How to get the records count by the datastage operator?

Johnny0638 · Post by **Johnny0638** » Wed May 27, 2015 9:23 am

Hi guys,

I have a requirement of filter the records by some conditions and check the recods count.

For example , I want to get the records which the conditions are column1 = 0 and column2 <>1, and the records count should more than 100.
I can use the filter operator to get the records with the conditions, but how to check whether the records count is more than 100?

if all the conditions are equal(=), i think i can use the group operator to get the countField, and check it . But if the conditions are not equal(eg:>, <, >=, <=, <>...), how to get the records count by the operator?

Which operator can get the records count from the dataset?

Any help please!
Thanks!

Johnny

chulett · Post by **chulett** » Wed May 27, 2015 9:27 am

Not really clear why you are mentioning "by the osh". Is this something you want to do inside a job (while it is running) or outside of a job after it completes?

Johnny0638 · Post by **Johnny0638** » Wed May 27, 2015 9:36 am

I want to realize this requirement by the osh script with operators , like this: import -> hash -> tsort -> filter -> group -> export

chulett · Post by **chulett** » Wed May 27, 2015 9:51 am

Sorry but I have to ask - why? One would typically do that by building the appropriate parallel DataStage job and letting the product generate that for you. Why do you feel the need to be down at that level building OSH scripts by hand? IMHO - leave that to the experts or people who come from the original product and have that skill set.

Johnny0638 · Post by **Johnny0638** » Wed May 27, 2015 10:05 am

Thanks for your reply.
But I have to write the original osh scripts in our develop environment.

qt_ky · Post by **qt_ky** » Wed May 27, 2015 10:09 am

That sounds odd. So, you're not using DataStage Designer at all?

ray.wurlod · Post by **ray.wurlod** » Wed May 27, 2015 4:16 pm

Johnny0638 wrote:Thanks for your reply.
But I have to write the original osh scripts in our develop environment.

Resist stupid requirements.

This task is far more easily done without osh. After all, it's not really a parallel operation. What you're really seeking to determine are link row counts. There are calls within the DataStage API for retrieving this information, or it can be garnered from the DataStage Operations Database if you are on version 8.7 or later.