►
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Intro: Kubernetes (PM) SIG - Ihor Dvoretskyi, CNCF & Stephen Augustus, VMware
Kubernetes is one of the most high-velocity open source projects in the world, and one of the most unique features of this community project - that it has it's own PM team and PM process. SIG-PM, originally established as a Product Management Group, today covers multiple aspects of Product, Program and Project Management of Kubernetes. In this session, the SIG-PM co-chairs will provide a brief overview of SIG-PM history and basic principles, the areas of interaction with the Kubernetes community, together with the information on how to start contributing to Kubernetes as a PM.
https://sched.co/MPhv
A
A
Interesting
things
that
we
run
into
in
doing
trying
to
do
this
for
open
source
communities
is
that
a
lot
of
the
times
the
the
way
these
structures
are
built
in
a
company
are
built
around
the
need
to
drive
some
products
to
sell
it?
Something
right.
So
a
lot
of
the
demands
that
are
put
on
APM
team
are
done
based
for
the
you
know,
based
on
generating
revenue
right
so
within
the
the
open
source
community.
A
We
have
this
really
interesting
problem
where
the
the
requirements
and
and
the
motivations
of
the
people
who
work
in
open
source
are
not
necessarily
the
same
right.
Sometimes
they
work
for
companies.
Sometimes
they
work.
Sometimes
there
are
individual
contributors
right
and
so
trying
to
balance
the
way
we
structure
something
programmatically
to
build
a
product
and
an
open-source
project
is
kind
of
it's
interesting
and
then
the
short
version
is
we
don't
have
this
solved
yet
right.
So
it's
always
great
to
have
an
opportunity
to
talk
to
people
about
how
you
know
we.
B
A
Really
interesting
conversations,
Jason
I
have
really
interesting
conversations
and
Claire,
who
is
one
of
our
release.
Team
leads
for
1:15
is
also
a
program
manager
at
Vivid
old.
So
getting
getting
all
these
minds
in
the
right
rooms
at
the
right
times
is
is
kind
of
critical
for
our
success
right
now.
So.
B
B
How
can
you
build
a
product
using
using
the
okra
source
code
base,
but
how
can
you
build
a
communities
as
a
product
that
will
be
available
for
everyone,
but
with
the
with
the
best
practices
that
came
from
the
mercial
business
world
and
at
the
beginning,
the
initial
phase
of
6:00
p.m.
we
had?
We
have
bootstrap
the
features
process,
which
was
the
which
was
like
the
basic
process
of
how
can
you,
as
the
project
contributor?
How
can
you
put
your
feature
to
the
Benares?
We
define
the
features
themselves
as
what
exactly
is
a
feature?
B
B
It
was
called
features
and
like
kind
of
product
backlog
at
that
point
of
time,
but
after
the
evolution
of
kubernetes
esterases
project
and
who
is
the
growing
codebase
with
the
growing
number
of
contributors,
we
figure
out
that
they'd
dead
partners
in
that
ways
how
how
these
that
this
product
could
be
built
not
necessarily
should
follow
the
same
procedures
and
the
same
processes
that
are
suitable
for
the
commercial
companies
with
the
commercial
products.
So
we
decided
to
build,
we
decided
to
build
CBM
and
to
focus
it
on
on,
like
a
few,
a
few
different
errors.
B
First
of
all,
a
few
folks
from
the
community
proposed
to
bootstrap
the
process
that
is
called
caps,
which
is
decubitus
enhancement
proposal
process
and
which
superseded
features
process
because
features
process
was
it
were
different
complicated
for
the
final
end,
because
caps
versus
is
the
mr.
royal
process
with
the
huge
lifecycle.
It's
not
just
a
single
feature
which,
which
could
be
like
just
an
alpha,
beta
or
stable
feature,
but
it
cap
cap
process
is
the
huge
process
that
involves
all
the
stages
of
there.
B
A
A
You
know
we
decided
to
do
a
rebrand
essentially
of
like
what
what
is
it?
What
is
a
feature
and
what
is
an
enhancement
I'm,
so
the
way
that
we
define
it
right
now
is
a
feature
is
a
unit
of
work
that
can
be
scoped
to
single
kubernetes
release
cycle
and
an
enhancement
is
essentially
a
body
of
work,
multiple
features,
but
what's
what's
supposed
to
be
captured
in
that
is
it's
it's
the
idea
that
we
can
take.
A
We
can
look
back
at
this
and
have
essentially
a
steel
thread
of
what
happened
throughout
multiple
release
cycles,
the
conversations
that
the
implementations
and
motivations,
the
the
actual
implementation
history,
the
reasons
we
did
and
didn't
do
certain
things
right
and
capture
this
in
github.
So
someone
who's
like
hey
I.
Have
this
great
idea
for
a
feature
or
I?
Have
this
great
idea
for
an
enhancement
knows
ahead
of
time?
I
can
I
can
read
through
these
things
right,
so
you
know
what
we
wanted
to
do
here
is
I.
A
Are
working
on
caps?
We're
also
trying
to
expand
to
these
essentially
like
what
does
it
mean
to
be
a
sub
project
within
sig
p.m.
and
you
know
some
of
the
ideas
that
we
came
up
with
and
you
know
they
kind
of
are
spawned
from
these
critical
needs
of
the
project,
and
you
know
one
of
the
next
top
ones
is:
is
project
management
right?
How
do
we
do?
How
do
we
do
issue
triage?
How
do
we
do
backlog
grooming?
A
Is
it
possible
to
figure
out
how
to
burn
this
huge
backlog
down
as
a
community
project
and
then
also,
if
we
can
do
that,
can
we
start
to
elicit
what
kubernetes
roadmap
can
we
start
to
describe
the
themes
around
kubernetes
and
and
starts
of
map
this
out
in
instead
of
just
saying,
like
hey
Claire?
What
are
we
doing
for
the
you
know,
1:15
right
and
saying
like
what
are
we
gonna
do
straight
up
till
1:20
right,
so
that's
active
work.
A
A
Capture
the
motivation
for
these
changes
right
be
able
to
socialize
the
idea
and
actually
make
sure
that
we
keep
the
history
of
this
right.
It's
actually
committing
these
ideas
to
github,
so
people
can
reference
them
in
the
future
right.
We
also
want
to
make
sure
that
we're
creating
an
artifact,
that's
actually
consumable
by
not
just
developers
right,
it's
commonly
for
you
to
understand.
What's
going
on
in
people's
code,
their
docks
right,
you
know.
A
So
the
idea,
for
you
know,
part
of
the
idea
for
cap
is
to
be
able
to
an
end
user
should
be
able
to
use
that
a
developer
should
be
able
to
use
it.
The
media
should
be
able
to
read
a
cap
and
generally
get
the
the
gist
of
what's
going
on.
In
that
feature.
Right
and,
and
then
you
know
having
having
these
kind
of
like
these
documents,
these
I
guess
you
know.
A
If
you
want
to
call
these
or
you
know,
what
have
you,
the
the
you
know,
they're
inspired
by
you
know
the
Python
peps
and
the
rust
RFC's
so
like
having
these
documents
that
allows
you
to
kind
of
like
map
the
history
of
the
project
and
and
also
like
kind
of
get
an
idea
of
what
the
themes
are
for
the
project
right.
So
that's
the
idea,
yeah.
B
Just
just
when
I
add
again
about
small
difference,
like
the
procedural
difference
between
caps
and
features.
Is
that
the
way
how
how
how
attractive
features
before
and
how
the
we're
tracking
caps
these
days
and
how
do
we
put
them
on
github,
for
example?
So
previously,
with
the
features
we
will
hit,
the
future
through
poll
and
all
the
features
were
tracked
as
that
issues,
and
if
someone,
for
example,
the
future
owner
liked
to
update
the
future
description
or
would
like
to
update
some
information
about
the
process
for
that,
like
their
github
interface,.
B
The
way
of
changing
some
ways:
caps,
we
have
ever
seen
early
boot
on
github
as
there
is
the
file.
So
even
if
someone
wants
to
put
any
any
of
any
update
to
to
these
files,
you
will
be
able
to
track
all
this
information
in
the
github
history,
because
of
did
so.
It
also
will
simplify
to
anybody
would
like
to
get
any
history
like
to
get
to
get
any
any
information
about
what.
A
Know
looks
really
interesting.
Right
now
is
that
we
were
kind
of
in
this
split
state,
using
enhancement,
issues
or
feature
tracking
issues,
as
they
were,
formerly
known
and
and
and
really
having
the
kept
processes
at
one
place
that
you,
you
look
for
things
right.
So,
to
give
you
a
quick
rundown
of
the
enhancements
lead
role,
which
is
part
of
the
release
team.
So
you
know
we
considered
like
the
enhancements
lead
and
the
communications
kind
of
like
extensions
of
sig
p.m.
A
A
But
you
know
kind
of
the
job
is
to
run
through
this
repo
that
is
150
or
so
enhancement
tracking
issues
right
Cole,
eat
them
on
a
spreadsheet
and
reach
out
to
each
of
the
individual
owners
to
kind
of
understand
what
the
status
is
right,
and
you
know
just
from
hearing
that
it's
kind
of
clear
that
that's
not
a
sustainable
process
at
all.
Right
and
that's
kind
of
like
the
job,
it's
one
person
is
ultimately
responsible
for,
but
and
occasionally
they
have
a
team
behind
them
all
right.
A
A
So
on
the
the
project
management
and
we
want
there-
there
are
a
few
things
that
we're
trying
to
achieve
right.
So
you
know,
first
and
foremost,
we
want
to
trivialize
the
ability
to
do
project
wide
issue
triage
right,
whether
it
be
issue
triage,
are
actually
like
burning
down
things
right,
so
you
know
we're
at
we're
at
the
place
where
I
you
know,
just
rough
numbers
p.m.
has
about
50
issues.
Open
sick
release
has
about
100.
A
Network
has
about
150
issues
open
right,
so
these
are
like
really
really
large
numbers
and
the
you
know,
and
that
and
that's
just
three
six
right
so
considering
considering
the
work
that
is
required
from
top
level
reviewer.
So
these
are
people
who
are,
if
you've
heard
of
like
the
owners
files
within
kubernetes
repos,
so
essentially
that
those
owners
files
are
the
way
to
delegate
reviews
right.
A
So
the
problem,
the
problem
with
these
top-level
owners
is,
you
know:
they're
they're,
the
people
responsible
for
merging
for
doing
API
reviews
in
kubernetes
for
merging
top-level
entry,
coronet
II
stuff
for
doing
issue
triage
for
understanding
enough
of
the
history
of
the
project
to
be
effective
in
all
of
those
roles,
and
it's
kind
of
clear
from
hearing
that
that
you
know
it's
easy
for
those
people
to
be
bandwidth
constraint
right.
So
those
are
people
in
say
cars.
A
Those
are
people
in
you
know
those
are
people
like
you
know
the
the
Tim
Hawkins
the
Brian
grants
the
Jordan
Liggins,
where
they're
there's
only
so
much
time
in
the
day
right,
so
we
need
process
around
burning
down
and
figuring
out.
What's
the
path
of
you
know
how?
How
does
an
issue
inter
cooper
Nettie's,
and
how
do
we
solve
it
quickly
right
so
we're.
A
A
Like
I
was
mentioning,
we
can
start
to
have
these
things
like
sake
planning
events
and
and
actually
do
roadmaps
for
kubernetes,
and
you
know
this
is
a
common
question
that
we
get
from
the
media
right.
You
know,
I,
think
you
know
the
fun
ones
are
what's
kubernetes,
2.0
gonna
look
like,
and
the
answer
is
no
one
knows
like
that.
Probably
will
never
exist.
A
B
A
B
A
B
Sue
the
CEF
end
user
community
is
not
the
body
of
CPM
itself,
so
second
seem
safe
and
use
the
communities.
The
party
whizzing
CN
CF.
You
may
remember
today's
keynote
by
Sheryl
Hahn,
who
is
a
director
of
ecosystem
at
CNCs
and
she's,
actually
driving
any
case'
community
at
sea
ends
here.
At
the
same
time,
the
end
user
community
as
East,
is
the
state
group
of
people
and
the
group
of
companies
who
can
concert
to
give
us
as
the
given
Etta's
community
feedback
about
how
they
consume
communities,
how
they
do
they
use
communities.
B
What
are
their
desired
features
non
deserted
teachers
and
sick
p.m.
who's
in
the
canoes
in
the
communities.
Community
is
exactly
the
body
that
is
responsible
for
interacting
with
their
end
user
communities
and,
for
example,
in
you,
for
last
year,
six
pm
conducted
the
end
user
survey
collecting
the
feedback
from
the
end
users.
How
can
you?
How
do
you.
B
B
Different
company
that
is
consuming
to
Banaras,
so
it
was
extremely
valuable
feedback
for
us,
as
the
offices
project,
to
understand
how
people
actually
use
this.
This
codebase.
This
community
is
producing,
and
also
recently,
for
example,
the
working
group
LTS
that
was
recently
constructed
this
community.
B
B
A
Think
the
LCS
point
is
kind
of
interesting
because
you
know
the
the
the
cycle
that
we're
currently
on,
where
we're
on
a
quarterly
release
cycle
essentially-
and
things
come
up
right,
you
know
we're
in
a
release
cycle
right
now
and
you
know
most
of
our
release.
Team
members
are
here,
it's
the
middle
of
queue
and
you
know
very
soon,
it'll
be
cubed
on
Shanghai
right.
So
there
are
a
lot
of
things
to
consider
in
like
trying
to
wrangle
the
right
people
in
the
right
places
at
the
right
time
to
get
these
things
done.
A
So
when
you
start
to
consider
something
like
LCS,
the
the
burden
is
is
greatly
increased
right
on
a
release
team
on
patch
release
managers,
branch
managers
right
because
the
maintenance
becomes
that
much
higher
rate.
Okay
right
now
we
have
a
you
know.
Right
now
we
have
a
if
we
select
a
patch
release
manager.
That
means
they
maintain
patches
for
as
long
as
that
version
of
kubernetes
exists
right.
So
if
you
sign
up
for
that
role,
you're
signing
up
for
a
year
of
work.
A
So
that's
something
to
be
aware
of,
and
that's
the
you
know,
part
of
the
reason
that,
on
the
sink
release
side
we're
working
on
creating
a
patch
release
management
team
as
opposed
to
just
a
person,
because
that's
not
great,
but
it's
so
some
of
these
things
that
come
up.
You
know
it
also
means
like.
If
we
change
the
way
we
do
release
cycles.
A
A
So
we
worked
closely
with
the
LTS
working
group
to
generate
that
survey
and
I
didn't
want
to
present
any
of
the
results
here
today,
because
they
also
have
a
session
and
that's
probably,
what
they're
gonna
be
doing
in
their
session,
so
go
to
the
LTS
working
group
session,
I
believe
it's
on
Thursday,
it's
like
afternoon,
Thursday
and
check
out
it's
a
great
session
to
be
yet
and
then
finally,
external
communications
right.
So
you
know
this
is
this:
is
kind
of
fuzzy
purposely
fuzzy?
A
What
we
want
to
be
able
to
do
is
you
know
there
are
lots
of.
There
are
lots
of
niceties
that
you
get
from
being
part
of
the
release
cycle.
You
know
dealing
with
release,
notes
and
and
generating
that
communication.
There's
a
there's
a
whole
like
CN
CF,
release
train.
That
happens
right.
So
it's
like
blog
that
it's
like
blog
post
webinars,
all
these
things,
podcasts
right
and
it's
all
centered
around
the
police
and
the
things
that
are
going
into
the
release
right.
A
So
what
we
noticed
is
that
there's
some
dissonance
and
between
the
things
that
are
going
into
the
release
and
the
things
that
happened
for
things
within
the
release
cycle
right,
which
are
not
necessarily
the
same
thing
right.
So
what
we're
hoping
to
do
eventually
is
is
provide
a
standard
way
of
giving
this
information
to
the
community.
Great
some
some
form
that
states
can
subscribe
to
your
working
groups
can
subscribe
to
to
be
able
to.
A
You
know,
celebrate
the
work
that
they've
done
and
then
also
like
call
for
volunteers,
or
you
know
just
generally
have
that
out
in
the
ethos
right,
so
you
know
again
whether
it's
the
blog
posts,
the
podcast
the
surveys
right,
but
also
like
generating
a
me
of
doing
this.
It
has
a
consistent
voice
right.
This
is
the
voice
of
like
the
kubernetes
project.
I.
A
Don't
think
that
we
have
that
I
think
we
have
lots
of
community
leaders
that
understand
the
goals
of
the
project,
but
not
necessarily
if
someone
was
to
you
know
if
someone
who's
come
to
the
project
fresh
and
say,
like
hey,
I
have
to
write
a
blog
post
right
like
this
is
exactly
the
template
that
you
need
or
if
you
need
to
you
know,
if
you
need
to
publicize
this
new
feature
that
you're
working
on,
like
you
have
a
forum
to
do
that
right.
So
we're
close,
we're
close,
but
a
lot
of
I.
A
B
At
the
same
time,
C
and
C,
if
is
there,
is
the
body
that
is
responsible
for
that,
but
it
shouldn't
be
the
single
body,
because
everybody
in
the
community,
every
every
single
person,
even
the
individual
or
any
other
company,
can
can
do
that
as
well,
like
promoting
promote
in
Jerusalem
Luise,
promote
and
Royce
an
issuance
of
defibrillators
community,
the
podcasts
webinars,
and
so
this
initiative
was
bootstrap
from
the
release
process.
But
these
days
the
the
goal
of
6:00
p.m.
is
to
is
to
make
it
independent
enough
from
there
really
strain
itself.
A
Cool
so
I
think
I
think
that
might
be
it
on
our
salon
and
so,
if
you're
interested
in
getting
involved
in
sig
p.m.
and
I
I
recognize
some
cases
from
the
same
goals
join
our
mailing
list
join
the
slack
channel.
So
if
you're
not
already
on
proven
80
slack,
that's
slack
that
cait's
dot
io
the
sig
PM
channel
is
hashtag
yeah
and
our
meetings,
our
bi-weekly.
B
A
So
it's
it's!
How
many
cells
can
you
juggle
at
the
same
time?
So
I
think
you
know
this
is
this
is
kind
of
indicative
of
the
the
same
bandwidth
constraint
things
that
we're
talking
about
in
the
project
right,
so
I
think
you
know,
because
caps
are
in
the
critical
path
for
everyone
submitting
changes
to
kubernetes.
It's
important
that
we
fix
that
we
make
that
make
sense
to
everyone.
So
I
think
we're
at
a
point
where
people
understand
the
value
of
caps
of
day
to
day.
A
There
is
some
friction
in
the
process
and
it's
you
know
it's
partially
around
deadlines.
It's
part
that
some
of
it's
around
formatting,
some
of
its
around
you
know.
Can
we
build
more
automation
around
this
process.
So
you
know
there
are
a
few.
A
few
different
things
to
tackle.
I.
Think
that
you
know
one
of
the
next
big
things
that
we
need
to
tackle
is
start
looking
at
project
management.
Very
deeply,
because
it's
clear
you
know
there
is
a
there's.
A
A
great
thread
from
Tim
Hawkins
started
this
one
off
and
it's
called
the
the
subject
line
is
deadlines
are
horrible
and
so
that
one
is
that
one's
a
great
read
and
you
know,
and
it's
clear
from
the
discussion
there.
You
know
these
are
all
what
I
like
to
call
like
project
elders
reading.
You
know
telling
you
telling
you
like
help
I'm
in
pain,
so
so
I
think
that
you
know
that's
the
next
most
important
thing
that
we
have
to
do.