topic Re: Unicode conversion and System Copy in Application Development and Automation Discussions

Unicode conversion and System Copy

Former Member — Thu, 14 Dec 2006 22:17:00 GMT

Hello

I am going to perform a unicode conversion. The procedure mentioned in SAP service marketplace is to perform system copy, perform unicode pre-conversion steps, perform export with R3load and then import.

My question is if I perform a system copy from prod to sandbox, then perform unicode pre-conversion steps and export PROD with R3load, which will convert it to unicode, then when I do import, doesn't it duplicate the already existing non-unicode sandbox instance which was copied in the first step from PROD.

To summarize my question, I mean should I perform a system copy of non-unicode PROD to sandbox, or should I just created a empty database unicode sandbox instance and then import unicode converted exp files of PROD into this sandbox?

Kindly help me in this matter.

Re: Unicode conversion and System Copy

Former Member — Wed, 20 Dec 2006 04:43:53 GMT

Hi,

You could theoretically do the export/conversion from your production system and build a sandbox. The problem with this option is that you have to be able to take the downtime of the export/conversion on the production system - this is why most people will do a copy to create the sandbox then do the conversion.

Another reason for taking the copy route is that you have to do the unicode preparation before the conversion and must people do not want to do this in production first as it can be a lot of work.

So I would suggest the copy to build the sandbox and then convert unless you are very familiar with the unicode conversion process.

I hope this helps,

Mike.

Re: Unicode conversion and System Copy

Former Member — Sat, 23 Dec 2006 14:26:56 GMT

Thanks michael for the update. I have one more question. The procedure in SAP Service market place is that install a unicode system on S2, export non-unicode data from S1, import to S2. My question is " does that mean just install Unicode software on S2, but do not install the database on S2". And then export non-unicode data from S1 which converts to unicode during export and then do the import to S2? Kindly help me in this issue.

Also I need to know our system is R3 + HR. Are there any additional steps involved while performing unicode conversion for HR system also. Please let me know.

Re: Unicode conversion and System Copy

Former Member — Sat, 23 Dec 2006 15:44:37 GMT

Hi Arun,

The install actually uses the export files from the export/conversion to build the new systems and populate its database. You should be able to find more information the handling of the export files in the unicode conversion guides (it essentially works like a system copy using the SAP tools but a conversion is done).

Apart from the conversion of the ABAPs to unicode, interfaces you might have to enable unicode and checking any print documents you might have I'm not aware of any specifics for HR.

Cheers,

Mike.

Re: Unicode conversion and System Copy

Former Member — Sun, 24 Dec 2006 03:15:03 GMT

Thanks for the update. Can you let me know one more thing. When I do a non-unicode BW 3.5 upgrade to BI 7.0 Unicode. Should I do a unicode conversion for BW 3.5 System first and then upgrade it to BI7.0. Or should I first upgrade BW 3.5 to BI 7.0 and then convert it to Unicode.

Re: Unicode conversion and System Copy

Former Member — Tue, 06 Feb 2007 20:54:17 GMT

Hi Arun,

It doesn't matter either way. If you do it before the upgrade to 2004s, then you probably only have an ABAP stack to think about. Converting after an upgrade just adds the BI JAVA (JAVA stack) handling - nothing too difficult but more steps.

Another tip is that you can convert from Oracle 10 (from 9.2) during a unicode conversion if you have not done so already. This saves you a separate task and more downtime.

Cheers,

Mike.

Re: Unicode conversion and System Copy

Former Member — Fri, 19 Oct 2007 10:03:01 GMT

Hello,

The final message from Mike on this thread just about answers my question but I want to confirm if:

1. It is definately possible to perform a BW 3.x -->BI7 upgrade <u>without</u> a UNICODE conversion.

2.In <u>non UNICODE</u> BI7 systems it is possible to install and run the Java stack functions without performance degradation

3. In BI systems that have been converted to UNICODE it is possible to continue to load data from non-unicode R/3 sources

4.When the UNICODE conversion takes place there is no increase in the overall size of the database.

Quick answers to the above would be very much appreciated!!

Cheers,