►
From YouTube: Meet our Contributors - Ask Us Anything 20180207
Description
Check out information on this program and other mentorship programs here: git.k8s.io/community/mentoring/meet-our-contributors.md
A
Alright,
hello,
everybody
welcome
to
our
first
streaming
edition
of
meet
our
contributors,
where
we
have,
let's
see
five
fabulous
contributors
on
the
line
if
I
or
six
rather
than
if
I
can
count
correctly.
This
series
has
been
created
as
part
of
kubernetes,
larger
mentoring
initiatives
and,
what's
better
than
having
six
plus
mentors
on
the
same
line
with
answering
questions
on
demand.
A
Today,
we're
going
to
be
taking
questions
from
the
Kuban
Etisalat
channel
for
meet
our
contributors
and
if
I
can
keep
up
Twitter,
hopefully
with
the
hashtag
of
Kas
MOC,
if
anybody's
out
there
following
along,
we
would
love
to
have
your
questions,
so
we
actually
also
like
to
open
up
this
session
in
the
future
for
potential
peer
code
reviews
as
well
as
code
walkthroughs.
So
if
anybody
listening
has
problems
that
they
would
like
a
second
pair
of
eyes
on
or
problems
where
they're
getting
stuck,
please
submit
those
ahead
of
time.
A
So
we
can
have
proper
folks
on
the
call
to
give
a
little
code
review
session
at
the
end.
But
today
we
actually
have
Ellen
on
the
line
and
we
are
gonna.
Do
a
little
small
outreach
interview
with
her
to
tell
everybody
about
the
RET
program
that
were
involved
with,
but
that'll
be
a
little
later
and
first
we're
gonna
go
through
some
intros
and
then
we're
going
to
get
right
into
the
questions
cool
all
right.
A
A
D
And
everyone,
my
name
is
see
her
on
me:
holder,
artsy,
I'm,
a
developer
advocate
at
cloud
native
computing
foundation
on
commercial
organization
that
holds
communities
as
an
open
source
project.
Kubernetes
community
I
am
involved
primarily
as
a
co-lead
of
seek
product
management.
All
some
deeply
involved
into
the
kubernetes
process.
E
Hi
excuse
me:
hi
I'm,
Aaron
Berger,
with
Samsung
SDS
I
am
co-founder
and
co-lead
of
the
kubernetes
sake.
Testing
I'm,
also
an
active
participant
in
say,
contributor,
experiencing
architecture,
scalability
and
I've
been
actively
involved
in
the
release
teams,
since
kubernetes
version,
1.4,
yeah
I
think
that's.
It.
F
A
G
I'm
Jorge
Castro
I
work
at
hep
do
mostly
participate
in,
say,
contributor,
experience
and
kind
of
do
community
tests,
one
of
the
big
things
I'm
working
on
this
cycle
is
the
kubernetes
contributor
guide.
So
if
you
search
for
that
in
your
search
engine,
it
should
pop
right
up
and
we're
in
the
middle
of
reworking
that
to
be
better.
So
if
you're
new
and
you
want
to
get
started
going
through
the
instructions
and
telling
us
what
needs
to
be
fixed
would
be
a
great
place
to
start.
A
Definitely
and
I
think
all
of
us
on
this
call
can
give
any
first-time
contributors
some
places
to
start,
especially
with
in
contributor
experience,
which
is
a
great
place.
So,
let's
jump
into
the
questions,
we
actually
have
one
question
with
a
flock
right
now
and
that's
from
that's
for
nikita
it's
from
an
inspiring
google
Summer
of
Code
and
turn
for
kubernetes,
which
is
awesome
that
we're
getting
the
word
out
in
the
student
community
and
they
wanted
some
guidance
on
just
how
to
start
and
approach
the
mentors
that
are
interested
in
the
program.
So
I.
B
Think
here
we
have
I
mean
I'd
like
to
divide
this
question
into
multiple
questions.
So
first
is:
how
do
you
start
contributing?
Second,
is
when
do
you
approach
the
mentors
and
third
is
how
do
you
approach
the
mentors?
So,
firstly,
if
I
mean
for
google
Summer
of
Code,
it
says
that
you
need
to
have
previous
contributions
to
the
project
to
get
selected.
But,
honestly,
you
don't
need
to
have
any
previous
contributions.
You
just
need
to
show
initiative
and
you
can
start
contributing
now
as
well.
B
Usually,
what
students
I've
seen
that
what
students
are
doing
right
now
is
that
they
try
to
find
an
issue
to
work
on
by
themselves
or
try
to
think
about
which
stick
to
join,
but
then
they
fail
miserably
at
that,
because
they're
not
I
mean
kubernetes
is
a
huge
project
and
you
don't
know
what
stick
to
join
or
which
what
I
should
be
and
well.
If
you
don't
have
mentoring,
then
the
issues
can
be
somewhat
hard
to
start
out
with
so
don't
jump
into
issues
by
yourself.
B
So
that's
that's
one
important
thing
that
don't
worry
or
don't
get
that
if
you
don't
find
an
issue
to
work
on,
and
so
this
is
the
when
of
when
you
should
be
contacting
the
mentors
about
how
you
should
be
contacting,
join
this
relevant
SiC
channel
on
slack
and
talk
to
the
mentor
and
just
introduce
yourself
tell
your
name
about
your
previous
experience
with
kubernetes
or
what?
If
what
have
you
done
up
until
now
like?
Have
you
read
the
documentation?
B
Have
you
looked
at
some
issues,
or
would
you
like
to
work
on
this
particular
project
or
something
like
that
and
ask
for
a
big
enough
friendly
issue
to
work
on
and
I'm?
Pretty
sure
someone
will
help
you
out.
So
that's
about
it.
That's
how
that's
how
you
ask
limiters
and
there's
one
point
that
I'd
like
to
specify.
B
Is
that
once
you
do
all
of
this,
the
most
important
thing
is
not
to
get
demotivated
because,
like
Ellen
had
previously
said
that
the
code
is
like
a
huge
spaghetti
and
it's
super
confusing,
so
you
are
gonna,
get
confused,
you
are
going
to
get
demotivated,
you're
gonna
feel
really
really
silly
at
times
and
you're
gonna
feel
shy
about
asking
questions
but
feel
free
to
ask
questions
with
this.
Everyone
who
works
on
kubernetes
knows
how
difficult
it
is
to
contribute
to
it
and
no
one's
gonna
judge
you.
B
H
B
That
could
be
possibly
it
may
be,
because
I've
mainly
worked
in
a
pyramid.
Chenier
and
I've.
Seen
projects
that
have
been
put
up
for
Google
servers.
Good
I
mean
we
appear
machinery,
so
yeah
I
mean
it's
gonna
require
some
mentoring
on
that
front,
but
I
think
I
think
Ellen
can
touch
up
on
the
6co
light
part
of
things.
F
I
might
be
wrong
because
the
code
I
work
with
seems
very
simple
to
me,
and
maybe
the
code
Nikita
I
worked
with
was
very
hard.
But
to
me
the
main
challenge
was
never
decode.
It
was
you
know
once
somebody
said
once
somebody
tells
me
like
okay,
this
code
is
supposed
to
do
this
and
such
and
such
a
way
for
these
and
that
those
objectives.
Then
it's
clear
and
it's
easy.
A
E
I'll,
just
so
as
a
one
of
the
problems
that
kubernetes
steering
committee
is
trying
to
work
on
is
making
sure
that
there's
a
clear
mapping
between
the
code,
the
organization
of
the
code
and
the
organization
of
the
people
who
own
it
so
we're
calling
this
like
sub
projects.
So,
for
example,
cig
apps
has
a
sub
project
called
the
workloads
API.
So
if
you're
interested
on
working
on
anything
related
to
you,
daemon
sets
deployments,
replicas
sets
jobs,
cron
jobs,
etc,
etc.
E
You
would
go
talk
to
the
people
who
are
running
the
workloads,
API
sub
project,
which
I
think
is
kenneth
owens
and
we're
working
on
getting
this
page.
This
information
documented
on
the
readme
for
cig
apps,
as
well
as
inside,
of
the
owners
files
for
all
of
the
code.
That's
related
to
this,
but
just
as
an
illustration
of
the
complexity
issue,
so
like
I'm
I'm,
a
member
of
the
steering
committee
I've
been
with
the
project
since
before
it
went
1.0,
and
somebody
suggested
that
the
workloads
API
sub
project
should
include
jobs
and
cron
jobs.
E
The
mapping
for
this,
which
brings
me
to
you,
like
just
I,
had
to
ask
people
and,
like
I've,
been
with
the
project
for
a
long
time
and
I'm
very
thankful
that
the
communities,
generally
speaking,
pretty
responsive
if
you
come
and
talk
to
them
as
a
human
being
you'll
get
another
human
being
to
talk
to
you.
I
personally
find
that
slack
is
usually
the
quickest
way
to
do
this.
E
If
somebody's
free
they'll
respond
to
you,
but
if
not
or
for
like
longer
conversations
that
I
can
expect
to
be
asynchronous,
I
usually
find
the
SIG's
they're
pretty
responsive
across
their
mailing
lists
too.
So
it's
it's
totally.
A
problem
like
somebody
from
the
Summer
of
Code,
will
have
as
well
as
somebody
who's
literally
trying
to
like
organize
the
entire
project.
So
the
complexity
is
sort
of
little
inherent
in
that
we
grew
extraordinarily
quickly
and
that's
something
we
all
recognize.
E
D
Just
like
that,
codebase
contributions
in
writing
code
is
not
the
only
area
where
you
can
get
started
to
his
contributions
to
cabin
areas,
because
Copernicus
is
not
only
their
technique.
Open-Source
project,
it's
the
big
open
source
project,
covers
most
all
the
areas
where
people
today
interact,
who
is
dis
systems
and,
for
example,
today
I'm
several
years
contributed
to
kubernetes.
At
the
same
time,
my
first
contribution
was
write
in
a
small
dog.
It
was
landed
on
a
website.
D
My
my
personal
experience,
I'm,
not
a
problem
I'm
not
to
develop
all
the
code,
but
I've
been
a
system
administrator
DevOps
engineer
three
or
four
years,
and
when
I
first
time
I
tried
to
avoid
kubernetes
as
a
project
it
can
satisfy.
My
needs.
I
found
that
some
information
in
the
documentation
website
is
lagging
and
it
was
terribly
easy
to
write
some
dog
from
myself,
submit
it
to
the
website
and
get
it
rendered,
and
today
it's
available
for
everyone
not
only
documentation.
D
If
you
are
again,
if
you're
not
a
program
and
if
you're
not
looking
for
the
code
contributions,
you
can
also
render
local
meetups
that
again
at
it
part
of
kubernetes
communicated
to
the
natives
ecosystem
you
can
play
and
during
their
surgeries,
where
you'll
be
involved
in
the
interest
process.
If
you
experience
project
a
product
manager,
CPM
is
for
you,
so
we
have
lots
of
arizim
through
input
of
contribution.
City
managers
community,
where
you're
not
on
the
program
and
experience,
will
be
available.
A
So
it
sounds
like
new
contributors
should
have
an
exploration
phase
essentially
and
go
and
see
what
sinks
are
out
there.
We're
working
groups
are
out
there
sinks
again,
our
special
interest
groups,
a
question
that
we're
getting
now
and
one
that
we
get
frequently
is
which
thig?
How
do
I
know
where
to
start
with
a
cig?
Is
there
something
about
my
personality
that
would
lead
me
towards
one
sig?
Does
anyone
want
to
take
that
question.
E
So
I
mean
I'll
just
start
off
with
the
whole
the
whole
name
of
the
sake.
It's
a
special
interest
group,
so
I
would
somewhat
start
with
like
what
is
the
area
that
you're
interested
in
specifically
so
like
I?
Think
CLI
is
a
pretty
great
place
to
contribute
personally
I
use
command-line
interfaces
in
anger
very,
very
frequently.
The
fastest
way
I
usually
interact
with
something.
So
if
I
wanted
to
make
that
experience,
better
I
would
definitely
join
sig.
E
Faster
is
automation
that
helps
reduce
the
friction
in
general,
so
we
have
a
large
amount
of
bots
and
stuff
that
makes
sure
that
labels
are
applied
consistently
and
that
tests
get
run,
etcetera,
etc,
and
there
are
sort
of
smaller,
more
lightweight
code
bases
to
which
you
can
contribute
from
a
cig
testing
perspective,
and
we
always
welcome
all
newcomers
and
sig
testing.
There
are
also
a
number
of
hard
automation
problems
that
contributor
experience
is
trying
to
shepherd.
I
will
defer
to
the
others
for
other
six
of
interests.
D
Yeah
I
would
I
would
describe
discretion
and
answer
this
question
and
be
in
a
more
general
way.
So,
first
of
all,
we
have
all
the
sixes.
We
have
thrown
two
thousand
of
succeeding
at
his
community,
but
most
of
them
are
categorized
into
two
major
unofficial,
but
major
groups.
Some
of
them
are
like
mostly
technical
and
own
in
some
code
base
and
another.
Another
piece
is
by
project
right
in
that
not
own
in
this
single
code
base,
but
it's
covering
some
specific
errors
in
communities
community.
So
we
have
multiple
six
that
are
mostly
touching.
D
Some
are
technical
components
like
ckp
machine
or
six
Li
or
signal
for
example.
So
all
these
special
interest
groups
in
given
areas
community
are
covering
some
specific
technical
errors,
some
errors
that
are
sold
in
some
technical
questions
of
developing
some
something
you
know
some
new
code
for
urban
areas
and
solving
some
operation
experience
for
the
existing.
At
the
same
time,
huge
number
of
six,
like
cig
dogs,
it
is
focused
on
documentation,
see
contributor
experience
that
just
focus
focused
on
on
board
in
the
new
contributors
and
making
the
life
easier
for
existing
contributors.
6:00
p.m.
D
A
Alright
and
I
think
you
actually
answered
the
next
question,
which
was
if
we
can
name
some
things
that
might
be
good
starts
outside
of
docs
for
beginners,
so
I
think
you
just
nailed
that
one
did
anybody
have
any
other
SIG's
to
add
to
that
list.
I
think
word.
She
looks
like
you
might
have
had
something
yeah.
G
Contributor
experience
like
I
said
well,
the
the
contributor
guides
always
looking
for
help.
We
have
a
bunch
of
documentation
that
teaches
people
things
about
kubernetes,
but
they're,
all
written
by
different
people
and
we're
churning
through
that
and
stuff.
But
mostly
you
need
to
just
have
organizational
skill
and,
like
I'm
learning
a
lot
about
how
things
work
just
by
going
through
that.
So
that's
a
good
place
to
start
and
there's
always
usually
in
in
meetings.
If
you
sit
through
a
cig
at
some
point,
usually
cichlids
will
call
out
hey
help
needed
here
or
sometimes.
G
If
you
look
in
the
SIG's
meeting
notes,
you
might
be
able
to
glean
tasks
that
are
still
not
assigned
to
people
and
I
know
some
SIG's
have
what
they
call
a
backlog.
So,
as
an
example
say,
cluster
lifecycle
Tim
to
spend
a
while
getting
that
backlog
sorted
so
that
it's
like
here's
stuff
that
we
all
want
to
do
that
isn't
assigned
to
anybody
and
I'm
sure
I'm
sure
other
SIG's
would
have
to
have
something
like
that:
I'm
not
familiar
with
any
right
off
the
top
my
head,
but
it
it
kind
of
helps
to
be
like.
G
A
E
Mean
like
so
I
can
speak
to
sig
testing,
for
example.
My
answer
to
you
would
be
like
go
look
at
the
issues
in
the
repo
that
we
own,
so
we
own
the
kubernetes
testing
for
repo
there's
an
open
issue,
count
of
something
like
200
and,
generally
speaking,
if
somebody
isn't
assigned
to
an
issue
that
means
nobody's
working
on
it
and
from
there
I
would
literally
just
say
like
what
sounds
most
interesting
to
you.
You
could
probably
welcome
the
help
on
just
about
anything
that
answer
gets
a
little
more
intimidating.
E
If
you
look
at
the
kubernetes
kubernetes
repo,
which
has
like
over
4,000
issues
right,
that's
maybe
where
you're
gonna
go,
ask
a
cig
more
specifically
or
look
at
the
Help
Wanted
label.
Although,
like
I've
heard
some
of
my
teammates
have
found
the
Help,
Wanted
label
isn't
being
curated
as
effectively
as
they
would
like.
E
So
if
there's
a
Help
Wanted
issue
that
looks
really
cool
to
you,
but
you're
not
sure
if
it's
being
worked,
I'm
gonna
kind
of
keep
saying
this
over
feel
free
to
like
go
to
slack
and
find
the
people
who
said
they
were
working
on
it.
But
haven't
done
anything
lately
or
the
person
who
originally
created
issue
or
the
sig
that
happens
to
have
their
label
on
the
issue,
go
asking
that
channel.
E
How
do
you
stand
up
and
manage
a
cluster
effectively
right,
so
kubernetes
anywhere
and
cops
and
getting
rid
of
the
horrible
bash
scripts
that
lives
in
the
cluster
subdirectory
of
kubernetes
are
all
things
that
fall
under
the
purview
of
cluster
lifecycle.
Similarly,
if
you're
interested
in
doing
cool
things
on
top
of
a
kubernetes,
so
you
kind
of
don't
care
how
the
kubernetes
cluster
was
stood
up
or
necessarily
what
its
capabilities
are.
But
you
want
to
be
able
to
deploy
things
on
top
of
it
that
make
management
and
applications
easier
or
automating.
E
More
things
easier,
cig
apps
will
probably
be
a
great
place
to
get
and
take
a
tour
of
some
of
the
sub
projects.
They're
like
helm
for
deploying
applications
or
large
sets
of
kubernetes
resources,
and
if
one's
charts
or
compose,
where
how
do
we
make
it
easier
to
onboard
people
from
the
docker
compose
world
into
kubernetes
applications.
E
If
you
were
interested
in
running
kubernetes
in
your
cloud
provider
of
choice,
so
you
have
a
lot
of
AWS
experience.
We
have
a
lot
of
Azure
experience
where
you
have
a
lot
of
Google
cloud
experience
and
you're
motivated
to
ask
why
kubernetes
doesn't
support
your
clouds
like
awesome,
new
feature,
whatever
that
our
sakes
dedicated
specifically
to
those
cloud
things,
and
so
even
just
offering.
Like
your
experience,
your
your
history
as
a
cluster
operator
could
be
useful
and
you
might
be
able
to
help
write
some
things
and
go
to
help
implement
some
of
that
stuff.
E
If
you're
interested
in
the
specific
components
like
cig
note
would
be
a
great
place.
If
you
wanted
to
talk
about
different
container
runtimes
for
your
node
of
choice
or
like
cig
storage,
would
be
a
great
place
to
go
if
you're
interested
in
in
getting
your
particular
storage
driver
to
work
well
with
kubernetes.
So
hopefully
that
was
a
little
bit
more
detail
of
an
answer
like
each
sake.
Sort
of
has
its
own
thing.
E
I
would
also
highly
recommend
just
like
jumping
in
there
slack
channels
and
looking
at
a
meeting
or
two
or
at
least
like
their
meeting
notes,
just
to
get
a
flavor
of
this
character.
Of
that
say
gift
you
don't
feel
like.
You
have
an
area
of
expertise,
but
you
just
kind
of
want
to
jump
in
someplace
and.
A
A
Thanks,
alright
and
then
we
actually
have
a
Jen,
a
pretty
general
question,
so
this
will
be
fun
why
kubernetes
contribution
so
why
be
a
contributor
I
would
love
to
hear
I
would
love
to
hear
Nikita's
thoughts
on
that?
Why
kubernetes?
Why
not
another
open
source
project?
You
know
what
what
landed
you
here
so.
B
B
A
F
For
me,
it's
because
well,
there
are
many
technical
challenges
out
there
and
things
there
are
that
are
important
and
things
that
aren't
as
important
and
I
think
what
kubernetes
is
doing
like
the
the
problem
cuber
Nettie's
is
trying
to
solve,
is
something
I
I
consider
important
and
I.
Think
it's
something
worth
spending
time
on
I.
Guess
it's
more
of
a
logical
thing.
It's
like
something
I
believe
in
I,
guess
so
to
me,
it's
more
like
that,
of
course,
I
was
learning.
I
was
letting
go
and
I
started,
making
friends
learning
go
and
turns
out.
F
Most
of
them
are
involved
in
kubernetes
anyway,
so
I
was
learning
kubernetes
even
without
specifically
wanting
to,
and
there
was
a
there's
a
the
communities
very
mixed
together,
like
the
NGO
community
and
the
communities
community,
so
I
started
I
did
the
cubic
corn
web
site
just
because
I
was
friends
with
Chris
and
Chris
Nova,
and
so
okay,
we
need
a
website.
So,
okay
I'll
do
it
and
you
know
that's
how
I
got
introduced
to
it,
but
a
more
in-depth
response
would
be
that
I
think
it's
a
problem
or
solving.
So
that's
why
I'm
here
Cheers!
A
C
It
was
all
like
complete
mess.
So
that's
why
I
started
contributing
C
class.
The
lifecycle
member
that
formed
well
actually
I
was
there
like,
during
the
very
early
formation
days,
obviously
plus
the
lifecycle
yeah,
so
I'm,
trying
to
improve
that
area.
Having
having
had
some
expertise
in
my
complete
management
and
general
like
deploying
complex
applications
from
on
a
set
of
machines
and
I
mean
that's
something,
Commerce's
does
itself,
but
for
deployments
itself,
it's
still
kind
of
hard,
so
yeah
there
is
more
important
to
be
done
in
a
cluster
lifecycle.
A
E
Definitely
let
me
know
if
the
business
logic
makes
sense
right
and
and
I
think
that
collectively
the
project
would
move
a
whole
lot
faster
if
we
were
more
confident
in
testing
a
more
granular,
faster
level,
as
opposed
to
thinking
that
we
needed
to
test.
The
entire
thing
is
one
big
black
box
before
we
were
fully
comfortable
signing
off
on
new
features
of
log
9,
and
this
is
a
great
way
that
I've
seen
some
new
contributors.
E
Add
value
to
the
project
is
they'll
come
in
and
they'll
notice,
an
area
that
hasn't
really
had
great
testing
coverage
and
they'll.
Add
tests
and,
like
I,
am
so
much
happier
if
you
add
unit
tests,
if
you
add
end-to-end
tests,
that
could
be
a
good
thing.
Don't
get
me
wrong,
but
it
also
could
mean
that
every
single
poll
request
is
going
to
take
like
five
minutes
longer
for
its
tests
to
pass
and
we're
already
at
the
point
where,
like
in
order
for
the
project
to
merge,
the
40
plus
Bowl
requests
a
day
that
it
does.
E
We
have
to
like
test
all
of
them
in
batches
of
5
or
10
it,
because
it
takes
about
an
hour
or
more
to
run
all
of
the
tests
for
a
given
pull
request.
So
if
we
got
that
time
down
by
writing,
more
unit
tests,
we're
taking
some
of
the
like
I,
don't
know
if
this
exists
today,
I'm
using
just
off
the
top
of
my
head,
the
scheduler,
for
example,
to
make
sure
that
the
scheduler
is
logic
for
its
predicate
and
it's
in
its
filters
right,
given
a
set
of
nodes
that
are
configured
some
certain
way.
E
Does
the
scheduler
put
the
pods
in
the
right
places
at
the
right
times,
like
that's
all
this
in
this
logic
that
doesn't
need
an
entire
kubernetes
cluster,
with
an
API
server
and
a
controller
manager
and
a
net
CD
cluster
and
virtual
machines
to
test?
That's,
that's
totally
doable
with
the
appropriate
set
of
box,
which
will
lead
me
to
plug
a
meeting
that
is
literally
happy
happening
right
now,
so
cig
testing
itself
has
a
sub
project
called
testing
Commons
that
we
just
started
up
recently.
E
You
care
more
about
how
the
thing
operates
than
I
do
I,
just
care
that
it
passed
sand
and
hopefully
it
passes
quickly
and
so
testing
Commons
is
our
sub
project
to
make
sure
that
we
actually
provide
the
frameworks
and
toolkits
to
sort
of
reliably
steer
people
towards
well
written
tests
using
the
right
set
of
tools.
So
that
would
be
a
great
place
to
show
up,
which
happens
like
right
now
on
Wednesday
at
7:30
Pacific
every
two
weeks,
so
yeah.
F
G
You're
talking
about
new
contributors,
also
new
people
have
an
eye
for
things
that
are
obvious,
that
we
might
not
right.
So
when,
when
we
were
mentioning
hey,
you
know
check
out
the
help
labels.
Someone
in
the
chat
was
like
well,
the
community
section
doesn't
have
a
lot
of
help
label
stuff
label.
That
I
was
like
Oh
duh
should
probably
go.
Do
that
and
I
was
good
as
I
was
going
through.
That
I
saw
one
that
was
like
document
better
how
to
set
up
your
laptop
for
local
kubernetes
development
right
and
that's
a
thing.
G
We're
like
UV,
ask
most
people
that
have
been
around
for
a
while
they've
had
that
set
up
for
three
years
or
something
and
they
don't
like
they
don't
think
about.
Oh
what?
How
does
a
new
person
grab
their
laptop
start
from
nothing
right?
And
we
say
things
like?
Oh
just
go,
install
doctor
go
to
this
go
to
this
go
to
this,
but
no
one's
actually
sat
down
and
looked
at
the
instructions
and
gone
through
them
and
make
sure
that
they
all
do
the
right
thing.
G
So
that's
I
just
wanted
to
mention
cuz,
you
know,
Aaron
was
talking
about
you,
people
writing
tests.
You
know
you
just
keep
in
mind.
You
have
eyeballs
that
that
we
can't-
and
it's
it's
totally
not
on
purpose
right,
like
sometimes
there's
just
slots.
I
was
like
that.
That
looks
like
it's
right.
You
know
what
I
mean
this
beginner
section
looks
like
it's
right
to
me.
I,
remember,
I,
remember
going
through
that,
and
it
was
okay
right,
whereas
a
new
person
might
type
in
the
first
command
and
it
might
be
an
error
right.
A
F
E
You've
made
it
takes
a
lot
of
time
to
run,
make
every
time
yeah
so
running.
Make
every
time
ends
up
if
I
remember
correctly,
like
doing
a
full,
quick
release
which
runs
the
unit
tests,
the
integration
tests
regenerates
all
the
API
code
builds
all
of
the
Linux
binaries
and
packages
them
up
into
a
tarball
and
I
think
builds
all
the
docker
images
as
well.
That
sounds
like
more
than
what
you
need
if
you're
just
testing
a
specific
piece
of
code,
so,
generally
speaking,
you
might
just
want
to
like
run
make
tests
or
make
test
integration.
E
If
you
know
the
area
of
code
that
you
have
changed,
you
could
probably
run
make
tests
with
the
what's
the
environment
variable
equal
to
the
package
that
you
have
changed
now.
That's
where
you're
starting
to
get
really
clever
as
a
human
right,
because
you
have
some
knowledge
of
the
codebase
I
personally,
like
that's
sort
of
what
CI
is
there
for
is
to
run
the
tests
for
you.
So
I
personally
have
no
problem
just
pushing
up
code.
E
G
E
We
just
turned
it
on
for
some
of
our
canary
builds
in
a
medical
test
infrastructure
and
it
cut
build
times
at
about
half.
So
it's
really
cool
but
like
I
would
classify
basil
as
pro
pro
pro
tier.
It's
really
kind
of
difficult
for
me
as
a
non
Googler
to
really
crock
all
of
it.
It's
come
a
lot
further
than
it
has,
but
it's
not
what
I
would
recommend
for
new
contributors
who
are
just
getting
started
because
I
personally,
don't
know
of
any
other
project
that
uses
basil
as
its
build
system.
A
All
right
also
for
folks,
listening
in
your
in
the
slack
Channel
copied
the
deck
that
I
think
did
you
create
that
for
how
to
find
SIG's
and
a
little
bit
more
of
a
cig,
deep
dive?
That
would
be
it's
very
valuable.
So,
if
anybody's
listening
in
and
once
that
information
visually,
it's
in
the
slack
channel
hash
meet
our
contributors.
A
E
I've
got
one
more
pro
tip
I
wanted
to
throw
out
there,
because
I
personally
find
that
I've
contributed
to
other
open-source
projects
and
had
my
poll
requests
commented
on
or
acted
on,
really
really
quickly.
I
have
found
that
oftentimes
in
the
kubernetes
project
people
can
get
frustrated
or
confused
as
to
why
there's
been
little
to
no
interaction
or
uptake
on
their
poll
requests
when
they're
used
to
it
happening
a
lot
faster,
so
I
just
kind
of
wanted
to
illustrate
a
few
things
and
talk
about
maybe
how
we
could
improve
that
situation.
E
E
Guess
I'll
go
to
this
one
now
sorry!
This
is
a
poor
example,
because
it's
already
been
approved.
I
want
to
find
one
that
hasn't
been
approved,
but
does
have
an
LG
TM
and
has
a
pull
request
to
master
well.
First
off
I'll
tell
you
this
if
you're
in
a
pull
request,
and
you
want
to
get
somebody's
attention
and
you
use
their
name.
E
This
may
be
what
their
notifications
page
looks
like
I
I
mean
I'll,
get
you
I'm
on
vacation
for
a
little
while,
so
I
have
1200
open
notifications,
I'm
participating
in
422
of
these
things,
I
tend
to
operate
on
the
order
of
sweeping
through
these,
like
once
a
day
when
I'm
actively
at
work,
we
also
have
a
tool
called
uber
Nader,
which
shows
us
a
pull
request.
Dashboard,
that's
a
little
more
informed
as
to
the
way
things
work
within
kubernetes.
E
E
It
might
take
a
little
while
to
get
his
attention
if
you're
purely
using
his
github
username
and
adding
him
and
expecting
him
to
respond
so
how
much
faster
way
would
be
to
find
somebody
other
than
lava
lamp
or
to
ping
him
directly
on
slack
so
I'm
talking.
This
has
nothing
to
do
with
me.
Scrolling
around
I'm
still
trying
to
find
a
good
Paul
request
for
the
pro
tip
I've
got
here,
but
basically,
if
you're
submitting
a
pull
request-
and
you
don't
have
a
lot
of
action
on
it,
but
you
do
know
the
sake.
E
E
So
at
once
this
says
your
poem
quest
has
been
approved.
It
basically
just
needs
to
pass
tests.
That's
not
technically
correct,
but
it's
close
enough.
So
here
you'll
notice,
the
bot
has
recommended
that
Bowie
should
be
the
person
that
you
assign
the
poll
request
to
using
this
command
and
you
might
assign
it
to
them
and
ask
like
why
aren't
they
responding?
This
is
this
is
ridiculous?
I'm
blocked
on
BOE
way
come
on.
Somebody
help
me
out
here
and
I
want
to
introduce
you
to
the
concept
of
owner
stocking
for
better
for
worse.
E
The
bot
tells
you
but
way
is
needed
because
he
lives
in
this
owners.
File
and
I
can
actually
click
on
this
owners.
File
and
I
can
see
under
the
list
of
approvers
that
if
Bo
is
really
like,
not
responsive,
maybe
he's
on
vacation
like
myself,
you
could
go
talk
to
any
of
these
other
people
and
then,
if
I
really
wanted
to
take
it
to
the
next
step.
I
could
like
go.
Take
a
look
at
these
peoples,
github
profiles
and
notice
that
oh
I
don't
come.
E
For
example,
I
know
Eric
probably
has
been,
but
oh
maybe
G
Marik
hasn't
been
contributing
to
the
kubernetes
project
in
like
three
months.
Maybe
he's
off
working
on
some
other
project
he's,
probably
not
the
person
for
me
to
go
talk
to,
but
maybe
I
see
that
vish
has
a
really
active,
github
profile
and
so
he's
still
really
involved
in
kubernetes.
D
I
was
just
yeah
just
just
or
not,
for
me,
I'd
also
recommend
not
to
start
with
asking
the
individuals
but
to
start
working
with
the
entire
group.
So
if
he
tried
to
contribute
to
some
error,
for
example,
it's
from
Aaron's
suggestion
was
the
contribution
to
GCE
error.
It
means
that
you
can
probably
start
this
working
with
the
GC
PC
GCP
select.
You
know
asking
some
people
to
interact
with
you
and
if
you
don't
have
any
specific
response
from
none
of
them,
you
can
reach
out
to
their
owners.
Yeah.
E
For
sure,
I
will
underscore
that,
because
I
don't
want
it
like
I
used
the
term
owners
file
stalking
a
bit
jokingly,
but
seriously
people
don't
like
to
be
like
nagged
out-of-the-blue.
All
the
time
right,
but
I
would
say
like
go
talk
to
the
snake.
First,
assuming
you
know
who
the
stick
is,
but
if
you're
really
confused
about
who
the
stick
is,
you
could
maybe
pick
one
of
ping
one
of
those
people
and
they
might
be
able
to
point
you
in
the
right
direction.
Again,
don't
respect
immediate
responses.
E
Not
all
of
these
people
live
on
slack
all
the
time.
Not
all
these
people
have
slack
handles
that
match
up
with
their
github
user
names,
but
it
is
a
good
way
to
remember
that
there
are
humans
on
the
other
end
of
this
and
some
of
the
humans,
don't
necessarily
use
github
as
their
sole
means
of
communication.
G
Yeah,
my
only
tip
is
slack
and
github
usernames
don't
always
match
and
also
on
slack,
it's
entirely
possible
for
someone
to
have
a
similar
late
mean,
like
slack,
doesn't,
have
unique
names.
So
if
you,
if
you
look
at
someone's
name
George
and
then
you
look
at
slack,
there
might
be
10
George's
like
you,
don't
want
to
accidentally
ping,
the
wrong
George
and
stuff,
so
sometimes
sometimes
in
github,
they'll
say
like
maybe
where
they
work,
and
then
you
can
kind
of
figure
out
what
their
email
addresses
or
something
like
that.
G
At
that
point,
I
just
try
to
revert
to
mailing
the
sig
mailing
list
is
always
good
because,
like
I'm
only
on
slack
during
work
hours
right
but
as
a
project
becomes
more
global,
sending
a
response
to
the
list
will
get
people
that
are
in
different
time
zones.
And
then
maybe
you
sent
it
at
the
end
of
your
day
and
then,
when
you
wake
up
the
next
day,
you
know
you'll
have
your
answer
as
opposed
to
slack.
G
A
I
think
these
are
all
great
pieces
of
advice,
especially
depending
on
the
individuals
personality
and
what
they
feel
comfortable
with
I
think
you
touched.
Probably
all
of
those
so
I
think
that's
that's
good.
All
right,
so
I
think
we're
gonna
wrap
up
today's
session
for
those
watching.
If
you
have
future
questions
ping
us
in
the
meet
or
contributors
Channel,
we'll
try
to
do
this.
Every
Wednesday
of
every
first
Wednesday
of
the
month,
I'd
love
to
do
this
more.
A
If
you're
a
contributor
watching
and
would
like
to
answer,
questions
or
participate
in
a
live
code,
peer
review
or
code
walk
through.
Please
get
in
contact
with
us
and
I
think
all
of
the
contributors
on
the
call
for
their
time,
and
also,
as
always,
your
upstream
contributions
to
this
amazing
technology.