BDLS for large tables

[Former Member]

Friends:

During BDLS I skipped some very big tables.

But I still need to convert them and at the same time I cannot afford to spend several days to convert.

What can I do to quickly convert them?

Thanks!

[Former Member]

One trick is use very often is to manually create an index and then run BDLS. This is an example for oracle databases (table COEP):

SQL> conn sapsr3

SQL> create index COEP_ZBDLS_1 on COEP (MANDT,LOGSYSO)

       tablespace PSAPSR3 parallel 16 nologging compress2

       storage (initial 10M next 10M maxextents 99999);

SQL> create index COEP_ZBDLS_2 on COEP (MANDT,LOGSYSP)

      tablespace PSAPSR3 parallel 16 nologging compress 2

      storage (initial 10M next 10M maxextents 99999);

Then you can run BDLS as always and drop the indexes afterwards. You might also be interested in this blog: Execute conversion of logical system names (BDLS) in short time and in parallel - Intermediate

Cheers Michael

[Former Member]

You can use old BDLS program RBDLS2LS for that. You can specify one or multiple tables there.This program creates RBDLS<Client-no.> job depending on your inputs so multiple jobs can be scheduled at one time for different tables.