View previous topic :: View next topic
|
Author |
Message |
ranjitbhingare
New User
Joined: 30 Nov 2005 Posts: 94 Location: PUNE
|
|
|
|
I have 2 clusters
1. PROD.CLUSTER : PRODUCTION
2. PROD.BACKUP : BACKUP TILL PREVIOUS DAY
I want to compare these 2 cluster key wise and extract records from PROD.CLUSTER which have changes.
Will this process run faster with program or can anyone suggest any utility by which we can achieve same faster ? |
|
Back to top |
|
|
William Thompson
Global Moderator
Joined: 18 Nov 2006 Posts: 3156 Location: Tucson AZ
|
|
|
|
It depends, how many bytes of data in the record do you want to compare? The whole record of just some specific fields? |
|
Back to top |
|
|
ranjitbhingare
New User
Joined: 30 Nov 2005 Posts: 94 Location: PUNE
|
|
|
|
I need to compare whole record.
I need to implement this on 14 production files.
e.g.
File :
KEY Record
100100112345ME Name Gender
100100112345MD DEP-NAME DEP-GENDER INDICATOR
.
.
Million records....
Now the production file will have same records with some changes.
I want to compare two clusters and extract changed records into a flat file.
Please suggest me an approach, whether I should code a cobol program or is there any utility which can help me to do this ?
Regards,
Ranjit |
|
Back to top |
|
|
murmohk1
Senior Member
Joined: 29 Jun 2006 Posts: 1436 Location: Bangalore,India
|
|
|
|
Compare using FILE AID/FILE MASTER in Batch Mode. |
|
Back to top |
|
|
ranjitbhingare
New User
Joined: 30 Nov 2005 Posts: 94 Location: PUNE
|
|
|
|
I have never used a file aid in a batch mode. Could you please let me know which document should I refer for it ?
Regards,
Ranjit |
|
Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007 Posts: 8797 Location: Welsh Wales
|
|
|
|
What about some programming logic to produce an audit trail of changes made to the cluster, as they are made.
Can I ask the requirement to see which records have changed. |
|
Back to top |
|
|
ranjitbhingare
New User
Joined: 30 Nov 2005 Posts: 94 Location: PUNE
|
|
|
|
I am creating an HLd for a project.
Its not possible to include audit trails as these clusters are updated by multiple online and batch programs daily.
We discussed about writing programs for each cluster to get the changed records.
I just want to know for millions of records which will be faster, cobol code for each cluster or using some utility to get the changed records.
We have a time limit for about an hour and half to do the entire process and this will be just an initial part of the process.
Please suggest accordingly...
thanks,
Ranjit |
|
Back to top |
|
|
ranjitbhingare
New User
Joined: 30 Nov 2005 Posts: 94 Location: PUNE
|
|
|
|
I tried fileaid compare but it does not give output in the same record format. It gives output like a report. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello,
If you have a custom requirement, i'd suggest custom code.
If you read each file in a COBOL program and compare, the performance will probably be similar to the file compare utilities and you will have more control of the process (as when new requirements arise). |
|
Back to top |
|
|
ranjitbhingare
New User
Joined: 30 Nov 2005 Posts: 94 Location: PUNE
|
|
|
|
Thanks a lot. Will go ahead with the cobol programs. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
You're welcome |
|
Back to top |
|
|
|