Portal | IBM Manuals | Downloads | Products | Refer | Info | Programs | JCLs | Forum Rules*| Site Map | Mainframe CD 
IBMMAINFRAMES.com - IBM Mainframe Support Forums Index
 
Register
 
IBMMAINFRAMES.com - IBM Mainframe Support Forums Index FAQ Search Memberlist Usergroups Profile Log in to check your private messages Log in
 
How to identify wrong data in millions of records?
Goto page Previous  1, 2
 
Post new topic   Reply to topic    IBMMAINFRAMES.com Support Forums -> Interview Questions
Author Message
Douglas Wilder

Active User


Joined: 28 Nov 2006
Posts: 206
Location: Deerfield IL

PostPosted: Tue Feb 05, 2008 10:19 pm    Post subject:
Reply with quote

Bad data in fields is no problem, if the practice that I was taught almost 30 years ago is followed. The first program in a system that uses data from outside the system must make sure no bad data get into the system. Check the program specs for this first program and it should tell you for each field what values are ok and what to do with data that is not good. To find the bad data look at the error reports and/or error files coming out of this program.
Back to top
View user's profile Send private message
References
dick scherrer

Global Moderator


Joined: 23 Nov 2006
Posts: 9180
Location: 221 B Baker St

PostPosted: Wed Feb 06, 2008 12:35 am    Post subject:
Reply with quote

Hello,

Quote:
How to identify wrong data . . .

This is an interview question from Perot Systems?

what logic we have to code if we use cobol pgm since file contains so many columns.
One answer is that until all of the validation rules for every field have been defined, it cannot be done.

Quote:
Again - this is the most ridiculous thread I've seen here.
Not all of us can can say this icon_smile.gif

Quote:
I gave a gag gift of a . . . . that simply said "DO IT".
Once bought 25 blank "keys", labeled them with "DWIM", and stuck them with 2-sided tape on some prominent place on selected user's keyboards. (DWIM - Do What I Meant).

d
Back to top
View user's profile Send private message
skkp2006

Active User


Joined: 14 Jul 2006
Posts: 67
Location: Chennai,India

PostPosted: Wed Feb 06, 2008 12:06 pm    Post subject: Reply to: How to identify wrong data in millions of records?
Reply with quote

Have counters defined in your program and everytime you process a record successfully increment that and display them in case of a abend will be a good start......

And as enrico-sorichetti said

Quote:
wrong data is any data which does not abide/respect the application specifications


sk
Back to top
View user's profile Send private message
Rahul_kumar
Warnings : 1

New User


Joined: 24 Jun 2006
Posts: 18

PostPosted: Sun Mar 23, 2008 9:53 pm    Post subject: Reply to: How to identify wrong data in millions of records?
Reply with quote

I thk we can use display statements in the program to find the statement for which the program abends.
or there is another method to identify the SOC7 error using OFFSET address but i don't know the same.Can anyone pls help.
Back to top
View user's profile Send private message
Douglas Wilder

Active User


Joined: 28 Nov 2006
Posts: 206
Location: Deerfield IL

PostPosted: Mon Mar 24, 2008 10:52 pm    Post subject:
Reply with quote

Years ago I supported a system, that when it abended, it would "snap dump", then the program would backout the changes for the current bad record, and write the bad record to an error file, then go on with the next record. I am not sure how they did that. A good edit program would have make that almost unneeded. But it is an interesting idea.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic    IBMMAINFRAMES.com Support Forums -> Interview Questions All times are GMT + 6 HoursGoto page Previous  1, 2
Page 2 of 2