Apply the filter conditions at source level to avoid unnecessary data flow in the job design.
Extract the only required columns from source.
Create large no.of datastage jobs with less no.of stages rather then less no.of jobs with more no.of stages.
Use dataset stage to store temporary data.
Avoid using the transformer stage as much as possible because it needs to have additional C++ compiler to execute the ETL Program.
Insted of using Transformer stage use copy stage to drop the unwanted columns or to send it to multiple stages.
use filter stage to apply conditions.
use surrogate key generator stage to generate sequence number.
prefer database sorts than data stage sorts.
use transformer stage to avoid aggregator stage.
Extract the only required columns from source.
Create large no.of datastage jobs with less no.of stages rather then less no.of jobs with more no.of stages.
Use dataset stage to store temporary data.
Avoid using the transformer stage as much as possible because it needs to have additional C++ compiler to execute the ETL Program.
Insted of using Transformer stage use copy stage to drop the unwanted columns or to send it to multiple stages.
use filter stage to apply conditions.
use surrogate key generator stage to generate sequence number.
prefer database sorts than data stage sorts.
use transformer stage to avoid aggregator stage.
No comments:
Post a Comment