CDPOS table replication with parallel initial load and filtered data.

[Former Member]

Dear Experts,

I would like to setup the CDPOS cluster table replication. The table has 2.000.000.000 records and I would like to replicate it using parallel load and filter on the data but without parallel replication. I got a bit confused with the setup possibilities.

Would be this a working solution?

-setup the IUUC_PRECALC_OBJ for CDCLS (physical table for CDPOS?) using

MTID:094

CONVOBJECT:CDCLS

NUMREC:400000000 - I want to have 5 calculation jobs running

- In the performance options I would setup the CDCLS with 5 parallel jobs and reading type 4

- in the DMC_ACSPL_SELECT table I would do the setup using only the ACS_PLAN 00000 because I do not want to use parallel replication in place.

The DMC_INDXCL is created in the source.

Do I need to prepare anything else? Could this setup work?

Thanks

Tamas

[Former Member · Accepted solution]

Hi expert,

I face a problem during access plan calculation. I have setup the DMC_ACSPL_SELECT entries like following

094   CDPOS  1 00000 OBJECTCLAS  0001    =    'VERKBELEG'    AND

094   CDPOS  1 00000 OBJECTID        0002    =    '0004838320'      AND

094   CDPOS  1 00000 TABNAME         0003    <>  'KONVC'

But I got a dump on the source: A dynamically specified column name is unknown.

Objectclas / Objectid / Tabname are part of CDPOS cluster.

Do you have any idea what could be the problem?`

When I run the FM /1CADMC/ACS_790000000003208 on the source using

OBJECTCLAS = 'VERKBELEG' AND OBJECTID = '0004838320' AND TABNAME <> 'KONVC' as im_it_sel_criteria I did not got error or dump.

Should I use the CDCLS instead of the cluster table CDPOS?

BR
Tamas

[Former Member · Accepted solution]

Issue solved!