Lookup

Subhasree · Post by **Subhasree** » Mon Jun 02, 2008 4:41 am

Hi All,

What is the difference between Normal and Sparse Lookup?
Im having a Source table containing 9 records and im doing a sparse lookup on target table containing 9 records.The lookup condition used is continue.The number of records from the output link is 387.How is it happening???
What is the concept of sparse lookup?In help i found that its doing a direct query on the database rather than from the buffer.But how that will be the cause for my job which is producing 387 records?

saikir · Post by **saikir** » Mon Jun 02, 2008 5:06 am

Hi,

In Sparse lookup for every incoming input row, a SQL is fired on the Reference table. This is effective if the number of incoming rows are very less and reference table has a large number of rows (1:100). Refer the DataStage Parallel Job Advanced Developer Guide for more info.

Sai

ray.wurlod · Post by **ray.wurlod** » Mon Jun 02, 2008 6:10 am

There is not enough information to answer the question. How many nodes are in your configuration file? What partitioning algorithms are used on each link in your job design? Do all nine source records have a matching key in the reference data?