SAP BW: Adding my own indexes to Cubes

I am still on BW 3.0. I have a cube I know SAP BW provides several database indexes on FACT and Dimension Tables, but I want my DBA to add some of my own indexes
for query performance.
1.     If I delete indexes using manage ->Performance Tab, will it delete my new indexes?
2.     If in my process chain I delete indexes using processing set (?) provided by BW, will it delete my indexes?
3.      Is it possible for me to create these indexes, say from an ABAP program, instead of asking DBA,
4.     Will it possible to integrate this ABAP code in my process chain (when deleting, recreating indexes).
Before I ask my DBA to add indexes, I just want to get an idea on what is involved in managing these indexes.
Prem

There are cases where BW default indexes do no meet performance rquirements, so if one can create additional indexes, it will be great. For xeample, Oracle has bitmap join indexes in Oracle 10g for database warehouses.
BW lets you create secondary indexes on ODS/DSO, but not on Cubes. That's why I am asking this question. One can ask DBA to create these indexes, but then one needs a way to retain those indexes.

Similar Messages

Adding navigational attributes to a cube

We have upgraded to BW 7.3. Previously when we added navigational attributes to a cube we need to activate the update rules from the ods but now it appears that we don't need to do that anymore. Is this the expected behavaior? We are using the 3.5 update rules.
thanks

HI,
If you are using 3.5, after selecting nav attr in the cube, the cube structure gets deactivated. The dependent update rules also gets deactivated. You will be forced to activate all again.
In BW 7.3, SAP would have given privilege to add without activation. It is good for us. So nothing to worry.
Regards,
Suman

Stop SAP BW to create primary index after transport

Hi there,
does anyone know how to stop SAP BW to create an index for transported cubes automatically?
Thanks
Oliver
Edited by: Oliver Schropp on Aug 19, 2011 4:04 PM

Hi,
Primary Index is SAP provided.
The primary index contains the key fields of the table and a pointer to the non-key fields of the table. The primary index is created automatically when the table is created in the database.
There is no way to automated deletion of it.
Thanks,

What is the difference between creating index on cube and infopkg in PC

Hi All
I have process chain in which after executing infopkg(load data infopkg),creating index on cube i.e Object type is Cube ,for which execution time is 1 hour,then after (subsequent step ) again create index at this time object type is "infopkg"
execute infopkg for which time is 2 minnutes,what is the diffrence between these two,if i reome create index from cube i can save 1 hour time,I have to reveiew this
chain for performance,plese post me your thoughts,it's argent,your help will be heighly appreciatable.Thanks in advance.
regards
EA

By default once u use create index process type Object type has Infopackage - change it to Cube tech name.
If its Cube - Indexes will be deleted or created for all the date in the cube.
Message was edited by:
Jr Roberto

Process Chain error while creating index on cube

Hi All,
I have two process chain P1 daily(except sunday) and P2 weekly(sunday), both have same process, i.e. cube C1 is loaded with same infopackage in both infopackage.
Now, on saturday after laoding the content index creating process on cube c1 is failed by process chain P1, but on sunday P2 have laoded C1 correctly and created indexes.
Now, my doubt is wheather i repeat process chain P1 from creating indexes or there is no need as indexes on cube are created by PC 2.
kindly suggest,
Regards,
Rajvinder

Hi Raj,
There is no need of running the process chain P1, the data has been bought by the process chain P2.
If there are any filters/data selection in infopackage level then you need to run the process chain P1.
Be make sure about the process chains and there infopackages/dtp's.
With Regards,
Ravi Kanth

Does SAP BI has it own database? If yes what is the name of the database?

Does SAP BI has it own database? If yes what is the name of the database?
Thanks!
York

If you authorization take a look at the followig tcodes:
1. DB6DBM
2.DBACOCKPIT.
3. DB02.
Ravi Thothadri

Adding our own function to sap standard function code.

HI,
My requirement is to add a functionality to icon(import) on alv application tool bar so that the customer get a popup message after data is transfered to excel.
is that possible.
else
i want to know the procedure to pass header of my alv when i try to download using gui_download.
here the problem is i am having 362 fields in my report.
so passsing header for these many fields is immposible.
when i use gui_download i am not getting header in my excel.
i am not good at ASML_====.
Pls help me.
Any helpful ans will be rewarded.

Declare two internal tables it_header for the header and it_downloads for data.
data:begin of it_header occurs 0,
string(100),
end of it_header.
<b>Perform Insert_header.
Call Function GUI_DOWNLOAD
exporting
filename = c:xyz.xls
filetype = DAT
tables
data_tab = it_download                 " itab with data
fieldnames = it_header                  " itab with header info.
exceptions
if sy-subrc <> 0.
endif</b>.
<b>Form Insert_header.     " You can put some logic ie some loop here for    * all the fields.
it_header-string = text-001.
append it_header.Clear it_header.
it_header-string = text-002.
append it_header.Clear it_header.
it_header-string = text-003.
append it_header.Clear it_header.
it_header-string = text-004.
append it_header.Clear it_header.
Endform</b>.

What is the effect of INDEXING on Cube? and on DSO?

Hi,
I am trying to figure out the effect of indexing on a Cube and DSO.
1. My findings so far indicates that Indexing is a database concept but so is partition but partitioning applies in the case of a Cube. Right?
2.What is the effect of Indexing on a cube? How is it related to the "database concept that I keep reading?
How is it implement in BI?
3.What is the effect of Indexing on a DSO? How is it related to the "database concept that I keep reading?
How is it implement in BI?
Thanks

Hi,
thanks for the detailed information. I will appreciate some help on these follow ups on your posting:
On your answer to 1:
I read that there are 2 partitions, logical and physical. It is the physical partitioning which takes place at the database level, so can I assume that when you said u201CEven DSO can be partitionedu201D you meant the physical partition? And that logical partitioning is not possible in DSOs?
On your answer to 2:
So in the above you suggested that indexes apply not only to cubes but also to DSO with the statement that u201C..secondary indexes can be built on DSO u2026u201D
If so, how come there is not feature to drop and rebuild indexes for data load to DSOs?
I will appreciate a hint on the difference between u201Cindexesu201D and u201Csecondary indexesu201D. I am ok with what u201Cindexesu201D is, is the use of u201Csecondaryu201D that threw me off.//
On your answer to 3:
i.
Based on the link, there does not appear to be a simply way to drop and rebuild indexes in DSOs. It appears an ABAPer needs to be involved. Isnu2019t there a simply way as in cubes, to click on buttons for the index implementation for DSOs?
ii.
It suggests one must choose between B-tree and Bit map indexes, right? Is this the function of the BI consultant or as DBA? If the BI consultant should be able to do this, where is it implemented in BI?
Can you explain what you meant by u201Ccardinality of a column is lowu201D?
iii.
I know how in rsa1, a Cube can be partitioned on 0calmonth and 0fisper. If this is done at the database level, then can it be done also for DSO? If so, where in BI is it done for DSOs?
You also noted that SAP DB does not support table partitioning but Oracle does.
I know in my environment, the BASIS consultants are always talking about Oracle so I am assuming we have Oracle and not SAP DB.
I never heard of SAP DB, is it a substitute for Oracle in SAP BI environments?
Is it a DB that comes directly from SAP; if so, why do companies not keep the total SAP package but prefer to use Oracle? Is it because the SAP DB does not support table partitioning?
Thanks

Is it allowed to Create your own indexes on not-udf fields

Hi all
I have a customer who ask if it is allowed to add you own clustered indexes on the SBO Tables on the Fields that are not UDF.. Example:
GO
CREATE NONCLUSTERED INDEX [<Name of Missing Index, sysname,>]
ON [dbo].[RDR1] ([LineStatus],[U_ET_1_3])
INCLUDE ([DocEntry],[LineNum],[TrgetEntry],[ItemCode],[Dscription],[Quantity],[Price])
GO
I have always assumed that such a thing is not legal, but can someone confirm this and perhaps point me to documentation or SAP Note that states this?

Hi Rasmus,
Thanks for posting in our SAP Business One forums.
As you already stated the use of index was not approved to be implemented in Business One databases.
However, it some specific cases, indexes had improved performance in a very effective way.
Note 1376447 - "Support scope for SAP Business One database index creation" allows to implement it under SAP' support approval.
Here an extract from the note.
While analysing the performance of a customer database if you discover that the addition of an index can improve the performance of the database kindly contact SAP Business One Product Support in order to get approval to add this index to your database.
Please note it is not possible for SAP Business One Product Support to fully test all modules in SAP Business One. As a result we are limited in the testing we can do to verify the overall performance. In such cases even if the index is approved, SAP is not responsible for the impact the index might have on the overall performance of the database.
With kind regards,
Juan Manuel Marrero
SAP Business One Forums Team

Problems adding new caracteristics ti 0IC_C03 Cube

Hi people:
We made a copy of standar cube 0IC_C03. We add some new time characteristics (0halfyear, etc..). De activation of the cubes is OK, but when we try to map the corresponding fields in the tranformation rules, the new fields doesn't exist. Like the group of transformation are standar, we try to add them manually, but the infoobject isn't seen.
Does this cube has a limitation for adding new infoobjects? Am I misssing some step I don't know?
Thank you in advanced.

Hi,
There is a relation between time characterstics ,
We have to take care of this relation in the transformation mapping where we are mapping source fields to target fields.
Please refer below link which will show you the relation between time characterstics.
http://help.sap.com/saphelp_nw70/helpdata/en/44/be8b631c743959e10000000a1553f6/content.htm
Hope this will help you...
Thanks,
Jitendra

Adding new field to Remote Cube

Hello,
We have an existing remote cube that we are trying to add a new field to in 7.0. I've already added the field to the datasource, unhidden it, and tested it in RSA3. Now i'm trying to bring it into BW and add it to the remote cube. I right clicked on the source system under the data target and chose Replicate Metadata however the new field does not show up anywhere. What am i doing wrong? Does a remote cube work differently? How do i get my new field over from R3?
Regards,
TMS

That's the problem. The R3 field does not show up in the datasource in BW so it can be moved over to the transfer structure. As you know when you replicate metadata, the new datasource fields appear and can be mapped into the transfer structure. That is not the case for this remote cube. The new field is not there.

Adding new fields to existing cube

Hi all,
We have a cube with data, now we added 2 new fields to the datasource. And want to add two more fields in the Cube to correspond to the new fields. What would be the impact on the data that is already there in the cube? Can I go ahead and add the fields in the cube and start extracting data? What I want to know is what would be the implications in such a scenario?
Any insight or help would be appreciated and fully rewarded.
Thanks,
Mav

Hi,
At least from the experience I had there were no performance issues. The only issue would be in terms of the time taken to transport. The larger the amount of data ,longer the transport time.
You can continue extraction of data for those 2 fields from that point on. No need to empty the data from the cube and reload again. Only thing is that the data in the past for these 2 fields may not be mapped to the existing data in the cube. If you want to do that dump the cube and reload. Otherwise the data will be mapped from the time you start extracting the new fields with the data in the cube.
Cheers,
Kedar

Problem with delete index of cube in process chain

HI ,i have a problem with my process chain,here on process name as delet index is ther which has been turned to red.when iam going inside of that in Display message iam getting the msg as"Indexes for InfoCube SMS_C06 have been deleted successfully".
since the process is in red next process is not been triggering,but i have to continue the process chain.
can anybody put some light into it.its very urgent........

Hi,
I ran into the same situation with one of my process chain that is part of a metachain. The process chain has the Delete Index that turned red. But the messages show that the indexes on all the cubes were deleted sucessfully.
As suggested, I did a repeat on the delete index variant, but no luck as it turned red again.
Anyone aware familiar with this issue?
Thanks,
Briston

Sap not following its own rules

i made a vendor recon account and assigned a sort key 010 to it. then i created a vendor master record and in that also i gave sort key as 010. now i made a po then grn then miro.
as a result of simulation in miro i got an entry vendor recon account cr with no po no. in purchasing doc column and gr/ir debit with po no. being displayed.
when i did fbl1n for this vendor account i didnot get anything in assignment field while it was expected that po no should be reflected in assignment field as sort key assigned to this vendor master record was 010
why is this happening SAP is not following its own rule that what ever is given in sort key what we see in assignment field
please explain with some solid proof

The system uses a standard sort sequence for displaying line items. Among other things, it sorts the items according to the content of the Allocation field. This field can be filled either manually or automatically (by the system) when a document line item is entered.
For this purpose, the system requires rules that determine which information is to be taken from the document header or from the document line item and placed in the field. The rules can be stored in the master record of an account which enables you to determine the standard sort sequence on an account-specific basis.
You have to define the rules for the layout of the assignment number field in the Implementation Guide (IMG) under Determine Standard Sorting for Line Items.
Hope you understand the concept.
Reward points if useful

Crystal Reports: access to SAP BW tables in own namespace

Hi BO experts
we are running XI3.1 SP2 including reporting tools, integration kit and so on. The installation and the integration to SAP BW is working properly for most szenarios.
Starting from Crystal Reports we would like to access a SAP BW table which was generated in an own namespace. By starting the database assistant we are connecting with the driver "SAP-tables,-clusters,or -functions" and see some tables in "DD (data lexicon)". But we can not find the self generated table in there. It is only possible to access the tables in namespace "/1CPMB/", "/B28/" and "/BI0/".
How can we access tables from another namespace? Are there any restrictions for accessing SAP BW tables, e.g. namespace, configuration, settings etc.to stick to?
Thank you very much for your suport in advance.
Kind regards.
Markus

Hi Markus,
just go to the Database Expert, select the connection to your BW server on the left window, press the right mouse button and choose Options from the context menu. In the field Table name LIKE: you can restrict the names of the displayed tables:
eg. assuming your namespace is /ZMYNSP then just give there /ZMYNSP% ( % is a wildcard)
Close then the Options window and choose Refresh using again the context menu (or just press F5)
CR can only display a limited number of tables when you expand your BW System and by using the Options panel as described above you can let the system display exaclty the tables you need.
Regards,
Stratos

SAP BW: Adding my own indexes to Cubes

Similar Messages

Maybe you are looking for