►
From YouTube: Kubernetes Meet Our Contributors February 2019 Session
Description
wWhen Slack seems like it’s going too fast, and you just need a quick answer from a human...
Meet Our Contributors gives you a monthly one-hour opportunity to ask questions about our upstream community, watch interviews with our contributors, and participate in peer code reviews.
Every first Wednesday of the month, check this out for more information: https://github.com/kubernetes/community/blob/master/mentoring/meet-our-contributors.md
A
Hi
everyone
I'm
Paris
I
work
at
Google
and
kubernetes
community
manager.
Here
today
you
are
watching
our
awesome
February
edition
of
meet
our
contributors.
We
have
a
couple
contributors
on
the
line
today
who
are
going
to
help
us
through
various
types
of
questions
that
you
have.
Those
questions
are
coming
into
in
to
us
from
slack,
particularly
the
needy
are
contributors
channel
also
Twitter
DMS,
if
you'd
like
to
be
anonymous,
also
slack
diems
if
you'd
like
to
be
anonymous,
a
lot
of
folks
prefer
the
anonymous,
not
amenity.
A
If
I
said
that
word
right
and
that's
totally
okay,
we
do
have
a
code
of
conduct.
We
appreciate
everybody
need
excellent
to
each
other.
The
mentors
please
feel
free
to
ask
each
other
questions
as
well,
because
your
growth
is
just
as
important
as
everyone
else's
growth
within
the
project.
We
already
have
a
couple
questions
queued
up,
so
we
have
an
awesome
schedule
ahead
for
you
today
and
let's
do
some
quick
intros
George
is
behind
the
screen
right
there.
A
C
B
A
Very
cool
and
just
for
those
listening
who
are
not
aware
of
our
jargon.
Sometimes
it's
thick,
so
I
apologize
but
special.
It
means
special
interest
group,
that's
how
we
divide
the
project,
its
codebase,
its
documents
as
artifacts
and
get
things
accomplished
and
moving
forward
in
those
certain
areas.
Gus
mentioned
networking,
that's
just
one
PS.
They
are
looking
for
new
contributors
and
then
dims.
Why
don't
you
introduce
yourself
next.
C
I
work
for
Hawaii
I
work
on
a
few
Six
Sigma
T's,
sig
testing
and
basically
some
of
the
cross-cutting
themes
that
I
tend
to
help
out
with
just
like
Gus
I
was
previously
working
on
OpenStack
as
well
and
I'm
taking
care
of
the
code
that
goes
wrote
down
as
well.
So
it's
you
know
it's
good
to
buddy
up
with
you
guys.
C
Was
your
entrance?
The
first
thing
that
I
worked
on
was
something
called
Magnum.
It's
a
deployer
for
cuban
artists
on
top
of
OpenStack,
so
OpenStack
Magnum
is
very
opinionated
about
how
things
like
Cuban
artists
would
run
on
OpenStack
itself.
So
that
was
the
first
introduction
to
Cuban
artists
and
then
I
started,
diving,
deeper
into
better
integrations
and
then
found
Gus's
work
and
then
followed
him
along
and
then
started
doing
a
few
other
things
in
Cuba,
Netta's.
A
B
Yes,
well
so
I
was
trying
to
learn.
Openstack
I
just
started
working
on
OpenStack
and
I
wanted
to
learn
all
about
it.
So
I
was
trying
to
put
it
all
together
from
first
principles,
kind
of
do
it
the
hard
way
kind
of
thing
so
I
could
learn
all
the
pieces
and
I
was
working
for
xbase
at
the
time,
so
the
resources
they
gave
me
with
the
Rackspace
public
cloud,
which
looks
it's
kind
of
an
open
stack
cloud.
B
So
I
was
trying
to
install
OpenStack
from
a
distance
the
OpenStack
pieces.
Just
look
like
web
servers
and
databases.
It's
just
Python
web
servers
and
my
sequel,
databases,
and
so
I
was
trying
to
install
these
whatever
it
is,
15
or
so
different
components
onto
some
machines
and
I
said
well.
I
know
how
to
do
this.
I
want
something.
That's
a
little
bit
like
you
know,
Borges
always
used
to
from
Google
and
kubernetes
had
only
just
sort
of
come
out.
So
I
was
like
I'll.
B
Try
this
new
kubernetes
think
to
do
that,
and
so
I
was
trying
to
do
OpenStack
on
kubernetes
on
OpenStack
way
way
way
before
any
of
the
pieces
were
ready
for
that
sort
of
thing.
So
it
didn't
work,
but
I
learned
an
awful
lot
about
all
the
different
projects
along
the
way
and
because
I
was
trying
to
run
committees
on
the
Rackspace
cloud.
I
ended
up
writing
the
cloud
provider
for
kubernetes,
which
is
the
chunk
of
code
in
communities.
That
knows
how
to
talk
to
OpenStack.
B
So
I
wrote
that
just
along
the
way,
just
to
get
my
little
experiment
working
and
then
forgot
about
it
and
then
about
a
year
after
that,
people
kept
sending
me
some
email
pointing
out
bugs
in
the
code
I'd
written
and
I
was
like
well
who
oh
people
are
using
this
right,
so
I
kind
of
accidentally
got
into
communities
development,
but
that
was
my
big.
My
big
PR
was
that
big
patch
was
that
contribution
and
that
went
into
committees,
0.5
or
something
back
in
2014
went.
C
I
am
I've
hopped,
various
open-source
organization
to
be
honest,
I
was
working
at
Apache
soft
4
nation
before
I
got
into
OpenStack
and
then
from
OpenStack
I
hopped
into
CNCs,
so
yeah
I've
been
doing
open
source
for
a
while,
but
I
don't
think
it's
traditional
in
the
sense
that
I
think
the
more
people
that
I
see
a
nowadays
are
people
who
are
trying
Cuba
natives
and
they
face
a
problem.
So
they
come
and
ask
and
see
the
code
and
you
know
and
then
gradually
become
contributors.
C
B
B
B
You
actually
kind
of
still
the
same
node
base.
It's
just
sideways
like
we
didn't,
have
stateful
sets
and
deployments
and
things
before,
but
the
basic
structure
is
still
the
same:
the
idea
of
HCD
and
putting
things
into
there
and
then
like
the
control
lips
that
the
structures
all
the
same.
It's
just
grown
there.
A
So
a
question
that
I
get
a
lot
from
current
contributors.
Nonetheless,
as
what's
your
best
advice
for
them
as
they
climb
the
contribution
ladder
and
for
those
that
don't
know
what
the
contribution
ladder
is
with
kubernetes,
it's
a
membership
process
that
we
have,
depending
on
how
much
contribution
you've
done
to
the
projects
thus
far
and
your
leadership
and
things
that
you've
brought
with
it.
Those
are
members
reviewers,
approvers,
sub-project
owners.
We
work
in
owners
files
within
github.
So
what
for
you
to
view
like?
B
Yeah,
my
I've
usually
just
fallen
into
free
software
projects
just
by
using
them
like
I
use
them
and
usually
I
use
them
in
a
slightly
unusual
way
like
trying
to
install
OpenStack
myself
from
scratch,
kind
of
thing
or
communities.
I've
got
here
running
on
a
little
Raspberry,
Pi
cluster,
just
slightly
unusual
situation,
and
then
you
discover
bugs
things
that
don't
work
quite
well
or
things
that
could
work
better
and
then
you
just
fix
them.
B
B
This
condition-
and
it
was
really
confusing
and
took
me
half
a
day
to
work
out
what
that
meant.
If
only
the
error
message
was
very
slightly
better.
That
would
have
saved
me
half
a
day.
So
there's
a
really
simple
patch
and
you
can
send
that
off
and
reviewed
emerged
and
just
do
that
and
just
keep
doing
that.
Keep
doing
that
and
then
eventually
you'll
get
more
familiar
with
the
codebase
and
take
on
bolder,
larger
changes.
C
Tend
to
kind
of
like
look
for
areas
where
I
have
some
prior
knowledge
I
used
to
take
care
of
Noah
docker,
so
I
I
still,
you
know
get
into
signal
related
stuff,
whether
it
is
issues
or
PRS
or
hanging
out
on
there.
You
know
making
meetings
so
we'll
find
out
some
areas
that
you
are
interested
in,
whether
it
is
storage
or
whether
it
is
you
know,
computer
related
stuff
or
whether
it
is
networking.
C
Then,
when
you
look
at
you
know,
issues
and
peers
being
filed
and
the
work
that
is
usually
going
on
just
you
really
get
a
sense
of
who's
doing
what
and
you
kind
of
try
to
follow
along
what
they're
doing
and
see
if
you
can
help
them
in
stuff.
That
is
happening
even
if
it's
a
quick
review
of
PR
that
somebody
is
doing,
you
know,
try
out
a
test
case
in
a
PR
that
somebody
else
is
working
on.
You
know
that
that
way,
you
learn.
C
Basically,
you
know
what
are
the
problems
and
then
there's
a
lot
of
history
that
you
can
gather
the
we
have
a
huge
number
of
issues
and
PR
s,
so
you
can
always
go
back
and
see.
Oh
somebody
tried
to
do
something
similar
to
what
you're
trying
to
do,
but
then
you
know
there
was
some
feedback
and
they
lost
interest,
or
there
was
another
way
to
do
it,
so
they
ended
up
doing
some
other
way.
So
there
is
a
lot
of
history
that
you
can
read
up
on
and
discover
for
yourself.
C
So
that
is
something
that
I
would
definitely
recommend
people
to
do
it
takes
some
time.
But
then,
when
you
put
in
some
amount
of
time,
then
things
become
easier.
There
is
a
learning
curve
involved,
and
you
know
the
only
way
to
deal
with
the
situation
is
like
drink
from
the
hose
kind
of
thing
right
so
and
keep
asking
questions.
There
is
no
no
wrong
question
hit
people
up
on
slack
turn
the
discussion
into
a
mailing
list
thread
and
you
know,
keep
asking
stuff
because
not
everything
is
written
down.
C
We
are
trying
to
do
a
good
job
with
writing
stuff
down,
but
it
takes
a
you
know.
There
are
still
things
in
people's
heads
and
you
know
context
is
everything
and
history.
You
know
historical
decisions,
you
can,
you
can
talk
to
people
and
they'll.
Tell
you
why
things
are
being
done
and
so
you
know
be
open.
I
know
it's
hard
for
some
of
us
who
are
introverts,
but
you
know
it
works
in
the
longer
run.
We.
A
Had
a
question
I
think
us
made
a
comment
that
says:
do
the
work?
What
how
does
a
current
contributor
find
out
that
find
out
what
work
needs
to
be
done,
especially
in
times
where
repos
don't
necessarily
use
Help,
Wanted
or
good
first
issue
labels?
What
are
some
tips
there
because
they
think
that
it
seems
pretty
ambiguous
and
they
want
to
know
exactly
how
they
can
make
immediate
impact.
B
Yeah,
if
you
so,
the
this
HCG
has
a
meeting
which
is
recorded
so
for
me,
I'm
in
Australia,
so
I'm
not
in
a
good
timezone
to
attend
a
lot
of
these
meetings.
They
usually
3
a.m.
or
something
for
me.
So
I
watch
a
lot
of
videos
for
them
and
it
gives
you
a
good
feel
for
what
the
focus
of
attention
of
their
group
is
at
the
moment
and
what
they're
working
on
and
what
they're
trying
to
get
done
for
the
next
upcoming
release.
B
Yeah
yeah,
so
so
I
tend
to
pick
I
tend
to
pick.
My
own
tasks
seems
that
I
can
work
on
myself,
so
it
tends
to
be
smaller
tends
to
be
like
I
said,
just
little
issues
that
I've
faced
personally
and
then
you
do
them
fix
them.
So
just
little
things
still
a
Grunch
work
like
I
said
error.
Messages
are
always
appreciated.
B
That
error
handling,
there's
always
some
something
somewhere
or
some
new
feature
that
doesn't
exist,
that
no
one
else
is
working
on.
Yet
just
you
could
start
working
with
it.
Basically
yeah.
There
is
a
lot
of
one
things
I
like
about
the
communities.
Community
is
they're,
all
very
friendly,
I
really
enjoy
working
with
the
people
and
the
attitude
everyone
wants
to
help.
Everyone
is
very
happy
to
help
pull
you
through
this
process.
So
you
know
you're
gonna
have
to
do
some
work
yourself.
You're
gonna
have
to
read
some
code.
B
C
And
so
there
are
several
ways
to
do
it.
One
example,
I
can
walk
you
through
is
like
towards
the
beginning
of
a
release.
If,
if
you
go
to
sick
storage,
for
example,
you
will
see
that
they
have
a
spreadsheet,
they
track
things
that
can
be
done,
and
what
can
we
do
this?
You
know
for
this
release.
What
can
we
do
for
the
next
release,
so
this
discussion,
on
which
ones
can
be
actually
done
depending
on
the
people
who
are
there
and
who
are
willing
to
work
on
stuff?
Then
there
is
one
based
on
those
discussions.
C
Then
there
are
enhancements
that
are
filed
for
the
airport
caps.
They
used
to
be
called
caps
now
they're
called
enhancements.
The
enhancement
repository
these
typically
a
lot
of
discussion
on
on
the
caps,
and
there
will
be
a
debate
over
okay.
Is
this
the
right
way
to
do
this?
Is
there
alternate
way
to
do
this?
What
are
the
problems
with
upgrade
downgrade
things
like
that?
So,
and
then
things
get
split
into
like
tasks
into
issues
and
the
PRS
and
people
start
filing
stuff
and
they
form
smaller
teams
like,
for
example,
in
storage.
C
Again,
there's
a
group:
that's
working
on
CSI
migration,
basically
going
from
you
know
right
now:
we
don't
lots
of
people,
don't
use
CSI,
but
then
CSI
is
something
that
we
would
like
to
push
people
towards.
So
how
do
we
migrate
existing
volumes
from
entry
providers
to
CSI
providers?
So
there
is
some
there's,
a
really
small
team,
there's
like
four
or
five
people,
so
there
there
are
P.
There
are
groups
like
that
working
on
stuff,
so
you
just
have
to
find
the
right
group
to
fit
it.
C
Sick
cluster
life
cycle
is
very
open
and
very
friendly
as
well.
Six
storage,
okay,
again
does
a
good
job
in
managing
I
know
how
to
allocate
the
work
and
how
to
get
things
done.
But
essentially,
when
you
hear
somebody
in
a
call
saying,
okay,
I
need
help
with
this
in
you
raise
your
hand-
and
you
say:
okay,
I
I,
don't
know
much,
but
I
can
help
I'll
read
up.
Will
you
be
able
to
help
me
so
raise
your
hand
and
in
that
usually
works.
A
Yes
and
some
other
things
too
asks,
if
there's
new
contributors
on
the
line
as
well.
So
that's
an
ample
opportunity
for
you
to
introduce
yourselves
if
you're
a
seat
chair
that
doesn't
do
that,
it's
listening.
You
should
all
right
guys
go
back
to
the
remote
team
thing
because
I
think
that's
so
critical
with
kubernetes
in
general,
because
a
lot
of
our
team
is
distributed
a
lot
in
the
United
States,
but
of
course
globally,
as
well
with
you
Dean
being
one
of
them.
A
B
Yeah
so
I
don't
know
with
any
organization
communications
the
biggest
cost
right,
so
so
the
limit
to
scalability
is
how
will
you
communicate
so
kubernetes
does
very
well
with
slacks
likes
a
big
part
of
communication
and
that
works
quite
well
around
all
all
time
zones
and
things
also
kubernetes
uses
a
lot
of
zoom
meetings
like
this
one
and
that
doesn't
work
so
well
they're
videoed,
which
is
good.
You
can
catch
up,
but
it's
becomes
a
one-way
communication.
At
that
point,
you
can't
contribute
back
to
the
seat
during
the
discussion,
but
you
can
watch
it
afterwards.
D
B
I've
worked
on
that
you
need
to
trust
you
a
person
a
lot,
or
at
least
you
need
to
know
that
they
know
about
the
project.
No,
the
point
we're
trying
to
raise
because
they're
going
to
need
to
field
questions
about
it,
so
that
requires
a
sort
of
second
person
on
your
project
or
whatever
to
be
that
that
works,
particularly
the
demo
style
that
works
well.
Like
I,
said:
I
had
some
people
demo.
Some
of
my
projects
at
cGAP
siaps,
has
a
regular.
C
B
Demo
slot
and
that
worked
fairly
well
because
it
was
less
about
discussion
and
more
about
just
presenting
the
thing
that
worked.
Ok,
if
it's
occasional
meetings,
if
there's
one
like,
if
there's
something
I'm
particularly
involved
in
or
particularly,
have
opinions
about
that
I
want
to
contribute
to
the
discussion.
They
don't
make
the
effort
I'll
just
you
know,
get
up
early
in
the
morning
and
attend
the
meeting.
That's
not
something
I
can
do
regularly,
and
so
that
works.
It's
it's
ok,
usually
also,
just
you
want
to
restrict
communication.
B
So
it
makes
a
big
difference
or
try
to
pick
tasks
that
are
sort
of
self-contained
that
you
don't
need
to
interact
heavily
with
someone
over
on
the
kubernetes.
That's
fine,
like
it's
free
software.
This
is
this:
a
normal
free
software,
stuff
people
take
a
change,
they
work
on
it
at
home
and
they
upload
a
patch
and
there's
fairly
slow
communication
around
that
patch
and
that's
normal
and
fine.
B
The
Kemp
process,
the
enhancement
proposals,
they're
they're,
more
of
a
discussion,
but
but
they're.
Also
not
you
need
to
prepare
a
document
ahead
of
time
and
then
there's
sort
of
rounds
of
people
good
comments
on
it.
Then
you
respond
to
them,
so
it
works
quite
well
remotely
you've
just
got
you
know.
The
conversation
takes
24
hours
to
put
a
bunch
of
comments
up
there
and
then
respond
to
the
comments
and
that's
ok.
They
don't
need
to
be
fast
conversations,
those
ones
so
that
works
quite
well.
So
all
in
all
it
works.
B
Fine
I
would
just
it
I
mean
it's
not
a
corporate
environment
I'm
not
standing
at
a
whiteboard
talking
about
something.
That's
just
not
the
way
we
work
as
a
community
and
because
of
that
it
works
quite
well
distributed
in
remote.
It
is
worthwhile
trying
to
if
you
can
afford
it
or
get
your
employer
to
pay
for
it.
It
is
definitely
worth
while
attending
a
cube
con
or
to
meeting
people
face
to
face
making
friends
building
relationships
that
really
really
helps
when
it.
Then
it
comes
time
to
asking
questions
or
even
getting
reviews
done.
C
I
wanted
to
pick
on
something
Gus
said,
which
was
the
call
when
you
come
to
cube
con.
Make
sure
that
you
attend
the
contributor
summit
that
happens
the
day
before
the
cube
card,
because
that
is
where
you
actually,
you
will
be
able
to
meet
everybody
who's
working
on
the
development
side,
because
once
Cube
can't
start,
you
won't
be
able
to
find
anyone.
Unless
you
make
an
effort
to
go
to
the
talks
that
they
are
giving-
and
you
know
that
they're
going
to
be
there,
otherwise
it's
it's,
you
can't
find
people.
A
C
Thank
you,
I
mean
just
this
morning
we
had
a
call
in
Kate's
infra,
which
is
the
Cuban.
It
is
so
basically
what's
happening.
Is
there's
a
lot
of
infrastructure,
that's
currently
owned
by
Google
that
we
are
trying
to
move
to
the
CNCs
owned.
You
know
whether
it
is
repositories
or
gke
clusters
or
whatnot.
So
there
is
a
lot
of
work
and
we
were
wondering
where
everybody
years,
and
so
you
know
we
need
help,
and
we
would
like
help
in
almost
everything
that
we
do.
A
A
C
I
can
I
can
give
you
an
example
right.
So
we
hired
of
this
library
called
G
log,
and
you
know
the
G
log
repository
was
not
accepting
any
full
request
from
anybody,
because
it
was
just
code
that
was
exposed
and
we
started
using
it
in
cuba,
knitters
and
and
they
were
issues.
So
we
were
patching
on
top
of
that
code
for
literally
two
three
years
and
we
had
like
five
to
ten
issues
and
at
least
20
email
threads
talking
about
this
problem
right.
C
The
other
problem
is
how
do
you
deal
with
the
command
line
parameters
and
things
like
that,
so
you've
got
to
think
through
all
kinds
of
stuff.
But
this
is
the
main
thing
that
I
had
to
do.
Was
research
into
the
main
email
threads
and
read
up
on
the
issues
and
tried
some
experiments
out,
and
then
you
come
up
with
the
solution.
C
A
Awesome
all
right-
and
you
meant
dims-
you
mentioned
earlier-
that
the
working
group
for
kubernetes
infrastructure
is
looking
for
folks.
Anybody
else
want
to
do
any
quick
plugs,
because
we
did
get
a
couple
of
questions
from
people
that
said
off
the
top
of
your
head.
What
are
cigs
that
you
know
that
are
directly
looking
for
help
Gus.
You
just
went
off
mute.
That
means
you
might
know
one
they're.
B
B
B
A
Networking
Mon,
if
anybody
listening
does
like
networking,
we
are
actually
going
to
signet
workings
meeting
tomorrow
to
talk
about
how
they
can
get
new
contributors
and
the
process
for
getting
them
new
contributors
and
expanding
their
current
contributor
pool.
So
if
this
does
sound
interesting
to
you,
their
meeting,
I
think,
is
tomorrow
at
most
likely.
2
p.m.
yes.
C
And
it
doesn't
have
to
be
new
feature
all
the
time
either.
There
is
always
need
for
updating
to
newer
libraries.
There
is
always
need
for
going
from
say
this
morning.
I
was
talking
to
direct
from
signal
and
he
was
saying
that
oh,
we
should
try
to
go
from
see
advisory,
see
groups,
we
want
to
see
groups
v2
right,
so
there
is
always
things
that
we
need
to
catch
up
on
and
update
the
existing
code.
Add
test
cases,
add
unit
tests,
add
functional
tests,
and
things
like
that.
So
there
is,
you
know
it.
B
Your
CSI
example
earlier
is
a
good
one,
there's
a
lot
of
change
happening
in
communities,
because
it's
very
fast
because
it's
now
a
reasonably
old
project,
there's
there's
a
number
of
transitions
from
from
something
older
to
something
newer
going
on.
A
lot
of
them
have
started,
and
you
know
that
the
new
frameworks
there
and
has
been
exercised,
but
someone
still
needs
to
go
along
and
migrate
the
old
parts
and
finish
up
kind
of
the
tale
of
that
migration.
B
So
CSI
their
storage,
plugins,
a
bunch
of
things
are
on
the
storage
plugins,
but
a
bunch
of
things
are
still
on
the
old
kind
of
way
we
used
to
do
storage
volumes
and
that's
that's
a
fairly
typical
thing.
So
there's
a
lot
of
places
where
you
can
just
jump
in
and
someone's
already
done,
but
kind
of
the
design
work.
They've
already
got
approval
for
that.
It's
already
started,
but
they
just
need
help
at
that
point,
then,
to
just
continue
the
migration
and
carry
out
this
sort
of
the
grunt
work
of
the
migration.
B
C
And
sometimes
just
removing
things
that
were
deprecated
and
you
know
that
has
past
the
duplication
period
is
a
big
help
to
existing
contributors,
because
you're
taking
care
of
something
that
you
know.
Otherwise
they
would
have
to
do
so
and
all
it
and
all
it
means
is
like
you
go
delete
code
and
make
sure
that
you
don't.
It
is.
A
All
right,
and
now
we
have
a
it-
looks
like
a
triple
partner
from
a
current
contributor
one.
Second,
let
me
get
to
that
I
think
the
first
part
was:
when
do
you
contribute
meaning
does
work
say
that
it's
cool
for
you
to
contribute
during
the
day?
Do
they
say
that
you
can
contribute
on
Friday?
Do
you
contribute
on
your
hours?
Is
it
like
the
mix
of
everything
like?
Are
you
an
adult
it
can
do
whatever
you
want
like
tell
me
what
what
is
up
with
you
and
you're
and
you're
kind
of
contributing
day-to-day
here.
B
I
heard
of
really
great
conference
talk
by
I
think
was
count
Sandler,
who
made
the
point
that
we'll
need
to
eat.
Somehow
we
need
to
turn
out
our
cleverness
and
our
awareness
of
technology.
We
have
to
turn
into
food
somehow
the
usual
way
to
do
that
is
through
companies
and
companies
lock
up
a
little
bit
of
intellectual
property.
They
do
things
internally
secretly
in
order
to
sell
it.
B
That's
how
you
get
a
little
bit
of
kind
of
money
from
this,
so
we
need
that
because
we
need
to
eat,
but
at
the
same
time
we
need
for
the
long-term
health
of
our
industry.
We
need
free
software,
we
need
community,
we
need
like
contributions
to
these
sort
of
shared
Commons
projects,
and
it
was
the
first
time
I'd
really
heard
that
kind
of
short
term
versus
long
term
trade-offs,
kind
of
explained
so
we'll
so
I.
Think
of
it
that
way.
B
I
think
of
I
need
to
contribute
to
work
because
work
pay
me
and
I
need
to
eat,
but
at
the
same
time,
I
need
to
contribute
to
these
big
interesting
free
software
projects
for
migrants,
the
long
term.
So
my
growth
for
my
career
growth-
for
you
know
not
this
job
may
be,
but
the
next
job
definitely
and
so
I've
always
stolen
a
little
bit
of
time
to
work
on
these
projects
even
within
within
my
employer.
I've
got
a
project.
B
I
need
to
work
on
in
the
next
two
weeks,
but
then
me
being
aware
of
whatever
that
commute
is
seeing.
Yours
is
that
will
help
my
employer
in
the
next
six
months,
so
I
need
to
kind
of
balance,
those
that
short
term
long
term,
trade
off
and
I
finds
and
I've
done
this
too,
when
I,
when
I'm
a
manager.
It's
very
easy
to
take
a
short-sighted
view,
because
you're
looking
down
you're
looking
just
at
your
project
and
you
sort
of
forget
that
there's
going
to
be
a
next
project.
B
So
as
an
employee,
looking
up
I
kind
of
have
to
sometimes
ignore
my
manager
and
say:
hey
I
know:
that's
what
you
want
me
to
do
and
I'm
working
on
that.
That's
cool
I
got
that,
but
at
the
same
time
I'm
also
taking
the
longer-term
view
and
looking
at
the
next
project
and
making
sure
that
we're
in
the
right
place
for
the
right.
You
know,
skills
and
the
rights
involved
in
the
right
teams,
for
where
that's
gonna
need
to
be
so.
B
So
for
me
that
means
I
do
most
of
my
work
during
work
time
for
for
my
employer
on
the
projects
that
me
to
do
and
for
me
that's
projects
around
the
periphery
of
kubernetes
I
work
on
a
Canary's
team,
so
I'm
doing
cubes,
FG,
sealed
secrets.
Vic
appear
a
bunch
of
these
projects
that
are
in
communities
but
they're,
not
in
the
kubernetes
project,
they're
kind
of
products
and
third-party
projects
around
the
outside.
B
So
then
my
kubernetes
work
tends
to
be
in
my
own
time
or
something
that
falls
out
of
some
of
that.
Naturally,
if
I
have
to
fix
a
bug
that
I've
encountered
at
work
and
for
me
I
run
a
cluster,
you
can't
really
see
it's
just
off
camera,
but
I've
got
a
stack
of
laptops
that
kubernetes
their
old
laptops
that
my
kids
have
dropped
and
smashed
the
screens
on
you.
C
B
Nodes
and
then
I
got
a
Raspberry
Pi
cluster
as
my
control
nodes
and
just
running
that
allows
me
to
learn
an
awful
lot
about
communities.
I
know
an
awful
lot
about
it,
just
because
I've
had
to
put
it
back
together
when
it's
broken
and
I've
had
to
nurse
it
through
upgrades
I've
had
to
do
all
these
things,
which,
if,
if
I
only
read
the
unit
tests,
I,
wouldn't
know
that
firsthand
as
to
how
how
kubernetes
actually
works
in
practice.
B
A
Actually,
pinged
in
the
middle
of
that
and
said:
how
do
we
get
more
companies
to
contribute?
I
mean
I,
think
there's
five
hundred
plus
that
contribute
now
based
off
of
the
stats.
I'm
assuming
I
mean
I'm
typing
to
them
right
now,
I'm,
assuming
that
they
need
larger
contributions,
but
or
maybe
they
don't
wait
for
the
ping.
But
what
are
what
are
some
would
some
advice
that
you
have
for
just
day-to-day
engineers
on
how
to
convince
their
employer
that
they
should
consider
upstream
up
streaming
some
of
their
work
right.
C
A
C
You
will
not
be
able
to
catch
up
quite
easily
and
you
will
hit
bumps
when
you
try
to
upgrade
your
existing
and
one
or
try
to
try
to
run
around
the
same
same
software
in
the
newer
clusters,
because
you
forgot
something
or
you
don't
know.
You
are
not
aware
of
some
of
the
things
that
have
landed.
For
example,
people
had
trouble
when
we
turned
on
our
back
right.
So
there
were
things
that
were
not
defaults.
C
A
B
C
I
can
tell
you
what
happened
with
me.
So
what
are
my
managers
ended
up
looking
at
some
of
the
work
I
was
doing
and
said.
Why
are
you
doing
this?
It's
we
don't
do
this
today
and
we
don't.
We
don't
really
care
about
it
right
now.
But
then,
three
months
later
it
comes
to
I
can
say:
oh
I'm
glad
you
did
the
work
that
you
were
doing
at
that
time,
because
we
are
going
to
use
it
in
our
12
product.
It's
like
they
don't
know
that
and
we
are
ahead
of
them.
C
So
there
is
always
a
teaching
component
involved
where
you
have
to
be
able
to
sell
your
work
to
your
employer.
Only
then
will
they
know.
You
know
if
they
don't
know
they're
going
to
assume
that
you're
you're
not
fully
utilizing
your
time
right.
So
you
have
to
you,
know,
educate
up
your
change
and
that's
anyway
useful.
C
You
know
for
your
career
as
well
right,
so
you
know
put
together
a
small
demo,
try
to
involve
your
colleagues,
and
you
know
if
you
have
have
a
brown
bag
every
every
two
weeks
where
you
are
sharing
something
it
might
be.
Just
a
10
minute
thing
and
you're
sharing
with
your
colleagues
or
if
you
see
a
good
video
like
there
was
a
nice
there
were
nice
videos
from
FOSDEM
where
people
were
talking
about
how
to
run
human
at
us
in
a
rootless
environment
or
something
like
that
send
them
send
those
kinds
of
things
around.
C
A
Awesome
I
gave
the
while
you
were
all
working
on
another
question:
I
gave
them
my
feedback
in
the
DM
and
then
I
said.
Of
course,
I
will
ask
you
all,
but
I
have
a
feeling
they're
going
to
say,
lady
ldr,
when
you
figure
it
out.
Let
us
know
I,
think
I.
Think
I
called
that
one
on
Gus's
on
Gus's
side
yeah
and
it's
I
think
my
TL
DR
was
exactly.
A
B
Yeah
I
gave
a
talk
about
this
years
ago
that
about
how
technology
goes.
You
know
from
from
crazy
to
interesting
to
boring
and
becomes
commodity,
and
so
you
know
the
the
old
things
we
used
to
talk
about
20
years
ago.
Was
you
know?
No
one.
No
one
made
money
selling
Linux,
but
they
made
money
using
Linux
used
to
be
the
the
phrase
that
we
throw
around
a
long
time
ago.
People
trying
to
work
it.
B
How
do
we
make
money
out
of
Linux
I
feel
like
kubernetes
is
doing
a
lot
of
the
same
things,
because
it's
kind
of
sprung
with
quite
a
mature
vision
where,
after
the
beginning-
and
it
went
straight
into
sort
of
a
community
project,
whether
with
a
free
license,
there
is
no
kubernetes
Inc,
there's
no
company,
that
kind
of
is
centrally
driving
it
and
try
to
turn
it
into
a
profit
thing,
and
this
is
good.
So
it's
it.
Committees
has
jumped
kind
of
straight
from
crazy
into
boring.
It's
now
commodity.
It's
now
infrastructure.
B
It's
now
just
just
part
of
what
we
can
all
build
on.
What
that
means
is
it's
actually
really
hard,
I?
Think
for
companies
to
make
money
from
kubernetes
they
instead
make
money
using
communities,
but
what
that
means
is
it's
hard
to
get
contributors
on
the
core
I
think
where,
as
everyone
wants
to
use
it,
to
run
their
application,
stack
of
course,
and
that
I
think
a
success
problem
I
think
that's
that's
why
it's
meant
to
be,
and.
A
B
B
Wish
list
is
swap
I
think
we
need
to
put
swap
back
on
I'm
annoyed
that
we
don't
have
swap
on
our
clusters
and
my
RAM
constrained
laptops
are
really
struggling
because
I
don't
have
swap
and
I
think
it's
actually
a
fairly
easy
problem
to
solve
and
I
think
it's
been
swept
under
the
rug
somewhat
because
of
the
Google
history.
Google
doesn't
run,
swap
linear
machines
and
people
assume,
like
that's
that's
the
way
it's
meant
to
be
without
actually
questioning
whether
it's
the
way
it
should
be.
B
C
C
C
It's
half
the
time
is
like
Brian
will
rattle
out,
like
12
things
and
like
Brian
or
something
because
otherwise
I
have
to
go
back.
Listen
to
the
video
again
and
it's
hard
to
keep
up
with
them.
And
that's
one
thing
that
I
love
about
this
community
is,
you
know
there
are
so
many
people
who
are
so
good
at
the
things
that
they
do,
including
you
guess
so
and
I
get
an
opportunity
to
learn
from
them.
So
that's
that's
what
I
like
here.
A
B
C
I
mean
make
sure
you
have
a
good
set
of
good
Gmail
filters.
That's
what
I
used,
so
you
can
divide
up
the
stuff
into
things
that
people
are
pinging
you
on.
You
know,
I
mean
you
know
this,
that's
what
I'm
just
saying
it
for
others,
so
so
II,
you
know
you
know
prioritize
like
the
slack
messages.
First
and
then
you
have
Gmail
where
people
are
actually
picked.
Pinging
you
coming
from
the
pull
requests
and
issues
so
things
like
that.
C
Just
organize
your
work
a
little
and
make
up
your
days
into
chunks
where
you're
working
on
different
things,
because
you
are
never
working
on
one
one
thing
in
Cuban
artists.
There
is
always
like
for
me,
for
example,
I
look
at
jobs
that
are
broken,
I
go!
Look
at
old
issues,
I,
look
at
incoming
here's
I!
Look
at
what
got
merged!
C
You
know
last
night,
our
you
know
which
has
a
potential
to
break
the
CIA
changes,
so
I
I
break
my
day
into
different
chunks
and
I
do
different
things,
and
just
because
of
that
it
feels
like
I'm
in
many
places,
but
I'm,
not
so
that
that's
one
way:
I
I
cope
with
the
fire
hose
from
kubernetes
the
fire
hose.
That
is.