►
From YouTube: Kubernetes Office Hours 20220420 (SIG Release Edition)
Description
Office Hours is a live stream where we answer live questions about Kubernetes from users on the YouTube channel. Office hours are a regularly scheduled meeting where people can bring topics to discuss with the greater community. They are great for answering questions, getting feedback on how you’re using Kubernetes, or to just passively learn by following along.
For more info: https://k8s.dev/events/office-hours
A
Double
checking
if
we
are
live
in
youtube,
wait
a
minute
and
we're
live
in
youtube.
Okay,
so
welcome
to
office
hours.
This
is
for
april.
We
do
it
monthly
and
we
have
a
great
set
of
panelists
today.
Some
of
them
are
we're
stealing
them
out
of
their
sigs
every
month.
A
So
this
this
this
month
is
stick
release
for
april
and
we'll
go
into
here
to
answer
your
questions
also
live
on
the
air
and
you
can
find
us
in
the
kubernetes
lag
in
the
office
hours
office
dash
hours,
slack
channel
in
there.
You
will
find
all
the
information
about
about
this
this
session.
A
So
let's
go
with
before
we
begin.
Let's
start
by
introducing
ourselves
so
I'll
go
first,
since
I'm
talking
I'm
carlos
antenna,
I
work
for
ibm,
ibm
cloud
as
an
architect
and
also
on
open
source,
currently
on
sick
release
as
a
shadow,
sick
release,
lead
for
125
and
also
another
open,
cncf
open
source
project.
Now
I
can
say
cncf
open
source
project,
k
native,
so
I'm
just
a
steering
committee
and
ux
lead
for
that
and
onboarding
canada
from
cncf.
So
that's
it.
Who
wants
to
be
next.
B
B
C
A
Right,
marco.
D
Hi,
I'm
a
software
engineer
at
surrender.
I've
been
helping
out
with
these
office
hours
for
a
while
now
always
find
them
valuable
for
myself.
So
I'm
happy
to
help
others
learn
as
I
have
so
that's
it.
E
Hey
this
is
veronica
lopez.
I
work
in
planet
scale,
but
I'm
also
part
of
sick
release
and
the
release
engineering
family
also
smarko.
I'm
a
release,
engineer
a
release,
manager,
sorry
and
yeah,
taking
care
of
several
releases
and
different
tasks
that
are
required.
Yeah.
F
A
Yeah,
it's
just
new
users
usually
get
confused
with
those
those
terms,
but
that's
that's
what
we're
here
for
so
before
we
start
here
are
the
ground
rules.
This
is
the
kubernetes
event,
so
the
code
of
conduct
is
in
effect,
please
be
excellent
to
each
other.
This
is
a
judgment-free
zone.
Everyone
had
to
start
somewhere.
A
So
please
help
out
your
body
by
having
supporting
environment
in
the
in
the
channel.
Also,
we
have
people
in
the
channel
while
we
do
our
best
and
answer
the
questions
and
the
panel
doesn't
have
access
to
your
cluster
so
so
live
debugging
is
off
topic.
We
cannot
share
kind
of
code
and
we
would
love
to
go
into
your
current
disclosures
and
help
you
out,
but
that's
there's
other
meetings
to
do
that,
but
do
our
best
together
moving
down
to
the
next
step,
so
you
can
figure
out
and
troubleshoot
based
on
your
questions.
A
Normally
we
provide
swag,
but
right
now
we're
not
doing
that.
Panelists
you're
encouraged
to
expand
on
the
answers
with
your
experience
and
pro
tips.
That's
that's
a
great
thing
to
have
people
with
experience
in
open
source
and
and
your
different
companies
audience.
You
can
help
by
passing
the
url
sharing
the
url
on
on
twitter
of
the
youtube
and
the
different
channels.
I
posted
on
linkedin
blogs
and
anything
like
that
post,
your
questions.
Usually
we
go
to
discuss.kubernetes.org
search
for
the
latest
questions,
so
usually
they
are
not
answered.
A
I
pick
a
few
of
those
you
can
also
help
out
by
tweeting
and
spreading.
That's
that,
like
I
said
this
panel
is
made
entirely
of
volunteers.
So
if
you
want
to
rotate
I'm
looking
we're
looking
for
people
always
to
come
out
every
month,
we
do
it
one
month,
one
one
time
in
a
month.
So
may
we're
going
to
be
scheduling
that
so
come
to
the
office,
hours
channel
ping,
nose
and
say
that
you
want
to
to
help
out
also
we're
always
looking
for
for
help.
A
So
with
that,
let's
get
started
with
the
sick
release,
presentation
or
show
and
tell
this
is
a
open
forum
not
not
for
money.
Veronica,
marco.
E
Yeah,
we
don't
have
like
a
formal
presentation.
We
just
wanted
to
share
things
with
you
all,
because
we
believe
that
the
more
relatable
we
are,
the
more
people
would
like
to
join
us
wishful
thinking.
I
don't
know
so
what
can
I
say?
We
have
been
working
on
the
release
team
for
a
while,
in
my
case
it
will
almost
be
three
years.
I
will
let
marco
answer
the
same
question
in
a
bit
and
I
don't
know
we
we
in
in
my
case
I
arrived
thanks
to
the
shadow
program.
E
Back
in
the
day
when
I
joined
it
was
to
be
very
way
easier
to
join
as
a
shadow,
because
there
were
like
very
few
people
applying
and
basically
it
was
like
if
you
want
to
join,
join,
unfortunately
and
unfortunately,
obviously
nah,
it
has
become
increasingly
harder,
but
that
doesn't
mean
that
we
don't
need
more
hands.
So,
oh,
what
I
can
say
is
that,
in
the
releasing
most
of
us
are
volunteers
there
there
are
a
lot
of
people
who
are
paid
to
do
open
source
in
their
companies.
E
Some
of
us
are
more
like
hybrid,
and
some
of
us
are
more
100
volunteers,
so
it
I
really
like
the
the
release
engineering
family,
because
it's
it's.
It's
really
a
really
nice
team
people
try
to
be
there
for
each
other.
More
than
any
other
team
that
I
have
ever
worked
at.
In
my
experience
and
also
the
technical
ability
is
not
compromised.
If
you
know
what
I'm
made,
which
it's
in
my
experience
again
tends
to
be
like
either
or-
and
here
there
is
both
the
downside.
E
Is
that
since
a
very
specific,
not
I
wouldn't
even
like
to
say
skill,
but
like
a
very
specific
interest,
has
to
exist
to
to
contribute
to
this
part
of
kubernetes.
There
are
some
parts
of
it
that
only
certain
people
know,
but
it's
not
because
they
are.
E
I
mean
they
are
extremely
smart,
but
yeah
and
talented
and
all
those
great
things,
but
it's
not
easy
for
anyone
talented
to
just
come
and
operate
the
things
like
easily
and
then
you're
done
right.
So
that's
probably
the
downside
and
where
we
get
a
lot
of
bottlenecks,
because
it's
by
definition
for
the
release
team.
So,
whenever
all
the
releases,
all
the
kubernetes
is
really
still
official
the
patches
the
release
candidates
alpha
beta,
all
of
them
go
through
us,
so
we
are
literally
the
last
part
before
it.
E
It
comes
live
before
we
give
birth
to
them.
So
there
are
very
specific
processes
that
require
a
lot
of
hands
and-
and
I
think
that
sometimes
most
of
times
we
are-
we
are
very
understaffed.
I
think
that
no,
we
are
extremely
understaffed,
which
I
will
finish
my
first
intervention
with
this,
because
I
think
it's
a
great
sign,
but
also
a
terrible
sign.
E
The
great
sign
that
we're
so
understaffed
is
because
people
we
we
still
have
the
same
amount
of
hands
and
probably
a
little
bit
more
but
more
interesting
technical
problems
have
our
rose
series.
I
don't
know
how
to
say
that
in
english,
but
anyway
we
have
more
interesting
technical
challenges,
these
days
other
than
just
giving
birth
to
new
kubernetes
releases,
which
is
one
of
the
biggest
misconceptions
that
people
have
probably
or
that
at
least
they
have
asked
directly
to
me
that
oh,
but
what
do
you
do?
The
release
team?
E
You
you
only
like
push
commands
to
release
things
and
like
yeah,
that's
part
of
it,
but
also
all
all
of
us
are
involved
into
creating
our
own
tooling.
You
know
so,
although
a
lot
of
infrastructure
teams
don't
or
people
think
that
infrastructure
teams
work
from
thing
with
things
that
come
from
the
ether,
but
someone
has
to
build
those
tools,
you
know
to
be
able
to
push
commands
and
that
that
that
make
things
happen.
E
So
we
are
like
self-service
team
in
terms
of
like
different
team
members,
create
different
tools
for
so
that
the
ultimate
release
process
can
just
be
clicking
on
some
commands
and
firing
a
couple
of
things
and
how
kubernetes
released.
E
So
that's
why,
as
more
people
get
involved
and
a
lot
of
the
security
and
supply
chain
topics
that
I'm
sure
everyone
is
talking
about
these
days
come
up
more
things
have
to
be
developed
behind
the
scenes,
yeah
for
every
release
to
be
secure,
efficient,
blah,
blah
blah
blah
blah.
So
a
lot
of
our
team
members
are
focused
on
that
and
we
have
less
time
and
less
hands
to
do.
The
simple
quote-unquote
big
quote:
unquote
the
simple
part
of
the
job.
E
So
if
anyone
wants
to
join,
we
really
need
you
right
now,
but
I
will
let
marco
go
ahead.
A
E
Sort
of
yeah
like
there
have
been
many
modifications
ever
since
I
came
in,
I
originally
started
in
another
like
sub
team.
That
is
called
ci
signal.
I
I
would
do
a
huge
under
service
by
describing
what
we
did
at
that
time,
but
I,
I
am
sure
that
a
lot
of
you
are
familiar
with
the
team
and
then
I
continued
being
involved
in
through
different
cycles
and
at
very
different
capacities.
E
I
don't
know
if
this
is
worth
mentioning,
but
I
hope
it
is
useful
for,
for
some
people
like
I
as
a
contributor,
I
have
been
in
different
hats
and
roles,
not
only
inside
early
sim,
but
also
in
the
day
job,
and
that
has
allowed
me
or
has
stopped
me
from
contributing
more
or
less
and
donating
more
time
or
less
time
or
even
disappearing
for
like
completely
for
a
couple
of
weeks
or
a
couple
of
months
or
whatever,
and
I
could
come
back
so
you
in
my
case,
and
I
think
that
marco
is
a
similar
case.
E
You
don't
have
to
be
chained
to
to
the
team
all
the
time
to
be
able
to
contribute
the
significant
and
amazing
things.
Obviously,
the
more
the
better,
but
if
someone
is
being
deterred
from
contributing
just
because
they
think
they
don't
have
the
time
or
like
they,
they
are
expected
to
be
chained
to
a
desk
or
the
computer
at
all
times.
It's
not
not
necessarily
true.
A
Okay,
okay,
marco,
go
ahead
and
tell
us
clarify
what
what
is
seek
release,
what
what
is
their
role
there
and
what
are
the?
What
are
the
things
that
are
compo
composing,
the
the
the
sig
release
and
the
type
of
the
type
of
things
that
a
contributor
would
do
if
they
join
right.
C
Yeah
so,
first
of
all,
I
have
to
say
a
huge
plus
one
to
everything.
Veronica
said:
I
think
that
yeah
I
have
been
also
at
siege
release
for
about
three
years,
maybe
a
little
bit
more.
If
I
remember
correctly,
I
started
around
when
1.14
was
around
so
quite
a
while
ago,
and
it
has
been
a
really
good
experience.
C
It
was
also
really
challenging,
but
I
think
that
the
team
is
very
supportive
and
very
good,
so
I
think
that
yeah,
it
has
been
a
really
good
experience
overall.
So
a
little
bit,
I
will
not
go
ahead
and
repeat
what
ferrari
said,
because
I
think
veronica
described
it
really
well,
but
a
little
bit
about
what
siege
release
does.
So
it
is
a
team
that
is
basically,
let's
say,
responsible
for
releases
that
delivering
them
getting
them
organized.
C
This
is
what
is
the
let's
say,
the
job
of
release
team,
to
make
sure
that
everything
is
in
a
good
shape
to
make
sure
that
it
is
adhering
to
the
process
to
make
sure
that
this
stuff
is
not
broken,
that
the
tests
are
green.
There
is
a
lot
of
responsibility
of
the
early
steam
itself
like
first
of
all,
secretly
is,
let's
say
there
are
multiple
teams
inside
the
sig
release.
There
is
the
release
team,
which
is
responsible
for
getting
the
current
release.
We
are
working,
for
example,
model
24
that
is
going
on
right
now.
C
D
C
C
And
let's
say
yeah:
that
is
basically
in
general
overview
it
and
the
release.
Team
itself
has
some
roles
or
sub
teams
like
ci
signal
like
the
enhancements
team
that
is
taking
care
of
caps
and
making
sure
their
production.
I
mean
production
ideas,
basic
architecture,
if
I'm
not
mistaken,
but
the
release
team
is
making
sure
that
everything
is
on
the
track
that
it
is
going
to
be.
That
is
not
going
to
miss
any
deadlines
or
any
other
imported
stuff.
C
So
there
is
also
that
there
is
monitoring
the
healthiness
of
the
release
in
a
sense
of
tests.
There
is
also
making
sure
and
monitoring
issues
and
prs
to
make
sure
that
everything
is
milestone
correctly,
to
make
sure
that
everything
has
the
correct
details
and
all
that,
so
there
is
pretty
much
a
lot
of
sub
things
in
the
release.
Teams
also
for
dogs,
for,
let's
say
communications
for
everything
needed
to
get
a
successful
release
out
of
the
doors.
So
it
is
a
bigger
team.
C
I
think
there
is
definitely
about
like,
maybe,
if
I'm
not
mistaken,
20
or
30
people
by
per
the
release
team.
So
definitely
a
lot
of
people
needed
to
get
released
successfully
out
of
the
door,
and
this
is
oh
yeah.
It
works
in
a
dead
way
and
then,
as
I
said,
really
century
thing
who
is
for
other
stuff
but
yeah
in
general
releases
and
releases
is
definitely
what
is
the
biggest
part
of
siege
release
and
creating
and
maintaining
tools
to
actually
make
that
possible.
A
I
mean
I've
seen
that
in
first
hand
I
can
tell
my
my
story.
I
wanted
to
contribute
into
into
the
release
process.
A
So,
as
you
mentioned,
the
the
sig
is
responsible
for
when,
for
example,
now
there's
going
to
be
a
new
version
of
kubernetes
right,
1.24
and
veronica
was
saying
that
the
the
sig
release
is
in
charge
of
making
sure
that
that
the
whole
kubernetes
is
currently
is
not
only
just
one
binary
gold
binary
that
we
compile
it's
multiple
docker
images,
multiple
tools,
it's
a
big,
it's
a
big
repo
repository,
but
it's
amazing
that
the
the
team,
the
sig,
manages
that
release
like
it.
A
A
Is
the
permanent
team
that
that
keeps
keeps
the
the
chain
the
supply
chain
running
and
and
then
the
the
release
team?
That
is
like
each
swaps
like
it
swaps,
in
on
every
release
right
like
124
and
125
and
126,
so
there
will
be
a
process
there.
So
I'm
familiar
with
the
release
team
aspect.
So
there's
I'm
looking
here
into,
and
this
is
something
that
people
can
I
can.
I
can
share.
Let's
see
if
I
can
share
this,
so
people
can
see
it.
D
A
Can
you
tell
us
a
little
bit
of
of
every
this
sub
component,
so
this
is
the
release
team
veronica?
What
a
little
bit
of
each
one
like
going
through
the
release
team
lead
through
communications.
This
is
what
happens
on
every
on
every
new
release
right
for
124,
there's
a
team
that
is
working
on
this
one
to
get
it
out
in
may.
By
may
3rd,
it
got
moved,
but
then
the
next
125
is
forming
and
there
will
be
a
different
set
of
roles.
A
E
Yeah,
that's
right
typically,
what
happens?
I
mean
it's
not
an
any
rule,
a
written
rule,
but
typically
what
happens
is
like
someone
who
was
working
as
a
shadow
or
or
helping
in
to
any
extent
and
in
one
of
these
sub
teams
like
enhancements
the
signal,
portrayed,
etc
for
the
next
cycle.
They
become
the
leader
of
of
that
subgroup
for
the
next
release.
It's
not
always
the
case,
but
if
that
that
is
what
typically
happens
and
yeah.
E
So
most
of
this
of
these
contributors,
these
days
are
selected
through
the
shadow
program
for
the
not
the
leads,
but
for
people
helping
in
any
capacity
on
on
these
tasks.
We
ask
you
to
have
a
little
bit
of
experience.
It
doesn't
have
to
be
completely
the
same
exact
experience,
but
some
notions
of
like
the
work
that
you
have
to
do,
but
our
goal
is
to
to
train
people
literally
to
to
to
to
shadow
and
yeah.
So
there's
a
lot
of
collaboration.
E
As
you
said,
it's
distributed
systems,
but
really
it's
a
huge
distributed
team
problem
or
a
situation
not
not
necessarily
a
problem
in
a
negative
way.
It's
just
like
challenges
to
be
told:
people
work
across
different
time
zones
with
different
responsibilities
and,
of
course,
with
deadlines
and
the
deadlines
you
see
on
the
websites
are
literally
like
the
final
deadlines
for
all
of
us
to
present
a
product,
but
behind
the
scenes
we
have
like
other
smaller
deadlines.
E
That
cannot
happen
if
another,
one
of
those
components
or
subteams,
don't
don't
release
things
like,
for
example,
for
the
docs
team
to
be
able
to
have
a
finished
product.
They
have
to
have
access
to
too
many
different
artifacts
way
beforehand,
so
that
they
they
can
write
about
it
and
they
can
document
things
and
and
and
communicate
things
and
stuff
later.
So
it's
it's
a
lot
of
of
thinking.
We
have
during
the
release
process,
let's
say
124
right
now.
E
We
have
weekly
meetings
where
all
the
sub
teams
give
their
updates
in
case
they
need
something
from
any
other
sub
team
or
or
whatever,
or
if
the.
If
the
tests
are
broken
and
they're
a
blocking
issue
for
for
them
to
continue
a
job,
I
don't
know
to
ask
for
help
etc
and
towards
the
end
of
the
cycle.
We
typically
have
either
a
daily
or
or
every
other
day,
meetings
to
be
more
on
top
of
things
and
yeah.
So,
for
example,
ci
signal
that
was
my
my
first
introduction
into
the
release.
E
E
So
back
in
the
day,
our
responsibility
was
to
be
aware
of
the
of
the
boards
of
the
test
grid
ports
for
blocking
issues
and
just
informing
issues,
so
you
would
have
to
be
aware
of
which
ones
are
blocking
issues
and
which
one
were
informing
issues
like
things
that
we
needed
to
pay
attention
to,
but
that
a
release
could
go
or
no
go
fast
forward.
E
To
now,
like
the
amazing
ci
signal
team,
give
the
the
general
release
team
a
formal
go
or
no
go
for
cutting
a
release,
yeah
which
in
the
past
was
not
like
that
or
not
formally
so
yeah-
that
that
that's
what
we
did
in
ci
signal,
but
again
for
every
subtask.
E
We
try
to
to
automate
the
process
or
doing
as
a
streamline
as
possible
to
to
avoid,
depending
on
a
single
person
or
like
a
single
point
of
failure.
Obviously
we
we
are
not
there
yet
for
every
every
task
for
different
reasons,
but
but
that
that
would
be
the
ideal
thing
to
do,
and
I
don't
know
bulk
triage
works
like
a
very
very
closely
to
to
ci
signal
and
a
lot
of
of
this
sub
teams
are
their
their
main
job
aside
from
reporting.
E
E
Yes,
exactly
so,
let's
say
that
the
ci
signal
team
detects
that
there
is
a
test
that
started
breaking
after
the
last
c
testing
contribution
or
pull
request,
so
they
track
that
down
and
they
they
reach
out
to
them
or
whatever
stick
is
responsible
and
and
they
work.
So
it's
a
solution.
E
And
so
it's
a
lot
of
communication
between
different
part
parts,
not
only
within
the
the
releasing,
but
with
other
teams
and
things.
A
Okay,
mark
mark
marcus,
just
a
question
for
you
in
terms
of
someone,
and
I
I
was
showing
a
minute
ago
that,
if
you
want
to
you,
know,
get
involved
with
the
community
a
lot
of
a
lot
of
in
the
conferences
you're
showing
keep
going.
There's
a
lot
of
people
that
join
and
ask
us
like.
I
want
to
start
an
open
source.
I
want
to
start
in
kubernetes.
How
do
I
get
involved?
Well,
joining
the
the
the
main
list
is
one
way
of
like
being
aware.
A
What's
going
on,
either
features
are
coming
or
bug
fixes
or
breaking
changes,
and
if
you
want
to
get
involved
in
in
something
like
your
want
to
learn
about
devops
pipelines,
how
to
release
code,
how
to
release
code
securely
or
or
what
what
it
takes
to
release
a
a
product
like
to
the
point
you
can
consider.
Kubernetes
is
treated
as
a
software
product
with
high
quality
right
and
high
speed,
agile
teams.
A
So
this
is
ray,
sending
the
announcement
of
the
form
to
apply
for
shadow
and
for
shadow
there's
a
a
role
book
on
the
on
the
website
that
you
will
see.
He
puts
the
links
in
here
of
what
what
are
the
requirements
and
some
some
of
them
have
low
requirements,
some
of
them,
those
you
like
what
are
expected
from
you,
so
you'll,
be
shadowing,
so
you'll
be
you'll,
be
in
training
mode
with
the
purpose
of
you
becoming
a
a
lead
if
you
want
or
moving
to
a
different
different
group.
A
So
you
just
click
the
form
here.
It's
a
google
form,
but
answer
a
few
questions
about
yourself.
What
are
the?
What
would
what
team
that
you
want
to
apply
and
then
see
if
you
get
selected
so
marco
do
do
people
need
to
be
an
expert
in
go
or
go
compiler
expert
to
join
one
of
these
teams?
Is
there
something
that
other
skills
veronica
was
mentioning
like
communication
skills
writing
skills
that
that
they
can?
A
They
may
join
some
of
these
these
aspects
and
help
out
some
people
think
that
they
have
to
be
a
a
programmer
or
an
expert
in
go
or
an
expert
in
in
certain
things
in
kubernetes
to
to
join
here,
and
maybe
that's
a
misperception
so
can?
Can
you
tell
us
like
what?
What
are
the
things
that
that
you
can
join
based
on
on
the
skills
and
interests
that
you
have
like
communications
or
maybe
dogs,.
C
Yeah,
that's
a
very
good
question
so
generally,
for
I
think
majority
of
release
team
roles.
You
don't
really
need
to
be
expert
in
go
or
programming
languages,
because
the
purpose
of
releasing
is
not
really
to
like
work
or
implement
suffix.
The
purpose
of
the
release
team
is
actually
to
organize
and
to
make
sure
that
the
release
is
on
track
and
that
everything
is
going
fine.
So
it
is
more
important
that
you
have
like,
let's
say
organizational
skills
that
you
can
organize
or
later
on
lead
the
team.
So
it
is
not
really
specific.
C
There
are
some
like
for
dogs.
It
is
helpful
that
you
know
if
you
are
maybe
with
writing,
but
even
then
it
is
in.
At
least
when
I
was
on
the
release
team.
It
was
not
up
to
you
to
ride
the
dogs
like
it
was.
Maybe
that
you
follow.
How
is
it
going
eventually
help?
C
You
travel
yours
or
something
like
that,
but
it
was
not
really
up
to
you
that
you
sit
down
and
buy
the
dogs,
but
you
was
there
to
help
to
organize
to
make
sure
that
developers
are
following
the
deadlines
to
make
sure
that
there
is
some
that
there
is
no
feature
that
is
missing,
documentation
and
similar.
So
it
really
boils
down
that
you
can
and
know
how
to
properly
organize
the
release
and
exactly
this
is
what
is
the
shadow
program
here?
C
You
can
actually
learn
that
it
is
not
something
that
is
super
hard
or
it
is
maybe
a
little
bit
challenging
in
the
beginning.
But
exactly
what
is
your?
What
are
you
going
to
learn
from
the
your
let's
say:
role
lead.
It
is
they're
going
to
show
you
how
actually
to
do
that.
All
that
if
you
eventually
take
that
role
later
on
in,
let's
say
upcoming
releases
or
something
like
that,
yeah.
A
And
you're,
my
I
heard
me
you,
I
know
you
were
you
said
that
you
were
a
release
lead.
I
just
want
to
take
take
your
input
on
on
this.
When
we
ask
you
a
question,
if
you
want
to
answer
so
how
how
what
are
the
social
aspect
of
joining
a
a
cig
in
terms
of
like
learning,
meeting
other
people,
networking
learning
from
others,
how
is
how's
the
social
aspect
of
it
like
this?
A
Are
all
these
people
rude
and
they
just
want
you
to
like
fix
bugs
and
just
just
state
your
your
scrum
status
is
where,
where
some
people
are
used
to
do
that,
how?
How
is
open
source
works
in
terms
of
like
a
leadership,
leadership
position
and
the
social
aspect.
F
I
think
that
it's
you
know
it's
going
to
vary
from
project
to
project
but
overwhelmingly.
I
feel
like
the
social
aspect
of
participating
in
the
release
team
and
and
really
like
that's
talking
to
folks
on
the
release
team.
You
know
so
under
kind
of
the
auspices
of
sig
release,
but
also
all
the
interactions
that
you
have
with
those
other
cigs.
So
all
of
these
roles
are
going
to
be
focused
on
a
lot
of
communications
and
a
lot
of
interaction
with
people
outside,
and
it
kind
of
depends
on
the
phase
of
the
release.
F
So
my
my
experience
prior
to
being
a
release
lead
was
in
bug,
triage
and
then
a
couple
of
cycles
in
enhancements,
and
I
think
in
all
of
those
things
you
know
the
bug
triage
stuff.
You
tend
to
be
a
little
bit
more
towards
the
back,
but
enhancements
especially
you
are
at
the
very
beginning
of
the
release,
you're
talking
to
a
lot
of
the
leads
from
the
other
cigs
kind
of
getting
a
good
understanding
about
like
what
they're
doing.
F
If
you
don't
have
that
experience
already
like
if
you're
a
new
contributor
and
you're
coming
and
say
you're
shadowing
enhancements,
it's
a
really
good
opportunity
to
to
learn
some
of
the
people
that
are
there
and
to
kind
of
understand
what
the
focus
of
the
sig
is
or
what
they're
going
to
be
working
on.
So,
if
maybe.
F
You're
interested
in
contributing
to
later,
you
definitely
can
get
a
lot
of
a
lot
of
good
pathways
to
continue
that
that
journey
and
and
talk
to
them.
I
think,
as
the
release
lead
one
of
the
important
things
that
you
know
your
your
overriding
goal
is
to
make
sure
that
the
release
comes
out
in
time
or
to
identify
when
it's
going
to
slip
and
like
make
sure
that
the
things
that
are
going
to
make
it
slip
get
unblocked.
So
like
the
go,
118
delay
right
now,.
C
F
A
lot
of
work
that
goes
into
figuring
out
like
what's
a
good
time
to
do
this.
What
what
things
are
we
going
to
get
lined
up
to
do
that,
but
even
backing
up
from
that?
It's
it's
looking
at
and
coordinating
with
cigs
like
oh
there's,
a
bug,
we
discovered
we're
going
to
have
to
fix
that
or
we
got
an
enhancement,
freeze
or
code
freeze
request
for
this
feature
that
we
we
discovered
like.
Oh,
it's
broken.
F
What
do
we
do
about
it
and
it
kind
of
comes
down
to
like
working
those
relationships
and
understanding
what
the
impact
is
and
what
the
downstream
kind
of
kind
of
impacts
our
you
know
from
introducing
that
kind
of
delay.
So
it's
definitely
a
very
social
kind
of
thing
where
it's
a
lot,
you
know,
I
think
I've
heard
steven
and
others
say
before
it's
people,
processes
and
tools
and
like
we
definitely
lean
very
heavily
into
the
people.
Part
of
that.
A
Yeah,
that's
that's
that's
my
my
experience.
At
least
I
I
I
was
a
little
bit
hesitant
to
apply
as
a
shadow,
but
I
I
think
I
I
give
it
a
try,
and
I
was
surprised
that
I
got
selected,
so
I
selected
release
notes
as
like.
Maybe
that's
that's
something
easy
that
I
can
get
get
into
with
the
you
know
a
certain
amount
of
time
free
time
and
I've
so
far
been
a
great
experience
meeting
you
know,
there's
there's
routine
weekly
meetings,
there's
a
schedule.
A
You
meet
what
I
learned
like
enhancement
handbook.
Do
you
realize
how
many
enhancements
or
caps
are
in
a
release?
I
didn't
know
I
was
thinking,
maybe
10,
there's
54
things
that
they
get
they
get
tracked
and
by
the
way
we
use
a
heavily
complex,
a
very
intelligent
spreadsheet,
to
do
to
do
that
that
tracking.
So
if
you
have
skills
in
spreadsheets
or
macros,
you
can
be
serving
here
and
and
also
I
was
mentioning
to
marco,
like
the
docs
and
communication
and
release
notes.
A
That's
something
that
that's
a
if
you
want
to
learn
more
about
kubernetes,
that's
a
space
where
you
can
like
help
like
there's
a
new
feature
coming
in
then
you
can
know
about
like
helping
working
with
that
seek
or
that
developer
like
let's
work
together
on
the
documentation
like
their
reference
or
the
manual
and
then
with
your
communications,
you'll
be
working
like
a
blog
post
right
there's,
usually
these
blog
posts
that
come
out
after
release
that
happens
there
also
communications
does
a
seminar,
a
webinar
of
what's
new
in
that
communication,
so
this
is
really
steam
and
just
to
close
out
a
little
bit
on
maronka
was
saying
on
on
the
release
engineering
aspect
that
we
have
great
folks
working
in
there
like
adolfo
and
others
working
you
mentioned
like
cutting
edge
or
new
things.
A
A
Engineer
may
be
a
good
place
to
learn
more
about
things
like
s
terms
like
s-bombs
or
signing
a
a
container
image
or
or
signing
the
s-bone
or
working
on,
like
lowering
the
the
the
registry
or
working
on
a
new
registry
proxy
to
have
those
images
closer
to
the
to
the
place
where
they're
pulling
right.
So
there's
a
some
some
work
in
there
that
you
can
actually
get
technical
and
help
out
veronica.
E
Yes,
so
I
think
everything
started
a
little
bit
with
the
s-bomb
and
the
build
the
bill
of
materials
just
that
as
it
was
without
any
further
ambition,
was
already
very
appreciated
by
some
folks,
at
least
that's
the
the
feedback
we
received
during
kubecon,
just
having
the
the
list
of
artifacts
and
items
that
conform,
a
release
of
whatever
kind
of
just
the
kubernetes
one.
We
heard
about
different
use
cases
that
we
were
not
even
contemplating
so
that's
pretty
cool,
but
then
it's
it
started
growing
really
fast.
E
With
things
like
the
salsa
framework
that
and
now
everyone
everyone
wanted
to
be
part
of
it's
it,
it's
a
whole
different.
It
deserves
its
own
different
talk
and
I'm
sure
that
I'll
talk
about
it
very
extensively,
especially
with
the
the
work
that
they
they
have
been
doing
at
chain
guard
so
yeah.
Basically,
in
very
quick
terms,
we
are
trying
to
to
secure
every
aspect
of
the
release.
E
This
is
in
very
broad,
friendly
terms.
This
is
not
formal,
so
don't
don't
don't
quote
me
on
very
specific
things,
but
yeah
the
the
main
goal
is
to
since
there
are
so
many
parts
of
what
what
goes
into
our
release
from
where
are
we
pulling
the
images
from
and
into
which
pockets
of
what
infrastructure
and
signature
processes
who
has
access
to?
What
repos?
Who
who
has
access
to
to
sign
the
packages
and
all
that,
like?
It's,
basically
identify
every
aspect?
E
Every
connection
from
a
to
b,
from
b
to
c,
where
corruption
can
happen,
corruption,
either
on
purpose
or
malicious,
or
just
because
computers
are
computers
right
so
for
to
verify
the
integrity
in
different
parts
and
levels.
We
have
things
like
the
salsa
framework
that
has
like,
I
believe,
four
stages.
E
I,
like
the
every
stage,
requires
different
levels
of
automation
and
and
verification.
E
As
far
as
I
understand
that
I
have
been
involved
is
the
the
level
one
and
level
two
are
fairly
easy
to
to
to
to
to
reach
in
terms
of
like
just
verifying
some
some
packages
and
origins,
but
things
start
getting
trickier
on
the
level
three
and
four
of
the
salsa
framework,
for
example.
Because,
then,
you
have
to
coordinate
and
not
only
the
the
packages
and
the
the
technical
parts,
but
you
also
have
to
coordinate
with
people
and
accesses.
It
becomes
a
little
bit
more
political,
especially
in
products
like
kubernetes
right.
E
This
doesn't
have
to
be
the
case
for
it
for
every
every
project,
but
the
good
news
is
that
you
get
amazing
benefits
from
just
following
step.
One
two
or
even
three:
not
every
project
has
to
to
comply
with
all
the
levels,
to
be
able
to
to
say
that
they
are
more
secure
enough,
but
yeah,
that's
the
the
guest
of
it
all
and
yeah
everyone
with
all
the
the
things
that
happened
in
2021,
for
example,
security
wise.
E
It
became
a
very
hot
topic,
obviously,
and
so
that
that's
what
what
we
have
been
doing.
So
the
signature
processes
have
changed,
for
example,
so
in
in
terms
of
our
very
specific
work
in
the
in
the
release
team.
When
now
the
the
tooling
that
some
of
our
team
members
had
already
written
for
in
in
the
past
had
to
be
refreshed
for
to
consider
like
the
signatures.
E
Now
the
the
signature
methods
for
to
comply
with
different
new
security,
signature
processes
and
it's
way
more
complicated
than
what
I'm
saying
right
now,
and
it
has
different
implications
into
the
final
release
process.
When
we
just
just
submit
the
command,
it
has
enormous
implications
but
yeah.
Sorry,
I
just
wanted
to
to
share
how
it
looks
like
for
us.
A
Thank
you,
yep,
thank
you,
marco
jeremy
and
and
veronica
for
that
introduction
to
to
seek
release.
I
I
know
we
don't
have
a
lot
of
time,
but
at
least
we
we
gave
a
a
glimpse
of
to
to
folks
watching
this,
that
you
have
it's
kind
of
like
double
double
dipping
in
a
cig
right.
You
can
start
with
this
release
theme
as
a
shadow
apply
for
it,
which
area
you
want
to
learn
whatever
you
do.
You
think
you
can
contribute,
or
you
can
just
you
know,
join
the
meetings
or
read
the
meeting
notes.
A
You
say,
google
doc
and
you
can
join
just
just
watch
and
learn
the
same
thing
for
release
engineering.
If
you
want
to
work
work
for
a
company
that
is
like
getting
into
these
signing
images
and
s-bump
and
starts
a
framework
and
secure
supply
chains,
and
you
want
to
learn
how
are
we
doing
in
cuba
how
we
are
applying
that
security
pipelines
in
kubernetes?
You
can
just
join
the
meetings
read
the
release.
Notes
joins
the
slack
and
try
to
understand
what
are
the
problems
there,
and
maybe
you
can
contribute
something
right.
A
It
could
be
ideas,
it
could
be
code,
it
could
be
like
managing
managing
the
branches
or
the
patches.
That
also
is
part
of
the
take
release,
so
just
want
to
conclude
with
with
that
that
portion,
but
don't
don't,
go
away,
we're
going.
A
Two
questions
go
ahead,
barcode.
D
Thank
you.
So
just
a
really
quick
question,
you
guys
being
part
of
the
the
release
team
for
1.24.
D
Are
there
any
enhancements
or
features
like
one
or
two
quick
things
you'd
like
to
highlight
for
people
for
the
upcoming
release.
A
And
communications,
the
communications
release
team
is
doing
a
lot
of
that
with
blogs
and
education
on
on
how
to
migrate
or
how
to
deal
with
it.
A
Any
other
one,
marco
veronica,
big
ones
for
124.
a
lot
of
deprecations.
I
know
for
a
fact,
so
it's
it's
a
heads.
A
There
we
have
to
do
the
the
pr
the
release
notes
for
so
rcc
dog
went
out
last
night,
so
rc0
124
is
out
and
yeah
we
usually
put
if
you
want
to
know-
and
that's
a
good
thing
worker
with
that
one
of
the
things
that
we
do
in
release
notes
we
put
at
the
top,
like
the
most
note
or
the
major
themes,
there's
a
section,
a
small
section
at
the
at
the
top
of
the
change
log
that
we
put
in
there
like
these
are
the
big
changes
and
then
other
ones
are
like
action
required.
A
F
F
A
I
can
tweet
that
under
under
the
announcement
of
this
okay,
so
we
want
to
do
a
few
questions.
Let's
move
on
to
that,
let
me
stop
sharing.
A
If
I
can
figure
out
how
to
use
this
thing
and
let's
see
what
we
gather
in
terms
of
of
questions,
I
think
yogi
picked
this
one
what's
around.
A
How
to
this
is
one
from
tony
the
caser:
how
to
control
concurrent
connections
to
the
pod
from
a
service
so
meaning
you
have
one
request,
and
you
don't
want
to
process
the
next
request
until
you
finish
the
the
previous
one,
so
this
was
paraphrasing.
This
is
a
general
question
relating
to
particular
service.
We
need
to
scale
and
access
in
a
specific
way.
The
service
of
the
pot
itself
is
a
single
threaded.
Threaded,
in
that
one
connection
can
be
established
at
a
time.
A
Specifically,
is
a
document
document
generation
service
based
around
openoffice,
which
can
only
process
a
single
request
at
a
time.
So
I'm
guessing
that
when
you
are
maybe
creating
a
pdf
or
creating
a
powerpoint,
creating
something
it
consumes
a
lot
of
memory.
So
if
there's
a
second
request
in
that
pod
boom
right
out
of
oom
and
and
you
cannot
process
so
it's
one
one
at
a
time-
and
I
think
I
saw
yogi
and
I
think
I
saw.
A
Team
tim
hawkins
answering
that
one
anyone
wants
to
take
that
one
like
how
do
you
control
one
request
at
a
time.
F
I
think
tim
hawkins
answer
is
probably
spot
on.
There's
there's
not
something:
there's
not
a
construct
in
kubernetes.
That
really,
you
know
natively
supports
that.
So
you
probably
want
to
put
some
sort
of
proxy
in
front
of
it
and
then
treat
each
request
that
comes
in
as
maybe
something
that
you
know
queues
or
maybe
spins
up
additional
instances
of
your
document
generation
thing
and
then
asynchronously
can
handle,
handle
the
output
from
that.
A
A
reality
uploading
the
project
so
yeah,
yogi
and
team
answer
that
question
in
those
those
do
ways
and
yogi
always
smarter
than
me
asking
like
you
know
why
or
what
are
the
the
corner
cases?
What
happens
when
you
don't
have
when
you
or
if
you
have
a
maximum
of
five
pods
and
you
have
the
six
requests?
What
do
you
would
do
with
that
request?
A
Do
you
tell
the
client
like
try
again
tomorrow
or
do
you
do
you
cue
that
into
that
letter,
q
right
and
process
it
tomorrow,
so
that
you're
getting
what's
going
down
the
path
of
like
think
about
queuing
like?
A
Maybe
you
want
to
tell
the
client
of
a
send
back
a
two
or
two
like
a
request
accepted
and
maybe
give
it
a
ticket
or
an
id
like
they
can
check
later
and
then
put
that
into
a
into
a
into
a
queue
to
do
async
so
event,
different
application
right
do
not
like
hold
an
open
connection
and
thing
was
on
the
other
side
like.
If,
if
depends
on
your
application,
you
can
hold
the
application
opening
with
a
proxy,
then
not
a
regular
proxy.
Like
nginx
forget
sake
right.
A
You
write
something
that
looks
like
a
proxy
right.
Your
programming
thing
that
you
you
request
to
a
pod
and
you
do
client
that
proxy
will
do
client
side
load
balancing
right.
So
you
have
the
dns
with
ip
address
of
the
pods,
and
then
you
know
which
part
you're
hitting
so
you're
keeping
track
of
that
of
that,
and
then
the
the
dumb
guy
at
the
end
for
carlos
antenna
said
k
native
is
a
plugin.
A
You
can
install
on
your
kubernetes
and
you
can
set
a
flag
of
concurrency
of
one
and
then
you
have
something
called
like
functions
as
a
service.
So
we
insert
a
site
proxy
called
the
qr
proxy
and
we
will
make
sure
that
one
request
at
a
time
is
it
gets
processed
into
your
into
your
file
with
the
side
benefit
that
can
scale
to
zero.
A
So
so
you
can
look
into
k
native
serving
because
that's
actually
openoffice
was
one
of
the
demos
that
happened
from
google
when
he
was
used
to
work
in
google
did
with
cloud
run,
and
it
was
that
that
use
case
like
they
want
only
one
request
at
a
time
or
in
concurrency.
It's
it's
for
an
application
that
doesn't
have
like
a
lot
of
high
throughput
right,
but
it's
something
that
that
you
can
do
with
k
k-native
serving
so
check
it
out.
A
Now
it's
a
cncf
project,
so
maybe
that's
that
one
any
any
other
ones
on
that
one!
It's
those!
These
type
of
questions
are
very
like
very
open
and
we
try
to
our
best
to
not
answer
like
oh,
it
depends,
but
I
think
those
those
two
two
answers
were
very
good
like
hold
the
connection,
queue
or
use,
something
that
is
going
to
implement,
actually
that,
but
it's
a
it's
a
an
extension
or
plug-in
that
does
that,
like
k-native.
A
If,
if
you
go
to
discuss
karen's
discussion
and
so
for
the
recent
questions,
you
will
find
these
ones,
and
this
is
it,
it
says.
A
big
bot
on
node,
I
like
to
f,
fell
over
a
pot
from
a
node
to
another
without
decline
to
lose.
The
connection,
is
this
possible
keeps
detail?
Avix.
Does
this?
A
Anyone
wants
to
take
that
one
if
they
want
to
evict
a
pod
from
one
node
to
another
node
without
losing
the
connection.
B
So
whenever
you
delete
a
pod,
it
sent
a
up
first
and
it's
given
around
10
seconds.
Someone
feel
free
to
correct
me
before
a
sick
kill
is
sent
to
it,
which
means
your
application
has
10
seconds
to
handle
any
existing
requests.
Stop
recepting,
stop
accepting
any
new
ones,
but
hopefully
you'll
have
another
pod
already
accepting
those
ones
anyway.
That
would
balance
your
english
object
can
pass
them
over
to
and
then
you
remove
the
old
hood.
A
But
the
connection
will
close
on
that
right
that
he
wants
to
avoid
that
that
closely
on
that
connection,
anyone
else
having
any
thoughts
I
I
would
say
that
this
is
a
pattern
that
is
usually
done
with
client
sign,
load,
balancing
again
or
like
grpc
or
something
like
you.
You
request
the
dns
or
the
service
name,
and
you
get
back
your
five
ip
addresses
of
the
pods
and
your
client
is
the
one
that
has
an
open
connection
to
to
one.
A
But
again
it's
like
it's
you're,
you're,
closing
a
connection
on
one
part
and
open
another
one,
but
it's
the
internal
clients
dealing
with
that
logic
right
and
not
on
the
outside,
but
yeah.
The
the
the
pots
are
two
different
containers.
Two
different
ip
addresses
there's
no
way
to
to
not
lose
that
connection.
Unless
you
have
something
in
the
middle
that
is
networking
and
holding
that
that
connection
open
and
redirecting.
A
A
Maybe
you
have
a
session
being
tracked
with
something
like
redis
or
something
like
that,
but
it's
usually
the
pattern
is
a
deployment
with
a
with
a
stylus
application
and
you
have
multiple
ones
so
in
one
pot
diets
you
know
pets
versus
cattle,
just
the
next
open,
a
new
connection
and
you
get
a
new,
a
new,
a
new
pod
and
usually
that's
how
you
do
upgrades
right,
you're
running
upgrades,
so
the
next
one
is
thai
ease.
A
Well,
this
one
I'm
going
to
plug
a
a
meme
that
I
did
the
other
day
about
finalize
finalizers,
so
remove
object
from
the
it
was
funny
when
I
was
like
really
talking
about
finalize,
I
was
joking
about
the
other
day.
Remove
object
from
delete
queue.
Finalize
your
questions.
I
have
an
object
with
the
finalizer
attached,
according
to
the
documentation,
once
the
object
is
deleted,
it's
actually
marked
for
deletion
and
will
be
deleted
once
the
finalizer
is
removed,
and
only
then
the
object
will
be
deleted.
A
So
in
input
anyone
who
wants
to
take
that
one
or
maybe
I
can
paraphrase
it
so
you
you
do,
keep
ctl
delete.
What
happens
is
a
deletion
timestamp
gets
put
on
that
object
and
then-
and
then
it's
like
marked
like
read
only
and
like
oh,
I
don't
want,
I
don't
want
to
delete
it.
Can
I
just
remove
the
the
delete
timestamp
before
it
gets
deleted,
because
maybe
the
finalizer
is
not
yet
removed?
C
Not
really,
I
mean
non-seo
market
for
deletion.
It
is
pretty
much
it.
I
if
I
remember
correctly
that
the
illusion
types
type
is
like
immutable,
so
once
it
is
set,
you
can't
change
it
anymore
and
that's
basically,
it
yeah
so
yeah
it
finalizes
will
keep
it,
but
many
controllers
on
installation
timestamp
is
said
that
they
will
try
to
delete
that
control,
that
that
resource
whenever
it
is
possible
and
that's
basically
it
you
can
pretty
much
recover
it.
It
is
best
to
let
it
go
and
then
create
a
new
one.
A
A
So
at
that
time
the
container
is
like
sick
term
and
there's
a
grass
period
right.
That
would
then
was
at
the
sick,
the
sick
kill,
but
but
the
pot
is
already
like
dead
and
the
body
is
in
terminated
state
and
the
finalize
is
just
sitting
there.
I
was
testing
that
this
last
night,
but
you
cannot
recover
because
the
pot
is
is
already
dead.
You
don't
sorry,
the
containers
is
gone
right,
it's
like
a
sick
term
and
then
the
object
is
just
sitting
there
in
that
cd.
A
Just
just
waiting
to
be
picked
up
and
you
cannot
recover
it.
I
at
least
I
know
that
how
it
works
with
with
a
pod,
but
it
could
be
any
object
but
from
the
documentation
yeah,
it
says
once
you
market
delete
the
delete
step
goes
there
that
the
whole
object
goes
kind
of
goes
into
re,
read
only
mode
right,
you
cannot.
You
cannot
recover
or
edit
that
back.
Even
if
you
try
to
delete
the
the
delete
timestamp
it
gets,
it
gets
put
back
in.
So
it's
it's
gone.
A
A
How
can
we
achieve
that
suppose?
We
have
two
name
spaces
when
s1
and
ns2
in
a
cluster,
and
if,
if
the
pod
within
ns1
has
a
claim
requesting
a
specific
storage
volume
and
then
the
cluster
dynamically
provisions,
that's
the
trick.
Dynamic
provision
is
a
volume
for
that
pvc,
so
pod
would
access
that
pv
and
some
data
is
stored
within
that
pv.
A
D
A
No,
you
cannot
do
that
wonder.
Oh,
so
I
think
I
answered
that
question.
I
remember
it
wasn't
late
last
night,
so
the
the
pv
has
a
claim
reference,
so
the
pv
can
only
be
claimed
by
one
pvc
so
once
that
pvc
cleans
that
pv
the
older
pvc
cannot
be
associated
to
that
pv,
so
yeah
it
needs
to
be
one
pv.
What
the
pattern!
The
the
pattern
usually
is.
If
you
want
to
share
data,
I
don't
know
it's
well.
You
want
to
share
data.
A
It's
like
the
pvc
can
be
associated
with
multiple
pods,
that
single
pvc
and
then
have
the
pv
bound.
A
Right
and
even-
and
even
if
it's
worse,
when
it's
dynamically
provisioned
right,
because
you
create
a
pvc
and
it
carves
a
volume,
and
it
gives
you
back
a
pv
and
then
you
create
the
second
pvc
and
it
carves
and
you
whatever
is
the
implementation
of
the
of
the
csi
right,
the
the
c,
the
csi
container,
storage
interface,
the
csi
plug-in
implementation
right,
curves
and
all
the
volume
of
data
and
associated
the
only
hacky
way
of
doing
this.
A
That
I
was
thinking-
and
this
is
what
I
wanted
to
hear
from
people-
was
if,
if
you
have
two
pvs
that
indirectly
map
to
the
same
data,
for
example
an
nfs
you,
but
that
would
be
not
dynamically
right,
so
well
kind
of,
like
you
have
two
pvs
and
that
actually
mounted
the
same
information.
Maybe
a
bucket
or
maybe
a
network
share,
nfs
or
a
bucket.
Then
the
two
pvc
will
actually
see
the
same
the
same
data,
but
it
would
need
to
support
something
like
read:
write,
many
or
no.
It
doesn't
matter.
A
That
was
the
only
way
I
was
thinking
like,
maybe,
but
that's
a
hacky
way
so
in
in
terms
of
of
different
namespaces.
The
recommendation
you
use
namespaces
like
a
12-factor,
app
representing
the
same
app
but
a
different
environment.
So
this
this
thing
about,
like
separating
pods
into
namespaces,
to
then
go
back
to
share
information
between
the
two
namespaces.
A
It
wasn't
the
purpose
of
having
namespaces
right.
It
was
to
have
the
separation
of
isolation
in
terms
of
like
stage
broad
production
right,
different
namespaces,
but
it's
the
same
app.
It's
just
different
environments
or
workspaces
yeah,
but
that
was
was
a
tricky
one.
I
was.
I
was
thinking
about
that
one
for
a
while
any
thoughts.
A
Yeah
pvcs
are,
are,
pvs
is
an
another
area
that
is
troublesome.
I
think
that's,
that's
it
actually
we're
on
time
any
any
questions
from
the
slack
or
any
other
questions.
From
from
the
comments.
D
To
the
people,
good
suggestion
regarding
the
finalizers
question
in
the
chat
to
use
valero
to
back
up
your
resources.
D
A
Yeah,
if
you
delete
something
by
mistake
right
how?
How
do
you
get
it
back
if
you
have,
if
you
have
backups
of
that,
that
information
yeah
you
you'll,
be
saving
those
resource
definitions
using
valero
on
a
on
a
schedule
right
just
backup,
if
you
don't
want
to
lose
it
back,
but
again,
immutable
emulative
infrastructure,
where
we
go
with
right.
A
So
if
something
you
lose
a
cluster
or
you
lose
a
namespace,
you
should
be
able
to
have
good
release
engineering
right
from
sig
release
to
have
a
cicd
pipeline
that
can
redeploy
your
application
with
the
new
version
and
fixes
and
cves
include
fixes,
included
and
patches
good.
I
wasn't
watching
the
the
this
this
lack,
but
yeah
it
was
a
good
good
people
joined
in
the
in
the
chat
brand
duffy
adolfo
yeah.
Thank
you
for
everyone
that
join
and
watch
live.
A
A
Amber
error
can
also
do
like
also
the
the
the
pvs
and
the
pvcs
related
to
that
to
that
data
around
that
sharing
information.
A
A
The
the
sick
release
should
be
very
proud
of
you
too,
coming
here
and
talking
to
folks
to
the
community
right
and
trying
to
no,
I
would
say
higher,
but
I
encourage
people
to
contribute
right
and
the
way
to
to
do
that
is
just
just
getting
involved
and
and
then
and
then
figure
out
what
your
interest,
I'm
gonna
put
it
right
right
there
like
it's,
not
it's
not
about
like
open
source.
A
It's
like
what
is
your
interest
and
if
it
matches
the
the
need
which
veronica
says,
there's
a
need
for
helping
hands
which
there's
a
lot
of
automation
and
there's
a
lot
of
ways
to
be
written
right.
So
if
you
want
to
practice
your
coding,
there's
also
opportunities
for
for
coding,
not
just
communication
and
release.
So
oh.
A
Of
people-
and
you
will
have
people
like
marco
and
veronica,
which
have
been
in
the
race
in
the
sikh
for
many
years.
They
are
mentors
right,
so
they're
there
to
mentor
the
next
generation
of
people
to
pick
up
this
work
and
and
do
the
next,
the
next
new
things
right.
So
we
are
we're
looking
for
for
to
to
be
for
people
to
that
want
to
be
mentored.
C
E
C
A
Okay,
well,
thank
you
so
much
and
thank
you
for
the
folks
watching
live
and
then
this
will
be
recorded,
so
you
can
always
reach
us
on
office
hours
and
you
can
reach
the
sig
release.
There's
a
sick
release.
There's
six
notes:
there's
six
enhancements!
There's
six
for
everyone
and
and
there's
the
the
channels
are
there.
I
and
I
I
was
showing
the
the
information
about
the
repo
the
ripple.
The
kid
repo
has.
All
the
information
is
kubernetes
slash,
sig
release
and
in
there
you
will
see
the
handbooks.
A
A
So
if
you
want
to
join
as
a
shadow
go
ahead
and
submit
the
form,
it's
a
google
form
very
easy
to
put
the
information
in
there
answer
a
few
questions
and
and
see.
If
you
get
selected.