SAP Note
954268 - Optimization of export: Unsorted unloading
Version 23 Validity: 15.11.2012 - active Language English
Header Data
Released On 15.11.2012 12:02:18
Release Status Released for Customer
Component BC-INS-MIG DB/OS Migrations
Priority Recommendations / Additional Info
Category Installation information
Symptom
During a system copy, you want to accelerate the export by unloading the data unsorted.
Other Terms
R3ldctl, system copy, SAPinst, DDLORA.TPL, DDLDB2.TPL, DDLDB6.TPL
Reason and Prerequisites
During a system copy, you can unload the data as sorted or as unsorted. By default, the system
unloads the data as sorted. This is controlled by the following entry in the DDL<db>.TPL
file: prikey: .... ORDER_BY_PKEY
Various factors determine whether unsorted unloading is possible:
l Target database
If the target database does not require any sorted data during the import, you can unload many
(but not all) of the tables unsorted (see below). If the target database is MaxDB, you must
export all of the tables as sorted.
If the target database is MSSQL, see SAP Note 1054852.
l Source database
If you use MaxDB as the source database, you can unload sorted data only. Do not override
this option when you export MaxDB.
l Table type:
Kernel or R3load lower than 6.40:
Tables of the 'Report' type (R), 'Nametab' type (N) and 'Screen' type (D) must always be
unloaded sorted. The table type is part of the STR file. It is stored in the fifth column of the
'att:' line of the relevant table.
Example for an ABCUSER1 table of the 'transparent' type (T):
tab: ABCUSER1
att: USER 0 ?A T all ABCUSER1~0 USER 0
Therefore, the following tables must always be unloaded as sorted (* means that all of the
tables that follow this pattern must be unloaded as sorted):
¡ D010*
¡ D02*
¡ DDNTF*
Kernel or R3load 6.40 >= 6.40:
Tables of the type "Nametab" (N) must always be unloaded as sorted.
Therefore, the following tables must always be unloaded as sorted:
¡ DDNTF*
l Code page conversion
If you convert the codepage, you must always unload cluster tables as sorted. The codepage is
converted for the following actions:
¡ non-Unicode --> Unicode
¡ MDMP --> Unicode
¡ Unicode (big endian) --> Unicode (little endian)
¡ Unicode (little endian) --> Unicode (big endian)
¡ EBCDIC --> ASCII
Caution: For a code page conversion, the conversion must be carried out during the export
to ensure that cluster tables are always exported immediately.
Solution
Prerequisites for using the unsorted unload:
l R3ldctl
The most up-to-date R3ldctl from SAP Service Marketplace, which generates two templates:
DDL<db>.TPL for "sorted" unpacking and DDL<db>_LRG.TPL for "unsorted" unpacking.
The following patch levels support unsorted unloading:
¡ Oracle: R3ldctl 6.40, as of patch level 7
¡ Oracle: R3ldctl 7.00, as of patch level 5
¡ DB2 z/OS : R3ldctl 7.00, as of patch level 9
¡ DB2 LUW : R3ldctl 7.00, as of patch level 6
Alternatively, you can create the template for the unsorted export manually.
l R3load
The latest R3load from the SAP Service Marketplace, which ensures that specific tables (type:
report, screen, nametab) and cluster tables are always unloaded sorted during a codepage
conversion - irrespective of the configuration in the DDL<db>.TPL file.
¡ R3load 6.40, as of patch level 55 or
as of the compile date February 10, 2006 2006
¡ R3load 7.00, as of patch level 10 or
as of the compile date February 10, 2006 2006
l Migration Monitor
You can use the new Migration Monitor to assign different DDL<db>*TPL templates to the packages.
¡ MIGMON.SAR 7.00, as of patch level 1
The Migration Monitor is downward-compatible. As new features and bug fixes are only
implemented in the latest version, SAP urgently recommends that you always use the latest
version available.
Release and availability:
The 'unsorted unload' feature is integrated in SAPinst as of SAP NetWeaver 2007 and can be
configured by means of the dialogs.
If you want to use the feature in earlier releases (that is, earlier than SAP NetWeaver 2007), you
must configure the unsorted export in the Migration Monitor manually. The feature should only be
used by a certified migration consultant or by an experienced migrator with in-depth knowledge of
the R3ldctl, R3load, Migration Monitor and STR splitter tools. Questions regarding the functionality
of the tools specified, integrating the feature into the system copy procedure, configuring the
Migration Monitor or splitting the STR files can only be dealt with as part of chargeable consulting
services.
ORACLE
Caution:
If you also use the DDLORA*TPL templates (generated by the patched R3ldctl versions, see above) for
importing data to an oracle system, you should not start the brconnect using the SAPinst. SAPinst
calls the brconnect with parameters with which the index statistics are also updated. However, this
is no longer necessary for imports with the new DDLORA*TPL templates. Therefore, before SAP
NetWeaver 2007 you should manually start brconnect with the following parameters after setting up
the target system:
brconnect -u / -c -o summary -f stats -o <DB schema> -t all -p <#threads> -f nocasc
Validity
Software Component From Rel. To Rel. And Subsequent
SAP_BASIS 640 640
700 700
References
This document refers to:
SAP Notes
1461056 Creating secondary index for R3ta table splitting
1319517 Unicode Collection Note
1066404 R3load, ORDER BY, Unicode conversion of cluster tables
1043380 Efficient Table Splitting for Oracle Databases
1040674 R3ta in combination with unsorted unload
1014782 FAQ: SAP MaxDB system copy
806554 FAQ: I/O-intensive database operations
This document is referenced by:
SAP Notes (7)
806554 FAQ: I/O-intensive database operations
1319517 Unicode Collection Note
1040674 R3ta in combination with unsorted unload
1043380 Efficient Table Splitting for Oracle Databases
1461056 Creating secondary index for R3ta table splitting
1066404 R3load, ORDER BY, Unicode conversion of cluster tables
1014782 FAQ: SAP MaxDB system copy
