bo*********@hot mail.com (James Neumann) wrote in
<77************ *************@p osting.google.c om>:
Is there any scope for presenting your client with a cost-benefit
analysis to the various scenarios you and others here have
mentioned?
Well, in the long run, I do think that it would be better for the
data to be in SQL Server, given the quantity of it and the kind of
use it gets. What happens is that it comes under very heavy use
during short periods of time. They have had a few corruptions of
the data file with A2K, but none since I got all workstations on
SR1a and Jet SP6 (that's well over a year now), and they never lost
even one byte of data even then -- it was just inconvenient.
IMHO, based on what you have described, I really cannot see the
justificatio n to convert. Why? Regardless of the choice of
conversion method, there would be a material expenditure of time
to do the conversion, and test it. . . .
Yes, but as a long-run investment, it may be worth something to
them.
. . . Also worth considering is the
cost associated with setting up a dedicated SQL Server box, if the
client is a small shop. . . .
They have the equipment available. And SQL Server (and maybe even
new equipment) is likely to be donated by, well, let's just say by
one of the largest software companies in the world. Ahem.
. . . If they are a department within a large
company, they may incur a significant overhead charge to have a
SQL Server database online. (In several cases, I've done projects
in Access as opposed to VB/SQL Server for this very reason.)
Online? Who said anything about online?
And no, it's not a large company. It's a small consulting
organization.
I haven't spoken to the technical aspects of this, for which I
apologize - others have covered this better than I could.
I've got all these aspects in hand already. They wouldn't be
considering the SQL Server conversion at this point if a certain
large company had not made overtures about donating software (and
maybe equipment) for a certain large project taking place in New
York City in early September 2004. Ahem.
They have had Small Business Server as long as I've worked with
them (since May 2000), and I've spoken to them about spending time
($$$) evaluating the cost/benefit ratio of converting the back end
to SQL Server (by testing a raw upsizing and seeing what caused
problems and what gave benefits). They've recently abandoned the
Small Business Server package (they weren't using anything but
Exchange) for a new Win2K Server without SBS (just Exchange).
The offer of the software donation has raised the issue again.
I have told them that it would be a major cost and a large project
to convert. But I didn't tell them it would be without any benefits
at all, which I do not believe.
In a certain sense, it's one of those CYA situations. I can imagine
that if they brought in an outside database "expert" they'd be told
that they were in horrid danger of losing all their data because
they are depending on a toy database, etc., etc. They really are on
the threshold of the point at which it really becomes cost- and
performance-effective to switch over to a SQL Server back end.
But I wouldn't have any problems myself if they stay with the Jet
back end for a while. I see no reason why things should suddenly
become unreliable. The time when they switched servers (last June)
would have been the the likely point for things to go haywire and
nothing happened at all. Indeed, the migration of the database
application to the new server was the one thing that went off
without a hitch (I just told the sysadmins to put the database in a
top-level share named the same as on the old server and give the
new server the same name as the old one).
I'd be happy either way. Indeed, I'd love to see this project
converted to SQL Server as it would finance my training in using
SQL Server beyond my current at-home dabbling in it. But I'm
certainly not going to push it just because of that.
--
David W. Fenton
http://www.bway.net/~dfenton
dfenton at bway dot net
http://www.bway.net/~dfassoc 
