Application Development Discussions
Join the discussions or start your own on all things application development, including tools and APIs, programming models, and keeping your skills sharp.
cancel
Showing results for 
Search instead for 
Did you mean: 

Index in GLPCA

Former Member
0 Kudos

Hi all,

We need to do extracts from GLPCA by an external connector (Qlickview) on a regular basis (daily or weekly) but we dont have any good indexes to use. Currently we are searching using fields RLDNR, RRCTY, RVERS, RYEAR, RBUKRS and perhaps POPER.

Would it be a good idea to put an index on CPUDT and CPUTM (creation date and time) to be able to extract data based on that info? How will performance be affected in the applications by adding another index?

We have 113 million rows and are adding approx 5 M per month. Will it take forever to create the new index?

Can I perhaps build the index "behind" SAP, ie directly in the database somehow?

Any other suggestion?

best regards,

Glenn

1 REPLY 1

Former Member
0 Kudos

Would it be a good idea to put an index on CPUDT and CPUTM (creation date and time) to be able to extract data based on that info?

If your needs are based on these fields, then certainly it would make your extarction faster.

How will performance be affected in the applications by adding another index?

Any SAP program that writes to this table will be impacted by this index.... this is b/c the new index needs to be updated (as well as the actual table). The impact is not enormous as long as you do not have a large number of indexes currently.

The "rule of thumb" for indexes... if the business truly needs the info processed in atimely fashion, you MUST create the index. If timeliness is not of great importance, then live without it.

We have 113 million rows and are adding approx 5 M per month. Will it take forever to create the new index?

It is recommended to create an index of this size during a "down" time. The creation will take a while... and an application that references the table will be halted until the new index is built. That is why a down time is best.

Can I perhaps build the index "behind" SAP, ie directly in the database somehow?

And yes, your DBAs can build index at the DB level. Your extractor will "find" this new index and use it. The negative is that your developers will not know that index actually exists b/c it will not be visible in SE11 for the table (unless you enter it later).

Hope this helps.