►
From YouTube: KBE Insider (E10) - Sascha Grunert
Description
KBE Insider interviews Sasha Grunert, senior software engineer at Red Hat. Sasha joined the open source community in 2018, contributing to open source projects as well as evangelizing Kubernetes-related technologies. In this KBE Insider episode we’ll focus on Sasha’s contributions to the open source projects such as Kubernetes, CRI-O, and his work on SIG Release.
A
Well,
hello:
everybody
welcome
to
another
episode
of
kubernetes
by
example,
insider
where
we
like
to
interview
the
actual
engineers
who
work
on
various
aspects
of
kubernetes,
so
that
we
can
get
a
sense
of
where
they
feel
like
the
project
is
going.
Obviously,
you
can
always
talk
to
product
managers
or
project
managers.
You
know
and
those
kinds
of
people.
A
However,
the
both
the
good
thing
and
the
bad
thing
about
open
source
is
often
the
engineers
have
direct
control
over
what
actually
lands,
and
so
we
find
it
useful
to
actually
talk
to
the
engineering
community
to
see
where
you
know
what
they're
passionate
about
what
they
want
to
see
happen
next,
because
I
think
it
gives
us
some
better
insights
into
whether
you
know
what
what's
going
to
happen.
Basically,
so
I
think
we're
on
our
10th
episode
and
I
would
like
to
welcome
our
prom
promoted.
A
I
don't
know
co-host
of
cevita
who
was
on
the
show.
I
want
to
say
episode.
Six
rings
a
bell,
but
you
know
savita.
Do
you
wanna?
Oh
sorry,
I
I
forgot
to
introduce
myself,
I'm
langdon
white.
I
am
currently
a
faculty
member
at
boston
university
in
the
school
of
or
the
faculty
of
computing
and
data
sciences.
I
teach
data
science
as
well
as
some
practical
classes,
doing
basically
engineering
software
and
machine
learning
projects
for
third
parties.
A
B
Sure
hi
hi
everyone,
my
name,
is
avitaro
gunatan.
I
am
thrilled
to
be
here
on
the
show.
I
think
I
was
on
episode,
5
6.
It
was
like
five
six
months
ago
and
it's
a
the
same
excitement
that
I
felt
sitting
on
the
other
side.
I
am
a
senior
software
engineer,
working
on
data
protection
area
and
I
also
contribute
to
upstream
kubernetes
and
conveyor
community.
B
C
Yeah,
hey
thanks
sarita.
Now,
thanks
for
having
me
on
the
show
and
nice
to
see
you
again,
I
mean
we
are.
We
already
know
each
other
from
the
release
team,
so
I'm
sasha
one
of
the
contributors.
I
consider
myself
as
a
contributor
to
kubernetes.
I
work
in
multiple
directions
on
the
project,
so
I'm
part
of
sig
release
acting
as
a
chair,
but
I'm
also
very
working
with
red
hat
on
the
lower
level
bits
of
kubernetes.
So
I
contribute
to
projects
like
the
container
on
time,
cryo
and
some
related
projects
as
well.
A
And
I
I
would
like
to
clarify,
I
appreciate
you
mentioning
you
know,
kubernetes
community
member,
because
really
that's
who
we
interview.
We
actually,
you
know
sometimes
it's
engineers,
but
you
know:
we've
had
people
who
are
kind
of
like
on
the
dev
advocacy
side,
we've
had
community
liaison
type
people
as
well.
I
just
have
the
old,
you
know
kind
of
open
source
contributor,
auto
language
right
around
engineering,
which
is
a
bad
habit,
but
one
of
these
days
I'll
I'll
fix
it.
C
Hidden
behind
this
huge
open
source
effort
and
many
folks
are
thriving
in
upstream
and
yeah,
and
this
got
me
into
open
source
actually,
and
I
was
more
and
more
interested
in
creating.
I
started
with
creating
my
own
projects
and
sure.
Surely
I
failed
because
I
feel
very
often
because
those
projects
yeah
nobody
really
had
interest
in
or
they
are
solving
a
particular
problem
which
is
not
really
which
is
out
of
scope
of
other
problems.
C
Then
I
started
to
understand
that
I
have
to
contribute
that
I
have
to
bring
myself
into
a
community,
contribute
to
them
and
then
get
something
back
in
terms
of
being
part
of
that
community
and
and
at
some
other
company
there
was
a
a
project
where
we
had
to
where
we
had
to
create
a
microservice
application
which
had
the
boundaries
that
it
should
be
highly
available
and
it
should
be
for
free
for
sure.
So
we
had
been
choosing
kubernetes
and
I
had
the
chance
to
work
myself
very
deeply
into
the
kubernetes
deck.
C
So
this
was
at
the
time
of
1.12
so
in
early,
not
in
the
early
stages
of
kubernetes,
but
it
was
some
years
ago
and
therefore
I
worked
myself
into
kubernetes.
We
created
an
application
stack
and
I
had
a
chance
to
teach
my
knowledge
to
my
colleagues,
and
this
was
really
an
amazing
experience
from
my
point
of
view.
C
But
in
the
end
it
wasn't
enough.
So
I
really
wanted
to
be
part
of
that
community
and
I
was
actively
searching
for
a
company
who
is
contributing
to
that
project
and
therefore
I
came
to
I
came
first.
I
came
to
souza
and
then
I
came
to
red
hat
and
I'm
really
happy
that
there
are
companies
around
those
projects
and
who
continuously
contribute
to
kubernetes.
A
So
I
wanted
to
ask
you
about
that.
The
well
first
of
all,
I
think,
I'm
pushing
25
years
so
yeah.
I
know
what
you
mean,
but
next
I
was
going
to
say
you
mentioned
community.
I
don't
know
seven
eight
times
in
that
statement.
What
like
what
is
it
about
a
community?
What's
what's
it
giving
you
that's
kind
of
over
and
above
you
know,
kind
of
writing
code
right
or
you
know,
or
getting
validation
for
building
software.
C
Yeah,
I
would
compare
to
to
actually
using
the
possibilities
the
internet
or
the
modern
internet
provides
to
us
right.
So
information
is
everywhere,
so
we
can
learn
everything
without
having
the
need
to
interact
with
people,
but
I
would
say,
having
the
opportunity
or
learning
by
coding.
Software,
for
example,
makes
it
way
more
easy
to
fastly,
learn
things
directly
from
people
who
already
know
it
and
that's
the
and
that's
the
real
benefit.
So
I
don't
see
it.
A
A
You
know
like
what
I
often
say
is
like
if
you,
if
you
watch
an
expert,
fail
watching
them
and
what
they
do
to
get
out
of
that
failure.
I
think,
teaches
you
as
much
or
more
about
like
how
to
be
a
good
programmer
than
than
almost
any
other
mechanism,
because
it's
it's
so
hard
to
teach
that,
and
it's
one
of
the
things
you
only
get
really
through
experience,
but
I
think
with
something
like
twitch.
You
get
a
lot
of
opportunities
to
kind
of
see
it
in
action.
A
I
was
actually
talking
about
this
with
a
medical
doctor.
The
other
day
and
medical
doctors
do
something
similar,
except
they
basically
fake
these
scenarios
right
and
then
they
go
through
and
it's
kind
of
a
continuous
training
exercise,
but
they
they
fake
them.
The
nice
thing
about
the
twitch
is
a
nobody.
You
know
we
can
do
it
for
real
and
nobody
has
to
like
die
or
be
ill.
You
know
to
to
learn
from
from
real
examples
in
a
sense,
so
I
was
gonna
hand
it
to
savita.
B
Yeah
yeah
sure,
so
I
wanted
to
add
something
to
what
sasha
said
like
it's.
Also,
the
so
for
the
kubernetes
community
is
a
little
special
for
me.
I
was
searching
for
the
place,
a
sense
where
I
would
belong
and
I
found
it
there
in
addition
to
all
the
innovation
and
stuff
like
that
on
your
bed.
It
is
same
way
for
so
many
folks
who
are
like
looking
for
that
kind
of
you
know
bonding,
even
though
I
haven't
seen
many
I
haven't
seen
cheshire.
B
I
haven't
seen
like
many
of
the
contributors
have
worked
with
like
two
years,
but
for
me
I
just
feel
like
it's.
My
happy
place
just
put
it
in
like
two
words,
so
I
do
have
one
question
for
sasha.
This
is
something
that
I
struggle
with.
I
know
like
you,
have
a
full-time
job.
You
are
a
maintainer
for
trial.
You
do
an
amazing
job
with
say
comp
profile
and
the
new
features
that's
getting
released
in
kubernetes.
I
think
it's
gone
beta
or
it
has
gone
better.
The
default
like
com
profile.
B
C
But
to
me
it's
it's
like
finding
or
focusing
for
a
dedicated
amount
of
time,
so
I
started
contributing
to
cryo,
but
I'm
completely
sure
that
I
can't
keep
up
the
same
pace
in
the
beginning,
which
I
had
in
cryo,
for
example,
when
starting
to
contribute
or
continuing
to
contribute
to
kubernetes
or
focusing
on
features.
So
I'm
really
happy,
for
example,
that
in
kubernetes
this
whole
release
cycles
exist.
So
we
can
have
dedicated
points
in
time
where
we
can
focus.
C
I
can
focus
on
enhancements
in
the
beginning
when
I'm
planning
for
enhancements
in
the
next
cycle,
for
example.
I
can,
for
example,
also
take
myself
off
of
those
enhancements
for
one
complete
release
cycle
and
then
focus
like
I
did
for
124
focus
on
the
container
image
signing
project.
So
I
I
just
can
only
focus
myself
on
a
bunch
of
projects
in
parallel
and
when
we
talk
about
this
on
a
weekly
basis,
then
only
on
one
or
two
per
week
so-
and
I
think
the
key
is
to
not
not
to
to
be
fast
everywhere.
C
But
I
think
the
key
is
to
make
progress
everywhere
over
a
certain
amount
of
time.
So
the
community
kind
of
trusts
you
that
you,
when
you
lay
down
some
work,
that
you
come
back
and
pick
it
up
and
if
you
have
committed
on
something-
and
you
also
come
later
on
and
pick
those
work
up
and
finish
it
and
bring
it-
bring
it
over
the
finishing
line,
which
is
a
what
one
of
the
biggest
problems
we
have
probably
in
kubernetes,
then
they
trust
you
that
you
will
do
is
also
for
other
parts
of
the
project.
C
C
A
C
I
mean
iterating
is
also
a
good
point
right,
so
not
everything
has
to
be
perfect
in
the
first
place
I
mean,
if
you
speak
about
enhancements
in
kubernetes,
then
they
have
to
be
on
a
certain
quality
level,
so
they
have
to
be
docs
and
tests,
and
things
like
that.
But
if
we
speak
about
the
container
image
signing
for
example,
then
we
knew
that
that
we
can
work
iteratively
over
the
whole
cycle
because
it
will
not
go
live
until
the
end
of
the
cycle.
Something
like
this.
A
A
C
Yeah,
I
really
so.
This
is
a
really
great
project.
From
my
point
of
view.
We
started
it
probably
two
years
ago
now
and
the
main
idea
was
that
we
have
one
issue
in
kubernetes
that
we
for
when
we
speak
about
second
profiles,
that
we
have
to
distribute
them
over
a
whole
cluster
that
all
profiles
have
to
be
available.
C
I
know
that
there
are
some
other
solutions
around
it:
how
to
solve
it,
but
the
overall
end-to-end
user
handling
is
not
that
optimal
and
it's
not
even
optimal
right
now,
and
there
were
some
approaches
in
solving
that
which
had
some
other
drawbacks,
and
I
was
just
sitting
together
with
those
folks
who
were
working
on
the
on
the
overall
solution,
and
then
we
started
to
think
about
hey.
Can
we
solve
this
as
an
out
of
three
enhancement
and
then
later
on,
decide
if
we
can
bring
it
back
into
kubernetes?
C
We
are
now
at
the
point
where
we
have
it
as
an
out
of
three
enhancement
and
we
have
a
bunch
of
features
in
it
which
which
are
great,
and
now
we
are
at
a
point
where
we
have
to
better
prove
those
features.
So,
for
example,
we
are
working
right
now
on
bringing
it
to
operator
hub,
io
and
then
later
on,
integrating
it
as
an
optional
feature
into
openshift,
to
get
more
user
to
get
a
bit
bigger
use
of
base
on
the
operator
and
and
the
overall.
C
So
it
solves
the
the
issue
of
not
only
distributing
second
profiles.
I
think
the
main
issue
with
zaccom
is
that
it
starts
with
creating
the
profiles
for
workloads,
because
people
don't
understand.
I
even
didn't
understand
it
in.
If
I
had
initially
read
some
articles
about
second,
how
it
works
under
the
hood
and
how
it
how
to
create
such
a
profile
right,
how
to
decide-
which
this
call
should
I
add,
or
for
which
architecture-
and
things
like
that.
So
we
are
work
evaluating
different
ways
of
recording
those
profiles
and
we
have
two
right
now.
C
We
can
pass,
for
example,
log
files,
and
then
we
have
also
utilizing
eppf
to
record
those
profiles.
So
these
those
are
optional
and
the
interesting
part
is
now
that
we
also
have
maintainers
which
really
work
tightly
together
with
c
linux
policies
and
they
integrated
the
recording
feature
for
the
linux
policies
as
well.
So
I
mean
the
project
is
growing
and
also
the
folks
working
on
app
armor,
for
example.
So
we
have
those
three
security
features.
A
You
know
another
example:
right
is
the
not
not
gcc,
but
I
think,
yeah
like
basically
the
translations
for
the
compiler.
I
think
it
was.
The
compiler
also
had
kind
of
a
similar
problem.
You
know
where,
basically,
you
had
like
every
translation
in
the
bucket
and
breaking
it
apart
is
you
know
a
whole
exercise
in
and
of
itself?
So
I
think
you
have
you
have
the
problem
kind
of
on
both
sides
of
that
equation,
both
it's
like.
How
do
you
get
them
out
there,
but
then
also?
A
How
do
you
get
out
there?
Only
the
kind
of
minimal
amount
of
stuff
that
you
want
to
land
in
the
target,
so
it's
been
a,
I
think
I
would
say
going
on.
You
know
seven
or
something
years
now.
You
know
a
major
effort
to
really
think
about
how
the
distribution
works,
to
redistribute
things
in
ways
that
we
can
target.
You
know
a
containerized
world.
A
I
just
say
it's
one
of
those
things
I
haven't
worked
on
it
myself,
all
that
much
I've
been
more
like
a
beneficiary
of
it,
but
I
find
it
fascinating
that
you
know
it's
it's
just
so
different
and
I
think
people
don't
realize
how
monolithic
our
operating
systems
have
actually
become,
even
though
they
have
continuous
delivery.
You
know
and
things
like
that,
because
you
know,
but
it's
still
monolith.
You
know.
C
C
A
B
Yeah,
so
I
wanted
to
add
that
the
second
profile
when
I
so
I
used
to
be
an
administrator
in
a
previous
life.
So
all
I
cared
about
is
like
having
something
that
would
work
without
having
to
know
everything
about
it,
because
if
I
start
digging
deeper
and
deeper,
then
I
wouldn't
be
able
to
support
any
features
for
the
platform
right.
B
I
was
deciding
and
working
and
supporting
a
platform
service
using
kubernetes,
and
these
little
things
are
too
wet
and
security
is
an
administrator's
nightmare,
everyone's
nightmare
and
it
has
to
be
implemented
right
and
when
I
look
at
the
sitcom
profile
operator,
all
I
see
is
that
it's
a
one-stop
shop
because
it
tackles
so
many
things.
So
many
profiles
and,
like
you
mentioned
like
three
of
those
things
app
armor
and
two
other
things
which
I
cannot
think
about.
B
The
one
thing
that
I
can
think
about
is
that
I
did
see
something
related
to
psp
and
now
that
we
have
duplicated
psp
or
we
are
going
to
duplicate
psp
in
1.25,
just
like
port
security
policies,
and
we
have
any
idea
to
support
part
security,
admission,
controller
and
pss
port
security
standards.
As
a
part
of
the
operator,
so
that
is
one
thing
that
I
wanted
to
know,
because
I
was
just
curious
not
that
I
work
in
the
admin
world
anymore,
but
I
like
to
like
you
know.
C
C
C
They're
already
duplicated
and
they
will
be
removed
yeah
in
favor
of
those
admission
controllers.
Yes,
yeah
I
mean
we
would
like
to
go
making,
for
example,
making
kubernetes
more
secure
out
of
the
box
is
something
we
probably
cannot
do
by
just
moving
features
from
the
security
profiles
operator
into
kubernetes,
because
those
features
are
already
so
complex
that
they
have
a
different.
So
we
would
never
get
them
in,
but
we
have
to.
We
have.
C
C
From
my
point
of
view
would
be
to
make
this
visible
to
the
end
user,
so
make
it
api
aware
and
provide
some
upgrade
and
downgrade
paths,
and
this
is
the
tricky
part,
so
we
could
probably
say:
okay,
we
apply
the
feature
only
to
new
workloads,
but
we
cannot
really
distinguish
new
workloads
from
rescheduled
workloads.
So
the
complexity
in
kubernetes
would
also
then
increase
a
lot
so,
but
but
really,
I
think.
C
C
That's
a
thing
where
I
am
I'm
not
sure.
If
the
I
mean
we
say,
we
support
a
bunch
of
releases
for
some
features
and
then
we
remove
it,
and
then
we
always
have
some
troubles
about
people
not
going
this
upgrade
path,
and
usually,
for
example,
openshift
solves
this
by
adding
custom
logic
on
top
of
it,
and
I'm
not
sure
how
can
my
question
would
be,
then
how
could
we
achieve
something
like
this
in
kubernetes
without
having
a
having
this
whole
upgrade
down
kit
logic?
A
So
for
our
viewers,
though,
like
we
should
definitely
it's
it's
something
to
be.
Excuse
me
very
aware
of
kind
of
in
the
125
time
frame
of
you
know.
This
may
impact
your
workloads.
This
is
something
you
should
be
following.
Otherwise
you
might
get
bitten.
What
do
you?
What
do
you
foresee
as
a
way
to
make
sure
the
community
is
aware
of
those
changes.
A
C
C
We
always
have
dedicated
packages,
but
we
can't
really-
and
we
have
need
some
testing-
that
which
ensure
that
the
packages
work
together
right
in
their
versions,
and
I
don't
think
that
kubernetes
is
made
to
go
this
route
and
that's,
I
think,
that's
also
why
we
mainly
decided
to
go
from
four
releases
to
three
releases
per
year,
because
the
architecture
thrives
the
way
how
we
release
it
and
that's
and
that's
the
way
how
it
is.
I
mean.
C
A
C
Yeah,
I
think
we
have
to
when
did
we
start
with
that?
I
mean
there
were
some
parallel
efforts
in
kubernetes,
which
we
consolidated
into
our
roadmap
and
vision,
and
this
vision
kinda
drives
the
our
overall
effort
in
release
engineering
and
the
release
engineering
sub
project
of
kubernetes,
and
one
part
of
that
is
the
salsa
compliance
and
the
yeah
the
sensor
complaints,
and
before
that
we
were
rewriting
our
whole
tooling
from
bash
into
some
golang
libraries
to
be
reusable
and
split
them
across
three
different
repositories.
C
Right
now-
and
this
gave
us
the
opportunity
to
after
that,
after
the
transition
to
do
some
refactoring
on
top
of
that
and
also
think
about
the
salsa
compliance
and
the
secure
supply
chain
overall
story-
and
this
is
also
the
headline
for
our
roadmap
and
vision.
So
it's
about
an
introspectible,
introspectible,
consumable
and
secure
supply
chain
and
the
secure
part
of
that
is
the
overall
salsa
compliance.
When
we
speak
about
container
image,
signing
artifact
signing
and
also
image
promotion,
all
those
topics
are
huge
for
itself
and
we
own
a
bunch
of
caps.
C
I
think
three
caps
around
this
topic
right
now,
but
the
idea
is
to
iteratively
start
working
on
it
and
we
started
with
container
image
signing
in
124,
which
is
now
implemented.
So
it's
great
it's
a
great
achievement
of
our
special
interest
group
from
my
point
of
view,
because
we
managed
to
get
all
folks
on
board
so
release
managers
who
said
they
struggle
with
finding
out
what
they
want
to
do.
B
I
am
so
excited
at
the
communities
release
engineering
effort
into
like
salsa,
and
I
also
like
do
a
little
bit
of
sub
project
running
with
security,
and
I
see
a
lot
of
folks
involved
from
six
security
and
there
is
this
cross
collaboration
between
singulis
and
seek
security
and
many
other
things
within
a
kubernetes
ecosystem.
So
for
those
who
are
watching
a
sig
is
a
special
interest
group
like
verticals
within
kubernetes
community,
where
eat
sig
focuses
on
one
component
and
they
work
together.
B
So
I
just
wanted
to
give
a
little
background
for
the
viewers,
so
I'm
excited-
and
there
is
also
that
the
recent
release
I
know
the
release,
121
rc,
the
release
candidate.
I
think
it's
supposedly
con
container.
It
supports
the
container
images.
The
container
images
have
been
signed
and
I
was
so
excited
to
see
that
you
can
verify
it.
So
there
was
this
tweet
by
one
of
my
good
friend
who
created
this
little
demo
and
you
can
see
it
in
action,
and
that
is
like
it
made
me
smile.
C
C
Do
we
have
to
provide
something
to
verify
the
signatures
or
what
can
we
do
and
but
at
the
end-
and
this
brings
me
back
to
my
contributions
to
cryo,
because
the
container
runtime
folks,
which
also
maintain
the
lower
level
bits
of
cryo
and
are
working
in
parallel
on
bringing
container
image
signing,
so
we
will
probably
see
it
again
in
cryo
and
then
we
can
close
the
loop
right
so
from
the
from
the
uppermost
container
image
signing
back
to
the
verification
of
the
signatures
directly
in
the
runtime.
So
this
is
really.
A
C
I
think
it
was
really
necessary.
We
kind
of
in
the
past
release
cycles,
we
kind
of
pressed
the
release
cycle
into
the
time
frame,
and
now
we
we
don't
gain
that
much.
We
gain
four
to
five
weeks
per
release,
probably-
and
this
gives
us
a
huge
relaxation
on
the
sick
release
side
and
also
on
the
development
side
I
already
can
for
my
personally.
I
can
plan
better
with
features
it
wasn't.
C
C
We
had
a
delay
this
time
about
two
weeks
because
of
the
golang
miner
update
and
now
we
are
at
a
point
we.
So
this
discussion
is
pretty
fresh,
but
we
have
to
reach
a
consensus
if
we
want
to
start,
for
example,
after
kubecon,
because
we
always
consider
the
time
in
kubecon
as
something
where
developers
are
not
available
at
all,
and
this
would
also
give
us
a
bit.
C
I
mean
it's
a
bit
pressuring
because
we
have
to
put
the
next
two
releases
until
the
end
of
the
year
and
we
don't
want
to
stop
in
december
20s
or
something
like
this.
So
mid-december
is
the
deadline
for
that,
but
we
can
also
go
for
a
shorter
release,
so
this
is
totally
totally
doable.
I
mean
it
gives
us
more
flexibility
in
planning.
That's
that's
really
great.
A
A
C
A
B
I
wanted
to
ask
like:
did
you
hear
anything
positive
about
moving
the
cadence
and
you
know
from
four
to
three:
was
a
shocker
for
so
many
folks
and
everyone's
getting
used
to
it.
Personally,
I
love
three
releases
per
year
because
I
cannot
finally
catch
up.
Otherwise
it
was
like.
I
was
just
running
behind
the
features,
without
even
understanding
what
they
are.
That's
how
it
felt
for
me,
but
from
other
company
perspective
another
developer
perspective.
B
I
understand
that
now
they
have
to
plan
so
like
ahead
and
then
implement,
because
if
they
miss
one
release
and
then
they
just
have
to
like
wait
for
another,
that
makes
the
featured
graduation
like
the
cycles
longer
and
it
can
push
the
graduation
graduations
the
feature
and
like
the
same
thing,
I'm
repeating
myself.
So
do
you
have
any
insight
to
that
or
like
did
you
hear
anything
positive
about
it,
negative
about
it?
If
you
could
share
with
us,
that
would
be
awesome.
C
Yeah,
I
I
just
heard
positive
things
to
be
honest,
but
maybe
that's
because
I'm
acting
as
a
chair
for
secretly
no
I'm
I'm
thinking
so
from
from
our
survey
perspective.
So
from
downstream
consumers.
We
only
got
positive
feedback
so
far,
because
it
seems
to
be
more
relaxing
right
now
and
it
gives
more
time
for
planning
and
retrospectives
and
things
like
that
and
on
the
other
side,
I
think
it
did
not
change
anything
to
the
cycle
itself.
C
When
it
comes
to
deadlines
and
the
stress
before
deadlines,
I
mean
we
also
have
those
product
readiness
reviews
which
have
to
be
done.
This
enhancements
deadline,
and
things
like
that.
So
I
think
it
changes
not
much
on
that
perspective
and
on
the
graduation
itself.
I
think
it
was.
It
was
good
because
I
I'm
not
sure
if
that's
already
done,
but
I
thought
that
sick
architecture
is
having
a
better
guidance
to
when
to
graduate
features
and
how
many
release
cycles.
We
want
to
wait
before
graduating
feature,
for
example.
C
What
recently
changed
is
that
better
releases
are
not
part
not
enabled
per
default
anymore,
because
they
are
not
considered
as
stable
and
and
a
decision
like
this
already
has
a
huge
impact
on
the
community
right.
So
because,
if
I
have
now
an
alpha
feature
and
I
plan
to
graduate
it
because
then
I
would
like
to
provide
it
to
my
downstream
consumers
and
it's
not
enabled
by
default
anymore,
so
I
have
to
enable
it
manually
now.
That's
one!
C
B
Thank
you
for
sharing
that
with
us,
and
I
also
want
to
give
another
shout
out
to
the
well
run,
release
team
shadow
program,
which
is
so
successful
that
I
hope
to
see
it
getting
adopted
by
other
other
sikhs
within
kubernetes
org,
which
is
already
happening.
I
see
a
lot
of
shadow
programs
and
I
think
it's
so
successful
that
it
can
be
adopted
across
many.
Many
other
was
the
open
source
projects
as
well.
C
B
A
Yeah,
I
was
gonna,
actually
take
it
slightly
differently,
so
upcoming
releases.
What
are
you
most
excited
to
see
land,
whether
you
know
whether
the
things
you
you
did
or
you
know
things
that
other
people
have
done?
What?
What
are
the
things
that
you
think
are
are
most
important
to
be
watching
kind
of
going
forward.
A
C
Yeah,
I
mean
that's
an
interesting
perspective
right
because,
from
a
developer
perspective,
I'm
already
in
planning
for
125
and
for
folks
who
want
to
upgrade,
they
are
probably
reading
the
release
notes
for
124.
So
I'm
really
excited
about
the
docker
shim
removal,
because
we
can
clean
up
some
code
afterwards
and
not
only
from
the
from
a
user's
perspective
but
from
a
code,
organizational
perspective
and
also
testing,
and
so
this
has
a
huge
impact
to
to
test
infra
and
things
like
that.
C
It's
great
yeah
for
124
for
sure
I
I
love
that
we
have
the
container
image
signing
in
that's
that's
also
a
great
feature
and
for
125
I'm
really
looking
forward
to
also
not,
for
example,
small
things
like
removing.
The
second
annotations
is
something
which
really
reduces
the
logic,
for
example,
when
it
comes
to
converting
those
annotations
to
the
profile
fields
and
the
other
way
around.
So
you
can
remove
a
bunch
of
code,
and
this
also
will
automatically
increase
the
security,
because
having
less
code
means
higher
security
right.
C
C
It
has
only
some
fields
have
been
added
over
the
past
years
and
that's
also
now
a
point
where
we
can
remove
the
support
for
one
v1
alpha
2
and
therefore
we
can
remove
huge
amount
of
conversion
code
in
between
those
versions
and
things
like
that.
So
that's
that's
something
where
I'm
really
looking
forward
to,
but
it's
probably
not
the
the
most
fanciest
features
so.
A
Yeah,
well
I
mean
that's
kind
of
that's
kind
of
why
I
asked
the
question
right
is
that
I
want
to
know
what
your
perspective
of
the
of
what's
kind
of
upcoming.
The
other
thing
I
wanted
to
bring
up
is,
as
I
know,
you
both
are
very
aware-
is
there's
a
certain
event
happening
in
spain
relatively
soon,
and
I
think
you're
both
pretty
involved.
I
wanted
to
be
involved
and
then
couldn't
actually
make
the
travel
happen,
and
so
can
you
tell
me
a
little
bit
about
kubecon.
A
C
Yeah,
unfortunately,
I
cannot
attend
in
person
as
well.
I
kind
of
I
mean
there
was
a
bit
of
back
and
forth
and
discussion,
and
then
I
decided
to
not
go
because
you
can
get
all
this
content.
Also
virtually
I
know
there
is
this
whole
get
together
and
things
like
that,
but
yeah
I'm
looking
forward
to
to
have
a
chance
in
in
north
america
by
the
end
of
the
year,
but
yeah
we
have.
I
have
three
talks,
but
it's
not
like
that.
C
C
The
second
talk
where
we,
where
I
can
present
how
we
can
craft
custom
second
profiles-
and
this
was
I'm
kind
of
proud
of
it,
because
I
did
the
submission
alone
and
it
went
into
the
conference
by
the
by
the
usual
track
right.
So
the
usual
reviewers
decided
to
let
give
this
talk
a
chance,
and
I
think
there
are
only
11
or
12
talks
from
all
submissions
now
part
of
the
overall
schedule,
which
is
really
amazing,
because
this
totally
trims
down
to
to
only
let
a
small
amount
of
percentage
go
through.
A
B
So
that
is
one
of
the
things
that
I
that
came
out
of
this
pandemic.
The
conference
is
moving
virtual.
I
learned
how
to
connect
with
people
virtually
and
yes
back
to
kubecon.
I
have
like
three
sessions,
but
one
of
them
is
a
maintenance
track
for
six
security
which
I
am
excited
about.
They
are
awesome
bunch
of
people
to
work
with.
They
create
a
very
warm
welcoming,
safe
space
and
I'm
so
proud
to
be
a
part
of
that
session
and
another
one
is
about
burnout.
B
Oh
no,
I
enjoy.
I
love
kubecon
I
get
to
meet
my
favorite
people.
I
get
to
meet
new
people,
so
I
won't
call
it
toxic
but
tiring,
trying
kubecon.
I
would
call
that
I'm
an
introvert,
so
it's
tiring
and
exhausting
and
for
after
kubecon
for
a
week.
I
don't
want
to
meet
or
talk
to
anyone,
so
definitely
climbing.
So
I
have
that
one
about
burnout.
Another
one
is
a
student's
panel,
so
I
I'm
happy
about
that.
B
B
When
I
was
student,
I
didn't
have
these
many
resources
and
now
I
see
like
how
much
we
can
help
and
give
back
and
start
nurturing
from
when
they
are
like.
You
know
like
when,
from
a
little
seedling,
that's
how
I
love
plants,
so
I'm
just
sticking
back
to
the
plant
analogy
if
that
makes
sense
so
like
nurture
them
like
provide
them
with
all
the
resources.
So
these
are
the
little
things
that
I
love
and
I'm
like
so
happy
to
be
a
part
of
and
I'm
a
little
bummed
sasha.
B
I
was
hoping
that
I
will
see
you
in
person,
but
there
is
always
another
cube
con
right.
This
is
the
beauty
of
coupon.
You
don't
have
to
feel
like
you're
missing
out.
There
is
always
another
one.
The
same
for
langdon.
I
was
excited
that
oh
he's
going
to
go
to
cube
carnegie,
but
no,
I
will
see
you
my
kubecon
emmy,
probably
so.
A
Maybe
we
could
have
kubecon
n
a
also
in
valencia
but
yeah.
I
totally
hear
you
yeah.
I
was
it.
It
would
have
been
nice
to
be
able
to
get
to
see
some
people,
you
know
being
back
to
traveling
a
little
bit
speaking
of
students.
I
would
also
you
know,
being
a
college
professor
that
was
part
of
that
was
part
of
why
I
joined
that's.
A
Why
I
kind
of
decided
to
get
into
academics
is
try
to
see
how
we
could
you
know
point
students
kind
of
in
the
right
direction
towards
you
know.
Software
engineering,
you
know
a
big
part
for
me-
is
actually
been
trying
to
build
out
on
ethics
and,
on
you
know,
making
sure
underrepresented
minorities
are,
you
know,
supported
and
encouraged.
A
So
those
are
some
really
important
things
to
me.
So
I
totally
agree.
So
we
are
we're
coming
up
on
time.
We
still
have
a
few
more
minutes,
but
I
did
want
to
ask
you
know
so
sasha
I
was
looking
around.
Usually
I
try
to
find
a
hobby
or
something
to
talk
about
on
the
show.
However,
you
seem
to
be
a
lot
like
me
in
that
you
get
the
only
hobbies
you
really
have
are
also
technology.
C
C
A
Yeah,
it's
it's
a
lot
of
fun
and
funnily
enough,
there's
actually
a
there's,
a
cosmopolitan
magazine,
which
you
know
is
largely
targeted.
As
a
you
know,
kind
of
a
women's
magazine
from
I
think
it's
1969
that
encourages
women
to
get
into
programming,
because
it's
just
like
having
a
dinner
party.
You
know
you
build
a
recipe
and
then
you
follow
the
recipe
and
it's
funny
how
much
like
how
accurate
that
feels
to
me,
and
I
think
that's
why
cooking
really
appeals
to
me.
A
C
A
B
No,
I
think
I
am
done
with
all
the
questions.
The
comment,
though,
is
that
I
love
eating.
When
you
were
talking
about
the
cooking
for
work,
because
I
cannot
follow
a
recipe.
I
cannot
follow
anything
to
the
tea.
I
struggle
and
every
time
I
end
up
making
something
new
and
I'll
taste
tell
me
at
least
my
husband
says
that,
because
you
know
he
has
to
live
with
me
every
day
so,
but
I
can
never
follow
so.
I
have
like
much
respect
like
when
someone
says.
Oh,
I
baked
this.
B
I
made
this
and
I
spent
like
four
hours
marinating
this
I'm
like
you
are
god
level
to
me.
I'm
impatient.
I
just
have
to
have
a
healthy
meal
and
like
preparation
time,
15
minutes
and
eating
time
is
like
inhaling
the
food,
so
I'll
be
like
done
at
like
five
minutes.
So,
but
the
only
food
that
I
love
I
take
time
is
pizza.
So
that's
there
yeah.
A
What's
funny
is
my
kids
have
actually
regularly
asked
me
is
like
why
didn't
you
ever
become
like
a
chef,
you
know
you're
a
pretty
good
cook.
You
enjoy
cooking
and
because
the
reason
is,
is
I
don't
want
to
make
things
the
same
way
every
single
time
which,
if
you're
a
good
chef,
that's
what
you
do
right
in
a
restaurant
right
as
you.
You
know
you
find
a
really
good
recipe
and
then
you
make
it
exactly
the
same
way
every
single
time,
so
I
don't
think
you're
in
the
the
wrong
place,
making
it
differently
each
time.
A
You
know
as
long
as
as
long
as
it's
going,
you
know
upward.
You
know
as
long
as
it's
getting
better,
but
I've
definitely
had
some
pretty
massive
fails
with
you
know,
making
it
up
as
I
go
along
all
right,
so
I
think
maybe
we
should
wrap
it.
There
we're
a
little
early,
but
I
think
that's,
okay
and
we
will
encourage
everyone
to.
A
A
It's
a
it's
another
good
way
to
get
involved
in
the
community,
both
from
a
perspective
of
trying
to
understand,
what's
happening
and
who
is
impacting
the
project
as
well
as
getting
an
idea
of.
What's
you
know
coming
forward
in
the
future,
we'll
probably
recap
a
bit
of
kubecon
in
our
next
episode,
and
I
think
that
was
it
but
yeah.
Hopefully
we're
gonna
have
a
nice
big
kbe
presence
at
the
next
cube
con
in
n,
a
you
know
and
we'll
be.