R3load, unsorted unload and cluster tables

[Former Member]

Dears,

I'm using the DistMon to export a  quite bigger database during an Unicode Conversion.

I need to speed up as much as possible the export process.

I'm using the latest R3load ( version R6.40/V1.4 Compiled Jun  1 2012 21:19:22) ; I read note 954268 and I'm aware of the controindications of using the unsorted export.

I would like to set the unsorted unload for all the packages and tables, except for the cluster tables and the packages containing cluster tables.

Till now I choose the unsorted manner  only for the splitted transparent tables  . All the other packages are in sorted manner.

The question is:

A part the SAPCLUST , which other package contains cluster tables ? Are the cluster tables all into the SAPCLUST package ?

regards

[former_member189725 · Accepted solution]

Yes , all the cluster tables are in the package SAPCLUST, and no other packages have cluster tables.

You can simply use the latest R3load and use unsorted unload. R3load is intelligent enough to figure out cluster tables which are to be exported sorted during a unicode conversion.

Regards

Ratnajit

[Former Member · Accepted solution]

Hi Ratnajit,

I filled the unsorted_export.txt file with all the tables supposed to be exported in that way before to run the prepare step , but at the end of distribution step I found the file ddl_mapping_export.txt wrong.

All the tables were under the item [ SORTED UNLOAD ].

Then I cleaned up all the stuff and repeated the prepare putting the item [ UNSORTED UNLOAD ] into the unsorted_export.txt, but again all my tables went under the item [ SORTED UNLOAD] into the file ddl_mapping_export.txt.

So at the end I fixed the ddl_mapping_export.txt manually before to run the export phase.

But I discovered another thing:

the file split_tab_sequential_imp.txt it's wrong too because contains n times the same split tables.

I mean suppose I have 180 split for the CDCLS table ; this file among others split tables contains the list from 1 to 180 of the CDCLS splits, than again the same list for this table and then again and again..

In you opinion it's a bug of the DistMon (it's the latest version) or I'm making some mistake ?

regards