[conspire] East Bay AWS usergroup meeting

[Michael Paoli]

> From: Texx <texxgadget at gmail.com>
> Subject: [conspire] East Bay AWS usergroup meeting
> Date: Thu, 4 Apr 2019 00:31:49 -0700

> This was a REAL users group meeting, the likes Ive not seen in years.
> This was an engaged crowd, quite knowlwdgeable to boot.

If you've not checked it out, you might also want to try
Open Late - sponsored by OpenDNS - on (egad) Meetup.com
They do good technical presentations ... but I've not been in
some year(s) (2ish or so?), so not sure what/how they're doing
at present.  Oh, and for better or worse, typically includes
free pizza and some beverage(s).  I think last time I went it
was a (pretty dang) technical talk on gpg - most specifically on
"version 2", and advanced/new features, troubleshooting, internals
and structure (also of PGP in general), etc.  Good stuff!  :-)

> The guy was hysterically funny.
> Who knew accountants could be fun??!!??
Oh, working with a good/excellent accountant/CFO can be *highly*
impressive.  Perhaps I missed out on the more entertaining bits,
but among my experiences - one CFO I worked directly under,
was exceedingly adept at what he did.  He was an extreme expert
in finding opportunities all over the place to save us money - and
practically and highly workably too - not some theoretical b.s. or
some cr*p that would have negative impacts and end up costing more
directly or indirectly than it would save.  Likewise he was also
quite good and finding us ways to increase net profits - everything
from where/how/why/when to adjust prices, to new revenue opportunities,
better marketing/advertising (and where/how to target those $s), etc. -
lots of good/impressive results under that CFO.
Likewise, once upon a time, my boss's boss was CFO of ... Wells Fargo.
I was in great position (flat hierarchy) in the company (me, worker bee,
nobody under me, and only 3 steps up the chain to CEO).  Anyway, at least
with what bits I did get to see, that particular CFO was excellent and
highly effective.  Of course with many big companies - as Wells Fargo
was then, and is even more so now - things can vary greatly depending
where one is within the organization (one's manager, what part of
company, located where ...).  At other times at Wells Fargo,
I could go up 5 levels of management, and with the exception of one
mostly remote manager (my boss at the time), I wouldn't even make it
off the same quadrant of the same floor ... and it was many more hops
to make it all the way up to CEO (something like 9 from me to CEO) ...
yes, I was way deeply buried in hierarchy in that other
position/group/location ... and even though it was still the same
company.

> The problem is that in the stampede to the cloud, people forget that the
> cost saving of the cloud is the ability to spin up resources on the spot
> and terminate them just as fast.
>
> If your machine is going to be up all the time, the cloud will be more
> expensive than leaving it running in a co lo cage.

That can be (partially?) mitigated within, e.g. AWS.  You likely heard,
but AWS - and likely others have similar - AWS has "reserved instances" -
you pay for 'em whether you use 'em or not ... but at a significantly lower
price than "spot" (pay as/when you use) instances.  So, you go with
reserved (at least in AWS) if you're going to use then all - or most all -
or even much, of the time.  Use spot just for stuff that will go up and
down a lot, or only be needed for shorter bursts.  And don't think
per application, think per infrastructure - what's needed to run
*everything* one has in AWS - at any given point in time - if it needs
be there all (or most) of the time ... reserved instances.
And I'm sure AWS also likes it too - for them, reserved instances is
effectively a "guaranteed" (~year prepaid/promised) income stream,
whereas all that "spot" stuff - it can go away at any time - juicy
margins while it's there, but no guarantees it will continue to be there.