►
From YouTube: 20200210 sig cluster lifecycle
Description
sig cluster lifecycle meeting for march 10th 2020
A
A
It's
a
good
question:
I
think
it
depends
upon
PR
because
usually
these
things
go
along
with
PR.
So
if
we're
doing
this
in
the
kubernetes
blog
for
awareness
that
might
be
beneficial
to
still
get
that
signal
out
there.
If
we're
not
doing
a
major
PR
push,
then
I
would
say,
wait
and
hold
because
usually
there's
a
lot
of
peer
that
goes
around
the
events.
B
B
You
know
we
have
to
also
think
about
the
questions,
because
we
didn't
collect
that
many
yet
I
also
invite
cops
and
mini
cube
with
the
other
projects
to
include
questions,
because
we
only
have
quasi
idea,
but
for
the
time
being
so,
if
you
have
any
specific
questions
for
the
projects
like
hey,
is
this
design
that
we
applied
last
releases
doing
work
for
you
or
not?
You
know,
project
specifics.
C
I
guess
it's
not
really
related
to
you
know
when,
when
we're
gonna
release
the
the
server
but
I
I
am
wondering
how
how
we're
going
to
share
the
results.
Cuz
I
I
from
from
from
my
experience
right
Simon.
As
you
view,
you
answer
a
survey
and
then
it
sort
of,
like
goodness,
the
results.
I,
don't
know.
Sometimes
you
forget
about
them
or
they
they
kind
of
disappear.
D
A
Time
we
actually
did
a
prison
or
live
earlier,
did
a
presentation
to
both
the
COO
medium
survey.
We
did
a
presentation
both
to
the
Canadian
group,
as
well
as
the
broader
sing
cluster
lifecycle,
I
think,
because
this
one
is
more
broad,
I,
think
presenting
to
this
group
the
results
and
the
details.
The
survey
for
us
to
be
able
to
take
a
look
at
should
probably
be
asked
for
post
survey.
Oh
I.
B
C
C
B
A
I
think
the
problem
is
we'd
have
to
use
a
third-party
service,
then
to
get
that
data
back
because
there's
all
kinds
of
like
privacy
constraints,
especially
with
gdpr.
So
last
time
we
just
use
like
Google
sheets-
and
you
know,
I,
don't
I
think
the
data
was
not.
There
was
no
specific
data
about
the
individual
person
submitting,
but
if
we
used
something
like
Survey
Monkey
to
do
our
surveys,
I
think
we
could
probably
be
able
to
post
back,
but
we
should
probably
coordinate
with
pitcher
of
X
to
verify.
That's
a
good
question.
F
Really
like
that
idea,
particularly
if
we
actually
try
to
tie
work
work,
we
are
going
to
do
to
it.
Like
you
know
that
will
say
you
know,
so
the
survey
showed
us
this
server
e
says.
Windows
is
super
important,
so
we're
all
going
to
like
go
for
Windows
or
whatever
it
is.
You
know
whatever
that
data
tells
us
to
do.
A
B
I'm
going
to
join
the
the
cig
meeting
organized
by
country,
banks
from
Thursday
I
think
the
topic
is
going
to
be
issue
triage,
but
if
there
are
topics
around
surveys,
I
can
also
discuss
those,
but
we
are
going
to
use
our
own
setup
for
the
survey.
We
are
not
going
to
use
Survey
Monkey
in
this
case,
but
yeah.
Maybe
some
of
this
process
like
you're,
saying
Tim,
it
should
be
documented
somewhere,
yeah.
A
A
B
This
is
just
a
quick
update
what
we
have
been
working
on
this
cycle
in
terms
of
the
more
interesting
items
we
basically
started
working
on
some
pooling
to
facilitate
the
move
of
qadian
to
the
new
repository
to
not
the
existing
issue,
tracker
poster
that
pick
up.
So
this
is
taking
a
lot
of
my
time
currently,
but
I
already
have
some
good
progress
in
terms
of
automating
tagging,
branching
him,
basically
maintaining
the
same
version
as
cavaties,
Kuwaitis
and
also
I,
have
to
update
the
cap
and
modify
sig
release
for
another
review
on
this
cap.
B
The
other
item
is
that
we
are
working
on
Windows
support
and
it's
hopefully
going
to
graduate
the
bit
that
this
cycle.
In
fact
the
chances
are
pretty
high.
At
this
point,
the
dogs
are
ready
to
mulch
and
also
we
are
pretty
much
already
testing
our
and
20s
syrup
for
Windows
incubated
with
the
Boeing
windows
/
linux
hostel
using
canadian
and
I
wanted
to
modify
especially
the
Costa
Rica
folks
that
we
have
a
look
at
the
existing
setups
for
TCP
and
AWS
and
banned
most
from
people.
B
X
people
who
now
works
for
VMware
expressed
that
he
can
do
much
simpler
if
he
writes
a
simple
bash
script
that
the
poison
GCP.
So
we
went
with
that
solution,
so
we
basically
wrote
our
own
infrastructure
deploy
our
premise,
which
is
surprisingly
only
something
like
200
lines
of
mash
compared
to
what
we
have
in
some
of
the
copy
providers,
which
is
much
larger
than
that.
So
maybe
in
the
future
we
can
transition
this.
B
F
G
Hi-Yah,
so
for
the
add-ons
project
we
submitted
for
the
google
Summer
of
Code
and
a
several
students
have
picked
up
on
that
and
are
quite
excited
about
it.
But
we
are
lacking
immediate,
well
triaged
work
that
they
are
asking
for.
So
we
are
a
bit
unexpected
unprepared
for
that,
because
it's
normally
a
summer
timeline,
so
just
trying
to
get
those
folks
connected
up
since
they're
asking
for
it,
and
then
we
have
a
planned
discussion
for
using
add-on,
installer
configuration
potentially
in
cops
since
Justin
and
folks
have
already
put
together.
G
What's
already
there
and
then
this
morning
for
component
standards,
we
did
a
project
management
review.
We've
been
lacking
p.m.
help,
and
so
you've
been
just
figuring
out
where
things
were
was
very
helpful,
appreciate
everyone
joining
and
just
guessing
through
things,
we
also
held
a
design
discussion
on
two
topics:
one
was
OB
Tech's
work
on
llamo,
v3
versus
v2
performance
and
a
collection
of
documents
separated
blobs,
and
then
there
was
the
animal
thing
and
then
there
was
also
quite
long
discussion
of
component
config,
with
instance,
specific
stuff
and
design
issues
in
cuba,
DM
and
we'll
post.
G
Basically,
you
know
up
and
coming
perhaps
in
the
next
release-
are
some
UX
changes
for
component
config
consumers
just
to
change
the
way
that
we're
kind
of
using
stuff
so
go
ahead
and
check
the
notes
there.
If
you
want
to
be
up
to
date
with
where
work
is
moving
for
component
standard,
any
questions
with
regard
to
add-on
or
component
centered
work.
I
do.
A
A
I
think
it
might
be
beneficial
to
this
to
maybe
go
through
a
state
of
the
state
of
where
addons
is
at
and
where
you
need
help
in
what
the
vision
is
for
the
next
release,
because
I
think
we've
gotten
to
the
point
now
where
we
have
enough
eggs
to
actually
drive
execution.
That's
always
a
useful
thing.
That's
sometimes
that's
enough
to
get
the
ball
rolling.
We've
seen
this
on
the
horizon
for
a
long
time,
but
I
think
the
frustration
has
reached
a
pinnacle,
so
they
did.
G
G
A
G
A
That
way,
we
can
actually
try
to
see
because
I
guarantee
we'll
probably
have
more
people
in
the
next
cycle,
to
try
and
address
some
of
the
backlog
items
and
start
to
unburden
on
some
of
those
things.
So
as
we
start
to,
as
we
start
to
pay
down
the
debt,
both
in
qadian
and
in
close
drape,
you
can
try
to
maybe
unite
the
ideas
together.
So.
B
B
G
G
Like
what
the
selection
process
was
like
and
if
this
was
going
to
be
a
many
to
one
or
many
to
many
situation
for
the
mentor
to
students,
ratio
and
the
yeah,
the
selection
process
has
not
begun
for
Summer
of
Code
right,
but
these
these
folks
have
identified
themselves
and
followed
all
of
our
community
channels
to
get
to
the
right
place
to
where
they
want
to
do
contribution
by
trying
to
probably
get
ahead
of
the
ball
yeah.
So
that's
a
yeah
these
these
folks
now
see
that
the
project
is
listed,
they've
been
in
they're
interested.
G
B
My
recommendation
there
is
going
to
be
to
always
ask,
but
what
about
probe
the
candidates,
the
students
in
terms
of
what
is
their
existing
knowledge
he,
for
instance,
if
somebody's
already
an
expert
in
networking,
you
can
try
assigning
him
to
something
related
to
a
Q
proxy
add-on
or
something
like
that
I
below.
Basically,
you
have
to
try
to
understand
what
they
know
already
the
students,
because
otherwise
you
might
be
throwing
resource
at
tasks
that
people
are
not
going
to
complete.
G
G
Yeah
it's
the
timeline
was
was
rather
generous
when
we
decided
to,
even
though
it
was
kind
of
a
last-minute
application
from
the
admin
side
of
things.
There's
a
lot
of
time
to
that.
But
as
soon
as
you
know,
everything
that
published
we
now
have
folks
who
are
joining
calls
asking
to
meet
people
asking
for
work.
So.
B
C
A
couple
minor
updates:
there's
a
certain
renewal
command.
Pr,
that's
going
to
merge
soon,
I,
don't
know
if
anybody
wants
to
take
a
look.
I.
Think.
Last
time
we
met
I
mentioned
that
there
were
some
unknowns
around
kuba
API
server
needing
to
be
restarted
once
once
the
certs
on
disks
were
were
changed,
so
I
don't
know
if
anybody
has
any
any
any
ideas
anything
they
want
to
share
about
that
on
the
pier
that
would
be
helpful.
D
C
Not
it's
not
a
blocker
for
the
PR,
so
that
should
merge
soon
and
the
other
I
think
some
of
it
would
be
nice
to
have
happen
in
the
short
term
would
be
a
binary
artifact
for
the
github
releases.
If
anybody
can
can
help-
or
you
know,
if
there's
if
there's
something
very
simple-
that
you
know
that
I
that
I
could
that
I
could
do
this
to
make
that
happen,
I
mean
I've.
C
F
Don't
know
that's
something:
I
can
definitely
work
with
you
on
I.
Do
that
fer,
because
this
is
a
binary
artifact
right,
not
an
image.
Yeah
hi
I
have
a
some
experience
with
because
of
comps,
which
has
the
same
use
case
and
I'm
trying
to
get
that
onto
the
okay.
The
official
at
working
group
gets
in
for
stuff.
A
F
A
H
Go
ahead
and
take
this
I
think
the
biggest
news
that
we
have
the
shares
that
we're
basically
running
towards
the
p1
alpha
3
release,
or
not
just
cluster
api,
but
quite
a
few
of
the
providers
as
well.
So
I
think
that's
why
we
don't
have
a
lot
of
the
inner
core
cluster
api
folks
here
today.
I
know.
Events
in
particular
is
working
towards
getting
us
there.
So
we're
looking
to
cut
the
initial
v1
out
for
3
release
of
core
cluster
API
version
0.30
later
today.
H
As
long
as
we
don't
hit
any
pending
blockers
between
now
and
then,
I
would
also
expect
to
see
b1
alpha,
3
oriented
releases
for
the
US
provider,
the
vSphere
provider
and
probably
the
azure
provider
within
the
next
few
days
as
well.
I
know
at
least
for
the
end
of
us
provider
we're
really
close
to
a
release.
I
think
we
only
have
one
or
two
more
potential
blocking
issues.
A
I
just
wanted
to
take
a
shirt
second
to
shout-out,
to
Guinea,
be
one
out
for
three
out
the
door.
That's
probably
the
biggest
step
forward,
I!
Think
in
the
history
of
Costa
API.
It's
for
realsies,
now,
I
think
right
before
it's
kind
of
like
you
know
you
can
integrate
with
it.
You
could
use
it
I!
Think
now
it's
like
it's
ready,
I
think
the
adoption
curve
is
gonna
go
tick,
but
we
need
to
do
all
the
PR
and
everything
else
that
goes
around
it.
A
G
A
Yeah
I
think
there's
there's
a
bunch
of
use
cases
and
user
scenarios
that
are
out
there.
I
think
you
know
getting
the
adoption
curve
rolling.
The
adoption
curve
now
and
focusing
on
stabilization
will
probably
be
super
useful
for
maybe
the
next
cycle
or
so,
and
it's
just
great
to
see
like
it
started
out
like
was
it
two
years
ago,
started
out
as
like
as
a
single
small
thing
two
years
ago
as
an
idea,
and
now
it's
like
it's
ready,
I
think
it's
ready
for
prime
time.
So
it's
exciting.
I
You
know
we
we
released
1.8
on
Friday,
actually
181
and
the
highlights
there.
Just
our
doctor
travels
a
lot
more
stable
and
we
now
have
from
a
cold
start.
We
can
start
kubernetes
and
your
vm
in
about
50
seconds,
so
we're
under
a
minute.
Finally,
on
a
cold
start
up
as
long
as
your
cache
is
warm.
So
that's
that's
the
the
highlights.
I
G
A
A
Pretty
much
all
right
are
there
any
other
updates
or
group
topics
that
folks
on
to
discuss?
Otherwise
we
can
get
29
minutes
back
from
the
eight
minutes
back,
that's
call
going
once
I'll
come
really
slow.
Some
people
always
interrupt
me
at
2:00,
going
twice:
okay,
three
times
alright.
Well,
that's
it
take
care,
everybody
stay
safe,
hope,
no
one
gets
coronavirus
and
if
you're,
if
you're
traveling,
you
know,
please
stay
safe,
hi.