Warning message at the end..dropping rows

sush26283 · Post by **sush26283** » Wed Dec 07, 2011 1:07 pm

I have a job and I need some help in recognizing the warning message that am getting...the job runs and completes but at the last step after all the trasformation and joins when the data passes through the last transformer before it goes into the table a lot of rows are getting dropped.

the log shows the following two warnings in a multiple way

Transformer_551,1: Field 'APT_Decimal::ErrorBase: From: divizor is 0 via exception..' from input dataset '0' is NULL. Record dropped.

Transformer_551,0: Field 'APT_Decimal::ErrorBase: From: divizor is 0 via exception..' from input dataset '0' is NULL. Record dropped.

can someone help me with that..

Thanks

ray.wurlod · Post by **ray.wurlod** » Wed Dec 07, 2011 3:05 pm

One message is from node #0, the other identical message is from node #1. You are running on a two node configuration, each performing identical processing on separate subsets of your data.

sush26283 · Post by **sush26283** » Wed Dec 07, 2011 4:01 pm

but so far in the design through out the flow there has been no error or warning...its only at this last step that the warning is coming up..i have also broken down the design in small steps to check for the warning but none..

what do you suggest me to do..

Thanks

ray.wurlod · Post by **ray.wurlod** » Wed Dec 07, 2011 4:44 pm

Detect which input columns to that Transformer stage contain nulls and which operation (stage variable, constraint expression, derivation expression) is having problems handling null because it contains a function call. Handle any such null values before they are passed to the function. Note that DSMacros also call functions.

Kirtikumar · Post by **Kirtikumar** » Fri Dec 09, 2011 1:00 am

Are you doing any division operations in the last transformer? And if so, are he columns used for division have zero or null in them?

kmsekhar · Post by **kmsekhar** » Fri Dec 09, 2011 1:43 am

few more points.. Usually In a transformer its better to handle all the nullable columns before we do any arthimatic calculations (ie., ADD,DIV,SUB,MUL,CONCAT,SUBSTRING etc...)
We have options to handle null using any of the functions

NULLToValue
IsNULL
.......
.......

sush26283 · Post by **sush26283** » Fri Dec 09, 2011 7:38 am

Thanks Guys...I found out the mistake I was doing...apparently the derivation was a faulty one that i was doing.