Wrong character set during system copy

[Former Member]

Hello!

I have got a problem with Character set after a heterogenous system copy. My source system is ERP 6 SR3 on AIX with a charset 1401 (Czech) and I want to make a test system on Linux using a heterogenous system copy.

I have done this procedure according the "System Copy for SAP Systems Based on SAP NetWeaver 7.0" guide and everything works fine but during a profile check (RZ10) I have noticed of this error "E:Language Key Czech and Character set 1100 do not go".

Then I have checked import logs and there is written "SAPAPPL1_6.log:(CNV) WARNING: conversion from 1401 to 1100 not possible".

My questions are:

- Where I can set/specify character set for destination system before/during system system copy?

- Why this information was not used by export/import tools?

- Should I do export/import once again or can I  switch from 1100 -> 1401 without loosing data?

Thanks to all,

Gracian

[JPReyes · Accepted solution]

Linux character set should be 4103 for the purpose of the heterogeneous system copy.

Is your source system unicode or non-unicode?

Regards

Juan

[Former Member · Accepted solution]

Hello Juan,

Thanks for your answer and no, our system is not unicode yet.  Could you tell me please, where I can set the character set for the import during system copy? I haven't noticed of this question...

Regards

Gracian

[JPReyes · Accepted solution]

Check RSCPINST in your source and target system and let us know what codepages are configured.

You can also check the NLS config in the same report and post the entries here

Regards, Juan

BTW: Could you tell me please, where I can set the character set for the import during system copy?

You have to set the codepage at the time of the export not the import... in the case of non-unicode the standard is 1100. thats why we are looking into RSCPINST, as you should have additional codepages for your additional languages.

Message was edited by: Juan Reyes

[Former Member · Accepted solution]

Thanks a lot,

The report shows me on the source and the destination system same values:

"Enter languages" EN English

"Enter country" CZ Czech Republic + Latin-1

Current NLS settings:

----------------------------------------

C o n f i g u r a t i o n   t y p e :

Non-Unicode System

Non-Unicode - Single Code Page configuration

C o d e   p a g e s :

1401   Code Page of Application Server

1404   Code Page of SAP GUI

D a t a b a s e   t a b l e   e n t r i e s :

TCP0C: Code pages and locale names

OK     Checked and no inconsistency found.

TCP0F: Frontend code pages

OK     Checked and no inconsistency found.

TCPDB: Database code pages

1401   SAP internal, like ISO 8859-2

TCP0D: Country code

CZ     Czech Republic

TCP0I: Processed language(s) entry

ACTIVE E

[JPReyes · Accepted solution]

During the Export the default value of the codepage is 1100, I have a feeling that you didn't change the default value. You can check that in the summary.html in the sapinst_instdir path.

Regards,

Juan

[Former Member · Accepted solution]

I have finally found the information! Not in export, but import settings Could you tell me please, if the 1401 is the right codepage when it is in the source system, or should I set it to 4103 as you mention above?

Thanks a lot!

Gracian

[JPReyes · Accepted solution]

I always set the codepage at export during a HSC,

In anycase you should use 1401, 4103 is for unicode (little endian)

Regards,  Juan