►
From YouTube: Intro: Helm - Matt Farina, Samsung SDS
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Intro: Helm - Matt Farina, Samsung SDS
Helm is a package manager that provides the tools to find, share, and use software built for Kubernetes. The package management features allow you to define, install, and upgrade applications while also handling dependencies on other applications. In this session you will learn about: - Why Helm and package management are useful to applications in Kubernetes - A brief history of Helm - The different parts of Helm and how they work together - Publicly sharing applications through charts - Where Helm is headed
To Learn More: https://sched.co/Grc8
A
So
here's
what
we're
gonna
cover,
because
I
like
to
give
a
little
upfront
first
we're
going
to
talk
about
what
is
helm
and
why
do
you
need
it?
Then
we're
going
to
get
into
where
the
helm
project
is
going
because
helm
as
we'll
talk
about
as
a
package
manager
and
what
that
means,
but
there's
more
to
it
than
that,
and
so
we're
going
to
talk
about
what
we've
got
where
we
came
from,
where
it's
going
in
and
how
we
got
there
and
then
we're
gonna
learn
about
helm
version
3.
A
If
anybody
was
in
the
keynote
I
think
they
talked
a
little
bit
about
it
helm
version
3
is
the
next
major
initiative.
We're
gonna
talk
about
where
you
can
learn
more
about
that
and,
of
course,
if
you're
interested
in
getting
involved,
you
can
learn
more
about
how
to
get
involved,
whether
it's
for
support
or
if
you
want
to
contribute
to
a
fun
open-source
project.
A
But
first
to
let
you
know,
hi
I'm,
Matt,
Farina
I
am
one
of
the
helm
mats.
It's
one
of
the
jokes.
We
have
going,
there's
actually
3
hell,
mats,
there's
Matt
Fisher,
also
known
as
bacon
gobbler.
If
you
look
at
the
helm,
CLI
releases
you'll
often
see
he's
the
person
who
did
the
release
of
those.
He
does
lots
of
our
release
management
but
he's
one
of
the
core
developers.
Then
there's
Matt
butcher,
who
is
known
as
techno
Sophos.
A
A
A
And
after
that,
of
course,
we
had
helm
version
two,
and
then
we
had
this
massive
growth
and
I'm
gonna
talk
more
about
that
in
a
minute.
But
how
version
two
had
all
of
this
growth?
And
we
learned
a
lot
because,
along
the
way
we
talked
to
you
and
people
like
you
who
are
out
in
the
community
using
how
dealing
with
the
problems
in
the
day
in
and
day
out
and
working
through
that
and
helm
during
that
time
was
up
against
other
tools.
There
were
other
package
managers
at
the
time.
A
A
There
are
tools
to
display
packages.
There
are
mirrors
for
people
who
want
to
mirror
it
in
their
own
prime
situations
or
to
have
it
local
or
part
of
their
CI
tool
chains.
Then
there's
other
automation,
tools
that
grab
them
scan
them
and
do
things
like
that,
so
it
all
contributes
to
it.
But
that's
a
lot
of
people
using
hell
and
that's
why
we
went
to
the
CNC
F.
That's
a
lot
to
scale.
We
have
a
lot
of
maintainer,
we
have
a
lot
of
users
and
there
were
things
that
we
needed
support
with
conferences.
A
A
You
have
team
to
support
us
because
they
spent
most
of
their
time
supporting
kubernetes,
and
so
that's
the
reason
we
had
to
move
to
the
CN
CF
because
helm
is
is
achieving
a
lot
of
growth
because
it
turns
out
we're
trying
to
build
something
useful
and
hopefully
we're
able
to
do
that
for
you
and
then,
of
course,
helm.
Is
a
package
manager,
I'm
gonna
break
this
down
a
little
bit,
but
it's
to
easily
share,
consume
and
manage
applications
defined
by
kubernetes
resources.
So
what
is
easy
easy?
A
It's
about
experience
right,
kubernetes
can
be
hard
and
we're
gonna
see
more
of
that,
because
we're
gonna
have
examples
here
in
a
minute
and
we're
gonna
look
through
certain
situations,
but
kubernetes
can
be
hard,
especially.
I
have
to
deal
with
a
lot
of
manifests.
Maybe
things
you're
not
familiar
with
that?
A
You
need
to
deploy
because
there's
business
logic
in
it
right
if
you're
gonna
use
a
database,
you
have
to
know
the
business
logic
of
a
database
in
order
to
in
order
to
fit
that
into
running
in
kubernetes
and
running
efficiently
at
scale
or
with
backups
and
things
of
that
nature.
And
so
how
do
you
make
that
work?
A
Right
kubernetes
looks
at
things
in
terms
of
resources:
you
might
have
a
deployment
or
a
stateful
set
or
a
secret
or
a
config
map
or
ingress
you're
gonna
have
a
lot
of
these
different
things,
but
when
we
think
about
things
in
terms
of
applications,
it's
a
lot
different
than
thinking
in
terms
of
resources.
You
might
increment
a
whole
application
together
rather
than
small
parts
you
think
of
it
as
a
bundle
of
resources,
and
then
how
do
you
control
it?
How
do
you
manage
it?
How
do
you
version
control
it?
A
A
So
how
do
you
do
that
with
the
assets
that
don't
and
then
how
do
you
do
that
when
you're
talking
about
a
package
as
a
whole
or
a
group
of
things
as
a
whole,
and
so
we
looked
at
you
know
these
kinds
of
things
of
how
do
you
deal
with
applications
this
way
and
of
course,
we
like
to
share
we're
an
open
source?
We
we
share
libraries,
if
you
look
at
things
like
Debian
and
package
management,
things
we'd,
love
to
share
our
packages
and
then
reuse
them.
A
You
know
in
Linux,
I
can
remember
the
last
time
I
installed
my
sequel
or
Maria.
It's
always
apt-get
install
her
yum
install
for
it
because
I
get
other
people's
knowledge
and
how
to
do
that,
and
so
we
like
to
share
and
we
like
to
consume,
and
then
we've
got
to
do
things
like
manage
like
how
do
you
upgrade?
How
do
you
update?
How
do
you
make
those
kinds
of
changes,
and
so
we
looked
at
a
package
manager,
that's
helping
enable
this
whole
experience
and,
of
course,
trying
to
make
it
easy
and
simple.
A
But
then
we
got
a
look
at
how
does
a
package
manager
fit
with
the
other
tools
around
it?
So
if
we
think
about
maybe
an
operating
system
package
manager,
a
lot
of
times,
they'll
have
a
configuration
manager
at
the
top,
and
then
it
uses
package
managers.
If
you
use
ansible
or
chef
or
puppet,
you
might
have
that
at
the
top,
and
then
you
might
use
apt
or
yum
package
management.
Then
you
got
binaries
and
config
and
operating
systems.
A
Things
like
that
right,
and
this
even
relates
to
things
like
programming
languages,
if
you're
using
NPM
that
might
sit
at
the
configuration
management
layer,
but
with
package
management.
You
know,
you've
got
your
package.
Json,
you've
got
your
source,
it's
running
in
node.
You've
got
that
kind
of
environment
of
these
layers
and
they
touched
on
a
little
bit
here.
But
this
is
where
you
would
do
with
a
typical,
maybe
Linux
setup,
where
you're
doing
these
kinds
of
things
and
you've
got
your
configuration.
That
goes
in
Etsy.
A
Your
elf
binary
is
because
I'm
figuring
linux
and
you've
got
these
different
parts
that
all
work
together,
well
home
kind
of
fits
into
that
same
space.
Right
with
helm,
helm
is
the
package
manager,
letter
and
images
are
those
binaries
that
you
have
to
run
and
you
might
have
many
of
them
in
order
to
orchestrate
something
together
and
then
there's
kubernetes
objects
that
describe.
How
do
you
do
that?
A
In
fact,
in
some
situations
you
might
even
say,
operators
can
fit
alongside
these
as
a
third
option
that
can
be
bundled
up
in
a
package
and
then
the
environment
they
run
in
is
kubernetes
and
at
the
top
I
put
configuration
managers.
Now
you
could
use
things
like
ansible.
In
fact,
ansible
has
a
helm
plugin
and
there
are
those
who
just
call
out
to
the
helm
binary
to
do
it
as
well
there
and
so
ansible
and
those
other
systems
can
also
work
with
it.
A
But
these
are
a
few
that
have
grown
up
around
the
helm
environment
that
people
use,
and
so
that's
why
I
point
these
out,
but
you
can
actually
insert
a
configuration
management
layer
on
top
of
it
to
get
everything
to
work
together.
So
that
way,
you
can
reproducibly
do
things
to
your
dev
environment,
staging
production,
multiple
production
environments,
and
how
do
you
scale
that
out,
while
still
keeping
a
manageable
package?
A
A
You
just
got
different
parts,
you
have
to
work
with,
and
so
what
happens
when
you
combine
all
those
together
to
do
this?
We've
done
this
in
one
of
our
repositories
and
to
go
deploy
this
you're
gonna
run
into
about
13,
kubernetes
objects
and,
of
course,
they're
grouped
around
the
database
and
WordPress.
How
do
you
manage
that?
Well,
that's
where
he'll
come
zin
and
charts,
and
so,
if
you
were
gonna,
go,
do
this
right
and
helm.
It's
actually
a
pretty
straightforward
thing.
A
In
order
to
break
this
down,
here's
the
command
Elm
install
stable
WordPress
in
this
case,
I'm,
actually
setting
a
parameter
pass
in.
So
if
I
were
going
to
go,
install,
say
my
sequel
on
one
of
the
things
that's
gonna.
Ask
me
for
is
what
is
the
root
password
I'm
installing
it
with
because
it
has
to
collect
information?
A
In
fact,
Debian
packages
and
package
managers
have
long
had
the
ability
to
collect
that
information
at
runtime
to
pass
into
the
package
so
notice
how
to
configure
things
helm
does
that
too,
and
so
in
this
case,
we're
collecting
a
piece
of
information
to
pass
into
the
package
to
know
how
to
do
it.
The
stable
here
is
the
stable
community.
A
Wordpress
needs
something
my
sequel
compliant
and
our
case
we're
using
Maria
DB
and
the
Maria
DB
chart
has
a
bunch
of
kubernetes
manifests
in
it,
and
those
manifests
have
an
interface
where
you
can
pass
things
into
it
and
then
the
WordPress
chart
depends
on
the
Maria
chart,
and
so,
when
home
goes
to
deal
with
it,
it
says:
okay,
I've
got
WordPress
I've
got
Maria
I've
got
Maria
at
these
versions.
How
do
I
go
install
that
I
get
that
installed?
A
A
Someone
helped
a
few
things.
You
probably
need
to
know
is
general
terms
is
there's
Hjelm,
which
is
the
command
line.
Client
tiller
is
a
it's.
It
runs
in
your
cluster.
It's
like
an
early
precursor
to
the
modern
controllers
you
have.
It
works
like
a
controller
running
in
your
cluster.
That
does
a
bunch
of
the
management.
There
then
there's
a
repository,
it's
kind
of
like
a
container
repository,
but
it's
for
sharing
and
storing
charts
and
sticking
charts
and
getting
them
out
in
a
chart
is
our
unit
of
a
package.
A
So
let's
look
at
what
a
unit
of
a
package
looks
like
and
we'll
look
at
the
WordPress
to
continue
this
example.
So
up
on
the
screen
here,
we've
got
a
few
packages.
The
readme
describes
what's
going
on,
but
then
there's
the
chart
gamal
and
that
describes
the
package
holds
metadata.
Is
there
an
icon?
Is
there
a
description?
Are
their
keywords?
Does
it
have
maintain
errs
a
lot
of
this
stuff
if
you've
done
package.json
or
some
of
these
others
you're
actually
gonna
see
those
same
kinds
of
fields
in
there,
so
you
can
deal
with.
A
How
do
you
manage
the
package?
How
do
you
find
out
details?
How
do
you
find
out
the
website
for
the
package?
So
you
can
learn
more
information,
it's
all
encapsulated
in
there.
Then
you've
got
the
requirements,
files
requirements
lock
any
ammo.
This
will
look
a
lot
like
your
requirements,
definition
in
NPM
or
glide
or
def
or
PHP
composer
or
many
of
the
other
systems.
It
has
the
same
feel
the
same
functionality.
A
It
was
designed
by
people
who
who've
done
package
management
with
these
other
tools,
and
so
it
has
that
same
feel,
except
it's
an
amo
file
rather
than
JSON.
Then
there's
the
templates
directory
helm
is
known
for
working
with
templates
and
template.
Izing
kubernetes
objects
in
order
to
pass
things
in
in
order
to
generate
them
with
logic
and
things
like
that
and
then
there's
values.
A
Values
is
free
for
me
ml
to
describe
passing
properties
in
with
defaults
that
you
can
use
inside
and
we'll
look
at
that
again
in
a
minute,
but
it's
the
ability
to
have
an
interface
where
here's,
where
you
collect
that
information,
whether
it's
a
database
password
or
a
name
like
hi.
You
know
cloud
native
con
here
is
where
you
describe
that
and
pass
that
in
that
can
then
be
used
inside
of
your
templates.
A
So
in
the
templates
directory
we
have
a
whole
bunch
of
things,
and
these
are
mostly
templates.
The
notes
text
file
is
a
template
that
notes
that
print
out
after
you
installed
it.
If
you've
used
homebrew
on
Mac,
you
might
be
familiar
with
out
spits
out
some
notes
on
descriptions.
It's
the
same
thing.
The
helpers,
TPL
file
is
template,
helper
functions
and
it
comes
with
a
few
in
a
default
chart.
A
When
you
use
home
to
create
it
or
you
can
go
ahead
and
create
your
own
and
add
them
there
and
then
the
rest
of
them
are
going
to
look
like
other
kubernetes
objects.
There's
a
deployment,
there's
an
ingress,
there's
a
PVC
there's
a
service.
These
are
other
templatized
kubernetes
objects
and
we'll
look
at
them
in
a
minute.
But
one
of
the
other
things
we
have
here
is
we
have
tests
and
you
can
actually
have
and
do
tests
that
sit
alongside
your
other
stuff.
A
So
if
we
were
going
to
go
take
that
we
could
say
helm
install
stable
WordPress,
you
know
name
WordPress
and
here
I'm,
giving
it
a
name.
If
you
don't
pass
in
a
name
to
call
it
helm,
will
automatically
generate
one
and
then
to
run
the
test.
I
just
say:
helm,
test,
WordPress
and
it's
gonna
go
run
the
test
and
give
me
the
feedback,
but
that's
how
you
can
run
tests
on
the
community
charts
repo.
We
actually
do
this
as
part
of
CI.
When
something
gets
to
a
certain
place.
A
In
there
we
will
install
a
chart
into
a
cluster.
We
will
run
the
tests
on
it,
get
the
results
make
sure
it
passed.
So,
let's
look
inside
a
template
file
inside
a
template
file.
You're
going
to
see
a
lot
of
things
here,
so
this
kind
of
looks
like
a
secret,
but
what
you're
actually
going
to
see
here
is
certain
things.
So,
if
values
ingress
enabled
well
in
that
values,
animal
file,
we
looked
at
a
little
while
ago.
There
is
a
Yemen
structure
of
ingress
and
underneath
it
it
has
enabled
and
it's
abilene.
A
So
it
can
be
true
or
false,
and
so
is
it
enabled
if
it
is
continue
with
this
file,
if
not
skip
it,
don't
generate
anything
right
and
then,
in
this
case,
because
it's
dealing
with
ingress
and
ingress,
this
is
for
TLS.
It's
got
secrets
that
can
actually
be
an
array,
and
so
we
range
over
that
and
then
those
things
become
available,
and
so
in
this
case
it
creates
a
secret
for
each
thing
and
then,
within
that
you're
gonna
see
where
it
actually
creates
based
on
a
dollar
sign
and
release
name
and
service.
A
A
All
of
that
gets
bundled
up
into
a
turbo,
and
in
this
case
here
you've
got
the
name
of
the
package,
its
semantic
version.
So,
if
you
look
at
it,
you
can
see
what
it
is.
You
can
open
it
up
and
you
can
inspect
it.
You
can
do
those
things
and
the
way
to
actually
convert.
It
is
really
simple:
helm
package
dot
if
you're
in
the
root
directory.
Otherwise,
it's
the
path
to
it
and
a
nice
little
tip
here
is,
if
you're
into
signing
right.
A
You
want
to
be
able
to
sign
and
then
verify
later
using
PGP,
because
you
want
to
know
who
it
came
from
and
validate
that
you
can
actually
add
a
sign
flag
to
it
and
then
many
of
the
other
commands
like
helm,
install,
have
a
verify
flag
and
so,
along
with
this.
Well,
that's
simple:
you
can
actually
lay
her
in
that
cryptographic.
Verification
along
the
way,
if
that's
something
you're
interested
in.
A
So
here's
a
handful
of
other
commands
because
it
should
look
pretty
simple,
so
you
can
do
home
repo.
Add
this
should
look.
A
lot
like
you
know,
get
remote,
add
write
to
add
a
new
repository
in
this
example.
It's
the
incubators,
the
name
we
give
it,
but
you
can
give
it
any
name
you
want,
and
then
you
just
specify
a
repository
location.
If
you
want
to
fetch
it
locally
right,
you
don't
want
to
install
it,
but
you
want
to
fetch
it
from
a
repository
and
inspect
it.
You
can
use
helm
fetch.
A
In
fact,
if
you
want
to
there's
a
version
flag,
you
can
pass
it
and
say
give
me
this
specific
version,
and
so
you
can
get
old
versions
from
that
repository.
If
you
want
to
install
it
again,
that's
helm,
install
if
you
want
to
update
a
running
instance.
Maybe
you've
changed
the
values
you
want
in
it.
Maybe
there's
an
update
that
works
with
semantic
versioning
to
what
you
have
and
you
want
to
update
to
a
new
version.
Maybe
you
know
we
have
WordPress
3.30
and
now
we're
looking
at
3.40
and
we
want
update
the
update
command.
A
Doesn't
it
does
it
all
together,
so
your
secrets
will
be
updated
and
your
secrets
will
be
updated
in
a
way
new
secrets
created
with
the
new
information
and
when
kubernetes
spins
down
the
old
deployments
and
stateful
sets
that
you
know
relied
in
the
old
one
when
those
are
gone.
It'll
clean
up
that
old
secret
as
well
for
you,
so
it
handles
your
upgrades
and,
of
course,
there's
actually
a
rollback
command,
because
sometimes
you
want
to
rollback.
A
You
deployed
something
oops
we
broke
something,
and
in
this
case
you
can
roll
back
and
it'll
roll
back
the
thing
as
a
whole
to
the
old
values,
to
the
old
configuration
to
the
old
version
and
it'll
make
that
work,
because
we're
looking
at
the
ability
to
manage
and
to
deal
with
upgrades
and
downgrades
and
all
those
things
now
there's
more
to
helm
than
just
the
helm,
CLI
that
we
walk
through
there's
a
whole
bunch
more
going
on
in
the
helm
ecosystem.
The
first
thing
we
talked
about
a
little
bit
is
the
stable
and
incubator
charts.
A
A
People
are
testing
that
repo
is
not
added
by
default,
but
if
you
install
helm
version
two
today,
you're
gonna
get
the
stable
charts
repo
they're
just
available
for
you
out
of
the
box
and
the
ability
to
access
it,
and
that's
where
many
of
the
charts
that
people
use
are
today
and
if
you
want
to
get
a
template
or
a
place
to
start
and
extend
that's
a
great
place
to
look
at
it.
We've
got
review
documents.
We've
got
common
patterns,
things
that
you
can
learn
from
things
that
you
can
use
in
your
own
charts.
A
A
We
had
some
open
source
projects
like
Kiel
that
don't
have
an
association
but
they're
they're,
doing
and
sharing
their
projects
and
the
ability
to
run
them
through
helmut
repositories
and
we're
going
to
expand
this,
because,
if
you
click
on
that
about
link
you're
going
to
learn
about
how
you
can
add
your
own
repositories
to
this.
In
fact,
this
is
one
of
those
projects
where
it's
fresh
and
the
UX
and
UI
can
use
some
work.
A
And
so
this
is
a
great
place
for
UX
and
UI
people
to
come,
join
in
and
help
us
continue
to
grow
that.
But
this
is
one
of
the
projects
that
we
maintain,
because
helm
is
thought
of
also
as
a
platform
just
a
CLI.
We
don't
want
to
be
that
NPM
central
repository.
We
really
want
to
enable
you
and
things
like
the
stable
and
incubator
repositories.
A
And
so
that's
one
of
the
projects
we
have
going
and
this
the
UI
is
built
on
top
of
monocular
and
monocular.
Is
you
can
install
yourself
for
your
own
repositories?
It
was
the
foundation
for
Kubb
apps
that
was
long
run
by
bitNami
and
still
is,
and
monocular
is
that
tool
you
can
run
it
on
Prem,
pointing
to
your
charts,
repos
and
have
a
visual
interface
with
commands
and
and
be
able
to
see
here's
here's
what
we've
got
available
for
us
now.
A
If
you
want
to
run
a
chart,
repository
of
your
own
helm
out
of
the
box
has
the
ability
to
create
a
repository.
It's
a
static
repository.
You
can
push
to
a
flat
web
server
or
put
into
object
storage
things
like
that.
If
you
want
something
with
more
juice,
more
power,
a
dynamic
server
that
can
do
these
things.
Chert
museum
is
a
project
under
helm.
A
That
does
that,
in
fact,
art
museum
is
bundled
with
harbor
one
of
the
other
CN
CF
projects
to
provide
chart
repositories
for
harbor
users,
because
there's
a
lot
of
different
ways
that
you
can
consume
and
use
that
and
then
there's
tools
to
help
you
write
I
just
talked
about
that
a
minute
ago
with
running
your
own
repositories
and
one
of
the
tools
we
have
is
called
chart
testing
a
pretty
straightforward
name.
It's
what
we
use
when
we
test
the
stable
and
incubator
charts
it's
what
we
ourselves
run
its
containerized.
A
It
allows
us
to
do
static
analysis
of
our
charts
to
look
for
schemas
things
like
that.
You
know:
does
it
pass
helm
commands?
Does
it
validate
schemas?
Does
it
follow
good
practices?
And
it's
got
a
lot
of
flags?
You
can
pass
it
and
you
can
run
it
yourself
on
your
own
repositories,
because,
even
though
we've
run
it
we've
packaged
it
as
a
container,
you
can
go
run
it
in
circle
or
Travis
or
or
Jenkins
or
other
CI
tools,
and
you
can
running
yourself
too.
A
In
fact,
many
people
do
that
today
already
and
then
there's
the
ecosystem,
because
the
home
project
does
so
much,
but
we've
actually
had
ways
to
to
extend-
and
you
know,
I
talked
a
little
bit
about
that
with
chart
museum.
We
want
to
go
beyond
just
the
core
CLI
and
a
lot
of
this.
We
want
to
set
people
up
to
be
successful
and
have
space
to
solve
their
own
problems
and
do
their
own
things.
So
we're
gonna
name
a
few
things.
A
He'll
actually
has
plugins
so
like
it
has
plugins
helm
has
plugins,
and
some
of
these
are
things
like
helm,
s3
or
GCS,
or
hum
github.
These
tools
actually
help
you
run
repositories
in
those
places
right.
So
how
much
three
knows
how
to
do:
s3
private
repositories
and
work
with
helm
and
there's
a
lot
of
those
things.
A
There's
tiller,
less
helm
v,
what
let
you
run
helm
v2
without
having
to
deal
with
tiller
running
in
your
cluster,
because
some
people
want
that
there's
helm
GPG,
which
will
use
your
systems
GPG
to
do
signing
and
verification
of
charts
and
try
to
make
that
process
easier,
and
these
are
just
to
name
a
few.
Then
there's
things
like
supporting
tools
around
it.
There's
a
drone
plugin,
there's
coop
apps,
there's
landscaper,
there's
charter
fire.
A
There's
all
these
projects
that
are
supporting
tools
around
home
and
then
there's
services
right,
kwai
has
the
app
registry,
where
you
can
put
your
home
charts
code.
Fresh,
has
managed
repositories,
there's
a
whole
bunch
of
services
that
are
growing
up
around
it,
and
we
want
to
enable
those.
In
fact,
these
repositories,
like
those
from
code
fresh.
A
You
can
have
those
in
the
hub
because
we're
trying
to
give
you
the
opportunity
to
use
the
services
and
still
expose
things,
and,
of
course,
if
you
want
to
find
out
more
about
what's
in
the
ecosystem
right,
we
have
a
website
for
that.
You
can
go
to
that
website
and
you
can
learn
more
about
what's
in
the
ecosystem
and
how
to
use
it.
So
where's
home
going
I
said
I'd
answer
that
so
the
first
thing
is
home,
v3,
alright,
so
helm
we
take
semantic
version
very
seriously.
I
know
kubernetes,
sometimes
like
breaks
and
changes.
A
P
is
between
minor
versions.
Helmet
rides
very
hard,
not
to
andwhen
helm.
V2
came
out
over
a
year
ago.
Long
over
a
year
ago,
we
said
all
right:
we're
not
going
to
break
our
API,
so
we're
gonna
follow
semantic
versioning.
We
will
add
features,
but
we've
learned
a
lot
over
time.
We've
learned
a
lot
from
interacting
with
you,
seeing
what
you
all
do
and
other
helm
users
do,
and
so
now
we're
looking
to
break
to
make
braking
changes
to
improve
on
helm,
to
refactor
things
to
bring
in
some
new
kubernetes
things.
A
Another
thing:
CN
CF,
graduation,
we're
gonna
push
for
graduation,
probably
in
the
next
year,
and
one
of
the
things
that
I
put
up
here.
So
if
you're
part
of
the
CNC
F,
they
do
a
lot
of
things
for
you,
and
one
of
the
things
here
is
they've
got
something
called
core
infrastructure.
They
actually
make
sure
you,
as
a
project,
are
following
a
lot
of
good
best
practices.
Now
anybody
can
do
this
all
right.
You
don't
have
to
be
a
CNC
F
project
to
do
the
core
infrastructure
stuff.
A
You
can
just
go
to
the
web
site
and
you
can
get
a
badge,
but
they
walk
you
through.
Do
you
have
two-factor
authentication
on
your
github
repos?
Do
you
have
security
contacts
all
of
these
things
that,
if
you're
going
to
be
a
stable,
mature
and
healthy
project,
they
walk
you
through
and
we're
going
to
go
through
that
and
we're
going
to
go
through
a
lot
of
other
things
that
help
push
us
to
graduation
and
help
make
sure
that
we're
a
good
project
and
we've
got
our
stuff
together.
We
want
to
grow
the
hub.
A
Yes,
okay,
so
I'll
break
you
for
a
question:
real,
quick,
oh
yes,
you
can
there's
a
couple
of
ways
to
do
it.
Tiller
has
a
G
RPC
server
and
the
helm,
client
and
other
tools
can
talk
to
it
and
have
those
things
and
interact
with
it.
That
way,
some
folks
have
actually
written
a
rest
front-end
that
can
work
in
front
of
that
I'm
forgetting
the
name
of
the
project
right
now,
but
you
can
put
a
rest
front
end
in
front
of
tiller
and
do
it
that
way.
A
Helm
also
has
a
package
directory,
and
so
the
helm,
client
imports
its
own
package
directory
to
do
that.
Stuff
and
people
will
write
other
tools
that
also
import
that
package
directory.
It's
one
of
those
things
where
we
take
semantic
version
seriously,
because
we
know
people
are
importing
our
go
code
base
and
using
it
to
write
their
own
applications,
and
then
you
can
do
the
same
thing
programmatically
by
importing
that
package
thing,
and
there
are
companies
that
have
built
their
own
tooling
and
management
around
that
doing,
just
that.
So,
yes,
you
can
all
right.
A
So
if
you
want
to
get
involved,
there's
a
few
ways
you
can
get
involved,
one
you
can
attend
a
meeting.
We
have
meetings
twice
a
week.
Tuesdays
is
where
we
talk
church,
just
charts
the
repos
building
them
yourself
every
other
week
we
have
an
office
hours.
We
can
just
ask
questions
if
you've
got
an
issue.
If
you've
got
a
problem,
you
can
come
talk
to
us
about
charts,
we're
happy
to
we.
Also
in
those
meetings
and
the
other
weeks.
A
We
have
kind
of
the
helm,
developer,
weekly
call
and
that's
where
we
talk
about
where
helm
is
going,
whether
it's
version
3
or
what
we've
done
in
the
last
week
or
some
crazy
idea,
we've
got
and
we
want
to
bounce
it
off
and
discuss
it
and
it's
a
place
where
the
core
developers
get
together,
and
you
know
anybody
else
can
come
and
introduce
themselves
and
see.
What's
going
on
and
even
learn
where
they
can
get
involved.
We've
had
lots
of
people
do
that,
even
recently
to
become
involved
in
even
helping
us
with
Version
three.
A
Now
you
file
an
issue
about
half
of
our
open
issues
right
now,
for
example,
our
support
requests
of
people
asking
questions
and
we
don't
close
those
until
the
issue
is
resolved
or
it's
been
unresponsive
for
a
very
long
time,
because
we
want
to
make
sure
it's
resolved
and
if
somebody
goes
searching
for
it,
they
can
find
an
answer
sometime.
So
that's
a
way
you
can
contribute
to
codon
Docs.
We
love
contributions
and
extending
the
community,
and
you
can
learn
out
how
to
do
all
of
this.
A
On
the
home
community
site,
like
kubernetes,
has
a
community
repo
we
do.
There
is
the
meeting.
If
you
want
to
know
when
meetings
are
there's
an
iCal
fee,
there's
you
know
find
out
how
to
contribute.
You
can
get
all
of
those
details
there
on
how
to
get
involved,
and
so
with
that
are
there
any
questions.
I
think
I
have
three
minutes
or
something
for
questions
there.
Any
questions.
Yes,
so
the
question
was:
is
tiller
going
away
in
Version
three?