Re: Data Corruption & 400 competitiveness

Rob

My reference to IBM improvements was that it seemed to me that when IBM first
comes out with some feature it does not always seem competitive with other
platforms, (except for the 400'ds massive stability & great security),
especially those that added that feature first, but then in later releases
IBM makes improvements & refinements, such that the feature matures.  It is
evident to me that the 400 is vastly superior to Windoz for general stability
& powerful DB, but I was wondering about pros & cons of 400 vs. IBM's
competition for companies that ONLY want it for serving a LAN as opposed to
also using the 400 for traditional business functions.

Some questions with respect to your current puzzle ... what are the odds that
you have users who forget to take tasks to end of job before they go home at
night & what are the odds that the tasks are not properly ended when the PC
is shut down or the 400 does its backup?

Have you done any tests in which you create a spread sheet, end the session,
then go back into the spread sheet, then abnormally terminate that session
using a variety of ways to accomplish that, to see what happens to the spread
sheet?

If I had the time, I might be doing that with some BPCS applications at
Global.  We have an intermittent data corruption problem right now with
doubled up material requirements for a few shop orders, which I can fix via
standard BPCS reorg if no one has had a shop order maintenance lock up trying
to alter it before I run the reorg.  This used to be once in a blue moon &
now we getting several hits a week.  So I  have to figure out which file(s)
corrupted, then track most likely software or user culprits.

For those of you on M_L who are not BPCS users, you may find it interesting
that BPCS has a menu devoted to fix-its ... things that are not supposed to
get corrupted, but in fact do so with great regularity & there are a bunch of
menu options to resolve common reoccurrences.  Most BPCS sites have added
their own menus of fix-its on top of this ... the last major one I added is
called cost reorg, to fix-it when the costs do not add up right.

We still run a manual SAVE 21 option off menu every evening & I have a
pre-save menu for my helpers to use when I am not there, so no one has to
remember commands & parameters.  Every night 1-3 users will each leave 1+
sessions signed on & I go look in their job stacks to see if they really at a
menu or in middle of some update program, because we do not have
http://www.precosis.com.au/rv1.htm (Remote View / Safe Ending) or
http://www.precosis.com.au/piu1.htm (Display Active BPCS Jobs) to make it
easy ... I try to bring everything down gently before SAVE 21.

Do you have multiple people updating this information, in which some of them
may be relative trainees in how to work the interfaces?

Our tool inventory is tied into labor reporting so that we get a usage report
on things essential to production like molds & applicators, such as scrap
percentage, which tells us which ones are in need of preventative maintenance.
The staff who maintain the tool system are not regular 400 users so we have
to run periodic query/400 to make sure they not made certain recurring human
errors.  It is not a problem with deleting things, but violating our naming
conventions with respect to coding things like item classes so they get
processed correctly.

MacWheel99@aol.com (Alister Wm Macintyre) (Al Mac)