Re: Analysis of transfers

Using the date could have misses if they occur within one second of when the
date changes, if you have real people working around that time.

Are any of your T transactions automated, like shop activity can have
automated transactions?

If multiple people are doing concurrent inventory transactions, then the
sequence number might not be reliable. If the same user is manually doing T
transactions, at the same time as someone is posting JIT620 mass updates
from labor transactions, between the from and to manual input, there might
be 100+ transactions posted by the JOBQ process, some of them impacting the
same item that the T transaction is being done with.

I don't suppose you have two or more sign ons with the same user-id
(different people, same sign on), concurrently doing same kind of
transaction?

Al Mac (WOW) = Alister William Macintyre
-----Original Message-----
From: bpcs-l-bounces@xxxxxxxxxxxx [mailto:bpcs-l-bounces@xxxxxxxxxxxx] On
Behalf Of Kevin G. Harper
Sent: Wednesday, February 27, 2013 1:49 PM
To: BPCS ERP System
Subject: Re: [BPCS-L] Analysis of transfers

Dick,

Is there any reason the sequence# (TSEQ) wouldn't be consecutive between
the "T-from" and the "T-to"? I believe on 8.3.4 the sequencing works fine.

If it is, then you should get accurate results by joining on TPROD, TTDTE
(excluding any highly unlikely cross-midnight transfers :) ), ABS(TQTY),
and a.TSEQ = b.TSEQ-1 .

Using the time could cause misses, even up to the hour (e.g. 2:59:59 /
3:00:00) so I would exclude that part.

Kevin

<snip>