Informatica 6.1 Performance issue.

[rhitesh02]

Hi,

Recently we have upgraded to 6.1 .. My mappings are running very slow compared to 5,1 version. In mappings , I have noticed that there are more than 2 aggregators are used.

Can any one help on improving performance of my mapping.

I have tried all sorts of thing at mapping level.. but I feel there will be some parameter which needs to be configured.

Please help.

Thanks,

Hitesh

 

[cprashanth]

Hi Hitesh,

Try using the sorted input option in source qualifier.
If sorted option is used then the Informatica server assumes that all data entering an Aggregator transformation are already sorted into groups that correspond to the Aggregator’s “Group By” ports. As a result, the server does not have to read the entire data source before performing calculations. As soon as the Aggregator detects a new unique group, it performs all of the calculations required for the current group, and then passes the current group’s record on to the next transformation.
Selecting Sorted Input often provides dramatic increases in Aggregator performance.

Hey r u the same Hitesh who is working for Patni?

-Prashanth

 

[raolp]

Try recalculating the index and data cache sizes of the aggregators based on the data-set size you are working with. This should help. Let me know if it works!
r.