►
From YouTube: Kubernetes Contributor Experience SIG 20180418
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
Where
you
are
in
the
world,
so
let's
get
this
going
this
morning.
If
you
could
all
add
your
names
to
the
attendee
section
on
the
notes
that
would
be
groovy.
Is
there
anybody
on
the
call
that
this
is
their
their
first
time
joining
or
hasn't
introduced
themselves
to
the
group?
Yet.
B
B
B
B
D
All
sub-project
owners,
if
you
could
update
the
slide
template
for
cube
con
if
you
do
not
update
a
slide,
the
slide
template
there's
a
high
probability.
We
will
not
talk
about
your
sub
project,
so
please
add
any
information
there.
You
do
not
have
to
worry
about
formatting
or
making
it
pretty
or
adding
memes
all
I
would
like
and
they'll
see
and
I
would,
like
is
words
on
a
slide
and
then
I'll
figure
out
the
rest.
So
please
update
that.
D
B
Current
contributor
track
session
loading
email
is
going
out.
Who
has
that
one.
D
D
I
did
research
other
voting
polls
for
us,
but
I,
don't
know
why
in
2018
voting
software
is
still
law
to
say
the
least,
but
so
we're
going
with
Sims
again,
most
likely
folks
will
have
one
week
to
vote
and
the
email
will
go
out
to
all
attendees,
including
new
contributors
in
case
they
want
to
take
a
peek
at
one
of
these
sessions.
That's
going
on
and
that's
it
questions
concerns
comments
with
current
contributor
track
at
the
summit.
A
No
updates
because
Joe
was
working
on
cat
5
and
he
went
to
the
Grand
Canyon
last
week,
so
he's
just
getting
back
and
spinning
up
and
I'm
the
next
one
is
kept.
Seven
thanks.
So
much
for
everyone's
feedback.
So
far,
I
have
a
prototype.
Url,
that's
all
set
up
there
take
a
father
link.
If
you
have
got
an
invite
for
me,
it's
basically
just
set
up,
but
we're
kind
of
kicking
the
tires
I've
set
up
month.
A
This
is
how
dedicated
I
am
I
set
up
mut,
so
I
could
test
the
email,
interaction
with
like
the
older
school
of
users
to
make
sure
that
we
get
decent
data.
Out
of
that
also
reached
out
to
the
ghost
community
gave
us
some
good
feedback
on
their
move
away
from
slack
that
I've
linked
in
to
the
cap,
which
is
very
interesting
feedback
as
far
as
they
just
totally
got
rid
of
real-time
communication,
which
seems
crazy
to
me,
but
it's
good
information
to
have
and
I'm
set
of
single
sign-on
and
stuff
and
that
all
works.
A
The
problem
is,
is
that
when
you
go
to
click-
and
you
said
github-
it's
like
this
application
owned
by
Castro
Jo,
so
it's
kind
of
under
all
my
personal
namespace,
so
I
sent
I,
sent
a
mail
to
Igor
to
figure
out
hey.
Is
there
someone
at
the
CN
CF,
who
has
like
officially
keeps
track
of
like
the
Google
Twitter
like
so
all
of
these
single
sign-on
things
have
like
a
client
ID
in
a
secret
key
and
I.
It
felt
weird
to
have
those
under
a
personal
space.
A
So
I
started
the
process
asking
on
what
that
looks
like
officially,
because
I'm
sure
someone
needs
to
do
like
credential
rotation
and
all
that
kind
of
good
stuff.
For
something
like
this
and
pending
the
kept
being
approved,
mostly
I'm,
just
ready
to
kind
of
run
the
prototype
for
awhile
see
what
we
can
glean
for
it.
Anyone
have
any
questions.
B
C
At
a
request
from
me
and
Tim
Lucas
set
up
AI
access
to
the
test
copy
of
the
toast
Chris
database.
That's
part
of
the
process
so
that
we
could
run
ad
hoc
queries
looking
for
certain
settings
and
maybe
set
up
some
dashboards
for
the
release
team,
because
it's
our
best
sort
of
source
of
compiled
data,
I've
I
spent.
Most
of
us
really
cycled
trying
to
work
with
you
had
well
yeah
realized
I
was
in
the
process
of
reinventing
dev
stats
to
get
it
to
work.
A
George,
oh
yeah,
I
have
a
question
for
you.
Josh
is
this:
something
is
this
something
that
you
would
consider
having
like
a
little
workflow
for
so
that
lets
say
sigelei
wanted
to
do
something
like
this.
Like
do
you
have
a
workflow
in
mind
where
it's
like
talk
to
Josh
and
then
he
goes
right.
Does
a
bunch
of
stuff
for
you
and
then
gives
you
the
data
or
no.
C
No,
this
is
this
is
a
well
I
guess.
It
really
depends
on
the
on
the
request
right
right,
because
if
the
request
is
I
want
to
do
some
exploratory
data,
mining
right
and
I
know
how
to
write
queries,
then
my
response
is
going
to
be
give
me
your
SSH
public,
key,
okay
and
I
will
set
up
access
for
you.
If
it's
hey,
can
you
look
in
the
dev
stats
data
for
X,
then
it'll
be
as
time
available
from
one
of
the
people.
Who
knows
how
to
do
that?
Yeah.
A
C
Don't
want
to
announce
it
as
a
public
resource
because
I
don't
think
either
Timur
I
have
time
to
handle
random
requests.
Although
you
know,
if
it's
understood
that
we're
not
going
to
turn
around
anything's
promptly,
it
wouldn't
be
bad
to
put
up
an
issue
or
something
so
people
knew
they'd
go
down.
I
mean
I
want
to
be
really
clear
that
if
they
happen
to
ask,
for
example,
during
code
freeze,
we're
really
not
going
to
get
back
to
them
for
three
weeks,
yeah,
the
so.
C
C
Okay,
the
I
mean
you
have
to
understand
a
fair
amount
about
this
database
to
get
useful
that
out
of
so
the
so
I'd
kind
of
expect
those
those
kinds
of
queries-
and
the
thing
is
I-
know
I-
can
take
those
kinds
of
requests.
You
know
during
times,
if
I'm
in
dev
stats
anyway,
looking
for
something
or
updating
a
dashboard,
then
I
can
run
somebody
else's
query.
It's
just
you
know.
Lena.
C
A
C
C
E
Have
like
an
issue
tracker
or
something?
So
it's
something
that's
not
strictly
speaking
necessary
for
exploratory
data
analysis,
but
if
you're
trying
to
prototype
some
kind
of
dashboard
against
live
data,
that's
where
it's
sort
of
necessary,
but
if
you're
just
trying
to
craft
and
Clery
yourself
to
figure
out
what
you
can
learn.
Historically,
you
should
be
very,
very
capable
of
doing
that
by
importing
a
database
locally
so
that
you're
not
wasting
Natsuki
resources.
Look.
You
said
this
was
a
test
database,
not
the
live
ones.
People
running
queries
against
this
are
going
to
be
impacting
right.
C
E
B
C
I'm
not
gonna,
be
able
to
help
them
yeah,
and
the
second
is
someone
to
go
through
a
brief
tour
of
the
public
testing
infrastructure.
That
is
the
test
that
we
run,
that
that
sig
testing
and
testing
in
front
run
on
kubernetes
and
that
sort
of
thing
and
how
to
look
at
the
most
tests
and
what
the
different
general
areas
are
test
mean.
A
F
Really
think
so,
so
the
ingress
repos,
my
understanding,
is
they're
one
a
split.
So
there
was
just
an
ingress
repo
that
handled
both
being
grass
controller
for
nginx
and
the
one
for
Google
compute,
and
then
they
split
those
apart
into
two
separate
repos
I.
Think
the
Google
compute
code
left
so
there's
like
evening-dress
GCE
and
now
what
was
left
was
ingress
nginx
and
some
of
the
people
who,
like
there
were
there's
like
a
couple
people
with
like
just
direct
right
access
to
the
repo
we're
manually,
merging
pr's
right.
G
F
G
F
The
bots
wouldn't
recognize
them.
They
were
asking
for
collaborator
access
to
just
that
repo,
but
I
think
we're
going
to
try
to
move
away
from
collaborator
access
like
we
removed
it
from
the
contributor
ladder,
all
that
kind
of
stuff
we're
trying
to
centralize
around
board
membership.
Okay
I
was
just
there
were
two
people
who
want
to
be
able
to
do
reviews
against
that
repo,
but
work
board
members.
Okay,.
E
Points
along
those
lines,
I
am
trying
to
be
very
repetitious
about
underscoring
that
there's
basically
very
little
requirement
to
become
a
member.
It's
not
like
it's
a
big
deal.
It's
just
a
sign
of
trust.
I.
Think
the
most
owners
thing
you
have
to
do
is
enable
two-factor
off
and
send
an
email
and
get
humans
to
+1
view,
but
there's
actually
no
requirement
for
pull
requests.
E
So
I'm
going
to
say
this
over
and
over
while
I'm
at
coupon
I've
said
that
during
the
community
meeting
I
say
it
in
interactions
with
people
directly
add
a
little
bit
more
color
to
why
membership
and
teams,
and
not
collaborators
directly.
It's
because
it's
really
really
difficult
to
audit
understand
who
has
what
kind
of
access
to
which
repos,
if
you
are
or
limber
only
admins,
can
actually
click
through
to
a
repo
and
see
who
is
a
collaborator
versus
a
member
of
a
team.
E
If
we
had
automation
that
drove
all
of
this
stuff
from
publicly
visible
files,
this
would
be
less
of
an
issue,
and
then
maybe
we
could
go
back
to
a
world
where
only
certain
people
have
only
certain
access
to
sort
of
area
clues.
But
for
now
is
just
be
easier
to
manage
who's
got
access
to
what
as
human
beings,
by
shuffling
everybody
into
consistently
named
teams
where
the
food
maintainer
z'
gives
people
right.
E
Access
to
the
food,
repo
and
foo
happens
gives
people
right
access
gives
people
having
access
to
the
food
repo
if
they
want
to
set
up
like
web
hooks
and
things
of
that
nature.
This
is
all
documented
in
the
community
repo
and
the
github
organizations
talk
to
me,
which
is
you
know,
a
little
bit
of
like
project
administrivia,
logistic
stuff
that
probably
shouldn't
necessarily
live
in
the
root
directory.
But
I
don't
know
that
I've
got
time
to
pick
up
that
torch.
E
A
It's
a
folder
or
someplace
yeah.
We,
we
should
talk
about
that.
A
cubicle,
cuz
I've
noticed
that
the
route
we
started
and
become
a
little
bit
untenable,
but
we
have
other
administrivia
in
other
places
where
it
might
make
sense
to
consolidate
those
things.
Ok,
so
the
way
we
can
help
is
just
continue
to
socialize
become
an
org
number.
It's
not
that
hard
cuz
it
before
someone
was
like
I
was
a
single
OpenStack
and
it
was
like
broken
like
this.
For
a
month,
I
was
like
what
what
happened?
Is
this
like
a
ok?
E
Fred
another
thing
that
I
would
hopefully
will
do
down
the
line.
Is
our
tooling
really
should
be
capable
of
a
lot
of
auditing
like
who's
it
in
a
hunter
file
that
isn't
assignable
or
isn't
an
ordinary
ver
know
he
had
like
big
umbrella
issues
for
this
back
when
lunch
github
was
just
wonderful
for
it,
but
we
don't
have
like
ongoing
auditing
on
that.
I
think
that's
something
that
we
could
probably
edit
ooh
proud
without
a
ton
of
difficulty.