►
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
Helm 3: Navigating To Distant Shores - Bridget Kromhout & Jessica Deen, Microsoft
Adjusting your spyglass & looking out over the water, you can see how useful a package manager like Helm is. Perhaps you’ve used it to manage the fractal complexity of packages on your Kubernetes clusters (without losing track of versions stashed in the hold). But Helm 3 is rumored to be different; you’re ready to get started on this exciting voyage as soon as you have some idea of what’s port & what’s starboard! Get clear answers to questions like “when do I use Go templates versus Lua?”, “how do I move my production Helm 2 installation to Helm 3 seamlessly?”, “what do I use now instead of tiller?” In this demo-fueled session, we’ll take you through differences from the Helm of yore, tips for a successful rollout or upgrade, & opportunities to shape the project’s future. The cloud native waters can be choppy but a technical deep dive powered by open source tooling will steer you right!
https://sched.co/MPXf
A
Right,
hello,
folks,
we'll
just
introduce
ourselves
I'm,
not
positive,
of
what
the
track
hoe
situation
is
here
all
right.
So
if
you
are
here
to
learn
about
helm
three,
you
are
in
the
right
place
if
you
are
not
in
the
right
place.
Good
luck!
Finding
anything
else
in
this
exciting
facility
a
little
bit
about
us.
You
have
to
have
the
intro
walk-in
stuff
happening.
Well,
people
are
all
finding
their
seats
by
the
way
if
there's
an
empty
seat.
A
Next
to
you,
if
you
raise
your
hand,
you
are
a
very
nice
person
because
lots
of
people
are
trying
to
find
somewhere
to
sit.
So
if
there's
an
empty
seat-
and
you
keep
your
hand
up
for
just
a
second,
then
people
can
actually
sit
in
chairs
as
opposed
to
standing
in
the
back,
which
would
be
lovely.
Okay,
hi,
I'm,
Bridget
and
I.
Do
a
tech
advocacy
at
Microsoft
I
live
in
Minneapolis
Minnesota
in
the
United
States,
we
are
in
the
middle
rate
below
Canada.
We
kind
of
hope.
B
Then
my
name
is
Jessica:
Dean
I
am
super
into
fitness
and
CrossFit,
but
I
also
love
Disney.
So
the
picture
of
me
smiling
with
an
Avengers
hoodie
on
puts
a
childhood
e
by
the
way
that
was
at
Disneyland
and
then
I
have
a
bb-8
that
rolls
around.
I
am
also
a
cloud
advocate
at
Microsoft,
so
aside
from
liking,
fitness
and
Disney,
and
all
that
I
also
very
much
enjoy
tech
code
and
hanging
out
with
other
engineers.
B
A
I'm
delighted
too
so
quick
outline
and
by
the
way
I
will
tell
you
that
usually
I'm
one
of
those
people
who
puts
one
word
in
a
picture
of
a
cat
on
a
slide
and
since,
for
some
reason
we
I
didn't
get
our
speaker
notes
up
correctly,
I'm
very
happy
that
these
slides
actually
have
lots
of
words
on
them.
So
we
might
actually
be
able
to
tell
you
the
story.
We
plan
to
tell
you.
This
is
your
spoiler
alert,
so
you
know
which
parts
you
can
and/or
nap
and/or
check,
Twitter
and/or
tweet
about
us.
A
If
you
like,
if
you
want
to
tweet
about
us,
the
information
is
at
the
bottom
of
every
slide,
but
we're
gonna
talk
about
first
of
all,
just
a
little
bit
of
mental
scaffolding
I'm
curious.
If,
if
we
can
get
a
show
of
hands
of
how
many
people
are
using
helm,
2
in
some
way,
so
I
would
say
the
vast
majority
of
the
room
like
80%
of
the
room.
B
A
Going
to
talk
about
what's
in
v3
overview,
breaking
changes,
everyone's
favorite
and
some
of
the
new
new
shiny
excitement
and
then
we'll
kind
of
say:
what
are
you
gonna
go
do
now
so
we'll
see
if
we
can
fit
all
that
in
35
minutes,
plus
a
demo,
alright,
so
I,
don't
know
why
I
am
jumping
so
far,
I'm
gonna!
Guess!
That's
because
I
don't
know
your
clicker
push
the
button
once
I
know
right
push
the
button
Frank
one
time,
not
multiple,
so
he'll.
A
Well,
if
we're
gonna
talk
about
helm,
which
you,
the
vast
majority
of
the
room
already
knows
perfectly
well,
is
a
delightful
package
manager
for
kubernetes
I
would
say
we
should
start
with.
Hey
containers
are
great.
They
do
solve
problems
for
us
right,
but
all
of
us
have
used
home
and
have
used
containers
and
know
that
they
super
do
not
solve
every
problem
that
we
have
because
they
introduce
a
whole
bunch
of
exciting
complexity.
A
B
I
mean
when
you,
when
it
comes
down
to
containers,
they
don't
solve
all
the
problems.
They
sometimes
maybe
create
more
problems,
and
then
we
create
more
solutions,
but
they
do
give
us
an
encapsulated
runtime
environment
and
gives
us
a
little
bit
more
overhead
to
to
not
only
manage
but
to
solve
this
complex
world
of
micro
services
and.
A
Ok,
so
we
all
agree
we're
here
at
coop
con.
We
definitely
want
to
orchestrate
our
containers.
I
actually
objected
a
little
bit.
I'm
like
orchestration
sounds
like
so
much
work
like
can't.
We
just
have
a
little
bit
of
light
jazz,
I,
don't
know
it's
just
it's
a
lot
of
work,
but
we
do
need
container
orchestration.
We
have
kubernetes
for
those
of
you
who
weren't
eating
cake
yesterday
or
the
day
before.
Whenever
that
was
it's
all
a
blur,
we
are
coming
up
on
the
five-year
anniversary,
which
is
a
little
bit
dramatic.
It's.
A
And
so
that's
the
fun
part
is
when
we're
like
great.
We
have
containers,
we
can
orchestrate
them.
Kubernetes
has
nothing
to
say
whatsoever
about
managing
our
applications
and
all
of
their
versioning,
and
we
could
have
an
exciting
microservices
murder-mystery
of
all
of
the
yeah
Mille
is
everywhere.
Do
we
have
to
have
a
Content
warning
for
yeah
Mille
like
tabs
or
spaces
I
mean
if.
A
You
I'm
so
sorry,
but
there's
a
lot
of
yeah
Milan
kubernetes,
but
but
if
we
want
to
manage
all
those
applications
at
your
home,
so
we
just
go
through
some
of
the
some
of
the
ways
just
kind
of
breaking
it
down
some
of
the
ways
that
hell
makes
things
easier
for
you
and
most
you
are
using
home
too.
So
you
already
know
it.
B
Helps
you
manage
your
complexity,
even
the
most
robust
and
complex
charts
you
can
or
most
complex
deployments
you
can
use
charts
to
streamline
that
you
can
obviously
version
control.
Your
releases
helm
simplified
a
lot,
but
there
was
still
I
mean
aside
from
offering
easy
updates,
I
assume
you're
going
through
all
the
bullet
points.
B
Aside
from
offering
all
the
easy
updates,
it
still
offered
a
great
deal
more
complexity
and
security
to
consider,
and
so
that
was
a
big
thing
that
the
helm
team
really
wanted
to
take
into
consideration
when
it
came
from
taking
a
look
at
what
we
did
right
in
helm,
2
and
then
thinking
about
helm.
3
was
really
focusing
on
security
and
streamlining
something
that
was
essentially
helm,
classic
enter
Google,
helm,
2
and
then
now,
where
are
we
going
in
home?
3?
Absolutely.
A
And
so,
like
again
sure
sharing
as
easy
as
sharing
is
great.
We
all
want
to
do
that.
Is
this
the
part
where
I
say
I,
don't
actually
believe
in
this
word.
I
have
news
for
us
all.
A
share
can't
turn
back
time.
Neither
can
you,
but
you
can
roll
forward
and
redeploy
a
version
that
will
actually
solve
the
problem
that
you
may
have
introduced
with
her
release.
We
spent
half
this
morning
doing
that
and.
B
A
Ok,
so
let's
talk
about
v3,
probably
the
starting
with
a
high-level
overview
before
we
start
jumping
into
specific
features,
specific
caveats,
I
would
say
this
has
been
a
long
time
coming
more
than
four
years,
and
this
is
coming
from
you.
This
is
coming
out
of
the
community
of
people
who
you
know
created
and
worked
with
home
to
yeah.
B
I
mean
how
many
people
have
heard
the
helm
three
rumors,
four
I,
don't
know
what
feels
like
the
past
nine
months
now
I
mean
the
removal
of
tillers,
big,
so
I'm
sure,
that's
probably
the
prominent
question
on
everyone's
mind
in
here,
but
also
what
does
that
mean?
How
does
it
work
the
same?
What
about
all
these
other,
maybe
rumors
and
spoilers,
so
that
we're
gonna
answer
really
pretty
much
right
now,
this
this
section
from
here
on
out
is
really
what
you're
in
the
room
for
so.
A
None
of
this
would
be
a
surprise,
but
you
know
the
probably
the
top
level
message
here
is
it's
much
simpler
and
safer,
which
getting
both
of
those
things
at
the
same
time
means
that
there's
a
lot
of
work
under
the
hood
to
make
it
feel
seamless
but
work
better,
and
that
does
mean
by
the
way
that
everyone
who
kind
of
pen
tests
Helmand
goes
tiller.
That
said,
juicy
target
like
excellent,
because
security
is
a
priority
and
that's
not
there
anymore.
A
B
Bridgette
commented
on
how
helm
three
really
came
from
listening
to
all
of
you.
Everyone
who
uses
it
and
a
lot
of
it
was
really
listening
to
how
you're,
using
and
implementing
helm
in
production
and
realizing
security
needs
that
weren't
being
met.
So
that's
where
it
is
more
secure
and
it's
focusing
on
how
to
really
adopt
and
consider
that
security
in
production,
real-world
scenarios
and.
A
It's
not
that
it
wasn't
possible
to
secure
before
it
absolutely
was.
It
was
just
a
lot
more
work
and
now
we're
going
with
security
as
a
primary
feature
by
default,
and
maybe
one
of
the
things
that
might
be
a
surprise
if
people
haven't
been
looking
at.
You
know
the
commit
commit
history
back
for
many
years.
Is
they
helm
literally
almost
predates
kubernetes
like
it
definitely
predated
a
whole
bunch
of
the
one
six
one,
seven
one
eight
features
that
we
consider
to
be
basic
and
necessary.
At
this
point,
yeah.
B
B
Of
days,
but
it's
still
it's
it's
almost
as
old
as
kubernetes,
it
kind
of
came
after
the
fact
that
was
home
classic
by
the
time
helps
you
came
around.
You
had
two
different
code
bases
that
were
really
being
merged
because
you
had
Google,
also
bringing
their
I
forget
what
their
name
of
it
was
limit
management
deployment,
manager,
yeah,
very
a
lot
of
creativity,
but
bringing
that
in
and
then
now
you
have
helm
too,
and
so
that
really
sets
us
up
for
where
we're
going
with
yeah.
A
Today
and
I
think
that
one
of
the
one
of
the
major
refactoring
was
the
fact
that
this
predated
role-based
access
control,
which
is
essential
and
fundamental
to
helm
at
this
point,
but
yeah
or
rather
to
kubernetes
at
this
point,
but
it
helped
me
actually
came
in
and
implemented
this
sort
of
thing
in
its
own
way,
as
well
as
C,
RDS,
so
I
think.
Probably
some
of
the
changes
that
are
necessary
and
helpful
is
like,
instead
of
having
a
custom
way
to
do.
B
Having
it
more
native
and
you
relying
on
what
kubernetes
already
does
and
doesn't
reinvent
a
second
option
or
a
second
way
of
doing
things
so
really
pulling
in
what
kubernetes
does
and
treating
helm
truly
as
what
it
was
designed
to
be,
which
is
a
package
manager.
It
is
designed
to
be
an
on-ramp
in
to
kubernetes,
so
utilize,
kubernetes
resource
objects
with
helm,
charts
for
simplification
and.
A
A
Think
that
honestly,
this
is
a
this
to
me
is
one
of
the
biggest
signs
of
project
maturity,
because
it's
very
easy
to
have
sprawl
and
feature
creep
and
add
lots
and
lots
of
things
and
choosing
to
take
the
time
to
edit
yourself
to
make
the
user
experience
better
is
a
lot
of
work
yeah,
and
so,
when
we
say
like
we're,
shipping
tiller
removal
as
one
of
the
major
features,
what
we're
shipping
is
we're
lowering
the
barrier
of
entry
for
contributors,
we're
shipping,
more
stability,
more
security
and
just
ease
of
use.
Yeah.
B
Speaking
of
these,
to
use
just
to
kind
of
see
how
fast
I
could
get
this
up
and
running
using
helm
on
a
brand-new
kubernetes
cluster
I
spun
up
a
cluster
yesterday
that
took
about
five
minutes
and
then
I
downloaded
helm
and
pretty
much
ran
helman
it
and
that
was
it.
I
didn't
have
to
create
service
accounts.
I
didn't
have
to
do.
Cluster
roll
binding
I
didn't
have
to
make
sure
that
tiller
was
ready
and
available.
B
It
was
ridiculously
simple
just
by
show
of
hands
how
many
people
in
the
room
are
familiar
with
draft
okay,
so
like
another
10
people,
so
draft
actually
works
in
combination
with
helm
and
simplifies
helm.
It
allows
you
to
easily
get
up
and
running
and
start
projects.
Draft
was
also
created
by
the
same
team
that
created
helm.
When
draft
first
came
out,
it
operated
very
similar
to
helm.
You
would
have
an
in
cluster
daemon
that
would
essentially
sit
there
and
translate
what
you're
passing
back
and
forth
probably
about
nine
months
ago.
B
I
want
to
say
that
daemon
was
removed
from
draft
and
draft
now
operates
the
same
way.
It's
a
client
only
tool.
So
for
those
of
you
in
the
room
that
have
used
draft,
if
you
haven't
check
out
draft,
create
its
on
Twitter,
it's
awesome.
It
will
help
you
with
helm,
but
for
those
of
you
in
the
room
who
are
familiar
with
it.
B
A
Absolutely
and
so
again
simpler
and
more
secure.
We
have
a
few
other
changes
that
we
want
to
highlight
just
because
they
may
be
at
first
a
little
confusing.
These
are
actually
we
have
aliases
so
that
if
you
type
it
the
old
way,
it's
fine,
it's
just
gonna.
Do
the
new
thing,
the
CLI.
You
have
slight
changes,
yeah.
B
And
you
can
do
so:
I
have
a
github
repo,
we're
gonna
show
at
the
end
and
I
created
a
script
that
I
don't
recommend
just
running
the
script
on
your
system.
I,
never
recommend
that
read
the
script
first,
please
come
on
trust.
You
right!
I,
I,
don't
advocate
that
at
all,
but
it'll
show
you
how
to
set
it
up
with
an
alias
and
the
demo
I'm
going
to
show
you
today,
the
helm,
alias
I'm
using
is
h3,
but
essentially
when
you're
running
the
helm
commands
you
can
do
helm,
3
or
h3.
B
If
you
use
my
suggestion
and
do
what
is
it
a
chore
help,
you
can
actually
see
all
the
new
commands
that
exist,
certain
commands
removed
serve
as
one
of
them,
mainly
because
we
felt
that
people
weren't
actually
using
it
really,
but
you
can
also
see
how
certain
commands
have
now
changed.
They're
now
more
native
to
package
management,
repository
understandings
rather
than
doing
a
helm,
fetch
I
do
a
helm
pull
rather
than
having
to
do
helm,
delete
purge,
so
I
could
make
sure
I
could
reinstall
using
the
same
name.
B
A
When
we
say
more
native,
we
specifically
mean
all
of
the
finger
macros
you
have
from
working
with
kubernetes
and
with
other
tools
in
the
space
just
to
kind
of
again
how
I'm
predated
a
lot
of
that.
But
now
comes
in
to
be
more
consistent,
yep
all
right!
So
we'll
talk
about
some
breaking
changes.
We
have
a
link
here.
A
It's
not
a
link
about
breaking
changes
specifically,
but
it's
a
link
for
contributors
about
the
things
that
were
expect
between
2
&
3,
so
that
you
can
get
an
idea
of
exactly
we're,
not
gonna
go
into
all
of
them,
though
again
there
is
more
detail
in
there
in
terms
of
what
kind
of
expectations
existed
and
are
now
being
shattered
and
changed.
But
we'll
talk
we'll
highlight
a
couple
of
the
things
that
you
will
notice
right
away
since
you're,
almost
all
used
to
home.
A
B
For
one
one
of
the
coolest
things
that
I
saw
was
that
helm
deployments
now
in
helm,
3
are
specific
to
namespace.
So
that
means,
if
I,
do
helm,
install
stable
WordPress
previously
by
default
it's
gonna
go
into
whatever
namespace
I
was
in,
but
I
wouldn't
be
able
to
name
not
the
same
name.
So
if
I
name
that,
obviously
I
don't
know
walking
panda
and
then
I
had
my
install
I
wouldn't
be
able
to
have
another
deployment,
that's
walking
panda
in
any
namespace.
B
A
You
it
will
give
you
a
slide
that
has
that
detail,
but
Before
we
jump
to
that
because
you're
right
that
is
important.
We
should
also
tell
them
if
you're
going
from
home
to
two
home
three
in
your
same
cluster,
you
can't
upgrade
and
home
to
release
directly
to
be
a
helm,
3
release.
You
can
certainly
redeploy,
but
if,
if
I'm
understanding
this
correctly,
the
migration
story
is
in
progress,
but
more
on
that,
as
the
Alpha
is
consumed
and
tested
by
people
and.
B
One
comment
for
that
is:
well:
you
can't
upgrade
a
helm
to
release
I,
also
work
very
closely
on
a
DevOps
team
and
you're,
obviously
a
coordinator
for
DevOps
days,
if
you're
doing
DevOps
best
practices
and
you're
keeping
your
charts
and
source
control
version,
controlled
you're
bumping
them.
Accordingly,
you
have
values
files
associated
with
your
projects
and
part
of
your
pipelines.
That
shouldn't
really
be
an
issue
from
a
theoretical
perspective.
You
should
be
able
to
just
do
h3
and/or
elm
tree.
B
Whatever
your
version
is
for
your
alias
and
redeploy
that
same
chart
with
the
same
values
file.
That's
gonna
operate
the
same
way,
but
if
you
have
an
existing
chart
and
you're
trying
to
upgrade
that
release,
that's
where
she's
clarifying
you
can't
directly
upgrade
right
now.
A
migration
story
is
in
the
works.
B
A
Is
a
good
thing,
because
this
is
collapsing,
simple
complexity
again
like
helm
to
the
chart,
dependency
management,
stuff
came
in
after
helm,
twos
file,
format
and
whatnot
was
already
in
place,
and
so
that's
why
this
was
often
a
separate
file.
It
doesn't
need
to
be
anymore,
and
so
the
team
put
a
lot
of
work
into
collapsing
that
complexity,
so
that
you
didn't
have
to
do
that
anymore
and
then.
B
A
B
Can
but
you
now
have
to
you,
have
to
specify
a
name
or
you
have
to
specify
a
flag
that
you
want
helm
to
generate
a
name
for
you.
It's
not
just
gonna.
Do
it
automatically,
which
I
think
is
good,
because,
theoretically
again,
if
you're,
considering
the
reason
why
helm
three
kind
of
came
about
was
also
taking
the
feedback
and
making
it
better
more
secure
for
production
based
deployments
in
production,
you're,
probably
going
to
have
it
name
a
release,
you
don't
want
a
production
serving
micro
service,
that's
walking
pan
and
it's
like.
A
B
A
There's
a
lot
of
if
this
is
a
space,
that's
interesting
to
you!
There's
a
lot
of
work
in
the
space
that
you
should
jump
in
I
mean
this
is
where
you
get
or
feedback
you
get.
You
know,
look
at
what's
in
github
and
we're
aiming
for
we'll
see
if
we
get
there
soon,
but
we're
aiming
for
you
know:
lots
of
choices,
pluggable
off
novel
artifact
types.
You
know
ability.
B
One
thing
she's
commented
on
contributing:
this
is
very
much
an
alpha
release
right.
This
is
alpha
3.0.
There
are
things
that
aren't
perfect.
One
of
the
things
I'm
going
to
show
you
in
here
today
is:
if
I
wanted
to
deploy
a
chart
to
a
namespace
right
now
the
namespace
flag
doesn't
work,
you
can
use
other
tools
to
see
which
namespace
and
still
deploy
and
that's
being
fixed,
but
it
is
an
alpha
version.
We
want
you
to
go,
use
it
and
contribute
and
be
a
part
of
it.
B
A
Didn't
plan
it,
but
it
was
perfect
for
you
to
have
a
bunch
of
talking
before
we
put
this
very
word
heavy
slide
up,
because
one
of
the
things
that
people
were
wondering
about
is
tell
me
about
this
Lua
thing
and
if
I'm,
not
a
Lua
expert,
does
that
mean
I
can't
really
use
how
many
more
I,
like
my
templates,
just
fine
and
like
actually
the
templates
still
work?
Just
fine
like
using
Lua,
is
completely
optional.
B
Examples
of
kind
of
Paul
like
in
helm,
you
would
have
late,
binding
errors.
You
can
simplify
and
streamline
your
chart
design.
In
fact,
when
you
use
Lua,
which
is
designed
as
a
programming
language,
obviously
you
can
simplify
the
amount
of
lines
in
your
deployment
itself.
If
you
have
a
chart
and
a
template,
that's
100
lines.
You
might
be
able
to
simplify
that
down
into
25
or
30,
so
there
are
benefits
for
that,
but
you
don't
have
to
use
it,
and
my
understanding
is
also
Lua
is
still
staggered
in
the
next
release.
Yeah.
A
Yeah
I
mean
it's
not
in
the
Alpha,
but
it's
coming
soon
TM
and
then
the
other
thing
to
consider
too,
is
that
one
of
the
benefits
that
you
get
is
being
able
to,
because
it's
a
real
programming
language
you
can
actually
validate.
You
know
your
charts
before
you
ship
them
and
wait.
So
that's
that's
kind
of
helpful.
You
can
put
those
conditionals
in
so
okay
library,
chair
support.
This
is
kind
of
exciting
yeah.
A
And
if
you're
thinking
to
yourself
well,
we
reusability
sounds
great,
but
I
work
at
a
giant
enterprise
and
we
have
very
important
policy
problems
to
solve
one
of
the
things
that
this
could
help
you
with
is.
You
could,
for
example,
say
here's
a
chart
that
has
some
you
know,
pod
control
stuff
built
into
it,
and
we're
gonna
make
every
one.
We
or
you
don't
import
this
particular
library
chart
so
that
we
enforce
this
policy
on
all
of
our
charts
without
having
to
hope
people
follow
the
guidelines.
So.
B
You
can
enforce
pod
security
as
a
result
is
one
example,
and
then
demo
time
so
now
I'm
going
to
go
ahead
and
I
guess
switch
over
to
my
system,
and
so
let
me
show
you
vs
code
here
so
I
have
a
helm,
SH
script,
which
pretty
much
just
has
a
bunch
of
helm.
3
commands
in
it
I'm
going
to
show
you
the
script
that
I
use
to
set
up
helm.
All
of
this
is
on
github
will
give
you
a
link
to
it,
but
essentially
I
set
a
helm
version
on
line
5.
B
You
can
see
right
here.
This
one
is
using
a
Mac.
You
could
change
the
version
to
whatever
the
version
is
it's
on
the
release,
page
of
the
helm,
download
I'm,
just
using
double
you
get
to
get
it.
I
extract
the
file
I
am
making
a
directory
in
my
temporary
directory.
The
using
the
temporary
directory
is
also
recommended
on
the
helm,
release,
notes
page
actually.
A
B
That's
one
thing
I
did
realize:
I
will
have
to
update
this
this
chart.
Actually
you
also
will
want
to
export
your
helm
home
configuration
to
go
also
to
a
temporary
directory.
So
you
don't
override
your
helm,
your
helm
home
files,
but
you
do
you
move
the
binary
over
into
a
temporary
folder
I
set
a
temporary
alias.
You
can
also
export
that.
So
that's
how
I
moved
helm
3
into
my
helm,
3,
alias
that
we're
going
to
go
ahead
and
see
I'm
going
to
switch
over
to
my
terminal
here.
B
Real
quick,
I'm
gonna
clear
this
top
screen.
No,
the
first
thing
is:
is
I
have
helm
two
and
helm
three
both
installed
on
the
system.
I
have
multiple
clusters.
Only
one
cluster
is
using
a
helm
3.
You
can
see
the
clusters
and
the
namespace
over
in
this
top
right
corner.
If
I
scroll
over
here
we're
gonna
go
ahead
and
type
helm
version,
that's
using
helm,
2
and
you'll
see
it
cannot
find
tiller.
B
But
if
I
do
H
3
version,
which
is
my
helm
3
instance
I,
don't
get
any
error
and
I
just
get
that
I'm
using
the
version
L
alpha.
So
I
can
still
have
both
versions
on
here
now,
I'm,
currently
in
the
default
namespace
on
a
cluster
that
has
helm,
3,
set
or
not
helm,
3
setup,
because
I
didn't
have
to
set
up
tiller,
but
it's
just
a
completely
clean
test.
Cluster.
B
And
it's
also
alpha
so
I
am
just
I
mean
this
is
very
much
kind
of
showing
you
how
to
get
started,
how
to
kind
of
start
playing
with
it.
If
I
do
a
helm,
LS
you're,
gonna,
see
I,
do
have
a
few
charts,
so
I
have
an
engine
X
ingress
chart
and
a
Jenkins
chart
that
I've
already
deployed
I'm
going
to
show
you
the
commands
that
I
used
to
deploy
it.
B
I
didn't
change
the
charts
whatsoever,
I
just
changed
the
alias
from
helm
to
h3
and
the
chart
works
the
same
way,
but
I
also
ran
that
command
against
the
default
namespace.
If
I
do
a
cou
been
s
and
I
change
to
my
cube
system,
namespace
I
also
deployed
the
cert
manager
chart
so
we'll
let
it
switch
over
can
see.
Cube
system
is
now
activated.
You
can
see
coop
system
over
here
on
the
right
by
the
way,
cube
NS
and
if
I
were
to
use
cube
context
to
change
my
context.
B
There
is
a
link
in
the
document
that
I'll
help
you
get
some
of
these
tools.
I
do
recommend
them
from
a
kubernetes
management
perspective.
But
now,
if
I
do
H
3
LS
you'll
see
that
I
have
a
dish
'''l
chart,
so
rather
previously,
when
you
would
do
helm,
LS
you'd
see
all
your
charts
and
all
the
random
namespaces.
Now
you
can
really
see
how
the
deployments
are
per
namespace.
I
can
switch
back
over
to
kuben
s
default
and
I'm
back
over
in
my
default
namespace,
the
commands
are
a
little
bit
slow.
B
Now,
if
I
zoom
out
we're
going
to
go
over
here,
I'm
going
to
show
you
my
home
commands
to
install
the
ingress.
I
literally
just
ran
helm,
3
upgrade
install
nginx
ingress
I
named
it,
I
could
choose
to
generate
the
name
and
then
I
chose
the
stable
upstream
chart.
If
I
were
to
copy
that
and
redeploy
it
just
upgrades
and
installs
it
the
same
way
that
it
normally
would
I'm
not
going
to
get
any
errors.
So
really
a
lot
of
the
work.
B
B
The
name
of
the
namespace,
so
I'm
running
h3
upgrade
Jenkins
install
I
have
an
install
living
in
a
coop,
Khan
namespace,
but
because
I'm
doing
this
on
stage,
I'm
leaving
the
name
of
the
deployment
the
same
but
I'm
specifically
deploying
it
to
a
coop
Khan
live
namespace,
you're,
gonna
notice.
When
I
run
it
it's
gonna
seem
pretty
uneventful,
because
it's
just
going
to
tell
me
that
it
doesn't
exist
at
an
insult
there's
not
going
to
be
any
output.
That's
spit
back
previously
in
home.
B
Apparently
now
that
did
deploy
so
literally
five
minutes
ago,
when
I
was
doing
that
it
does
not
give
you
this
long
list.
If
you
don't
get
the
long
list,
that's
by
design,
because
they
actually
pulled
all
the
code
out
of
it
to
give
you
a
spit
back
as
far
as
all
the
objects
and
they're
going
to
redesign
that,
if
you
try
to
parse
this
code
and
parse,
this
this
example
into
I
just
want
the
name.
Space
I
just
want
the
names.
B
I
just
want
the
version,
it's
really
hard
to
parse
that
code,
and
so
that
is
something
that
is
also
on
the
roadmap
for
changing
that
and
streamlining
the
way
that
you
can
actually
parse
the
spit
back
or
the
output
from
your
from
your
running
commands
and
just
to
show
you
how
the
deployment
actually
works.
If
I
do
ok,
get
pods
and
I
put
that
in
coop,
Khan
live
I,
don't
see
anything,
that's
fine,
we'll
go
ahead
and
do
K
get
pods.
Where
I
deployed
out
my
Jenkins
deployment,
you
can
see.
B
I
deployed
that
38
minutes
ago.
I
literally,
did
it
as
a
backup,
just
in
case
2
minutes
on
stage.
I
wanted
to
make
sure
that
everything
was
working.
It's
always
hard
when
you're
live
with
conference
Wi-Fi,
but
I
do
have
a
Jenkins
deployment
up
and
running.
So
if
I
do
cube,
cuddle
port
forward
I'm
going
to
port
forward
over
to
8080,
which
is
my
working
deployment
right
now,
I'll
switch
over
to
Google,
Chrome
and
I'll,
refresh
I'm,
already
signed
in
and
I
did
that
that
was
the
deployment
that
I
did
moments
ago.
B
We
can
zoom
in
and
see
localhost
I
signed
in
and
was
able
to
export
will
clear
out
this
top
screen
here.
I
ran
the
print
command
the
same
command.
That
I
would
normally
run
from
a
Jenkins
chart
to
get
my
password.
It's
ok
if
this
shows
on-screen
because
I'm
deleting
this
afterwards,
but
that's
the
password
that
I
use
to
go
ahead
and
sign
in
so
this
deployment
was
done
from
helm.
B
3
I
didn't
have
to
make
many
changes
aside
from
modifying
my
values
file
as
I
would
any
other
deployment
and
helm
3
is
going
to
operate
the
same
way.
So
a
lot
of
the
changes
that
we
reviewed
are
really
changes
that
are
under
the
hood
from
an
actual
use
perspective,
it's
very
similar
if
I
do
a
helm
or
H
3
H
you'll
see
how
some
of
those
commands
that
I
told
you
are
different.
So
now
I
can
do
helm
uninstall.
B
B
B
B
We
are
gonna,
give
you
a
link
to
the
repo
I,
the
set
up,
helm,
SH
or
the
shell
script,
that
I
created,
which
I
will
modify
it's
something.
That's
super
brand
new
they're
making
changes
as
we
speak,
but
again
it.
This
will
help
you
get
started
and
that
helm
chart
will
help
you
get
started
or
the
helm
script,
but
the
biggest
change
that
you
can
see
in
helm,
one
that
you're
gonna
get
used
to
is
the
fact
that
tiller
is
the
one
alpha.
A
Two,
oh,
but
before
we
move
off
of
alpha
one,
the
library
chart
stuff
that
we
started
talking
about
where
you
can
maybe
use
that
to
enforce
some
policy
is
available.
Now
the
way
secrets
are
stored.
If
the
secret
storage
back-end
is
available
to
test,
they
started
the
groundwork
for
the
OCI
stuff,
ACI.
B
And
there
are
also
chart
commands,
you
can
take
a
look
at
that
our
exploratory
right
now.
In
fact,
when
you
go
to
the
helm
releases
page,
they
give
you
pretty
much
a
little
overview
of
some
of
the
stuff
that
we've
talked
about.
But
it
tells
you
where
we're
out
with
some
of
the
new
commands
and
some
of
the
new
things
that
the
helm
team
is
testing.
B
A
B
Having
a
yeah,
especially
having
Lua
as
a
plug-in,
that's
definitely
on
alpha
three
and
again
a
lot
of
this.
This
is
all
what's
on
the
roadmap
right
now.
This
can
change.
There
are
certain
things
that
were
on
the
roadmap
for
alpha
one
that
changed.
We
need
your
feedback,
so
that's
going
to
lead
us
I
think
into
the
next
slide,
which
is
get
involved
in
order
to
really
have
a
product.
That's
going
to
be
able
to
deliver
value
and
that's
going
to
do
what
you
needed
to
do
in
production
and
in
your
workloads.
B
A
B
Also
go
to
the
helm:
three
docks
it's
at
the
v3
dot,
he'll
miss
H,
Summa
parts
of
the
docks,
you'll
notice
are
still
talking
about
home.
To
this
really
is
alpha.
This
really
is
brand
new,
but
if
you
see
something
that
should
be
clarified,
if
you
like
contributing
also
the
docks,
you
can
also
contribute.
I
think
it's
on
the
version
three
branch,
but
you
can
also
contribute
to
the
docks
as
well.
We
we
essentially
want
to
listen.
We
want
to
work
with
all
the
members
and
communities,
so
we
invite
you
to
get
involved
and
if.
A
You're,
like
pull
requests
accepted.
That
sounds
so
hostile
I
just
want
to
talk
to
some
people.
You
there's
a
community
called
literally
every
week.
You
can
see
the
link
on
home,
dad
SH
and
we're
very
interested
in
like
the
new
use
cases
and
workflows
that
you
come
out
with
given
the
changes,
because
we
want
this
to
work
for
you
and
if
you
would
like
to
spend
a
lot
of
time
with
actual
helm,
people
for
like
a
two
day
event
in
wonderful
Amsterdam
in
the
fall
which
is
gorgeous.
A
A
Absolutely
so
that's
the
place
to
be
if
you
want
to
be
doing
in
person,
helm,
project
stuff
and
we
do
have
a
blog
post
that
is
listed
with
all
of
the
details
for
that.
If
you
want
to
learn
more
about
all
of
these
things,
I
mean
we
have
the
days
lab.
Stat
io
page
points
to
a
lot
of
the
tooling
in
this
space,
including
draft
other
tools
that
just
could
talk
about
and
then.
B
You
can
also
go
check
out
all
the
scripts
I
tried,
there's
a
scripts
folder.
It
actually
has
a
set
up
aks
if
you
wanted
to
get
started
with
setting
up
a
cluster
and
then
set
up
helm,
which
is
the
essentially
the
SH
file.
The
shell
script
file
that
I
went
through.
That
really
just
shows
you
how
to
set
it
up.
Put
it
into
a
temp
directory.
I
will
update
that
with
some
of
the
export
commands
and
kind
of
streamline
it,
but
it's
a
great
kind
of
little
okay.
A
Of
course,
since
we
work
at
Microsoft,
we
can
tell
you
that
there
are
some
great
Doc's
available
for
our
manage
kubernetes
solution,
if
you're
into
that,
and
if
you
just
want
to
learn
more
that
short
link
there
for
the
Microsoft
learn
pages
specifically
around
kubernetes
have
some
really
good.
Like
light
box
videos
of
Brendon
burns,
you
know
explaining
concepts
so
show
that
all
to
your
team
who
are
trying
to
on
board
on
this
stuff,
yep
and.