IBM Information Server

[aladap12]

Hi,

I have one Issue,Everyday I need to process 1 Sequential file.
First I need to check the size of the fille. If the records in the file less than million records then I have to Stop the process, else I have to continue the process.

The process Involves,I need to compare Today's Input File with Yesterday's Processed file and I have to capture
1.Missed,(which are in Yesterday's file but not in today's file)
2.New (which are in Today's file but not in Yesterday's file )
3.Changed records.

Secondly I have to Notice Managers about this three types of Information.
Later I have to merge three files again and I have to use this file again for next day processing.

Could you please suggest me better solution for resolving these Issues
Appriciate your time and help----------VKR



A

[chulett]

First, please do a better job with your choice of subjects, they should represent the nature of the question or problem you have. Something generic like 'Hi' or 'IBM Information Server' makes searching difficult.

Secondly, do you have the MKSToolkit installed on your DataStage server? I'm curious if you have access to 'UNIX' functions or if we're stuck with Windows batch type stuff to help here.

[aladap12]

I am not sure about MKS Tool Kit.
Can you are any one suggest me how would i tell data stage if my input file does have less than 1 milion records.

In the IBM Information server 8x, server edition.. windows platform..

Please let me know

VKR

First, please do a better job with your choice of subjects, they should represent the nature of the question or problem you have. Something generic like 'Hi' or 'IBM Information Server' makes searching difficult.

Secondly, do you have the MKSToolkit installed on your DataStage server? I'm curious if you have access to 'UNIX' functions or if we're stuck with Windows batch type stuff to help here.

[chulett]

OK, for a Windows solutions this should help.

[ray.wurlod]

Do you have Enterprise Edition? There are some useful change detection stages available in parallel jobs which, in conjunction with the Funnel stage, address all your questions.

Change detection in server jobs will require either using an operating system utility such as diff or using a Merge stage (full outer join). A Link Collector stage could merge rows from multiple streams.

In either case detection and email can be done easily with a server routine. Server routines can be invoked from parallel jobs (if they are built as before/after subroutines) or in the controlling sequence.