►
Description
Kubernetes Public Steering Committee Meeting for 20230306
A
Foreign
welcome
to
the
March
6th
kubernetes
steering
committee
public
meeting
some
house
rules.
This
meeting
is
recorded,
so
please
be
aware
of
what
you
say
and
do
in
this
meeting.
Also,
whenever
we
are
having
conversation,
please
do
abide
by
the
kubernetes
code
of
conduct
committee,
kubernetes
code
of
conduct
and
which
just
essentially
means
like
please
be
next
to
everyone,
and
with
that,
let's
start
the
meeting
before
that
do.
B
A
Thank
you
Tim.
Let
me
just
share
my
screen.
C
A
I
think
the
notes
would
be
also
on
the
screen
and
I
will
share
a
copy
in
our
chat
again,
so
that
people
can
open
up
the
notes.
So
yeah
we'll
go
through
the
agenda.
But
if
you
have
anything
to
add
to
this
agenda,
please
feel
free
to
do
so.
You
know.
C
A
A
First
I
think
Tim
already
has
done
an
update
on
this
kubernetes
steering
and
code
of
conduct
committee
resolution
thing.
Tim.
Do
you
have
anything
that
you
want
to
say
in
a.
A
Thank
you
so
much.
The
next
item
I
think
was
on
Ben.
It
was
about
the
Visa
letters
that
we
might
write
to
support
applicants
for
various
kinds
of
visas,
be
it
for
coupon
or
any
other
reason.
I,
don't
see
men
in
the
meeting,
so
we'll
probably
take
it
like
async.
A
Next
up
is
the
pr
to
split
chairs
and
Deals
I.
Think
Bob,
that's
on
you
do
you
have.
D
Yeah,
the
the
pr
is
out,
it
could
still
use
a
fit
more
Ax
from
steering
before
start
socializing
It
Wide
I
did
try
and
keep
it
very
tightly
scoped
to
just
the
split.
There
are
definitely
some
things
that
need
to
be
done
as
follow-up
items.
D
Honestly,
the
the
Sig
governance
stock
is
kind
of
a
mess
in
terms
of
formatting
and
there's
also
some
things
that
are
very
out
of
date
that
just
no
one
is
doing
like
no
one
is
writing
a
cap
to
create
a
subproject
that
we
should
just
remove
to
or
update
to
make
sure
it's
actually
in
line
with
what
we're
doing
today.
A
Cool
I
missed
another
Point,
which
was
the
subclassic
leads
stuff
that
is
on
me,
I'm,
mostly
going
to
work
on
it
this
week
and
have
it
dropped
ready
for
everyone
to
look
at
there's
an
update
from
my
site
about
annual
reports.
A
Kristoff.
Are
you
on
point
on
this
or.
A
Of
everyone
we
all
are
with
respect
to
getting
information
from
the
states.
So
I
had
a
question
regarding
this,
so
if
any
of
the
Sig
right
now
wants
to
spin
down
what's
the
process,
so
we
have
it
documented
other
than
maybe
asking
approval
from
us
and
spinning
down
the
docs
and
sending
some
comms
to
the
community.
I
was
wondering
like
what
else
do
we
have
if
somebody
wants
to
spin
down
should.
D
A
Didn't
awesome
that
answers
the
question
right
now.
Maybe
we
can
noodle
in
the
docs.
If
we
don't
have
something
we
can
write
some
steps
based
on
the
discussion
that
we
had
right
now
cool.
Are
there
any
questions
on
annual
reports?
I
see
a
lot
of
people
who
are
from
the
general
Community
joining
like.
Do
you
have
any
questions
on
the
annual
reports
that
you
need
to
write
as
part
of
your
Sig
or
working
group.
A
No
questions,
then,
you
can
reach
out
to
your
license
over
various
ways
like
email
or
reach
out
to
them
on
Slack.
That's
the
best
way
to
reach
awesome.
I,
don't
think
we
have
any
more
follow-up
items.
So,
let's
hop
on
to
the
next
one,
which
is
kubernetes
at
the
open,
infra
Summit.
G
Hello,
sorry,
so
this
is
I
added
this,
so
I
brought
it
up
in
the
state
contributor
experience
meeting
last
week,
but
basically
some
of
you
might
be
familiar
with
the
the
project
team
Gathering
that
we
have
been
having
virtually
every
six-ish
months
within
the
open
info
Community
a
lot
of
openstack
projects.
G
Services
gather
there
to
talk
about
technical
things,
kind
of
plan
for
the
next
release
and
focus
on
integration
with
other
projects,
and
we
will
actually
be
having
our
first
in-person
one
during
the
open
info
Summit,
which
is
June
13th
through
15th
and
I,
wanted
to
make
sure
that
the
kubernetes
community
was
aware
that
you
are
definitely
invited.
G
We
would
just
want
to
know
by
I
think
I
said
April
2nd
if
you're
interested
in
participating-
and
all
of
that
is
outlined
in
the
issue
that
I
made
and
I
also
sent
the
issue
to
the
mailing
list
and
I
also
dropped
it
in
provider.
Openstack
slack
channel
so
just
wanted
to
raise
it
with
you
all
as
well.
We
will
be
having
a
virtual
one
before
then,
but
I
think
it
is
it's
the
very
last
week
of
March
actually
holy
moly.
G
That's
like
coming
out
very
quick,
but
we
would
really
love
to
see
you
at
the
in-person
one
if
you're
interested.
So
if
you
have
any
questions,
please
feel
free
to
reach
out
to
me.
A
Thank
you
Brendan
for
that
update.
E
B
A
Cool
I
think
we
can
have
the
open
discussion
item
or
no.
H
Okay,
so
I
think
James
sent
an
email
about
getting
clarification
regarding
content
hosts
in
documentary
infrastructure.
So
I'm
curious.
If
there's
like
some
conversation
happening
about
this
or
early
feedback.
H
D
I
think
this
honestly,
like
we
probably
need
to
look
closer
at
the
charter
and
make
sure
it's
explicitly
scoped
in
there
to
like
I
I
I
have
not
looked
at
it.
I'm!
Sorry,
it's
there
I,
don't
really
have
a
good
excuse
for
it,
but
that
would
be
where
I
think,
like
the
refinement
of
that
needs
to
happen.
E
So
let
me
let's
try
to
dig
up
this,
this
email,
because
it's
a
worthwhile
discussion
and
I
think
you
know
the
the
the
question
around:
where
are
the
boundaries
for
infrastructure
and
and
who
can
make
decisions?
I
think
it
does.
You
know,
at
least
to
date,
has
been
you
know,
a
steering
conversation,
and
perhaps
we
can.
We
can
look
at
what
it
means
to
loosen
the
reins
a
little
bit.
H
I
B
Still
just
back
up
on
the
charter,
like
the
Tim's
email
like
when
I,
when
I
read
it
back,
then
it
was
sort
of
it
felt
like
condition
on.
If
you
think
there's
a
problem
is
I
I
glance
at
the
charter
back
then,
and
glancing
at
it
now
I
feel
like
it
can
cover
the
need.
But
it
is
somebody
worried
about
a
specific
thing
like
we.
We
need
a
specific
sentence
about
a
specific
thing
in
it,
because
it
would
maybe
be
easier
if
we
were
more
concrete.
If
there's
a
directory.
H
Yeah
I
think
the
the
issue
we
got
specific.
We
we
freeze
out
guest.gci.io
I,
think
it's
about
to
happen
next
month
and
one
possibility
is
to
start
clean
up
that
old
registry.
By
deleting
some
specific
images,
I
will
pick,
for
example,
the
Azure
CSI
drivers,
I,
would
say,
for
example,
those
drivers
are
essential
to
ETS
cluster
and
by
our
application
essential
to
Azure
customers.
H
E
So
I
would
I
would
say
that
you
know
I
think
that
the
content
that
you
know
the
the
service,
the
the
service
that
Kate's
infra
provides,
has
customers
and
the
customers
are
effectively
responsible
for
their
content.
So
you
know
I
I.
Ideally
the
the
conversations
that
are
happening
about
images
getting
deleted,
content
being
deleted
retention
policies
being
set
should
should
have
should
be
a
broader
discussion
across
all
of
those
customers
as
well.
D
D
B
The
example
there
like
eks
is
a
commercial
offering
that
Amazon
supports.
B
I
B
I
Not
pulling
this
directly
EPS
isn't
pulling
this
directly,
but
there
are
users
running
clusters
on
Amazon
that
are
using
the
AWS
EBS
CSI
driver,
which
is
one
of
the
Apes,
is
listed.
This
isn't
the
only
example,
so
more
generally
I
think
sick
hates
them
for
a.
We
need
some
understanding
to
ask
for
understanding
as
we're
looking
at
not
being
able
to
stay
within
budget
this
year,
the
Sig
is
discussing
ways
that
we
can
cut
costs
so
far.
I
I
So
as
an
example,
if
we
step
back
from
specific
images,
if
we
reach
a
point
in
the
year
where
the
like
the
obvious
move,
is
we
can't
afford
to
run
everything
the
whole
Project's
going
to
shut
down
we'll
shut
down
the
old
registry
instead,
at
that
point,
maybe
it's
infeasible
to
actually
go
to
like
every
Sig
and
say:
hey.
Your
images
are
going
away.
How
do
you
feel
about
that?
I
Everything
is
going
away.
Can
the
Sig
do
that
to
need
like
steering
sign
off?
How
does
that
work
or
like
in
a
recent
case,
we've
identified
that,
while
we'd
like
to
support
scale
testing
it's
one
of
the
few
obvious
cost
levers
we
have
where
we
can
get
a
big
chunk
of
money
back
if
you
like,
turn
off
or
reduce
the
frequency
of
the
scale
tests.
I
That's
like
a
single
button.
You
can
press
and
get
a
ton
of
funding
back,
not
enough,
but
a
lot.
So
thus
far
with
my
testing
hat
I've
been
like
voluntarily
getting
scalability
to
go
through
their
tests
and
determine
what
we
can
reduce,
but
if
that's
insufficient
since
provides
the
service,
can
they
just
like
turn
them
off.
H
I
Shutting
off
you
know
who
who
decides
like
if
scalability
hadn't
voluntarily
reduced
their
testing
and
we're
still
trying
to
get
the
the
cost
down
like
Kincaid
SIM
for
just
turn
it
off
or
do
they
need
steering
side
off
or
how
does
that
work?.
I
We're
look:
the
Sig
is
looking
at
specific
images
as
a
a
less
damaging
solution,
but
we
could
very
well
wind
up
at
a
point
later
this
year,
where
you
know
just
reducing
specific
things
isn't
sufficient.
If
the
sort
of
thing
doesn't
happen,
I
don't
think
it's
super
clear
right
now
how
that
decision
gets
made.
I
I
E
This
is
a
great
question.
I
just
want
to
acknowledge
that
I
think
I
think
it
doesn't
have
an
easy
answer.
As
a
and
I
think
you
know
for
speaking
as
a
a
member
who
is,
is
also
a
a
lead
for
a
sake.
I
I
would
not
be
excited
about
that
without
some.
Some
form
of
discussion
and
I
know
that
we're
not
implying
that
there
would
be
no
discussion
or
we're
talking
about
who
ultimately
holds
the
the
the
button
or
the
the
pen
for
it.
Tim
I'm
I'm
gonna
ramble.
B
I
think,
instead
of
me
thinking
that
I
can
read
it
in
various
parts,
probably
what
we
need.
Oh
that
wasn't
the
link
in
I'll
come
back
here
in
the
charter.
There's
a
artifact
hosting
section,
I!
Think
what
we're
basically
saying
is.
We
need
to
expand
that
section
to
have
a
retention
policy
within
scope
and
may
be
within
the
process
of
putting
that
within
scope
in
the
charter
we
establish.
C
I
Like
I
think
it's
more
General
than
that
like
and
we
keep
like
focusing
on
like
specific
content,
but
if,
like
so,
let's
use
an
example,
that's
similar
in
the
sigs
with
sub-projects.
The
TLs
are
in
theory
for
six
I.
Have
cherry
and
teal
split.
I
I
That
could
be
things
like
scalability
testing
services
or
you
know
really
any
other
infrastructure
that
we're
paying
for
and
the
question
at
hand
that
we
should
also
visit
is
that
there
is
very
current
discussion
of
deleting
very
specific
content
because
it
has
a
higher
like
risk
reward
ratio
than
turning
off
whole
services.
But
I
don't
think
either
of
those
are
clear
and
I
would
think
that
the
turning
off
specific
content
would
be
a
subset
of
the
who
can
shut
off
services.
And
how
does
that
get
approved?
I
And
of
course,
in
all
these
cases
we
would
expect
some
discussion
with
participants
and
and
letting
people
know,
but
someone
has
to
be
able
to
decide
because
I
don't
think,
there's
ever
not
going
to
be
controversy
when
you
turn
things
off
I'm
relatively
happy
with
the
scale
example
where
they
voluntarily
are
going
through
and
and
cutting
things
with
us.
But
you
know,
I
can
imagine
other
cases
where
you're
like
I,
don't
don't
eat
myself
and
hates
Emperors
like,
but
we
have
to
fund
things
somehow
and
who
decides.
D
One
thing
I
do
want
to
just
call
it
for
this
specific
situation
too,
removing
stuff
from
Keats-
or
you
know,
case.gcr.io.
D
D
I
I
E
H
H
E
I
would
say
that
you
know
some
of
these
points
are
hit,
and
you
know
again
to
to
your
point
that
project
infrastructure,
budget
component
of
the
of
the
charter
I
think
it's
worth
digging
into
the
specific
points
and
adding
some
specificity
around
making
sure
that
we're
reaching
out
to
stakeholder
groups.
I
I
know
that
that's
implicit
and
and
you're
in.
E
H
A
I
So
I
want
to
point
out
that
the
like,
maybe
a
little
bit
more
context
on
the
specific
discussion,
though
some
analysis
done.
That
shows
that
there
are
a
handful
of
images
that
compromise
the
majority
of
the
bandwidth.
And
if
we
look
at
these
images,
you
know
no
current
references
are
to
this
old
location.
We've
been
trying
to
identify
ways
that
we
can
head
off
the
project
shutting
down
entirely
while
minimizing
damage.
I
So
the
Sig
has
pulled
almost
every
other
obvious
lever,
short
of
something
like
shutting
down
the
old
registry,
entirely,
there's
a
list
of
a
little
over
a
dozen
images
that
are
in
enormous
portion
of
the
bandwidth
there's
a
long
tail
of
something
like
24,
where
none
of
the
images
are
individually
like
multiple
percent,
but
there's
a
number
of
images
that
are
as
large
as
something
like
15
of
bandwidth.
I
So
the
current
discussion
is
exclude
core
images
that
are
used
to
actually
bootstrap
nodes
to,
because
that's
particularly
difficult
for
users
to
repair
and
include
the
other
top
images
that
leaves
you
with
about
11
images
that
are
something
like
59
of
the
traffic.
If
you
cut
those
images
out
they're
still
available
at
the
new
host,
we
may
have
enormous
savings
that
avoid
having
to
cut
anything
else,
but
there's
not
precedent
and
just
yanking
these
out,
even
if
they
are
available
somewhere
else.
I
We're
also
looking
at
other
avenues,
but
that
is
a
very
real
discussion
and
that's
the
like
more
specific,
where
that's
headed
I
do
think
that
we
should
be
reaching
out
to
the
owners
of
these
specific
images.
But
it
isn't
clear
if
it
needs
to
like
block
on
their
permission
or
steering.
F
Is
sick
hates
in
for
asking
for
the
power
to
do
this,
like
what
what
specific
governance
issue
are.
We
are
we
talking
about.
I,
don't
want
to
you
know,
I
know,
there's
a
lot
of
technical
complexity
here,
I
don't
want
to
bring
that
into
a
governance.
Discussion.
I
specifically
am
trying
to
understand
like
what
governance
changes.
Do
we
need
to
make
to
give
someone
the
right
authority
to
be
able
to
make
these
kind
of
decisions
I.
I
I
I
would
also
say
have
like
skimming
over
the
charter
again,
that
it
looks
to
me
like.
Maybe
it
isn't
called
out
for
things
that
aren't
artifact
hosting
and
that's
probably
like
later
in
the
year
that
may
be
relevant
depending
on
how
well
these
other
things
go.
F
Okay,
so
there
there
might
be
some
specific
Charter
changes
to
Sig
Kates
and
fraud
in
order
to
specifically
call
out
that
say.
Kate's
infra
has
the
authority
to
end
up
making
these
kind
of
recommendations,
making
these
kind
of
decisions
and
that
kind
of
stuff,
in
consultation
with
the
sigs
that
are
actually
deploying
these
artifacts.
I
Yeah
also
I,
guess
I'd
point
out
that
we
don't
like
well,
we
sort
of
have
like
laws.
We
don't
really
have
like
a
court.
This
group
seems
to
sort
of
double,
as
both
so
I
think.
People
want
to
be
sure
this
time
that
they're
reading
this
right
and
that
and
that
either
Kate
Jennifer
can
do
this
or
if
they
need
to
be
doing
something
else.
I
What
do
they
need
to
be
doing
so
they
can
get
on
that,
because
another
thing
here
is
any
of
these
moves
that
we're
making
that
are
like
of
lower
order,
drastic,
like
cutting
back
scale
tests.
They
have
to
happen
as
soon
as
possible
because
we
need
cumulative
savings.
F
You
know
we
guess
we
should
go
and
clarify
in
the
charter,
but
my
current
interpretation
and
reading
of
it
is
that
stick
Kate's
in
for
a
can
make
those
decisions.
Does
anybody
from
steering
on
the
call
right
now
have
any
objections
to
that,
because
what
I
would
might
maybe
suggest
two
sick
Kate's
infra
is
like
start
operating
like
you,
do
have
this
Authority
and
then
we
can
go
in
the
so
that
we
can
do
these
kind
of
things
in
parallel,
so
that
you
can
start
whatever
processes.
F
Whatever
recommendations
you'd
want
to
make
as
far
as
shutting
those
things
down
and
like
engaging
with
the
other
stakeholder
groups
as
needed,
and
we
can,
in
parallel
work
on
the
charter,
changes
to
clarify
that
in
language.
E
Yeah,
so
I
I
think
I
would
but
mostly
agree
there
and
and
that
I
I
think
I
feel
a
good
next
step
is
crafting
such
policy.
It's
dictated
in
the
project
infrastructure
budget,
section,
yeah,
I'm,.
H
H
So
even
the
job
we're
running
like
can
I
have
the
power
to
move
some
job
from
one
Club
provider
to
another
one
because
I
feel
like
this
is
helping
saving
budget,
because
in
the
in
James
email
whoa,
there
was
like
twitching
call
out
moving
to
a
copywriter
dealing
with
artifacts
and
also
maybe
get
rid
of
specific
introduce
change
in
the
code
base
that
help
ensure
we
meant,
and
we
had
the
infrastructure
under
budget.
I
There's
a
little
bit
tone
back
there.
I
would
say
that
you
kept
the
power
to
theme
help
budget
is
used,
but
like
same
how
like
like
job
scheduling,
is
something
that
testing.
Doesn't.
You
can
tell
sick
testing
that
this
is
too
expensive
or,
and
you
know,
needs
to
be
migrated
or
something,
but
even
from
a
suggesting
perspective,
we
don't
just
we
typically
don't
unilaterally
change
jobs.
It's
still
going
to
involve
the
the
owning
Sig.
I
We
can
shut
things
off
because
they're
too
expensive.
We
can
suggest
that
another
cloud
is
available
as
an
alternative
I,
don't
think
we
can,
if
it's
not
directly
equivalent
I,
don't
think
just
move
it.
It
depends
on
what
you
mean
by
migrating,
it
so
like,
if
we're
just
talking
about
the
execution
environment
and
that's
fine,
but
if
we're
talking
about
like
switching
into
in
test
providers
or
something
that's
that's
not
the
same
thing
also
before
we
totally
switch
that
so
back
on
that
does
anyone
have
objections?
I
I
have
one
miter,
one
Kristoff.
The
only
problem,
I
see
with
just
totally
doing
this
in
parallel
and
not
figuring
it
out
now.
Is
that
the
one
of
the
very
obvious
early
steps
is
going
to
be
announcing
towards
end
users
and
I?
Think
that
has
the
potential
for
reputational
impact
on
the
project
once
we
declare
that
we're
going
to
remove
these
images
from
people
that
damage
is
done,
and
you
can't
you
can't
just
walk
that
back
so
I
think
we
should.
I
Everyone
should
feel
that
that
is
fine
now
and
not
figure
that
out
later,
because
it
was
going
to
be
a
very
early
step
in
the
process.
I
think.
F
I
would
actually
put
it
the
opposite
like
it
should
be
a
late
step
in
the
process.
It
should
be
one
of
the
actual
final
steps
in
the
process
before
actually
doing
the
pulling,
because
you
want
to
have
confidence
in
what
the
plan
is
before
you
announce
the
plan
publicly.
So
all
the
pre-planning
steps,
as
far
as
identifying
which
images
are
problematic,
going
and
engaging
with
the
stakeholders
and
the
sigs
to
find
out
impact.
All
of
that
kind
of
stuff
should
be
happening
before
there's
any
public
announcements.
I
I
agree,
but
you
have
to
make
sure
that
that's
what's
actually
happening
and
I've
seen
conversation.
The
other
way
that
we
should
like
announce
now,
and
that
is
my
biggest
concern
and
the
reason
that
I
would
want
other
folks
weighing
it
and
I
think
that
sort
of
announcement
is
going
to
look
bad.
Even
if
we
have
to
do
it.
And
you
know
it
affects
the
reputation
of
the
project.
F
F
Before
we
go
making
public
announcements
like
that,
you
know:
we've
we've
got,
we've
got
folks
who
are
very
heavily
involved
in
cicades
infra
here
on
the
call
today
and
I
would
say
to
you.
Let's
make
sure
that
we
have
that
plan
before
we're
announcing
publicly
like
we
again.
We
we,
as
a
project,
have
learned
that
multiple
times
in
our
history
that,
like
it's
better
to
have
a
a
very
firm
plan
internally
before
we
go,
engage
externally.
E
So
I
think
that
you
know
I,
agree
and
disagree
in
both
directions.
I
think
it's
like
a
matter
of
semantics
for
the
project.
We
tout
that
our
Communications
are
public.
This
meeting
is
public.
This
conversation
is
already
happening.
This
conversation
can
already
be
seen
on
slack
threads
in
various
places.
E
I
will
go
back
to
what
is
stated
in
the
the
project
infrastructure
budget
section
of
the
charter
crafting
such
policy
in
is
done
in
collaboration
with
steering
committee
owns,
project
spending
and
Sig
architecture
owns
kubernetes
definition,
not
sure
I
agree
with
the
Sig
architecture
component,
but
I
I
do
think
a
a
policy
needs
to
come
forth
that
we
can
review
as
a
community.
E
There
will
be
some
sort.
There
will
be
some
combination
of
stakeholders
across
steering
across
case
infra
across
other
affected
parties,
but
I
think
we're
saying
we
need
to
bring
the
plan.
There's
no
I,
don't
see
there
being
boundaries
between
internal
and
external
communication,
since
we,
the
the
boundaries
are
fuzzy
in
general.
I
But
there
is
a
difference
between
us
discussing
these
things
versus
US,
announcing
that
we're
going
forward
with
a
change.
For
example,
there's
been
some
discussion
of
what,
if
we
just
announced
that
you
know
some
things
will
be
removed.
My
personal
take
is
that
if
we
can
do
a
much
more
concrete
list,
we
should
narrow
down
that
list
first
and
and
be
specific
and
and
clarify
to
people
that
we're
not
taking
away
everything.
I
But
that's
me
and
not
everyone
I,
think
that
this
is
the
part
where
I
think
you
know.
Maybe
if
we're
saying
that,
if
we
say
sounds
like
you're
just
right
off
the
charter
says
it's
in
conjunction
with
steering,
that's
where
I'd
say
set
the
plan
bring
it
to
us
and
let's
confirm
before
we
announce
I,
also
wonder
if
we
might
want
to
put
something
in
there
because
of
the
damage
we've
had
with
cons
before
about
engaging
whatever
our
most
appropriate
comms
group
is
maybe
contrabx
yeah
contributor.
I
I
It's
also
worth
pointing
out
that
if
this
stuff
doesn't
go
well,
we
I
almost
certainly
will
be
revisiting
things
like
wholesale
turning
off
the
service,
as
opposed
to
like
specific
deletion
or
we
may
be
visiting
okay,
the
scale
team
voluntarily
reduced
things,
but
we
need
more
overhead
and
that's
where
it
starts
to
get
less
clear.
I
Yeah
like
like
who
exactly
owns
the
scheduling,
whatever
I
would
say
that
Kate's
Empress
is
the
one
holding
the
budget,
if
not
Kate's
him
for
end
steering
and
that
that
needs
to
have
room
to
include
things
like
you
can
no
longer
make
this
spend
by
whatever
means
that
is
enacted.
E
So
I
think
there's
also
a
more
money
related
discussion
around
what
is
in
your
budget,
what
is
projected
spend
from
the
Sig
versus
what
is
what
we
can
give
to
the
Sig
or
what
somebody
can
give
to
the
say
great.
So
when
we're
saying
you
know,
when
you
know,
if
we
look
at
last
year
right,
someone
dived
in
and
saved
us
right,
basically,
is
that
technically
part
of
your
budget
is
that
technically
part
of
K10
for
his
budget.
B
E
I
think
what
I'm
saying
is
is
if
there
is
a
if
there
is
a
specific
amount
of
money
to
be
given
in
credits
or.
However,
however,
it's
it's,
it's
going
to
be
dispositioned
to
a
to
a
Sig
for
for
a
project
and
the
project
happening
to
be
project
infrastructure.
In
this
case
right.
Should
they
go
past
it?
What
happens
then.
I
Oh
I
see
I
think
right,
that's
a
good
conversation,
but
that's
maybe
more
of
a
next
year.
Conversation
so
taking
scale
is
a
great
example.
Under
normal
circumstances,
my
personal
argument
would
be
if
I
were
laying
out
our
budget
from
scratch.
I
would
absolutely
allocate
approximately
this
much
towards
scale
testing
with
no
qualms.
It's
the
sort
of
thing
that
we
can't
get
off
the
shelf.
That
has
benefit
to
the
quality
of
the
project.
You
know
not
just
for
large
clusters.
I
We
can
see
performance
which
scales
down,
but
this
year
we're
headed
towards
the
cliff
and
in
the
fullness
of
time
the
project
has
to
make
changes
that
don't
have
the
most
drastic
impact
and
to
stay
within
the
budget
and
that
may
not
necessarily
map
to,
but
we
think
your
Sig
is
being
unreasonable.
It's
more
like
your
Sig
happens
to
have
something
that
I
can
in
two
seconds
cut
the
cost
on,
and
we
have
to
find
somewhere
to
cut
cost
and
we
can't
spend
all
year
doing
it
if
it
were
not.
I
For
that
case,
then,
I
don't
think
that's
how
we're
approaching
it
I,
don't
think
we're
saying
that
like
scale
has
disproportionate
cost
or
the
like
image.
Well
image
hosting
does
Sig
testings
had
some
things
that
we've
cut
where
maybe
we
have,
but
it
isn't
necessarily
the
case
that
we're
cutting
things
because
the
it's
just
disproportionate
we're
cutting
things
because
we
can,
if
we
have
to
cut
something
and
the
expediency,
is
a
real
problem.
I
I
think
there's
a
lot
of
money
wasted
in
cigs
having
CI
they're,
not
monitoring,
but
it's
much
harder
to
say
how
much
has
that
total
money?
And
could
you
run
around
and
find
all
those
things
and
cut
them
than
it
is
to
take
a
look
at
scale
tests
and
say?
Well,
there
is
a
big
button
and
cut
it
if
we
can
get
to
a
place
where
the
budget's
more
healthy,
then
I
think
we
should
have
better
policies
around
how
it's
apportioned
right
now,
I
think
we're
kind
of
more
in
an
emergency.
H
D
Don't
think
so
because
that's
really
hard
to
gauge,
like
I,
think
it's
good
to
have
that
the
information,
if
we
can
we
if
we
can
get
it
but
I,
don't
think
we
necessarily
want
to
set
hard
budgets
for
you
know
every
Sig,
it's
like.
E
No
no
I
mean
I,
mean
I,
mean
Sig
capes
in
for
a
in
the
general
case.
Right.
If
we
looked
at
this
in
the
general
case,
it
would
I.
I
would
imagine
this
is
under
steering's
purview
right
to
go
after
so
again,
I'm
gonna
go
back
to
I
think
we
need
a
policy
specifically
for
the
interest
stuff
and-
and
we
should
sit
down
together
and
review
the
policy
and
enact
it.
I
B
I
guess
when,
when
I,
when
I
talk
about
that,
it's
like
it's
not
as
simple
as
calling
it
just
waste
like
I,
think
that
those
those
end
users
are
in
places
and
we
as
a
project,
need
to
interact
with
those
places
to
encourage
patterns
of
local
usage
as
opposed
to
just
sort
of
this
magic
free
stuff.
Up
in
the
the
open
source,
Community
Cloud.
H
F
So
one
sorry
I'm
just
trying
to
think
how
to
word
this
right
now,
like
we've
identified,
we
have
an
issue
here.
We
know
we
need
to
take
some
some
steps
on
this
issue
from
a
governance
standpoint.
F
I
feel
like
it
is
in
Sig
Kate's
infra's
Court
to
do
this
like
the
way
I'm
currently
reading.
This
I
would
expect
of
all
the
different
sigs
that
we
have
in
the
project
that
this
is
a
responsibility
that
would
fall
to
Sig
Kate's
infra,
to
you
know,
research.
The
options
you
know
come
up
with
Solutions
and
ultimately,
like
be
the
authority
that
makes
the
decision
in
this
kind
of
case,
if
that's
not
true
or
if
there's
Charter
language
that
needs
to
be
updated.
F
It
keeps
coming
up
and
steering,
and
then
it
doesn't
get
actioned
and
it's
personally
I'm
trying
to
I
I'm
feeling
a
little
bit
frustrated
about
it,
because
it's
like
what
more
do
we
need
to
actually
make
decisions
on
this
and
someone
to
take
leadership
on
this
particular
issue
as
opposed
to
again
swirling
around
and
around
around,
like
what
are
the
actual
next
steps
that
we
need
to
make
happen
to
delegate
the
authority
needed
to
somebody
to
make
these
decisions.
I
I
think
we've
clarified
that
in
that
call
and
I
want
to
point
out
that
the
past
discussions
that
happened
stop
happening
because
we
had
a
huge
injection
of
emergency
funds
and
we
expected
uses
to
Trail
off
and
then
early
this
year
we
found
that
the
opposite
happened,
which
is
why
we're
back
where
we
are.
It's
unfortunate
and
I
think
that
we've
I
think
we've
basically
hammered
throughout
this
whole.
I
A
I
think
both
of
those
are
doable,
and
one
thing
that
we
can
do
going
forward
is,
if
need
to
be,
we
can
have
a
joint
steering
and
Catering
for
a
meeting
if
we
need
to
lecture
record
the
action
items
from
each
of
the
groups
so
that
we
make
progress
and
if
we
just
write
things,
what
is
expected
from
each
of
the
groups
for
from
what
I
hear
right
now
is
we're
expecting
cares
infra
to
write
up
that
policy
if
it
is
not
yet
clear
and
then
sharing
will
send
it
off.
Is
that
everyone?
E
A
Good
I
think
that's
a
very
important
and
maybe,
if
needed,
we
don't
need
it
for
next
month's
meeting.
If
we
need
to
talk
about
it
and
get
things
rolling,
you
should
have
a
call
even
before
that.
E
To
Ben's
point
this
is
you
know
it
will
continue
to
crush
us
the
longer
we
wait
to
do
something
about
it.
So
yeah
as
as
soon
as
as
soon
as
we
have
policy
in
hand,
I
think
we
should.
We
should
review
it
async
as
much
as
possible
and
meet
if
we
need
to.
D
Yeah
next
month
is
also
kubecon
and
I.
Think
that
clashes
with
one
of
our
meetings,
how.