Yet Another Ninja wrote:
On 6/25/2009 11:27 PM, John Rudd wrote:
On Thu, Jun 25, 2009 at 10:09, Chris
Hoogendyk<hoogen...@bio.umass.edu> wrote:
Gone are the days when you totally avoided upgrades because of the
time,
hassle and risk involved.
Time and hassle, maybe. Risk, no. Risk is not a binary, it's a
balancing act. Live updates don't remove risk, they simply alter the
risk balance. There will always be applications and environments
where risk is high enough that will cause you to wait.
For example, your 2 minutes of downtime... on wall street that could
cost you millions of dollars of stalled or canceled transactions.
(well, not lately, but before the crash...) So, your CFO will ask
you: is the risk of upgrading vs not upgrading worth a couple million
dollars? If the upgrade isn't worth it, then they will likely choose
to avoid it. Like I said "if isn't broken, don't upgrade", which
translates to "don't upgrade until the cost of not upgrading exceeds
the lost revenue of your outage window".
(and redundant systems may OR MAY NOT mitigate that)
can we get back to Spamassassin and a sane update cycle context? .-)
nah. I think we should get back to SORBS bites, and so does res, and so
does so and so, etc. ;-)
actually, my point was that there is not much excuse for not having a
more up-to-date perl these days, so yeah, go ahead and boot 5.6.x. If
there are legacy or OS things that require the older perl, you can
actully have your cake and eat it too. My Solaris 9 installs still have
/usr/bin/perl, which is 5.6.1, and the OS stuff from Solaris can still
use that. I have 5.8.7 in /usr/local/bin/perl on the Solaris 9 systems,
and SpamAssassin uses that. It's easy to manage $PATH and the #! lines
of scripts.
So, go for it.
--
---------------
Chris Hoogendyk
-
O__ ---- Systems Administrator
c/ /'_ --- Biology & Geology Departments
(*) \(*) -- 140 Morrill Science Center
~~~~~~~~~~ - University of Massachusetts, Amherst
<hoogen...@bio.umass.edu>
---------------
Erdös 4
