►
From YouTube: Kubernetes SIG K8s Infra - 20230201
Description
A
B
D
A
E
E
A
F
A
A
Okay,
no
one
so
I.
Think
next
item
is
building
report
again,
since
we
have
the
first
day
of
February
I
think
we
can
basically
directly
check
span
for
last
month
unless
something
so
something
in
the
daily
spend.
A
A
A
A
Science,
December,
so
I
think
it's
it's
like
took
a
month.
I
opened
a
bridge
about
this
and
I
asked
Jim
angel
from
Google
to
embed
to
investigate
this.
We
had
a
conversation
about
this
yesterday.
You
should
basically
check
the
link
in
the
issue
explaining
the
situation.
So
there's
like
a
new
service
in
gcpe.
We
get
it.
We
basically
get.
We
got
bill
and
we
also
get
a
discount,
but
that
discount
is
not
reflect
in
the
building
dashboard.
F
C
You
sorry
scroll
up,
I
just
want
the
number.
G
X
yeah
I'm,
sorry
I
was
just
I'm
like,
as
you
know,
I
did
the
original
report
in
the
original.
Like
bigquery
analysis,
do
you
have
to
do?
Do
the
right
people
have
access
to
that
now
or
is
there
something
like
do
you?
Are
you
asking
for
Access
or
me
to
fix
it
or
like
what
like?
What's
the
right
way
to
do
this.
A
Relate
to
bigquery
storing
information
is
more
like.
Basically,
the
information
we
extract
for
bigquery
is
now
misaligned,
because
there's
a
new
spend
there's
a
new
line
of
project
in
bigquery,
so
we
need
to
basically
update
the
query
in
data
studio.
So
if
you
can
basically
give
access
to
anyone
interest
to
modify
the
dashboard,
so
we
can
have
something
reflecting
that's
going
to
be
fun.
That's
going
to
be
awesome
of.
If
you
want
to
take
a
look
at
the
core
we
use
currently.
G
I'll,
take
a
quick
look,
I
think
I
added
you
I,
think
to
the
the
report
and
I
can
add
more
people
if
need
be,
and
one
thing
we
do
want
to
be
careful
of
is
just
because
it's
discounted
doesn't
mean
it
isn't
coming
from
our
like.
All
of
our
spend
is
discounted
right,
so
we
want
to
make
sure
that
the
discount
isn't
coming
out
of
our
the
biggest
bucket
yeah.
A
H
H
H
Usual
behavior
for
all
their
new
products
like
it's,
not
helpful
first,
but
it's
a
free
iPhone,
so
I
wouldn't
worry
too
much
about
it.
A
No,
it's
not
an
issue.
Okay,
let
me
let
me
be
clear:
it's
not!
We
are
not
built
for
that.
That's
what
I'm
saying
we
are
not
there
for
that.
We
basically
get
so
what's
happening
is
basically
we
get
to
we.
We
are
built
for
those
service,
but
at
the
same
time
we
get
a
discount
that
nullify
them
and
that's
reflect
directly
in
the
billing
report
in
basically
the
gcp
billing
report,
so
we
charge
one
one
K
and
we
get
the
discount
for
1K.
That's
what's
happening.
A
Depending
on
which,
which
project
is
in
fact,
it
could
be
5K,
something
I,
think
and
December
it
was
2K.
January
was
4K.
A
Foreign
is
now
confusing
because
it's
not
showing
up
when
I
check
internal
in
the
gcp
console,
what's
happening,
I
say
bidding
for
December
but
is
not
impacting
the
overall
bill.
But
when
I
check
in
that
building
in
display
report
is
this
spread
basically
there's
a
Delta
happening
I
can
even
let
me
let
me
one.
A
Yeah
I
did
that
there's
no
there's
no
doubt.
I
A
We
used
to
do
that
when
team
okay
was
around,
but
science
is
now
busy
with
other
things.
We
decide
to
not
do
that
anymore.
I
Okay,
well,
I
would
suggest
it
might
be
worth
a
a
brief
moment
in
the
meeting
when
we're
going
over
this
report
to
pull
up
the
actual,
like
account
page,
that
shows
the
the
like
burn
rate
and
how
much
is
left,
because
that
should
be
the
like.
That
should
be
accurate
and
should
show
us
how
much
is
left.
A
I
A
A
And
this
is
the
one
thing
we
were
talking
about:
I
think
APA,
opening
issue
about
that
that
basically
redirects
all
the
non-google
traffic
to
S3,
yeah
I
think
we
should
try
to
do
that
in
the
in
coming
weeks.
C
Yes,
I
do
so
in
the
agenda.
The
plan
is
we
go
over
billing
and
then
we
look
at
AI
review.
I.
Think
we
can,
aside
from
anything,
that's
like
pertinent
or
immediate
I,
have
that
billing
review
kind
of
towards
what
Ben
was
saying?
Let's
just
look
at
how
many
credits
we
have
left
and
see
if
there
is
some
sort
of
a
trend
between
you
know
if
things
are
going
down
or
up,
but
I
have
more
to
talk
on
that
in
a
minute.
C
But
when
we
go
over
the
AI
review,
one
thing
I
want
to
make
sure
of
is
we
have
identified
not
who
is
necessarily
doing
what
specifically,
but
we
have
a
lot
of
moving
parts
and
not
just
AIS
but
more
initiatives
that
we
need
to
kind
of
track
better.
A
good
example
would
be
artifacts.cates.io
in
my
head.
Justin
is
kind
of
the
initiative
owner
because
he
has
done
a
lot
of
the
initial
work.
C
He
has
started
writing
Porsche
and
whatnot,
but
we
might
be
able
to
delegate
a
bunch
of
that
work
towards
coopermatic,
but
still
have
Justin,
be
the
initiative
owner
verbiage
to
me
doesn't
really
matter.
I
just
want
to
make
sure
that
it's
clear,
like
we
have
a
bunch
of
initiatives
and
then
under
the
initiatives
or
the
AIS
that
we're
trying
to
delegate
out
I
bring
this
up
because,
through
various
conversations
it
seems
like
everyone
has.
You
know,
ideas
about.
What's
going
on
and
I
am
all
for
ideas
and
I.
C
That's
that's
really
it
for
this.
This
part
of
it.
C
And
this
kind
of
goes
back
to
like
the
PM
thing.
Honestly,
we
talked
about
doing
more,
like
PM
work,
or
at
least
me
kind
of
taking
it
over
late
last
week.
I
think
Arno
and
I
am
willing
to
do
that
and
kind
of
break
things
up
and
fill
out
that
board
more,
but
also
I
then
immediately
got
onto
a
plane
this
week
and
you
know
that
doesn't
isn't
really
conducive
to
work.
So
that'll
that'll
be
a
next
week.
Gp
problem,
I
guess.
G
I
I
I
think
that's
as
one
of
the
initiative
but
I
remember.
We
called
me,
but
people
I
think
that's.
G
My
title
and
I
think
that
sounds
I.
Think
that
sounds
great
and
I'm
happy
to
work
with
coopermatic
on
that
and
I
think
I
think
having
yeah
more
more
clear
like
what
are
we
working
on
in
like
what's
the
blocker
editing
point
would
be,
would
be
awesome
for
getting
things
done
faster.
J
And
for
us
it's
also
easier
because
then
we
have
a
dedicated
person
that
we
can
talk
to
to
the
specific
issues.
So
it's
not
so
much
looking
for
the
details
for
ourselves.
C
Yeah
Ben
I
see
your
hand.
I
Sort
of
a
side
thing,
but
you
mentioned
that
like
what
we
should
or
shouldn't
get
I
just
wanna
clarify
that
we
should
still
be
looking
at
the
at
that
data,
Studio
or
looker
report
that
we
have,
because
we
do
want
to
know
when
there's
Trends
in
specific
Services
moving,
but
I
think
we
should
also
look
at
the
just
here's
the
money
in
the
account,
because
that's
a
more
definitive
thing
and
something
that
we
need
to
keep
an
eye
on
as
well.
I
But
if
we
want
to
know
like
where
you
know
where
is
it
going?
The
the
data
Studio
report
gives
us
a
good
idea
of
like
okay,
we're
seeing
increased
usage
of
artifact
registry.
That's
expected
because
the
residue
is
Shifting
over,
but
if
we
see
cost
go
up
Suddenly
on
like
5K
node
scale.
Testing
is
100
more
expensive
this
month
and
like
that,
we
need
to
flag
that.
C
Yeah
I
think
I
think
the
one
thing
that
I
want
to
see
a
little
bit
better
is
I
think
the
reporting
was.
The
data.
Studio
report
was
good
when
we
were
first,
you
know
starting
out,
but
I
really
would
like
to
see
some
sort
of
reporting
that
also
shows
like
again
AWS
spend,
so
we
can
at
least
see
where
the
needle
is
going
right.
C
This
is
something
that
I'm
looking
to
ask
hippie
and
II
to
work
on
a
little
bit
more,
so
that
would
actually
be
another
initiative
under
there,
mainly
because
they're
doing
a
bunch
of
the
kind
of
accounting
and
administer
via
work
right
now
and
going
from
that
to
figuring
out
better
reporting
and
kind
of
the
phenopsy
type
of
work
that
we
need
is
a
pretty
good
transition,
I
think
and
then
kubermatic
can
work
on
the
the
prowork,
artifacts.cates.io
work,
etc.
A
So
the
idea
is
to
basically
shift
those
workloads,
so
we
reduce
the
overall
cost
of
the
gcp
infrastructure
and
the
gcp
budget.
So
they
are,
we
I
think
we
last
week
we
have
some
kind
of
priority
meeting
and
we
identified
basically
four
workloads.
We
need
to
migrate,
so
I
put
ownership
on
the
docs
for
when
I
say
ownership,
point
of
contact
for
each
of
the
initiative
we
identified
given.
Is
it
okay
for
you?
Oh.
C
Point
of
contact
is
probably
the
better
the
better
phrase.
It's
just
like
that.
They're
the
person
that,
if
there
are
questions
we
should
be
asking
them
they're
the
ones
that
are
kind
of
driving
it
and
have
the
initial
idea.
It's
just
you
know
we
need
to
share
the
load
a
bit
more,
so
they're
the
ones
you
know
leading
the
charge
on
whatever
that
initiative
is,
and
we
direct
questions
towards
them.
A
Okay,
so
I
put
the
name
of
the
people
splint
of
contact
for
those,
obviously
for
workloads
or
for
track
of
Roads.
So
if
you
have
any
exciting
card
question
reach
out
to
those
people
in
public
place,.
A
C
A
A
A
To
have
minus
one
plus
one,
and
this
on
that
suggestion,
okay,
R
plus
one
for
Mario
plus
one
from
j3
I,
will
try
to
keep
you
as
people
in
mind
about
this
I.
C
I
think
the
weekly
Cadence
is
actually
going
to
be
better
one
of
the
problems.
I
think
we
initially
have
run
into
is
this
being
bi-weekly.
We
can't
necessarily
hit
the
ground
running
without
having
out
of
band
meetings,
so
as
long
as
these
are
separate,
but
still
just
as
open
meetings.
Absolutely
and
yes
time
zones
are
hard.
I
would
be
perfectly
fine,
biasing
a
little
bit
more
towards
the
EU
just
because
this
one
is
a
little
bit
more
us-centric
and
then
the
U.S
folks,
you
know
we'll
try
and
hit
it
as
we
can.
A
Okay,
so
I'm
gonna
try
to
set
a
meeting
starting
next
week,
a
weekly
meeting
almost
dedicated
to
basically
how
we
bootstrap
AWS
and
how
we
can
use
it
quickly
and
I
have
some
conversation
about
how
we
and
take
great
Cooper
Martin
in
the
in
basically
how
we
align
work
with
diplomatics
about
what
we're
doing.
C
To
answer
the
question
question
is:
should
we
use
GitHub
task
tracking
and
planning
tools?
Absolutely
we
do
have
a
project
board
up
it's
just
kind
of
going
back
to
what
I
said
it
needs
to
be
populated
between
travel
and
other
stuff.
The
plan
is
for
it
to
be
updated
next
week
and
that
should
have
all
of
the
initiatives
that
we've
been
talking
about
and
we
can
start
tracking,
more
granular,
AIS
yeah.
A
We
might
have
me
Triple
boards,
depending
on
how
we
go
fast,
yeah
to
be
to
be
completely
transparent.
I
think
we
need
to
growth
only
for
basically
AWS
usage.
The
rest
I've
already
have
issue
umbrella
issue,
small
task
and
I
think
it's
kind
of
straightforward
to
do
those
migration
for
some
of
them,
so
I
think
my
I
think
the
main
concern
is
basically
hey
how
we
yeah
the.
C
Ci
stuff,
yeah,
I,
I
kind
of
have
to
disagree
in
just
because
of
my
position
that
tracking
board.
We
have
a
lot
of
eyes
on
this
project
now
that
tracking
board
isn't,
unfortunately,
isn't
just
for
us
and
the
folks
doing
the
work.
It's
the
place
where
we
can
point
higher
up
folks,
too
and
say
this
is
where
we're
at,
and
this
is
what
we
have
identified
left.
C
C
A
Is
on
the
meeting
items
because
I
see
Jiffy
against
talking
about
priority
list?
Do
we
want
to
skip
that
or
Jiffy?
Do
you
want
to
say
something.
C
No,
it's
it's
more,
just
recognizing
that
in
this
meeting
we
came
up
with
a
priority
list
last
week
in
a
meeting
with
the
Sig
leads-
and
this
is
this
is
what
we
have
decided
is
like
our
Guiding
Light
for
at
least
the
next
three
months,
if
not
more
so,
identify
that
recognize
like
based
on
those
priorities.
Well,
that's
a
lot
to
read
based
on
those
priorities.
Where
do
the
initiatives
lie
and
then
we
can
then
kind
of
delegate
tasks
and
see
where
we
need
to
put
people
accordingly.
C
While
I
do
that,
I
see,
Mario
has
his
hand
raised
before
you
talk.
Mario
I
want
to
read
quick
what
Jay
said
Jay's
concerned
about
the
visibility
into
ongoing
AWS
costs
when
he
left
November
the
in
November.
His
understanding
was
the
agreement.
I
mean
I'm,
assuming
you
typed
this,
because
you
can't
talk.
C
You
can
interrupt
me,
but
my
understanding
was
the
agreement
with
AWS
was
for
ongoing
credit
support
for
Kate's
infra
on
AWS
and
that
the
credits
would
be
used
and
that
we
said
Kate's
in,
for
it
could
demonstrate
that
those
credits
were
going
to
be
consumed.
C
So
I
hate
to
go
too
much
longer
without
being
able
to
show
the
consumption
of
those
credits
that
is
I
got
you
yeah
I,
understand
loud
leaf,
blower.
C
Yeah
I
I
do
have
one
thing:
I
want
to
say
this
goes
back
to.
We
need
better
reporting
and
that's
something
that
we're
gonna.
You
know
throw
hippie
and
II
at
because
they're
already
doing
account
administrivia,
so
they
can.
They
can
help
build
reports
and
focus
more
on
the
Finn
Ops
Mario.
You
had
your
hand
raised
sorry.
J
Yes,
I
no
worries
so
for
the
priority
list
that
was
discussed.
You
can
also
read
it
up
in
the
select
Channel,
where
dim
started
the
thread
where
all
of
the
action
points
are
written
into
and
some
some
proposals
already
came
up
regarding
the
section
list,
just
that
we
don't
forget
this.
A
J
H
Talk
about
it,
so
one
of
the
initiatives
that
I'm
working
on
is
freezing
the
old
registry.
The
intent
of
that
work
is
to
signals
with
me
today
this
first
three,
you
shouldn't
be
using
any
more.
Let's
go
and
use
this
new
one,
we're
very
confident
that
it
works
so
that
kept
merged
this
morning
next
evening,
so
yeah
as
part
of
that
cap
I
need
to
send
an
email,
I
drafted
an
email
that
says
hey
it's
frozen.
This
is
what
it
means
fee
for
a
contributor.
H
This
is
what
it
means
for
an
end
user.
The
TL
DR
tldr
is
about
1.27,
won't
be
there
anymore
or
any
of
the
patch
releases
of
paper,
which
should
be
fine.
H
Yeah
yeah
I'm
at
Planet
of
security
guard
as
well.
Yes,
if
you
scroll
all
the
way
down
a
little
bit
further
further.
H
H
H
Yeah
all
right,
those
are
the
timelines
right.
The
rest
of
April
it'll
be
Frozen,
so
that's
a
date
between
the
first
rc
and
the
last
RC,
so
I'm
going
to
use
that
time
to
make
all
the
technical
changes
to
the
Power
jobs
and
all
the
tooling.
A
So
basically,
I
put
the
draft
of
the
communication,
because
I
expect
to
send
that
by
Tuesday
next
week,
not
many
basically,
my
match
should
be
sent
this
by
Tuesday
next
week.
So
what
I'm
asking
to
people
in
this
course
is
basically
do
a
review
until
Monday
February
6.
So
we
can
shoot
out
this
message,
Tuesday
and
spread
the
word
about
this,
so
that
gave
roughly
60
days
for
people
to
be
aware
about
refreshing,
we're
freezing
the
old
registry.
B
H
Something
we
can
look
at
in
the
future
is
maybe
revisit
every
direction
conversation
because
the
registry
will
be
functionally
useless
after
what
20
after
one
23
goes
here
and
I.
Believe.
A
A
I
By
me,
yeah
I
mean
we
may
have
to
can
we
may
have
to
consider
measures
at
some
point.
Just
I
mean
we
only
have
so
much
funds
and
we're
already
pretty
over
on
that.
But
we're
I
mean
we're
going
to
be
supporting
these
versions
for
a
while
anyhow,
so
I
don't
think
that
has
to
be
decided
today
and
we're
going
to
regardless
of
whether
or
not
we
shut
it
down.
We
need
to
take
a
relatively
aggressive
posture
to
push
people
to
to
migrate.
Now,
whether
or
not
it
it's
still
possible
to
use.
H
I
Like
I
couldn't
tell
you,
even
if
I
wanted
to
what
the
plans
are
for
like
gcr's
future,
but
I
mean
in
any
case
at
some
point,
we'll
the
project
will
want
to
stop.
I
It
usually
just
goes
down
paying
to
like
host
these
things,
redundantly
that
no
one's
using
or
that
there's
very
limited
usage
on
when
we
have
when
you
can
fetch
it
from
the
new
location,
something
else
we
should
talk
about
at
some
point
that
we
shouldn't
go
into
here,
but
to
put
on
people's
radar
to
think
about
is
we
have
just
migrated
all
images,
Forward
Forever,
and
we
have
things
like
the
image
promoter,
signing
images
where
the
scope
of
what
we
need
to
host
only
goes
up.
I
I
A
I
Not
asking
to
solve
that
in
this
meeting,
but
I
would
like
people
to
start
thinking
about
it,
because
we
have
a
really
good
opportunity
in
phasing
people
over
to
the
new
registry
to
reset
expectations.
I
So
communicating
communicating
at
least
to
intentions
sooner
is
good
whether
or
not
we
even
get
around
to
implementing
it.
So
we
don't
need
to
solve
this
now,
but
I
think
people
should
start
thinking
about
like
what
do
you
think
would
be
reasonable
there,
and
and
can
we
follow
up
on
that
and
maybe
say
that?
Okay,
besides
phasing
out
the
old
registry
with
the
with
the
new
registry,
we
carried
forward
many
many
many
years
of
publishing
images.
I
Do
we
really
need
to
be
still
hosting
like
kubernetes
1.0,
on
the
on
the
new
one
or
or
someone's
random
image
back
when
people
were
manual
pushing
and
there's
like
dirty
tags
and
stuff?
There's
there's
content
that
probably
shouldn't
exist
anymore
and
we
don't
have
a
policy.
So
it's
less
of
a
cost
issue
and
more
of
a
like
are
some
of
our
infrastructure
struggles
a
little
bit
with
that
for
expanding
amount
of
content
to
process,
and
one
answer
is
well.
I
We
should
just
we
should
just
get
rid
of
old
content
instead
of
trying
to
fix
that
the
the
the
image
signing
stuff
in
the
image
promoter
has
made
it
a
little
bit
challenging
for
us
to
have
all
the
artifact
registry
regions
that
we
did
for
cost
reasons,
and
part
of
it
is
because
it's
append
only
indefinitely.
A
Okay,
I
think
the
one
thing
like
we
should
basically
open
some
kind
of
issue
to
track
this
in
the
KK
repo
and
have
the
right
people
to
talk
about
this.
Like
I'm
thinking
about
people
like
sick,
sick,
Arc
and
yeah,
especially
specifically,
seek
out
to
discussing
NC
testing
like
how
we
establish
image
retention.
I
In
her
tear
point,
if
you
were
using
kubernetes
1.0
and
production
like
it,
never
existed
at
the
same
point,
it's
pretty
weird
for
us
to
support
you
migrating
forward
to
registry.kates.io
to
continue
to
use
an
extremely
old
release
and
it
puts
a
little
bit
of
a
strain
on
some
of
our
infrastructure
process
and
ever
expanding
set
of
images.
A
J
A
B
A
Okay,
so
I'm
gonna
just
give
the
ads
up
here,
so
we
just
found
for
people
maybe
new
and
for
the
recording
in
October
there
was
an
announcement
about
getting
three
millions
credits
for
the
kubernetes
depression,
there's
no
expiration
date
on
those
credits.
So
we
don't
know
and
we
don't
get
the
treatments
right
away.
We
get
three
350
and
thousand
dollar
each
month
or
I
think
any
every
three
months.
I
think
something
like.
C
That
yeah
so
yeah,
it's
250k
each
month,
which
amounts
to
a
three
three
million
dollar
donation
over
the
course
of
12
months.
Okay,
but
the
one
thing
that
I
will
also
append
to
that
is,
we
have
been
pretty
pretty
clearly
told
use
them
or
lose
them.
So,
whatever
we
don't
use
this
year,
we
should
not
expect
the
full,
like
Grant
back
next
year.
So,
okay.
A
A
So
right
now
we
did
migration
of
three
accounts
from
the
CHF
organization
to
the
new
kubernetes
organization.
We
did
one
yesterday
with
EP.
Thank
you
for
that.
Ep
I
think
riyan
and
AP
is
also
working
on
moving
the
icon.
We
use
for
CI
to
basically
the
new
organization.
That's
gonna
up
I
think
that's
going
to
happen
in
the
next
week
or
something
like
that.
K
The
AWS
account
migration
should
be
done,
hopefully,
by
the
end
of
this
week,
I'm
not
keeping
some
daily
limit
issues,
but
working
diligently.
B
Yeah
so
again,
if
there's
anything
in
me
or
anywhere
else,
let
me
know.
A
D
D
Were
one
of
the
things
that
we're
looking
for
is
to
connect
with
our
account
manager
from
the
cncf
to
AWS
for
some
stuff
that
we're
having
some
issues
on
in
migration?
D
E
F
F
Yeah,
the
it's
it's
no
longer
seems
to
be
like
urgent
anymore,
there's,
not
much
motivation
in
Jordan
who's.
Looking
at
it
isn't
sure
why
this
is
happening
at
all
right.
There.
A
I
A
C
F
A
I
C
C
A
F
I
Sorry,
this
fell
through
there's
just
been
a
lot
yeah.
A
A
G
F
A
A
G
I
mean
that's
fine,
I
think
the
bottleneck
is
in
for
changes
right
so
like
I
hope
that
we
can
work
with
kubernet
to
to
get
those
info
changes
made
quickly
like
cloudfront
would
still
be
an
infra
change
right,
we're
not
we're
not
blocked
on
the
proxy
servers.
G
We're
blocked
on
like
terraform
making
everyone
feel
very
unsafe
and
like
it
would
be
great
if
we
had
something
that
made
people
feel
comfortable
so
that
we
could
apply
things
quickly
but
yeah
everything
is,
you
know,
sort
of
One,
Step,
Above
script,
ops,
so.
A
B
G
I
mean
we
still
have,
so
we
have
to
get
the
the
artifacts
onto
S3
anyway,
right
so
I
have
there
is
a
separate
PR
to
do
the
promotion
there
that's
merged
and
we're
we're
sort
of
doing
better.
On
the
code
front
we
can,
we
can
get
the
artifacts
onto
S3
as
well.
We
can,
there
are
there's,
there
are
infra
changes
needed
to
create
those
buckets
or
a
bucket
yeah.
G
So,
like
I,
feel
like
we're
not
blocked
on
code,
we
are
blocked
on
infra
changes
and
when
we
there
will
be
infra
changes
for
cloudfront,
there
will
be
infer
changes
for.
A
G
The
for
the
redirector
service,
which
is
like
the
same
redirection
service,
we're
using
on
images
I,
would
like
to
use
the
I'd
like
to
to
like
continue
with
the
code.
If
we
want
to
if,
for
some
reason,
we
think
we
can
set
up
a
cloudform
distribution
manually
or
something
which
is
not
what
we've
done
in
the
past,
then
we
could
just
do
that
and
I'm
fine
with
that.
But
yeah
like
right
now,
as
far
as
I
can
see,
the
bottleneck
is
the
time
to
apply
terraform
changes.
Basically,
okay,.
A
A
A
I
What
I,
that's
all
I'm
trying
to
say,
is
I.
Don't
because
the
only
thing
we're
blocked
on
is
the
infrared
I
think
we
should.
We
should
actually
buy
storage
just
manual.
We
don't
even
need
to
check
all
this
in
to
get
or
we
don't
need
to
do
the
most
complicated,
for
we
also
don't
need
to
say
that
we
can't
do
Porsche
in
the
future,
but
we
have
a
big
chunk
of
spin
that
we
can
take
off
gcp
and
put
on
Amazon
solve
two
problems.
I
I
We
have
to
move,
spend
and
what
like,
if
we're,
blocked
on
needing
access
to
terraform
or
whatever
we
should
just
just
do
it
like
if
you
have
to
click
buttons,
click
buttons
there
this
we
can
always
come
back
later
and
terraform
it
and
add
Porsche
and
whatever,
but
like
the
minimum
viable
thing
here,
should
I
mean
you
shouldn't
wait
for
the
next
meeting
either.
It
should
really
be
as
soon
as
we
can
get
someone
on
it
so
who
Whoever
has
access
to
that?
C
A
G
Race,
so
I
I
feel
like
we
have.
As
this
group,
we
have
tried
to
like
move
away
from
manual
things
because
they've
bitten
Us
in
the
past.
What
I
propose,
therefore,
is
why
don't
I
apply
the
terraform?
That's
it's
a
it's
one
step
above
manual
like
if
someone's
gonna
play
it
manually.
It
might
as
well
be
me
and
I
might
as
well
use
the
terraform
and,
like
at
least
have
some
hope
of
getting
on
to
like
you
know,
a
maintainable
path.
A
G
I
I
G
I
I
One
point
is
just
whatever
we're
doing.
We
really
don't
need
to
be
worried
about
the
only
thing.
Long
term
here
is
that
when
we
don't
move,
we
keep
we
keep
burning
through
those
credits.
It
is
not
going
to
be
a
long-term
Tech
debt.
If
someone
disagrees
with
the
approach
you
want
to
take
now,
because
it's
going
to
be
super
easy
to
make
to
iterate
on,
like
one
Project's
binary
hosting
on
one
DNS
entry
on
one
crud
HTTP
download
host-
that's
that's
not
gonna,
be
something
that
bites
the
future.
I
A
A
A
G
G
A
A
Okay,
after
I
mean
I'm
gonna
Grant
you
access
to
a
project.
G
A
Okay,
okay,
we
out
we
have
at
time
so
any
last
things.
H
C
H
Yeah,
that's
fine
by
me
as
well,
so
you
probably
want
to
reuse
the
one
of
the
guitar
projects
spin
up
cardboard
on
there.
Yeah.
H
A
Okay,
I
think
we
could
thank
you,
everyone
for
coming
and
see
you
in
the
next
meeting.