DSXchange

Ravi I would not use a Pivot Stage. I would do 12 lookups to the same table. If you put your tables into to hash files and do the lookups from there. Hash file lookups are the fastest from my experience. Pivot Stages have been very slow in my experience. Thanks Kim. Kim Duke DwNav - Navigate ETL Bus...

I have tried using the example you gave but it still did not work. I would like to say though that i have not used the routing yet in a real job but i have been running the "TEST" in the routine. I just wanted to make sure it works fine first on the routine test, then use it in a real job. Tunde. --...

You could put your source data through a Pivot Stage to translate each of you input rows (with ID + 20 columns) to 20 rows (with ID + 1 column) and then do a single lookup using the output of the pivot. We do this to translate input account balance records which look like: Account_code, year, jan_am...

Hi Ray, you are right. For inserting (array size 1, transaction size 0) into an empty table OCI is twice faster as ODBC. Udo Ray Wurlod 16.07.03 09:19 Bitte antworten an datastage-users An: datastage-users@oliver.com Kopie: Thema: Re: Performance comparation with OCI vs. ODBC ODBC is usually pretty ...

Udo, I cant explain in technically, but Ive seen a study some years ago, which compared generaly ODBC vs. native. Theyve found the factor ranging between 0.7 and 1.3, so about the same youve found out. There is a prejustice which you can often hear, but which is not true ODBC = Microsoft standard Mi...

ODBC is usually pretty good on SELECT. Its when youre trying to to INSERT or UPDATE that its not quite so performant. -- __________________________________________________________ Sign-up for your own FREE Personalized E-mail at Mail.com m CareerBuilder.com has over 400,000 jobs. Be smarter about yo...

Hi all, I compare the performance between OCI-Plugin and ODBC-PlugIn (Oracle Wire Protocol). I expect that native access via OCI is faster as ODBC access. But the realness shows me an other result. With OCI I read 4500 rows/sec and with ODBC 6000 rows/sec from a table. Both Jobs read the same column...

How to create hashed file with multivalued column(s) (DataStage 4.x): In the Manager client, create a new table definition for the hashed file and make sure that the "metadata supports multi valued fields" checkbox is checked. Fill in the appropriate column defintions. Scroll the Columns grid to the...

Jessica, if you have a multivalued attribute that has values you wish you parse, you can use a regular hash file stage as long as you do one-for-one lookups. You can reference the attribute as the entire multivalued attribute, you just will have to extract from the multivalued variable what youre in...

Hi Ray, thanks for your reply We try to build a reeferenec lookup to return mutilple rows and wa want to use the hashed file with a multi-valued field but we are not sure how to create hashed file with a multi-valued field and also how do we turn "normalized on" capability to extract multiple rows f...

When a hashed file is created with more than one key column, the internal mechanism continues to use a single physical key column (called @ID), this is the value used by the hashing algorithm. Provided a different separator character is not specified (which is only possible via CREATE TABLE in a UV ...

Not sure what youre asking here. Perhaps you could describe more exactly what it is you are trying to accomplish. If, for example, you need a reference lookup to return multiple rows (that is, your search is not by primary key but by secondary key), you define the secondary key column(s) in your Tra...

Hi, What is the easiest way of building an input file with 1 record and retrieve many records from HASHFILE. example: want to extract many transactions records within given period of an input intermediary-number Thanks Jessica Nguyen IMPORTANT NOTICE This communication including any file attachments...

Tunde You called it HashFileLookup("Myfile", "5:@TM:30", 1) or HashFileLookup("Myfile", "5":@TM:"30", 1). The first way is wrong. I have not tried this. I know it will work. Thanks Kim. Kim Duke DwNav - Navigate ETL Business Rules www.Duke-Consulting.com -----Original Message----- From: Babatunde.Eb...

i tried what you suggested but it did not work. I created the data with just one key and it worked. but when i created it two key and my input was 5:@TM:30, It did not work. when you tried it did it work? Tunde. -----Original Message----- From: Kim Duke [mailto:Kim_G_Duke@Yahoo.com] Sent: Tuesday, J...