Unique ID

WOW , Thanks everyone for the prompt response(s)

I find Rob's suggestion most suitable.

Now, there are several (16) different transaction files that are pulled
together into the DW
and I use a timestamp in each record to extract only the records that were
changed/added since the previous extraction.

Since there are several files involved I assume that option 1 (GENERATED)
and option 2 (sequence) are not viable.
Option 3 (UUID) seems IMHO to be the right solution.
Now,
a. how unique is the UUID? - there are close to a 100 million records all
in all
b. how do I generate a UUID in the RPGLE programs that add records to the
various
files?

(again) Thank you all

Gad

message: 3
date: Tue, 1 May 2012 12:44:28 -0400
from: rob@xxxxxxxxx
subject: Re: Unique ID

I suggest one of four options:
1 - Using an identity column. See GENERATED at

http://publib.boulder.ibm.com/infocenter/iseries/v7r1m0/topic/db2/rbafzhctabl.htm
2 - Using a 'sequence'.

http://publib.boulder.ibm.com/infocenter/iseries/v7r1m0/topic/db2/rbafzcsequence.htm
3 - Using a UUID.

http://publib.boulder.ibm.com/infocenter/iseries/v7r1m0/topic/rzatk/GENUUID.htm


Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com