21/OCT/08>------------------------- Error during export/import with jload Symptom:
Jload aborts with error: SEVERE: java.lang.OutOfMemoryError Solution:
5. Stop SAPinst.
6. In the control.xml file in the installation directory, search for
\
7. Replace \ var jdkArgs = installer. onOS400() ? [\\
switch( new SystemMgt().getSystemInfo(). getOSFAName() ) {
8. Save your entries and restart SAPinst.
--------------------------
PI: Follow-Up Activities Exchange Profile Note that it is essential that you change the connection for the following steps of this section. Without adjusting the connection parameters, all changes in this section affect the exchange profile of the source system.
--------------------------
PI: Follow-Up Activities Integration Server
After you maintained RFC server groups, perform the following steps:
9. Maintain RFC destination AI_MONITORING_SERVER in the ABAP environment.
This destination only needs to be maintained if performance data
collected by Process Monitoring Infrastructure is displayed in the
Runtime Workbench.
10. Refresh host name buffer.
After you have maintained the required RFC destinations, the host name
buffer needs to be reset. Call transaction SM51 and choose Goto -->
Host Name Buffer --> Reset --> Entire System.
-----------------------CHANGED
----------------------UPDATED IN THE CURRENT GUIDE!---------------------
--------------------------
In chapter 6.1 \Procedure Using the Migration Monitor\the use of R3load's socket option is restricted to non-Unicode systems. This is not correct, you can use the socket option, if your R3load has at least patch level 73 (compile date: 12.01.2008)
---------------------------
All DB specific procedures: Stop SAP system during export of Java stack
Symptom:
If you perform a system copy of a dual stack system with database specific methods, during the export of the Java stack, SAPinst does not prompt you to stop the system.
Solution:
Stop the system (except for the database) before staring the export.
---------------------------
Error in R3load logfiles:
DbSl Trace: Error 18 in stmt_execute() from oci_execute_stmt(), orpc=7 DbSl Trace: ORA-18 occurred when executing SQL stmt (parse error offset=0)
(IMP) ERROR: DbSlExeModify/DbSlLobPutPiece failed rc = 99, table \ (SQL error 18)
Error message returned by DbSl:
ORA-00018: maximum number of sessions exceeded (DB) INFO: disconnected from DB
Solution:
Increase the parameters in
unix: $ORACLE_HOME/dbs/init
and restart the database and then continue the database load.
----------------------------
Table splitting: perform export/import with MigMon
The export and import has to be performed with the Migration Monitor when the table splitting feature is used.
----------------------------
If you plan to use Advanced Configuration options during the SAP System Database Import, make sure you have installed the most current version of SAPinst to avoid performance problems during the dialog phase. You can find the latest version of SAPinst at http://service.sap.com.
----------------------------
--------------------------
Symptom:
You have a Unicode system and table UMGPOCNV contains one or several entries. Solution:
See SAP Note 1119611.
-------------------------
ABAP: UNICODE conversion error Symptom:
During the import into a UNICODE system the following error occurs (for example in the SAPCLUST.log):
myCluster (63.2.Imp): 2085: (Warning:) inconsistent field names(source):physical field K1N05 appears as logic K1N5. myCluster (63.2.Imp): 2086: (Warning:) further investigation recommended myCluster (63.2.Imp): 1924: error when checking key field consistency for logic table TACOPC .
myCluster (63.2.Imp): 1927: logic table is canonical. myCluster (63.2.Imp): 1930: received return code 2 from c3_uc_check_key_field_descr_consistency.
myCluster (63.2.Imp): 1224: unable to retrieve nametab info for logic table TACOPC . myCluster (63.2.Imp): 8032: unable to acquire nametab info for logic table TACOPC . myCluster (63.2.Imp): 2807: failed to convert cluster data of cluster item. myCluster: CLU4 **00001**
): 319: error during conversion of cluster item.
myCluster (63.2.Imp): 322: affected physical table is CLU4. (CNV) ERROR: code page conversion failed rc = 2
---------------------------------------------------------- |
| RSCP - Error
| Error from: Codepage handling (RSCP) | code: 128 RSCPENOOBJ No such object | Dummy module without real rscpmc[34]
| module: rscpmm no: 2 line: 75 T100: TS008 | TSL01: CPV p3: Dummy-IPC p4: rscpmc4_init
`----------------------------------------------------------
Cause:
This problem is caused by incorrect data which should have been removed
from the source system before the export.
Solution:
There are two possible workarounds:
1. Modify DDL
search for the keyword \\
2. Modify the TSK file (most probably SAPCLUST.TSK) BEFORE R3load import is(re-)started.
search for the lines starting with \VER_CLUSTR I\remove the lines.
-------------------------
ABAP/Package splitting: extracting SPLIT.SAR terminates with error msg. If package splitting is configured, the archive SPLIT.SAR has to be extracted. The extraction terminates with an error, although the archive has been extracted completely:
SAPCAR: error reading from
Ignore the error message and repeat the SAPinst step which has stopped with the error. As the archive has already been extracted, SAPinst will not try again to extract it and therefore will continue with the next step.
-------------------------
Java: Server unable to start when migrating from platforms with IBM JDKs
When migrating from platforms with IBM JDKs (e.g AIX) to a platform with e.g. Sun JDK, the application server Java may fail to start after the system copy with the following error in dev_bootstrap (which can be found in $(DIR_INSTANCE)/work):
Unrecognized option: -Xj9
[Thr 2960] *** ERROR => JHVM_LoadJavaVM: Cannot create Java VM (rc=-1) [
[Thr 2960] *** ERROR => Cannot load Java VM (server) (rc=-1) [jlnchxxi.c