DB13).
20. Check logon groups and assignment of application servers to logon groups (SMLG).
21. Check the connection to SAPNet - R/3 Frontend (OSS1).
22. Check self-defined external commands (SM69).
23. Check the thresholds (RZ06).
24. Check entries of the following tables in all involved systems:
�� TXCOM (SM54)
�� THOSTS (SM55)
25. Check the logical system names: Refer to the discussion of logical system names in
the preparation section of this guide [Page 14]. If your circumstance is such that logical
system names must be changed in the system that results from the copy, then perform
this logical system name change at this time per OSS notes 103228 and 544509. Your
corporate logical system naming strategy should be observed when making this
change.
BW customers: If you have copied a BW system, read SAP Note 325525.
26. Check for every client in your SAP system the detail settings (client role, changes and
transports for client-dependent objects, changes for client-independent objects,
protection level, restrictions) (SCC4).
27. Check if you can delete clients that will no longer be used in the target system (SCC5).
28. Check the contexts and segments of remote application servers for the SAP Monitoring
Infrastructure if required (RZ21).
29. Configure the domain controller in the Transport Management System (TMS) by using
transaction code STMS.
30. Post-processing concerning customer objects:
...
If customer objects are not original in the new system, modify the corresponding entries
in table TADIR.
If you encounter problems modifying a customer development class using transaction
SMTS or SM31, try to use the option Validate (ENTER) instead of the option Save to
save your changes.
31. ABAP Program Loads
The ABAP loads are platform-dependent programs which are generated during runtime
and which are stored in database tables. They are not exported when you use the
R3load procedure to copy your SAP system. The ABAP loads are generated in the target
system when they are first used. This may, however, reduce production system
6 Final Activities
86
performance. To avoid this, you can use transaction SGEN to generate the missing
loads.
Load generation requires a large amount of system resources. You should schedule the
generation job to run overnight.
For a detailed description of the features, see the online documentation in transaction
SGEN by choosing Information on the SAP Load Generator, or in the Job Monitor by
choosing Job Monitor.
32. If you did not change the database when copying the system, you have to start
program RS_BW_POST_MIGRATION in the background with variant
SAP&POSTMGR. Otherwise, if you changed the database when copying the system,
you have to start program RS_BW_POST_MIGRATION in the background with variant
SAP&POSTMGRDB. Program RS_BW_POST_MIGRATION performs necessary
modifications on DB specific objects (mainly BW objects)..
Checking the Target System
The following actions are suitable for checking the consistency of the target system:
...
1. Perform initial consistency check (SM28).
2. Check the system log on all application servers (SM21).
3. Check the consistency of the database (DB02).
4. Perform server check (SM51).
5. Test transactions frequently used by the customer.
6. FI customers: Run the job SAPF190 (accounting reconciliation) and compare the
results to those gained on the source system before the system copy (Accounting →
Financial Accounting → General ledger → Periodic Processing → Closing →
Check/count → Comparison).
7. FI customers: Run the jobs RFUMSV00 (tax on sales/purchases), RAGITT01 (asset
history sheet), RAZUGA01 (asset acquisitions), RAABGA01 (fixed asset retirements) and
compare the results to those gained on the source system before the system copy.
8. CO customers: Run the report group 1SIP and compare the results to those gained on
the source system before the system copy.
7 Additional Information
87
7 Additional Information
7.1 Remote Installation with SAPinst
88
7.1 Remote Installation with SAPinst
Purpose
You can run the SAPinst GUI in standalone mode to perform a remote installation.
This enables you to install an SAP system on another host (the remote host) while monitoring
the installation with the SAPinst GUI on your local Windows or UNIX computer (the local
host).
Prerequisites
Make sure that you have performed the preparation activities for your local host
(SAPinst GUI host) and your remote host.
For more information, see Installation Preparations in this documentation.
Both computers are in the same network and can ping each other.
To test this:
�� Log on to your remote host and enter the command ping <local host>.
�� Log on to the local host and enter the command ping <remote host>.
SAPinst ports
SAPinst uses the port 21212 during the installation for communication with the SAPinst
GUI. If one of these ports is already used by another service, SAPinst aborts the
installation with an appropriate error message.
In this case, you must start SAPinst or the SAPinst GUI from the command prompt as
follows:
In the following commands, <free_port_number> defines an unused port
number. Since SAPinst also uses <free_port_number> + 1, this must
also be free.
For example, if you enter 60000 as <free_port_number>, SAPinst uses
the port 60000.
�� UNIX:
�� SAPinst: ./sapinst
SAPINST_DIALOG_PORT=<free_port_number>
�� SAPinst GUI: ./sapinstgui.sh -port <free_port_number>
�� Windows:
�� SAPinst: sapinst SAPINST_DIALOG_PORT=<free_port_number>
�� SAPinst GUI: sapinstgui.bat -port <free_port_number>
Process Flow
...
1. You start the SAPinst server on your remote host.
2. You start the SAPinst GUI on your local host.
3. You perform the installation using the SAPinst GUI.
For more information, see:
7.1 Remote Installation with SAPinst
89
Starting SAPinst on the Remote Host [Page 89]
Starting SAPinst GUI on the Local Host [Page 90]
7.1.1 Starting SAPinst on the Remote Host
Use
You use this procedure to run SAPinst on the remote host when you want to run SAPinst as
a remote installation [Page 87]. The remote host is the host where you want to install the SAP
system.
Prerequisites
You have prepared your system for SAPinst, that is you must have set the DISPLAY
environment variable to <host_name>:0.0, where <host_name> is the host on which the
SAPinst GUI will be displayed.
Procedure
Your Remote Host Runs on a Windows Platform
...
1. Log on to your remote host as a user who is a member of the local administration
group.
2. Insert the installation DVD in your DVD drive.
3. Start SAPinst from the SAP Installation Master DVD:
Enter the following commands:
<DVD drive>:\IM<x>_<OS>\SAPinst\NT\<OS>\sapinst.exe
SAPINST_START_GUI=false
4. SAPinst now gets started without the SAPinst GUI and waits for the connection to the
SAPinst GUI. That is, you see the following at the command prompt:
guiengine: no GUI connected; waiting for a connection on host
<host_name>, port <port_number> to continue with the installation
5. Start the SAPinst GUI on your local host, as described in Starting SAPinst GUI on the
Local Host [Page 90].
Your Remote Host Runs on a UNIX Platform
...
1. Log on to your remote host as user root.
2. Mount the SAP Installation Master DVD.
3.
Mount the DVD locally. We do not recommend using Network File System
(NFS).
4. Start SAPinst from the SAP Installation Master DVD:
Using the default installation directory
Enter the following commands:
cd <SAP_Installation_Master_DVD>/IM<x>_<OS>/SAPINST/UNIX/<OS>
./sapinst SAPINST_START_GUI=false
7.1 Remote Installation with SAPinst
90
5. SAPinst now gets started without the SAPinst GUI and waits for the connection to the
SAPinst GUI. That is, you see the following at the command prompt:
guiengine: no GUI connected; waiting for a connection on host
<host_name>, port <port_number> to continue with the installation
6. Start the SAPinst GUI on your local host, as described in Starting SAPinst GUI on the
Local Host [Page 90].
7.1.2 Starting SAPinst GUI on the Local Host
Use
You use this procedure to run SAPinst GUI on the local host when you want to run SAPinst
as a remote installation [Page 87]. The local host is the host where you want to control the
installation with the SAPinst GUI.
Prerequisites
You have prepared your system for SAPinst, that is you must have set the DISPLAY
environment variable to <host_name>:0.0, where <host_name> is the host on which the
SAPinst GUI will be displayed.
Procedure
Your Local Host Runs on a Windows Platform
...
1. Log on to your local Windows host.
2. Insert the installation DVD into your DVD drive.
3. Change to the following directory:
cd <DVD drive>:\IM<x>_<OS>\SAPinst\NT\<OS>
4. Start the SAPinst GUI in one of the following ways:
�� With additional parameters:
Enter the following from the Windows command line:
startinstgui.bat host -<host_name> -port <port_number>
<host_name> is the host name of the installation host
<port_number> is the same port as SAPinst uses on the remote host
�� Without additional parameters:
i. Enter the follwing from the Windows command line:
startinstgui.bat
The SAPinst GUI starts and tries to connect to SAPinst on the local host.
However, normally there is no SAPinst running on the local host.
Therefore, the SAPinst GUI cannot connect and the SAP Installation GUI
Connection dialog appears.
ii. Enter the host name of the Installation Host and the same Port as
SAPinst uses on the remote host and choose Log on.
5. Perform the installation from your local host.
7.1 Remote Installation with SAPinst
91
Your Local Host Runs on a UNIX Platform
...
1. Log on to your local UNIX host as user root.
2. Mount your installation DVD
Mount the DVD locally. We do not recommend using Network File System
(NFS).
3. Change to the following directory:
cd <SAP_Installation_Master_DVD>/IM<x>_<OS>/SAPINST/UNIX/<OS>
4. Start the SAPinst Gui in one of the following ways:
�� With additional parameters:
Enter the following from the UNIX command line:
./startInstGui.sh host <host_name> -port <port_number>
<host_name> is the host name of the installation host
port_number> is the same port as SAPinst uses on the remote host
�� Without additional parameters:
i. Enter the following from the UNIX command line:
startinstgui.sh
The SAPinst GUI starts and tries to connect to SAPinst on the local host.
However, normally there is no SAPinst running on the local host.
Therefore, the SAPinst GUI cannot connect and the SAP Installation GUI
Connection dialog appears.
ii. Enter the host name of the Installation Host and the same Port as
SAPinst uses on the remote host and choose Log on.
5. Perform the installation from your local host.
7.1 Remote Installation with SAPinst
92
7.2 Interrupted Installation with SAPinst
Use
The SAP system installation might be interrupted for one of the following reasons:
An error occurred during the processing phase:
SAPinst does not abort the installation in error situations. If an error occurs during the
processing phase, the installation will hold and a dialog box appears. The dialog box
contains a short description about the choices listed in the following table as well as a
path to a log file that contains detailed information about the error.
You interrupted the installation by choosing Cancel.
The following table describes the options in the dialog box:
If you
choose..
.
Meaning
View Log A frame appears with history information (log files) about the steps that you
performed last . These log files contain a short description of the error that has
occurred. The dialog box will remain in the background, so you can choose one of
the following two options (Retry or Stop) after viewing the log information.
Retry Since SAPinst records the installation progress in the keydb.xml file, you can
continue the installation by choosing Retry.
The installation continues from the point of failure without repeating any
of the previous steps.
We recommend that you view the entries in the log files, try to solve the problem
and then choose Retry.
If the same or a different error occurs again, the dialog box will appear again.
Stop The dialog box and the SAPinst GUI will be closed. The installation stops but
SAPinst records the installation progress in the keydb.xml file. Thus, you can
continue the installation from point of failure without repeating any of the previous
steps. See the procedure below.
Reset You must restart from the beginning, that is, with the default keydb.xml file.
You must delete the previous installation before you restart SAPinst.
For more information about deleting a SAP Web AS installation, see
Deletion of an SAP System Installation in the documentation
Component Installation Guide - SAP Web Application Server Java
6.40 SR 1 on <OS>: <DB>, Part II Installation and Post-
Installation.
7.1 Remote Installation with SAPinst
93
UNIX only: You can also terminate SAPinst by choosing Ctrl+C. However, we
do not recommend that you use Ctrl+C, because this kills the process
immediately .
Procedure
The following procedures describe the steps to restart an installation, which you stopped by
choosing Stop, or to continue an interrupted installation after an error situation.
Log on to your remote host as a user who is a member of the local administration group.
...
1. Insert the installation DVD in your DVD drive.
2. Enter the following commands from the Windows command prompt:
cd <DVD drive>:\IM<x>_<OS>\SAPinst\NT\<OS>
sapinst.exe
3. From the tree structure in the Welcome screen, select the installation task that you
want to continue and choose Next.
If there is only one component to install, SAPinst directly displays the dialog
What do you want to do? without presenting the Welcome screen.
The What do you want to do? screen appears.
4. In the What do you want to do? screen, decide between the following alternatives and
choose OK.
Alternative Behavior
Run a new
Installation
The installation will not be continued.
Instead, SAPinst deletes the mentioned installation directory for the chosen
installation service and starts the installation from the beginning.
The log files from the old installation are put into a backup directory with the
following naming convention:
<log_day_month_year_hours_minutes_seconds> (for example,
log_01_Oct_2003_13_47_56).
Continue old
installation
The installation that was interrupted is continued from the point of failure.
UNIX
Log on to your local UNIX host as user root.
5. Mount your installation DVD.
Mount the DVD locally. We do not recommend using Network File System
(NFS).
6. Enter the following commands:
cd <SAP_Installation_DVD>/IM<x>_<OS>/SAPINST/UNIX/<OS>
7.1 Remote Installation with SAPinst
94
./sapinst
7. From the tree structure in the Welcome screen, select the installation task that you
want to continue and choose Next.
If there is only one component to install, SAPinst directly displays the dialog
What do you want to do? without presenting the Welcome screen.
The What do you want to do? screen appears.
8. In the What do you want to do? screen, decide between the following alternatives and
choose OK.
Alternative Behavior
Run a new
Installation
The installation will not be continued.
Instead, SAPinst deletes the mentioned installation directory for the chosen
installation service and starts the installation from the beginning.
The log files from the old installation are put into a backup directory with the
following naming convention:
<log_day_month_year_hours_minutes_seconds> (for example,
log_01_Oct_2003_13_47_56).
Continue
old
installation
The installation that was interrupted is continued from the point of failure.
regards
karthik
reward me if usefull
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
hi sachin,
ok.i will provide you some more information.
If you are already using the backwards-compatible 6.40 kernel, you can use the 6.40 migration tools. They are discussed in detail in the homogeneous and heterogeneous system copy guide which can be downloaded at http://service.sap.com/instguides*.The manual explains the prerequisites, the step-by-step procedure and the post-processing tasks.
Please note that you need a certified migration consultant to assist you with heterogeneous system copies
Informix Administration and Housekeeping for SAP R/3
Level: Introductory
Preventing potential problems in the R/3 Informix Dynamic Server (IDS) database environment is as important as fixing problems. Preventive administration is achieved by constantly managing and monitoring certain aspects of the database system for details such as optimal response times for database access and update activities, database growth, and database integrity.
Introduction
The key to supporting and maintaining any SAP R/3 environment is to be proactive in the maintenance and support of the system. Preventing potential problems in the R/3 Informix® Dynamic Server® (IDS) database environment is as important as fixing problems. Preventive administration is achieved by constantly managing and monitoring certain aspects of the database system for details such as optimal response times for database access and update activities, database growth, and database integrity. The following Tasks and ctivities section shows various functions that need to be performed at specific times by the Informix database administrator.
This article will discuss the following:
Tasks and activities
Tools
Client copies
Database copies
Database growth
Tasks and activities
The tasks and activities described in this section are the activities the Informix database administrator needs to perform in order to manage and maintain an optimal R/3 Informix database environment and the required administration tasks needed to support the R/3 environment. These do not include other critical items that the R/3 administrator performs in their normal routine to administer the R/3 system, such as checking that the R/3 system is up and accessible and performing optimally.
Performing these tasks are recommended for supporting a production system. You may decide that some of the system tasks are not required in all system environments, such as a sandbox, or that the frequency indicated is not appropriate for all the systems in your landscape. Some decisions, such as the topics listed under Initial Database Setup and Maintenance Planning, need to be part of your system architecture and design. These are general recommendations that should be implemented in the environments that are appropriate for your site. Remember that a development environment is your production environment if it is the only environment being used. It can be critical to the timeline of the project if a development environment becomes unavailable to users during the development phase.
Initial database setup and maintenance planning
Your answers to the questions in Table 1 below can affect how you set up the Informix environment after the initial installation of the R/3 application. Most of these questions deal with the logical log setup, archive and backup requirements, and database growth. It is important to ask management and the functional teams these questions AND TO GET ANSWERS. The answers will vary by the SAP System ID (SID). You should work with the R/3 administrator and the UNIX administrator at your site to develop a cohesive strategy for some of these operational issues
regards
'
karthik.
hi sachin,'
this is the continuation for the above info...
Table 1. Database Setup and Planning Questions
Daily Checks and Tasks
Check that the database is on line and available.
If R/3 is available then, normally, the R/3 Informix database is on line and available.
Execute the onstat - command to verify that the Informix database is on line, and execute the onstat -u command to verify that the user sapr3 is connected to the database.
Use SAPDBA to verify that the user sapr3 is connected.
Check that all the dbspaces are on line, as appropriate, using the onstat -d command. All dbspaces being on line and available is also dependent on the $ONCONFIG ONDBSPACEDOWN parameter setting.
If the ONDBSPACEDOWN is set to a value other than 1 (abort), there may be dbspaces that are marked as down when this system is on line.
Perform consistency checks of the Informix catalog and reserve tables using oncheck -cc and oncheck -cr.
Check the status of any nightly scheduled R/3 background or UNIX cron jobs using the R/3 transactions DB13 or SM37, or UNIX commands.
Verify that the logical logs are backed up and freed sufficiently.
Use the R/3 transactions DB12, DB13 or SM37 if you are using onarchive.
Execute onstat -l to verify that the logical logs are being freed sufficiently and onstat -m to verify that the logical logs are being backed up successfully.
If using a third-party storage management tool, check the reports from that tool.
Perform a database archive and verify the database archive has completed successfully.
Execute the onstat -m command to verify that an Archive has completed successfully.
If using a third-party storage management tool, check the reports from that tool.
Check the online message log for problem messages using the onstat -m command or the R/3 transaction ST04.
Check for mail messages in Informix UNIX mail.
Check for Informix-related messages in /var/adm/messages file (the filename may vary by operating system).
Monitor that the Informix system is performing optimally using the R/3 transaction ST04.
Monitor R/3 queries, as needed, using R/3 transactions ST04 and ST05.
Review the SAP R/3 database (call and response) timings using R/3 transactions ST03 and ST04.
Execute UPDATE STATISTICS for a specific deviation level or for specific tables using SAPDBA or R/3 transactions DB13, DB20.
Use the R/3 transaction TU01 to determine the tables that have the highest update activity.
Be as proactive as possible in investigating on line bottlenecks and possible issues.
Weekly Checks and Tasks
Perform weekly database archives and logical log backups. These can be scheduled using R/3 transaction DB13 if your backup and archive method is onarchive.
Verify that the weekly database archive and the logical log backups have completed successfully.
Use R/3 transactions DB12, DB13 or SM37 if you are using onarchive.
Execute onstat -l to verify that the logical logs are being freed sufficiently and onstat -m to verify that an archive has completed successfully and that the logical logs are being backed up successfully.
If using a third party storage management tool, check the reports from that tool.
Check that all the dbspaces are less than 80% full (or your selected critical point level) using SAPDBA or the R/3 transaction DB02.
Monitor the database growth using the R/3 transaction DB02 or SAPDBA.
Release unused memory using the onmode -F command.
Perform database consistency checks using SAPDBA or the Informix oncheck utility.
Send archive and backup media to offsite storage as appropriate to your site storage requirements.
Check and remove Informix dumps, if not needed by Informix technical support, from the Informix dump directory.
If using onarchive, remove /tmp/ARC* and /tmp/oncat* files.
Monthly or Quarterly Checks and Tasks
Recycle tape backup media to proper tape retention cycles.
Cleanup the $INFORMIXDIR message and console files.
Schedule table reorganizations using SAPDBA, as necessary.
Execute UPDATE STATISTICS (force) for the entire database, if applicable, using SAPDBA or the R/3 transaction DB13.
Perform database system checks using SAPDBA, the R/3 transactions DB13 or DB16, or the R/3 line command infcfgcheck.
Prior to Heavy Load Activity and Client Transports
Investigate if changing logical logs to /dev/null is an option.
Increase the number of logical logs if necessary (even if they are set to /dev/null).
Increase the number of LOCKS in $ONCONFIG.
Perform a level 0 archive.
After Heavy Load Activity, Client Transports, and Client Deletes
Execute UPDATE STATISTICS (force) for all the updated tables.
Check that all the dbspaces are on line (see note under Daily Checks and Tasks regarding dbspaces being on line).
Check that all the dbspaces are less than 80% full (or your selected critical point level).
Reduce the number of LOCKS in $ONCONFIG (if increased).
Change the LTAPEDEV parameter value back to the original setting if it was changed prior to the activity.
Perform a level 0 archive.
As-Required Activities
Perform an Informix database software upgrade.
Perform a database copy.
Perform a database restore.
Tools
Many of the activities indicated above can be accomplished in a multiple of different ways and with different tools: SAPDBA, R/3 Computing Center Management System (CCMS), and Informix utilities. There is functionality overlap between the different methods and they all complement each other. The question is which tool should be used for which function? The rule of thumb is to use R/3 for monitoring and SAPDBA and the Informix tools for modifying database parameters and the database. But, as always, the best tool to use is the tool that you are most comfortable using. Table 2 details which tools can be utilized for various activities.
Table 2. Available tools
The R/3 transactions offer the consistency and familiarity of the R/3 interface and are sufficient for monitoring the Informix database environment when the R/3 system is available. SAPDBA and the Informix tools can be used for routine and non-routine database administrative tasks and can be used when R/3 is unavailable.
Client copies
There are times when you need to create a client that looks just like another client in the same SID or in a different SID. Performing this task is accomplished using the client copy transactions in R/3. The R/3 administrator will most likely perform the actual client copy transactions and procedures. The database administrator needs to be aware of some items that will affect the database and the environments.
The system from which you are copying is called the source client and the system to which you are copying is called the target system. When you copy one client to another within the same SID, the source and target systems are the same. This is shown in Figure 1.
Figure 1: Client Copy within the Same SID
If you are copying a production client to another client in a different SID there will be production data in a non-production SID. This may cause security issues that should be addressed.
Sufficient space should be allocated in the dbspaces in the target system to complete the client copy. If you are just copying configuration data, the space requirements are much less than if you are copying master and transaction data. Generally, for configuration data only, approximately 300 megabytes of free space is required throughout the dbspaces. Master and transaction data always go together, as one cannot be copied without the other. Regarding copying master and transaction data, the larger the client from which you are copying, the more space that will be required in the target system. Copying master and transaction data space requirements will depend on the amount of data in the source client. Master and transaction data normally will increase the size of psapbtab and psapstab dbspaces. To estimate the amount of space required for a specific client copy, a test client copy can be performed prior to the actual client copy.
Client-independent objects such as ABAP/4 programs, tables, and structures will not be copied during a client copy.
Both the target and the source system should not be used during the client copy (to prevent database inconsistencies and locks). Larger client copies can consume large amounts of processing time and may impact the use of the source and target systems for a greater length of time.
If the client that is being created already existed in the SID and is being overlaid, you will need to increase the number of LOCKS in the $ONCONFIG file and, if possible, set the LTAPEDEV parameter value to /dev/null so as not to run out of logical logs. Or, you can increase the number of logical logs if the LTAPEDEV parameter cannot be set to /dev/null. In essence, re-creating a client is first deleting all the data for the client and then inserting all the data for the client. You may want to suggest to the R/3 administrator using R3TRANS to delete the client first to cut down the time of the actual client copy.
Review the most current OSS notes regarding client copies.
Additional information about client copies can be found in the SAP help documentation.
Database copies
There are times when a client copy is not sufficient to copy data from one SID to another. Sometimes you need a copy of all of the client-independent and client-dependent objects, or you need to copy multiple clients, or you need a new test or training system. The easiest way to do this is by copying the entire Informix R/3 database. Using the database copy procedure, the target system database is totally replaced with a copy of the source system database. When the copy is complete the target system database name will be the same as the source system database name. You should not create a production system using a database copy; use the proper transport methods for creating production environments.
Figure 2: Database Copy
There are several things to consider when planning a database copy to populate an R/3 database:
Where are you copying the database?
Are you overlaying another SID database?
Do you have a backup of the original SID database so you can restore the database, if needed?
Do you have hardware and resources to create a new R/3 SID and database?
The database disk requirements from the source system must be the same on the target system.
Is R/3 already installed on the target system? If not, it should be installed first.
Is this a one-time occurrence or does the possibility exist that this will happen often?
If you are copying the production system, will there be security issues due to the type of data in the system?
Will the new system be part of the R/3 landscape and need to be incorporated in the transport process?
If you are changing the hardware platform or migrating from another database to Informix, hence a heterogeneous system copy, you will need to request a migration key from SAP for certain R/3 releases.
The Informix database version must be the same for the source and target database systems in the homogeneous database copy.
A new SAP license will be needed for the target SID.
There are different ways to copy an R/3 Informix database. You can use SAP tools or Informix tools. Two methods are noted below:
Homogeneous database copy using SAP R/3 tools. This is the only method supported by SAP.
Informix database archive of the source SID database, and Informix database restore to the target SID database.
The procedures described here should be sufficient to complete the database copy. In addition, there are various Informix utilities to copy the data from one system to another that are not discussed here.
For All Methods
Determine the disk space requirements for the target database system from the source database system. Obtain a copy of the SAPDBA dbspace information or a copy of the onstat -d from the source database system. Create the same dbspace and chunk information on the target system as that of the source system.
An R/3 installation needs to be completed on the target system. Although you do not need to install the database delivered with the R/3 installation kit, it is recommended that you do. Complete the target system installation, including the database build, and test the target system. Verify that the target system works properly prior to overlaying the database.
Run UPDATE STATISTICS and data consistency checks on the source Informix R/3 database. Copying inconsistent or erroneous data will only give you two copies of inconsistent and erroneous data.
Review the current OSS notes regarding database copies, homogeneous system copies, or heterogeneous systems copies (e.g., 89698, 89188).
SAP Database Copy
The SAP database copy uses the R3INST utility and is similar to a standard R/3 installation. When performing an R/3 installation you can choose the Load SAP-Data from existing SAP R/3-System option. This will prevent the SAP delivered database on the CD-ROMs included in the installation kit from being built. To stress the point, it is recommended to install the database from the installation CDs so that you can verify that the R/3 system is properly installed.
The steps below make up the SAP database copy process:
Export the database source and create the export control files. Creating the control files and exporting the database is performed using R3INST for R/3 releases prior to 4.x and R3SETUP for R/3 releases 4.x and after.
Transfer the database export and control files from the source SID system to the target SID system. This can be done using file transfer protocol (FTP), copying the files from one machine to another, or by sharing the files across the machines using network file services (NFS). It is usually better to copy the files using FTP rather than using the network to access the files.
Load the exported source SID database into the target SID. The database import is performed using R3INST for R/3 releases prior to 4.x and R3SETUP for R/3 releases 4.x and after.
Create an archive of the new database.
Start R/3 and perform an installation and consistency check using the R/3 transaction SICK. The R/3 administrator should review the General Subsequent Actions section in the R/3 Homogeneous System Copy manual for the R/3 release you are working with and perform any necessary tasks and reconfigurations.
Please refer to the R/3 Homogeneous System Copy manual for your R/3 release for step-by-step details of the first three items detailed above.
Informix Database Restore Method
When working with very large databases, the database restore method is one of the fastest ways to copy one R/3 system to another R/3 system. Using this method means the target R/3 SID and database name will be the same as the source system name. This also means that the source system and the target system must be two separate physical machines. The target system database will be exactly the same size as the source system database and the dbspace and chunk information must match exactly between the two systems. The following are the steps for the Informix database restore copying method:
Create a level 0 archive of the source system Informix database.
Create the dbspaces and chunks on the target system with the target system SID name.
If the target SID name does not match the source SID name, create a symbolic link in the /informix directory for the target SID to match the source SID.
su - informix
cd /informix
ln -s targetSID sourceSID
Install the target R/3 system using R3INST if it is not already installed.
Alter the $ONCONFIG and sqlhosts files as necessary.
Restore the level 0 archive to the target system.
If you do not want the target SID to have the same name as the source SID you can rename the database.
dbaccess-
rename database source SID to target SID ;
ctrlC
Renaming the database will not change the internal Informix reserve pages that contain the original source SID path names. To change the internal Informix reserve pages, Informix technical support needs to be involved (using the Informix technical support tbchksap utility). You cannot delete the symbolic link in the /informix directory unless the internal reserve pages match the SID name. You will also not be able to use SAPDBA to perform dbspace extensions. The Informix utility, onspaces, or onmonitor, must be used to create, add, and drop chunks and dbspaces.
Run UPDATE STATISTICS and data consistency checks of the new database (if this had not been recently done on the source database).
Create an archive of the new database.
Start R/3 and perform an installation and consistency check using the R/3 transactions SICK and DB02 --> Check. The R/3 administrator should review the General Subsequent Actions section in the R/3 Homogeneous System Copy manual for the R/3 release you are working with and perform any necessary tasks and reconfigurations.
Database growth
Database layout, size and growth have a large effect on the overall performance of the database. Though you are concerned with database layout, size and growth in all your R/3 environments, usually your production environment is the most important environment for these factors. Sometimes a sandbox or development R/3 environment is installed with the installation default values for dbspace sizes just to get the environment up, running, and available. This should never be the case for a production R/3 environment.
One of the objectives in laying out the database dbspaces across the available disks and controllers is to provide adequate space for even growth of the dbspaces across the disks and controllers so as not to introduce I/O bottlenecks or disk hot spots. Database growth should be regularly monitored using the R/3 transaction DB02 or SAPDBA. Examine the dbspace and table history for fast growing tables and be as proactive as possible in allocating additional chunks, implementing table fragmentation, or detaching indexes, as required. Monitor dbspace usage for I/O bottlenecks and move dbspaces to other disks if values are skewed or if bottlenecks are found.
Summary
Many of the administration and housekeeping functions related to an R/3 environment are not very different from the same tasks required in other Informix application environments. The difference is the size of the R/3 application itself and the manner in which administration and monitoring are performed in an R/3 environment
regards
karthik
reward me points ...if usefull
This message was moderated.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.