►
From YouTube: SIG App Delivery Operator Working Group 2020-09-30
Description
SIG App Delivery Operator Working Group 2020-09-30
B
A
B
A
A
A
Yeah,
it's
I'm
in
this
really
funky
room.
The
house
was
built
40
years
ago,
but
they
built
this
greenhouse
and
it's
part
of
the
house,
but
it's
a
greenhouse,
which
is
why
you
see
stucco
behind
me.
So
it
looks
like
the
outside
of
a
house
and
what
you're
seeing
is
the
reflection
from
these
windows.
Okay,.
B
A
Yeah,
you
know
it's,
we,
my
husband
and
I
will
spend
more
time
up
here
when
he
retires.
He
works
in
aerospace
and
he
works
on
secure
programs,
and
so
he
can't
work
remotely
and
so
normally
ammo.
Normally
I
travel
a
lot,
and
so
I
come
here
on
the
weekends
in
between
and
stuff.
I
have
been
here
for
like
three
months
out
of
the
last
five
months,
because
I
can-
and
so
I
spend
like
three
weeks
up
here
and
then
I
go
back
to
california
for
three
weeks
and
it's
just
a
nice
change
of
scenery.
B
B
It's
a
city,
oh
in
austria,
between
salzburg
and
vienna,
but
I
can
have
the
mountains
and
the
lake
pretty
much
right
next
or
pretty
close
there
and
the
less
you
go
to
the
office
unless
it
matters,
because
even
if
I'm
driving
to
the
office
say
twice
or
three
times
a
week,
it's
just
a
one-hour
drive
maximum,
which
for
us
type
of
commute,
is
not
that
bad
for
us
to
wear,
not
ideal
but
totally
doable.
If
you
wanted
to
do
it,
so
things
are
definitely
mixing
up
here.
A
A
B
B
Updates
that
was
interesting
for
you.
I
went
through
the
entire
document
on
operate
list.
Again,
I
just
replied
to
some
notes.
I
was
expecting
some
already
over
I'm
going
to
remove
the
related
technology
section.
I
think
it's
just
causing
too
much
confusion
and,
quite
frankly,
as
we
discussed
the
last
time,
it's
not
really
what
the
initial
charter
was
about.
B
B
C
B
Some
people
will
have
opinions
on
this,
I'm
personally
struggling
a
bit
using
something
like
a
higher
level.
Abstraction
like
you
would
need
this
to
control
my
underlying
amazon
cloud
resources.
I
like
the
idea
behind
it,
but
I
don't
see
it
as
an
application.
Runtime
really,
but
that's
might
be
interesting.
This
discussion
may
want
to
include
this
because
one
of
the
things
about
monitoring
in
there
was
there's
a
reconciliation
loop
which
doesn't
exist
if
you
control
resources
outside
of
kubernetes.
Unless
you
implement
it
yourself.
B
So
the
the
the
monitoring
point
in
there
was
obviously
like
there's.
Not
the
monitoring
part
of
the
operator
is
on
traditional
monitoring.
It's
it's
the
reconcile
loop.
That
tells
you
whether
something
has
changed
right
and
this
doesn't
exist
if
I
monitor
something
outside
of
kubernetes.
So
if
I
have
my
ec2
resources,
there
is
no
reconcile
loop.
C
A
B
B
A
B
A
A
B
C
B
We
define
it
because
that's
where
it
changes,
I'm
not
saying
you
can't
do
it.
I
was
just
struggling
doing
it
on
the
same
cluster.
So
for
me
just
the
example
of
like
having
an
eks
cluster
that
is
controlling
its
own
underlying
ec2
infrastructure,
feels
kind
of
weird
to
me,
because
the
abstraction
is
going
in
the
wrong
in
the
wrong
direction
than
there
yeah,
because,
if
you're
getting
more
getting
more
to
a
lower
level
of
abstraction
than
to
hire
one,
but
overall
doesn't
mean
that
you
can't
do
it.
A
B
A
B
B
C
B
So
this,
I
think
it's
going
to
be
way
more,
like
a
collaborative
effort
we'll
get
in
there,
it's
more
solving
application
delivery
problems
in
kubernetes
by
example,
more
or
less
tell
us
your
problems.
This
is
how
you
can
solve
them
and
also
help
like
with
this
demo
application
development.
So
I
I'll,
I
think,
in
the
next
meeting
I'll
crowdsource,
this
a
bit
more
say.
Okay,
this
is
the
talk
like
obviously
hearing
now
we
have
submitted
it
because
the
the
sig
chairs
are
submitting
it.
B
A
C
A
B
A
B
C
B
B
B
And
by
the
way,
if
I
upgraded
how
I
cannot
blue
green
deploy,
how
do
I
tell
my
environment?
Can
we
need
a
blue
green
deploy?
I
cannot
roll
back
so
there's
a
lot
in
the
space
that
yep
people
should
be
looking
at,
but
there
was
a
great
talk.
I
might
actually
get
her
to
speak
again
on
these
topics
and
I
think
one.