Usage of R3ta during Database Export

[Former Member]

Hi everyone,

I have some questions about the usage of R3ta to split some tables during an export/import process.

I'm doing a Unicode migration of a ECC 6.0 system running on Oracle 10g / Win2003 x86_64.

I have done the Table Split Preparation step via sapinst and it finished ok. This step created a subdirectory DATA in my export directory and created numerous .WHR files and a whr.txt file.

After that I started the export process once again via sapinst. I choosed to split the STR files, extracting the 250 largest tables with a package size of 1 GB. I checked the parameter split predefined tables and mentioned the file whr.txt created in the first step as the file containing the tables to be split.

But in the end, the tables in the files whr.txt are still exported sequentially (but are well split into 1GB pieces)

Did I forget something during the process ?

I am using the latest R3ta, R3load, dboraslib.dll, R3szchk and R3ldctl.

[Former Member]

Hi Sebastion

How many parallel processes (Number of Parallel Jobs) did you specify? If it was only one, then all packages will be exported one after another.

Regards

Michael

[Former Member]

Hi Michael,

indeed, I did not tell about the number of R3load processes used. There were 6 of them in parallel.

The last table (and the one that takes most of the time to be exported) should have been handled in parallel but only 1 R3Load was active at the time (6 before)

[Former Member]

Problem solved.

For whatever reason, the files for table splitting were created in a wrong directory. After having copied them to the DATA directory and restared the export process, everything went fine.