TI taking much time to load 7 Rows

[anoops81]

Hi All ,

I have a TI process which loads just 7 rows from the data base . The Data is pulled from DB without any delay . But when I try to load it , It is taking 10 minutes to load the data.

Below is the code from the Data Tab

Code: Select all

V_asset = Cellgets('Stage_Debt_Forecasting',collateral,INDEX_NAME,'Accounts',Vbsla,QTR,vYear,SCENARIO,v_product_code,INVESTMENT_NUMBER,vDTS,'corp_co
de',VCurrency,vCurrency_type,'Asset Manager');

cellputs(V_asset, 'Debt_Forecasting',collateral,INDEX_NAME,'Accounts',Vbsla,QTR,vYear,SCENARIO,v_product_code,INVESTMENT_NUMBER,vDTS,'corp_co
de',VCurrency,vCurrency_type,'Asset Manager')

It is just taking data from 'Stage_Debt_Forecasting' cube and loading to 'Debt_Forecasting'.

We are on 9.5.2


Regards
Anoop

[tomok]

How do you know it's only 7 rows. What is your data source for the TI process? Is it a cube view? How do you know the view is only going to return 7 records? Do you have zero suppression set on the view?

[anoops81]

Hi Tomok ,

We have a SQL Query which loads the the two cubes with just 7 rows of data. I have used that SQL as data source . It is not cube view.


Regards
Anoop

[Gregor Koch]

Hi
As you are saying you are aware that the code you show is not even writing the data that you get from the SQL source.
You are reading and from 'Stage_Debt_Forecasting' cube and writing to 'Debt_Forecasting' cube. Why do you need the SQL source?

When trying to get advice it is always best to show the whole code from all tabs, as some people actually take the time to read all of it and then are able to give you a good answer without having to go back and forth.

At this point my guess is that either your read ('Stage_Debt_Forecasting' cube not performing well, being overfed) takes more time than it should or your write.
Why the CellGetS would take long is beyond me as no consolidation should be required to do this.

Hash all code in the data tab and to a simple Asciioutput to eliminate any issues with the source (there is a very recent thread on this as well).
If that runs well, decide whether you want to load from the SQL or from the cube and proceed accordingly.
If the latter is the case 'un hash' the line that reads from the cube and Asciioutput that to see which of the to cube actions actually slows you down.
Is there a feeder in the 'Debt_Forecasting' cube that is base on the measure 'Asset manager'?

[anoops81]

Hi Gregor,

Sorry for the delayed response.

Actually client wants to see the string data at consolidated level of two dimensions Account and Corp_code, so we created a stage cube to which string data will be loaded It has only one element for Account dim and corp Dimensions ("Accounts","corp_code"). These element are actually the top consolidation for account dim and corp_code dim in the main cube.

Code: Select all

V_asset = Cellgets('Stage_Debt_Forecasting',collateral,INDEX_NAME,'Accounts',Vbsla,QTR,vYear,SCENARIO,v_product_code,INVESTMENT_NUMBER,vDTS,'corp_co
de',VCurrency,vCurrency_type,'Asset Manager');

cellputs(V_asset, 'Debt_Forecasting',collateral,INDEX_NAME,'Accounts',Vbsla,QTR,vYear,SCENARIO,v_product_code,INVESTMENT_NUMBER,vDTS,'corp_co
de',VCurrency,vCurrency_type,'Asset Manager')

When we try to pull the string data from stage cube and put it in consolidated view for Accounts and Corp code via TI, it is taking time.
The data is getting loaded and the process is completing successfully.It is taking approx 2 min to load a single string measure. We did some testing and found that 'cellputs' is taking much time to finish.

I have removed the SQL source and kept the Stage_Cube view as data source for this String load. No rules are there for string measure which we are trying to load .

I think the issue is with loading into consolidated level .Anyone have tried loading data into consolidation before ? Does it take much time or it is another BUG in TM1 9.5.2 ?? !!!!


Regards,
Anoop


I believe cellputs

[moby91]

What about the "Use Unicode" setting in your TI process ?
Is it enabled or disabled ?


See the thread

http://www.tm1forum.com/viewtopic.php?f=3&t=6481#p27048
Re: TI - SQL Query with JOIN or View, Slow to Execute

[anoops81]

Hi Moby,

I am loading from a cube view. I think unicode check is available only for ODBC Data source .


Regards,
Anoop

[David Usherwood]

Two areas to attack:
1 Is your source cube Stage_Debt_Forecasting properly fed?
2 I'm not a fan, but would FeedStrings be worth using?

[anoops81]

Hi David,

We were able to load the same data at leaf level within seconds. So I believe that the data is fine in stage_cube.
We are having issues with feedstrings.String values are not apprearing after enabling suppress zero,strings just vanishes. We have raise this issue as a bug with IBM ,they created an APAR for it , it has been months and still they are working on it [:(] .


Regards
Anoop