on 01-27-2010 10:26 AM
Dears,
I'm doing a Sys. Copy using R3load method. As there are very huge tables I'm using the table splitting feature.
I'm thinking to use the script of note 1043380 to accelerate the creation of the WHR files before to perform the export of the DB.
Some of you already used it ?
I need some clarification about this script.
The note says do not use the special keyword ROWID for the partitioned tables; it says the script has no meaning for them so suggest to use an arbitrary name. In which way can I check if the table I'm going to split is partitioned or not ?
regards
Are you using sapinst to do ROWID table splitting?
One option to check table type is through SE11 - enter table name - Display button - technical settings button - next to table name you can find what kind of table is that
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Basically only BI systems use partitioned tables so far. During the export procedure you are asked to run the report SMIGR_CREATE_DDL. If it does not generate any <TABART>.SQL files, then you don't have partitioned tables.
And by the way, you don't use the table splitter package to speed up the table splitting in the first place. You need to employ ROWID splitting, which can run a LOT faster (exporting data), than the access via an index.
Cheers Michael
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Tnks for the feedbacks.
I need to clarify if I can use the script of note 1043380 using the ROWID clause for all the transparent tables I want to split (cluster tables included ) or It's better to use in any case as "Colunm Name" the Column suggested into the R3TA_hints.txt file.
Are there controindications using the ROWID clause for all the transparent tables ?
If I use the ROWID I do not have anymore to create the secondary indexes and this save a lot of time.
Regards
You cannot use ROWID on clustered tables:
SQL> exec table_splitter.ranges('CDCLS','ROWID','4');
BEGIN table_splitter.ranges('CDCLS','ROWID','4'); END;
*
ERROR at line 1:
ORA-20100: It is not possible, to split cluster tables by ROWID!
Cheers Michael
Edit: but i recommend using ROWID split for all other tables.
Hi All,
In my coming project, i has to use table splitter to export the database as the size is around 7 TB. I gone through some of the documents and forums but did not find the step by step activities to be follow to use this (i.e. table splitter) technique.
Request you to let me know the source where i can geet the step by step activities and if you have one please let me know.
Regards
User | Count |
---|---|
85 | |
10 | |
10 | |
9 | |
7 | |
6 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.