[linux-elitists] What to do about cluebatting such companies, that require possibly *YEARS* old Distros

Rick Bradley roundeye@roundeye.net
Wed Jan 26 14:03:38 PST 2005

* Aaron Sherman (ajs@ajs.com) [050126 16:36]:
> > Then again, I suppose we're talking about improperly administered
> > environments almost by definition in this thread, so, um, carry on.
> If we're talking about successful companies, then I debate your use of
> the word "improperly". The company is not in business to upgrade an OS.
> They are in business to make money. Demonstrate the cost/benefit
> analysis that shows them making more money using kernel 2.6 everywhere
> and I'm sure they'll move faster.

Differ.  I know of a large state hospital which successfully operates in
the way I'm describing.  I also know of a multi-billion dollar private
financial company which operates this way -- who actually not only
"upgrades" their software easily (and can revert almost instantly), but
actually picked up and moved their entire operation from one facility to
another with zero downtime, while continuing to run high-availability
production, testing, development, and research environments spread over
multiple user bases, without a hitch.  I also have had lengthy
conversations with the now-retired process manager for a large
publicly-traded corporation with production environments in nearly every
state in the U.S., and I can assure you that, despite the fact that
assmonkeys exist in every corporation over size 15 employees, that even
the largest corporations, when properly managed, can deploy a roll-out.
These are all large-scale roll-outs, granted, but the principles can be
greatly compressed to advantage for medium and small-scale roll-outs,
given the proper awareness of the techniques.

Now, many businesses don't need to have production environments, in
which case, I agree, who gives a shit whether they're running the
leetist new k3rn4l.  If I'm not trying to sell them "business in a box"
I'm personally not concerned about them (and I'm not).  Go develop
"business in a box", secure and auto-deploy the upgrades and quit whining
about how Billy's Autobody is never going to patch his box.  REALLY?!?
Stop the fucking presses.  If you want to solve that guy's problems you
turn business systems into appliances, sell them for $39.99 a month and
automate the process of patching.

Otherwise, we're dealing with people who have a stake in their systems'
reliability, security, and efficiency.  Availability is a direct
bottom-line item (as it is even for Billy's Autobody, though he'll
probably never know it), and these people are much more likely to
appreciate that.

For those people, where the systems can't make process superfluous some
awareness of production methodology is necessary.  Let me reiterate
that, where it matters, if production outages are a matter of course,
then somebody's incompetent (or, perhaps pissed off, but then it's
reasonable to question the quality of management).

 http://www.rickbradley.com    MUPRN: 184
                       |  think it is best to
   random email haiku  |  config eth0 to be live and
                       |  eth1 as private. Well.........

More information about the linux-elitists mailing list