On Wed, Dec 15, 2010 at 11:41:45AM -0500, Jason Qualkenbush wrote:
> I'm being asked to build a 32bit system.

There are times when 32-bit systems are appropriate; in particular, some
(many?) applications run faster on a 32-bit system. There are other times
when a 64-bit system is appropriate. Neither is "good" or "bad" or
"clean" or "dirty" or even "future-proof" or "future-ready."

If you can't come up with a specific reason to prefer one over the other
then it doesn't matter which one you pick. In such a situation, having
fewer libraries to install is about a good a reason as any to go with
the 32-bit system.

> I'm irritated that I'm being forced to build this thing as 32bit.

I think you're letting your feelings get in the way of your professional
judgment. Your boss doesn't seem to have a solid technical reason for
wanting a 32-bit system, but at the same time, you don't have a solid
technical reason for wanting a 64-bit system -- yet you get annoyed at
your boss and not at yourself. Lacking technical evidence, your desire
for a 64-bit system is no better (or worse) than your boss' desire for a
32-bit system.

> It's a RSyslog, Cacti, Nagios system

I can't think of a good reason to prefer 32 or 64 bits here; there may
be a reason but I don't know it. Failing such a reason, I'd build
the machine to run in 32-bit mode, run performance tests, rebuild it as
a 64-bit system, run those same performance tests, then compare the
results and let that inform your final decision. If I couldn't do that
I'd probably go with a 32-bit system only because it has a longer track
record.

Again, what matters is evidence, not feelings( or emotions, if you
prefer); if you don't have the former then getting upset about the
latter is a waste of time (not to mention somewhat unprofessional).

Adam
_______________________________________________
Discuss mailing list
[email protected]
https://lists.lopsa.org/cgi-bin/mailman/listinfo/discuss
This list provided by the League of Professional System Administrators
 http://lopsa.org/

Reply via email to