►
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
An Introduction to Helm - Matt Farina, Samsung SDS & Josh Dolitsky, Blood Orange
CNCF Helm is a package manager that provides the tools to find, share, and install and manage software in Kubernetes. In this session you will learn the basics about using Helm. We will look at the makeup of charts, the packages for Helm. From there we will explore the ways to share and consume charts. To complete the lifecycle, we will look at installing and managing charts in a Kubernetes cluster. Helm is a project that was first introduced in 2015 and was part of Kubernetes prior to the creation of the CNCF. It entered the CNCF as part of Kubernetes and grew to eventually became a sister project, alongside Kubernetes, in the CNCF.
https://sched.co/UajI
A
So
we
went
through
and
created
our
slide
deck
and
we
created
this
wonderful
slide
deck
of
stuff.
We're
gonna
share
with
you,
and
we
got
to
the
end
and
Josh
says
to
me:
do
you
realize
we
didn't
put
the
helm
logo
anywhere
in
the
slide
deck?
And
so
we've
got
one
slide.
This
is
the
helm
logo
for
the
project
we're
going
to
talk
about
this
is
our
logo
and
so
hi
I'm
Matt
I'm,
one
of
the
hell
maintainer
Zand.
A
This
is
Josh,
he's
the
one
of
the
other
hell,
maintainer
x',
there's
several
of
us
and
we're
gonna
walk
you
through
helm
today,
so
I
thought
it
would
be
great
to
talk
about
where
helm
came
from
to
understand
why
we
have
it.
Yeah
helmet
was
the
product
of
a
company
called
day--is
helm
version.
One
day
has
had
a
platform
as
a
service
called
workflow
and
just
with
platform
as
a
service
you'd
deploy
a
12
factor,
application
and
you'd
run
it
somewhere
and
they
built
it
so
that
you
could
run
it
on
top
of
kubernetes.
A
But
as
with
all
platform
as-a-service,
you
don't
want
to
just
have
a
12
factor
app.
You
need
some
place
to
store
your
stuff,
my
sequel
Postgres,
something
like
that
and
they
needed
a
way
to
install
that
into
your
kubernetes
cluster
as
well,
and
thus
helm
was
born.
Helm
was
born
that
gave
you
the
ability
to
install
that
other
thing
to
go
along
with
the
paths
now.
As
you
can
see,
this
is
an
archive
site.
There's
the
Internet
Archive
to
show
it
because
day
isn't
around
anymore.
A
This
is
one
of
those
things
where
helm
lived
on
and
became.
Successful
and
des
is,
is
sadly
not
here.
So
what
is
helm
helm
is
a
package
manager
makes
sense
enough
said.
Probably
not,
but
helm
is
a
package
manager
for
kubernetes
and,
as
I
said
this
morning,
it
is
kind
of
like
apt
or
yum
or
home
brew
for
kubernetes.
But
what
is
a
package
manager?
I
grabbed
a
couple
of
definitions
here,
one
is
off
of
Wikipedia
and
the
other
is
off
of
the
helm
website.
A
You
can
see
the
helm,
one
is
kind
of
targeted
at
how
helm
does
things
and
the
Wikipedia
one
talks
a
bit
about
how
things
happen.
It
does
installing
upgrading
configuring,
it's
the
the
technical
bits
of
what's
going
on
when
you
use
a
package
manager
and
the
Wikipedia
one
is
based
apt,
which
it
was
inspiration
for
home,
but
it
doesn't
get
to
the
core
essence
of
it.
I,
don't
think,
and
so
I
came
up
with.
What
I
think
is
my
own
definition
here,
and
this
is,
you
know,
a
package
manager
takes
you've,
got
knowledge.
A
They
definitely
don't
know
the
business
logic
about
Postgres
and
they
can
install
it
and
they
can
run
it
and
it
works,
and
it's
simple
and
it's
of
the
ability
to
take
that
complex
business
logic
packaged
it
up,
so
somebody
else
doesn't
need
to,
and
you
can
have
a
few
people
who
have
a
lot
of
knowledge
and
they
can
share
it.
So
many
many
people
who
don't
have
that
knowledge
can
go
use
something
and
that's
where
package
managers
are
really
powerful.
A
So,
let's
take
a
look
at
a
simple
update,
an
apt
update,
an
apt
example
right.
You
can
install
Maria
this
way
with
apps
on
a
boon
too.
It's
simple
right.
You
do
app
update
and
that'll
update
your
knowledge
about
your
packages
and
then
you
install
it
and
it
just
works.
Now
we
could
go
a
little
bit
more
complicated.
Here's
Postgres,
because
something
like
a
boon
to
doesn't
know
about
Postgres
you've
got
to
do
the
first
couple
of
commands
here.
A
Where
you
add
knowledge,
you've
got
to
deal
with
certificates
because
they
have
encryption
in
security
and
so
you've
got
to
grab
that.
And
then
you
grab
knowledge
about
the
package
for
Postgres
and
then
the
same
thing
happens.
You
do
an
update
to
get
the
latest
details
after
you've
added
a
a
repository,
and
then
you
can
install
it.
It's
pretty
simple
and
you've
got
four
lines.
You
can
paste
into
a
terminal
and
it
just
works.
A
So
let's
look
at
what
it
would
look
like
with
helm,
Helden
two
simple
lines
right
in
this
case:
we're
adding
a
chart
repo,
it's
the
bit
nami
one
in
this
case
and
then
you're
installing
Maria
DB
from
the
bitNami
repo.
It's
simple
now
it's
installed
into
your
cluster
and
it
works
that's
what
home
does
it
took
something
complicated?
This
would
be
a
stateful
set,
you're,
probably
going
to
have
some
secrets.
You're
gonna
have
other
details
and
kubernetes
manifests
and
it
makes
it
simple
to
install
and
just
have
working.
A
But
before
we
dig
into
helm
a
little
bit,
I
wanted
to
ask
a
question:
is
helm,
trustworthy
right,
you're
gonna
see
a
lot
of
things
at
the
conference
this
week,
a
lot
of
new
projects
and
products,
and
things
like
that
and
we're
gonna
definitely
walk
through
home
here.
In
a
minute,
but
I
thought
it
was
a
good
question
to
ask:
is
it
trustworthy?
A
Is
it
worth
using
and
so
I
wanted
to
walk
through
a
few
things
that
kind
of
highlight
that
the
first
is,
as
you
can
see
from
the
first
helm
security
report
here,
they
gave
us
a
glowing
recommendation.
In
fact,
this
was
one
of
a
fantastic
process
to
go
through
to
learn
about
the
security
implications,
to
have
them
review
the
code
and
the
process,
and
it's
something
we
will
repeat
but
I
love.
This
line
in
the
middle
helm
can
be
recommended
for
public
deployment.
We
never
expected
such
a
glowing
recommendation
from
a
security
auditor.
A
The
second
thing
I
wanted
to
talk
about
was
Helms
longevity.
If
you
actually
look
at
this,
we've
got
a
bunch
of
events
on
here
and
you
can
see.
Helms
started
all
the
way
back
in
2015
and
since
then,
we've
had
all
kinds
of
events
along
the
way
there
was
home
version.
1
then
helm
version,
we've
had
helm
summits,
we've
had
massive
growth,
helm
joined
the
CNC
F
other
than
being
a
project
with
kubernetes,
and
what
this
speaks
to
is
longevity?
A
Lots
of
products
and
projects
in
this
space
come
and
go.
Helm
is
one
of
those
projects
that
has
amazed
me
with
its
longevity
and
how
long
it's
been
around
for
and
that
speaks
to,
hopefully,
in
the
future
it'll
be
around,
and
of
course,
this
morning
you
probably
saw
we've
passed
a
million
downloads
per
month
and
what
I
find
most
interesting
about
this
is
that
it
speaks
to
reproducibility
right.
A
We
can
reset
reproducibility
yeah,
get
helm
out
to
places
all
over
the
internet
to
see
I
systems
like
circle
CI
or
to
countries
all
over
the
world
like
China
and
wherever
we
can
deliver
helm
to
all
of
these
places.
Reproducibly
every
month
we
use
semantic
versioning,
and
we
actually
take
this
really
really
seriously,
because
we
realize
lots
of
things
change
and
we
want
something
stable
that
you
can
count
on
and
so
with
helm.
You
know
semantic
versioning,
you
have
to
break
API.
A
So
any
time
you
break
an
API,
you
increment,
the
major
when
you
add
features
you
increment
the
minor
and
when
you
just
have
bug,
fixes
or
or
non
feature
changes.
You
can
increment
the
patch,
and
we
take
this
very
very
seriously,
because
we
know
that
if
somebody
is
incrementing,
the
minor
and
breaking
say
go
API,
so
that
can
be
really
frustrating
and
we
don't
want
to
do
that
in
the
experience
and
so
helm,
3
was
just
released
and
you'll
see
that
our
Go
api's
aren't
gonna
change
until
home.
A
4
you'll
find
that
the
command
lines
aren't
gonna
be
broken
until
helm
4.
When
maybe
we
do
that,
you
can
trust
those
things
from
helm
and,
of
course,
there's
helm
to
support.
We
know
that
helm.
You
know
it
takes
the
time
for
people
to
change,
and
so
what
do
we
do?
We
have
good
support
windows
because
we
want
to
support
everybody
who's
using
helm,
give
them
time
to
make
the
change.
A
Not
you
know,
with
everything
else
going
on
not
have
people
feel
rushed
or
stressed,
or
anything
else,
because
we're
there
to
support
you
and
here's
one
of
the
neat
things
we
do.
We've
added
release
candidates.
Whenever
we
have
a
major
or
minor
version,
we
do
release
candidates,
so
you
can
take
the
release
for
a
spin
before
it
comes
out
and
give
us
feedback,
because
we
realize
that
release
candidates
and
giving
everybody
an
opportunity
to
look
up
or
release
before
it
comes
out
and
just
upgrading
without
thinking
some
of
you
may
not
want
to.
A
In
fact,
we've
had
more
release.
Candidates
come
out
than
releases
since
we
started
doing
this,
and
if
you
want,
we
have
a
power
users
mailing
list
where
we
will
email
you
every
time
a
release.
Candidates
come
out
and
we're
going
to
do
this
go
through
this
cycle.
So
that
way
you
can
see
and
follow
along,
because
we
want
to
notify
you.
We
want
to
give
our
power
users
or
people
who
want
to
engage
an
opportunity
to
engage.
At
that
point.
A
A
We
really
do
have
lots
of
people
who
are
engaged
and
we're
there
for
the
long
haul
and
of
course
we
sign
a
release
if
you're
into
security
and
we've
got
several
security
features
in
home,
including
things
like
provenance
and
signing
charts,
but
we
actually
signed
our
every
stable
release,
so
you
can
validate
that.
It
came
from
one
of
the
hell
maintainer
x'.
A
You
can
see
that
it
came
from
us,
and
so
we
value
these
kinds
of
things
that
hopefully
lead
to
a
stable,
long
project
that
will
continue
long
into
the
future
and
that
you
can
trust
and
of
course
we
don't
just
deliver
for
Mac
and
Linux,
which
many
systems
do.
We
also
deliver
for
Windows,
because
we
know
Windows
is
incredibly
popular
with
lots
of
developers.
Some
of
them
don't
want
to
admit
it.
It
may
not
always
be
the
coolest
thing.
A
B
Thanks
hi
very
happy
to
be
here:
I
realized
what
most
people
are
over
here.
So
Oh
talk
like
this,
so
I'm
gonna
talk
about
what
what
exactly
helm
is,
how
you
would
use
it
and
give
some
detailed
examples,
so
helm
is
made
up
of
packages
that
are
referred
to
as
charts.
So,
if
you
hear
the
word
helm
chart
we're
really
just
talking
about
a
package
that
helm
knows
about.
B
B
Additionally,
you
can,
in
this
file,
specified
dependencies,
so
there
might
be
some
other
chart
that
your
that
your
package
relies
on
so
here's
an
example
Mariya
DB,
so
you
might
have
your
application.
That
requires
that
there's
a
database
running
you
can
actually
specify
in
this
file
give
me
these
external
dependencies
and
it
will
deploy
that
as
part
of
your
installation.
B
This
this
file
is
called
values.
Yeah
Mille.
This
file
basically
specifies
the
default
settings
for
your
chart.
So
things
like
what
is
the
container
image
that
your
that
your
deployment
is
going
to
use?
What
type
of
service
is
a
cluster
IP,
a
load,
balancer,
etc
and
any
other
type
of
configurable
parameters,
the
default
values
of
which
go
into
this
file
inside
your
home
turn?
There
is
a
directory
in
helm,
charts
called
templates.
Anything
under
that
directory
is
treated
as
a
yeah
mol
template.
So
what
happens?
Is
you
can
see
all
these
curly
braces?
B
B
Here's
a
special
file
that
sits
next
to
the
templates
called
helpers.
This
is
an
optional
file.
If
you
do
the
helm
create
command.
You
get
an
example
of
this,
but
this
is
for
sections
of
yam
all
that
you
may
want
to
repeat
all
over
the
place,
so
you
want
a
specific
set
of
labels
on
all
of
your
resources
for
that
application.
For
example,
you
can
set
these
these
sort
of
functions
in
this
templates
file
called
helpers
TPL,
and
then
you
can
use
this
all
over
in
your
chart
and
then
also
test.
B
So
any
any
file
that
has
a
test
in
the
name
is
treated
is
treated
as
a
helm,
test
file
and
there's
actually
a
helm
command.
Where,
once
you've
installed
a
release,
you
can
run
helm
test.
It
will
look
at
these
files
and
run
some
sort
of
check.
For
example,
is
my
database
up
things
like
that
and
then
finally
notes?
Txt,
this
is
a
friendly,
also
template
able
file,
but
this
will
basically,
when
the
end
user.
When
the
operator
installs
your
chart,
says
here's
how
to
use
the
application,
you
can
put
a
ASCII
turtle.
B
If
you
want,
you
can
do
it
every
one,
so
yeah
so
like
I
was
saying
the
templates
are
treated
as
dynamic,
Yambol,
amel
templates,
and
what
this
does
is
it
prevents
that
you're
duplicating
the
same
code
everywhere.
So
imagine
if
you
had
10
micro
services
and
you're
running
the
same
you're,
writing
the
same
deployment.
Animal
and
service
die
animal
for
all
these
things.
You
could
make
that
into
a
helm,
sharp
and
deploy
it
with
a
different
docker
image,
and
things
like
that
and
values.
B
Values
are
the
configuration
for
the
charts
and
basically,
you
can
put
them
into
a
file
values.
Yeah
most
the
defaults.
You
can
have
a
custom
values
file
which
we'll
talk
about
in
a
second
or
you
can
set
them
one
by
one.
So
the
combination
of
templates
and
values,
the
result
of
combining
these
two
together
will
give
us,
hopefully
valid
kubernetes
yeah
Mille.
That
is
then
applied
and
installed
into
your
cluster.
B
B
So
you
might
want
to,
for
example,
do
different
things,
deploy
your
application
a
little
bit
differently
in
development
than
you
would
in
production.
For
example,
you
might
want
a
node
port
service
running
locally,
but
you
actually
want
a
load
balancer
when
you're
running
an
AWS
or
some
you
know,
target
production
environment,
and
so
that's
where
custom
values
come
in
and
what
this
does
is.
It
allows
you
to
provide
custom
custom
configurations
for
for
your
chart
and
provides
dynamic
results
and
there's
several
ways
to
do
that.
B
Is
it
collapses
from
right
to
left
so
anything
on
the
right
is
giving
you
the
lip,
the
latest
and
it
kind
of
collapses
down
and
then
the
final
one
being
the
defaults
and
those
are
the
values,
apply
it
against
your
installation,
so
how
I'm
actually
manages
the
lifecycle
of
your
application?
It's
not
just
this
templating
engine,
but
it
actually
gives
you
gives
you
tools
to
manage
the
installation
and
the
removal
and
to
query
the
status.
So,
for
example,
we
have
a
helm
status
command.
B
You
run
that
against
something
you've
just
installed
and
helm
will
return
to
you
whether
it's
healthy
what
revision
it's
on
when
it
was
installed
and
we'll
give
you
the
notes
back
as
well
how
list
so
one
of
the
nice
things
about
how
in
comparison
to
just
running,
cube,
CTL
or
just
running
another
tool
that
just
installed
the
gamal
helm,
actually
stores
something
in
the
cluster
that
mat
that
manages
the
revision
of
that
of
that
release.
So
we
have
a
record
of
everything.
That's
been
installed
with
helm.
B
B
You
may
have
updated
your
chart,
so
you
can
see
my
app
directory.
You
may
have
updated
a
file
in
there
or
you're
upgrading
to
a
new
chart
version
or
in
most
cases,
you're
updating
the
configuration.
So
let's
say
you
have
an
installation
of
my
micro
service
and
you've
released
a
new
docker
tag
or
container
image
tag.
I
should
say,
and
you've
bumped
that
version
you
can
apply
it
against
the
existing
release
and
helm
will
make
sure
that
that
the
deployment
or
whatever
the
resource
was
that
was
referencing.
B
B
There
is
a
rollback
command
so
since
pel
manages
every
single
revision
of
every
single
release,
if
you
have
done
an
upgrade
and
now
you're
on
revision
2,
you
can
say
helm
rollback
my
release,
1
to
the
first
revision
and
we'll
take
back.
It
will
take
care
of
getting
you
to
that
state
that
working
state
of
revision,
number
1
and
then.
B
Finally,
if
you
want
to
completely
uninstall
your
application
from
home,
there's
a
there's
a
command
for
that
as
well
and
that's
helm,
delete,
and
this
will
anything
that
was
installed
that
was
under
that
templates
directory
that's
been
installed
in
the
kubernetes
helm
will
make
sure
that
it's
removed.
So
this
is
this
is
one
of
the
things
that
is
really
powerful
about
helm.
Is
that
you've
installed
a
bunch
of
gamal
into
your
cluster?
B
That's
part
of
a
specific
application
if
you're
just
doing
a
massive
cube,
CT
I'll
apply
and
there's
all
these
things
running,
and
you
want
to
get
rid
of
that.
It's
very
hard
to
keep
track
of
what
you're
trying
to
remove
and
so
helm
is
a
very
good
way
to
put
this
thing
into
a
box
put
it
in,
upgrade
it
and
then
take
it
out
of
the
cluster.
A
Right
so
one
thing:
I
noticed
people
taking
photos
of
the
slides
up
here
as
you
went
along
and
that's
great,
but
I
also
want
to
let
you
know
that
if
you
go
to
the
schedule
and
you
pull
up
this
session,
you
can
download
a
PDF
of
all
of
these
slides
as
well.
So
if
you
want
to
get
them
after
the
fact,
you
can
always
come
back
and
get
them.
A
So,
let's
talk
about
helm
and
there's
actually
more
to
helm
than
the
helm,
client
helm
just
like
kubernetes,
and
some
of
these
other
things
has
multiple
projects,
because
we
really
want
to
make
people
successful
using
the
client
using
packages
moving
things
around
and
it
turns
out
there's
more.
We
can
do
to
it
than
just
have
a
client,
and
the
first
thing
we
have
is
we
have
the
helm
hub.
The
helm
hub
is
a
place.
You
can
go
hub
home,
dot,
SH
that
will
help
you
find
Church
to
install
it's
a
place
where
you
can
search.
A
You
can
view
you
can
read
up
on
many
charts
from
many
people
there
distributed
all
over
the
place,
they're
not
just
hosted
by
the
home
project.
They're
hosted
by
companies
who
are
are
actually
building
the
projects
and
products
that
you
want
to
use
and
they're
hosting
these
and
well
they're
distributed
the
helm
hub
is
a
central
search
location
where
you
can
discover
those,
and
not
only
can
you
discover
them
but
like
in
this
example,
you
can
read
about
them.
So
here
we've
got
the
description,
that's
in
the
readme
of
the
package
and
on
the
right.
A
If
you
look,
you'll
actually
see,
there's
commands
on
how
you
can
add
the
chart
repository
that
it
is
and
you
can
go
ahead
and
install
it
and
if
you
see
on
the
right
there's
even
things
about
different
versions.
So
if
you
want
to
read
some
of
the
the
details
of
the
different
versions,
you
can
click
those
links
on
the
right
there
to
the
versions
and
read
more
about
past
versions,
and
you
can
see
what's
going
on
here,
but
the
helm
hub
isn't
just
for
some
targeted
set
of
companies.
A
It's
actually
for
anybody
who
wants
to
go
ahead
and
make
charts
and
share
them
with
others.
We
want
to
make
this
easy
easy
to
discover
and
easy
to
share.
So
if
you
head
over
to
get
Hubb,
you
can
see
here's
all
of
the
charts
and
here's
all
the
companies
and
people
and
their
repositories,
and
you
can
add
your
own
and
it's
as
simple
as
a
pull
request.
You
can
also
learn
about
hosting
your
own
chart
repository
there
and
how
you
could
do
that
before.
You
know
that
stuff,
you
need
to
get
started
before
you
share.
A
And
helm,
the
home
hub
is
actually
running
something
called
monocular.
Monocular
is
another
project
hosted
by
helm,
and
this
project
is
used
to
power
that,
but
it
doesn't
have
to
do
that
say:
you've
got
a
company
or
an
organization
and
you've
got
your
own
internal
charts
and
you
want
to
make
them
searchable.
Discoverable
put
a
nice
UI
on
it.
You
can
run
this
yourself
right
in
your
own
kubernetes
cluster
and
point-eight
your
chart
repositories,
and
so
this
is
one
of
the
other
things
we
have
that
stuff.
A
That
powers,
the
helm
hub
is
available
to
you,
and
if
you
want
to
run
a
chart
repository,
we
have
another
project
called
chart.
Museum
helmet
self.
The
helm
client
can
do
a
static
chart
repository
you
can
use
commands
that
will
generate
the
parts
you
need
the
index
ID
a
mole
that
transfers
information.
You
can
read
more
about
this
at
the
home
website
stuff
like
that.
A
But
if
you
want
something,
that's
more
dynamic,
something
that
you
might
put
in
front
of
object,
storage,
where
you
can
push
and
pull
charts
from
the
way
you
would
images
chart
museum
will
let
you
do
that
and
it
will
generate
the
index
that
yamo
it
sits
in
front
of
it.
That's
one
of
the
other
projects
that
we
have,
because
we
want
to
make
it
easy
to
have
chart
repositories,
whether
they're,
static
or
dynamic,
with
a
server
in
front.
A
We've
got
tools
for
that
and,
of
course,
if
you've
got
charts,
you
probably
want
to
test
them
right.
You're
gonna
hear
a
lot
of
talk
here
about
CI
CD
about
testing,
and
we
firmly
believe
in
that
we
do
constant
testing
with
everything
that
we
have
in
home
and
we
provide
a
tool.
We
have
our
own
chart
repository
with
stable
and
incubator,
and
this
is
the
tool
that
we
use
and
we
developed
it
in
the
way
that
helm
needs
it
to
test
our
own
charts
it'll
test
enamel.
A
It
has
the
ability
to
install
charts
and
actually
test
that
they're
functioning
using
tests
that
can
be
packaged
in
charts.
It
will
do
linting
on
charts
to
make
sure
that
you're
in
Asians
are
consistent.
It
has
all
kinds
of
tools
and
we've
made
this
available,
so
lots
of
people
can
use
it.
It's
just
one
of
the
other
projects
we
have
and
companies
are
using
this
today
on
their
chart
repositories
and
of
course
we
talked
a
little
bit
today.
A
You
learned
about
helm
3,
but
if
you
are
using
helm
2
and
you
want
to
make
that
jump
from
helm,
2
to
helm,
3
we've
got
another
project
and
that
project
will
help
you
do
that
migration
in
helm
2.
You
would
have
heard
a
little
bit
about
tiller
that
thing
that
I
said
this
morning
is
gone
and
there
are
certain
ways
that
data
is
stored
and
we
want
to
make
it
easy
to
make
that
transition
and
we
have
helm
2
2
3
plug
in
that
will
do
that,
for
you.
A
It
helps
you
along
the
way
it
has
features
like
dry
run,
it'll
help.
You
run
two
systems
side
by
side.
If
you
want
test
things
out,
it
does
a
lot
of
those
things
and
then
we've
got
stuff
like
chart.
Releaser
now
chart
release
are
here
is
a
project
that
you
can
use
that
will
help
you
release
your
charts.
When
you
want
it,
you
can
tie
it
to
CI
and
it's
one
of
the
other
projects
we
have
to
help.
You
run
your
own
chart
repositories
alright,
but
we
go
beyond
helm
beyond
just
the
helm
project
right.
A
We
as
a
project,
have
it
but
there's
a
whole
ecosystem
around
helm
of
stuff
that
you
can
use,
and
here
I've
got
a
few
things.
There's
flux,
which
is
a
that
that's
symbol
in
the
top
left
is
flux.
It's
another
CN
CF
project,
it's
one
of
the
new
sandbox
ones.
It
works
with
home.
Next
to
it,
there's
drone
drone
is
a
CI
system
that
works
with
home.
Airship
is
something
from
the
OpenStack
project.
It
works
with
helm
and
helm
file
on
the
right
there.
A
Helm
file
is
a
configuration
management
system
that
can
sit
on
top
of
helm
and
help
you
deploy
and
manage
your
configuration,
there's
a
whole
ecosystem
of
tools
out
there
built
around
helm.
That
will
help
you
with
packages
managing
it
and
running
that
installation,
especially
as
you
want
to
get
more
complicated
with
what
you
do
now
helm
did
something
neat
in
home
3.
We
have
an
experiment.
A
In
fact,
we
added
a
new
system
to
helm
3
to
let
us
do
experiments
where
the
API
isn't
stable
yet,
but
some
of
this
is
because
we
can't
the
experiment
we
have
right.
Now
is
the
ability
to
use
OCI
repositories.
That's
where
you
push
and
pull
your
images
from
to
store
helm
charts.
This
is
a
new
thing
coming.
A
The
reason
it
is
an
experiment
is
we're
still
working
with
the
open
container
initiative,
the
OCI
to
to
work
out
the
spec
to
work
out
the
bumps,
because
storing
artifacts
other
than
container
images
is
new
for
them
and
we
wanted
to
release
home
3.
So
it's
an
experiment,
and
so
things
may
change
there
and
lots
of
places
don't
support
it
yet,
but
we're
working
with
the
OCI
so
that,
hopefully,
in
the
future,
it's
a
stable
feature
that
you
can
push
and
pull
your
charts
from
OCI
containers.
A
A
3
was
just
released
last
week
so
make
sure
that
you
check
that
it's
for
the
right
release
of
helm
as
you're
browsing
reading
articles
you're
on
stackoverflow
you're
in
places
like
that,
because
you'll
find
that
there
are
changes
in
differences
just
between
versions
so
be
careful
when
you're
out
there.
Looking
that
it's
for
the
right
version
of
helm,
so
here's
how
you
can
get
engaged.
We
love
engagement
with
home.
We
love
issues,
feedback
showing
up
at
meetings.
You
can
see.
We
have
slack
rooms,
we've
got
a
mailing
list,
Twitter
YouTube!
A
You
can
follow
us
on
social
media.
You
can
engage
with
the
project.
We
welcome
that.
The
image
up
there
is
of
our
meetings,
because
all
of
our
meetings
are
recorded
and
they're
put
up
on
YouTube
and
you'll
actually
see.
This
was
done
on
Halloween,
and
so
you
can
see
Josh
and
a
costume
up
there.
It's.
A
A
hot
dog,
alright,
the
next
thing
is:
we
have
a
booth
so
tonight
when
you're
doing
the
crawl
or
when
you're
here
many
of
the
scenes
you
have
projects
have
a
booth.
So
if
you've
got
questions
and
I
expect,
you
do
please
feel
free
to
come
by
our
booth
and
ask
there's
going
to
be
somebody
there.
Most
of
the
time
please
come
by
ask
questions
engage
with
us.
A
Tomorrow
we
have
the
helm
three
deep
dive,
and
so,
if
you
want
to
dig
deep
on
helm
three,
maybe
you
came
here
and
you're
like
this
session,
isn't
for
me
I'm
looking
for
more,
this
is
probably
the
session
for
you.
A
couple
of
the
helm,
maintainer
czar
gonna
go
deep
on
home,
three
to
explain,
what's
changed
and
what
it
means
to
you
and
if
you're
new
and
you
want
to
learn,
go
ahead
and
check
it
out.
A
A
Risk
encryption
for
secrets-
yes,
that's
probably
not
for
helm
secrets
in
kubernetes.
You
can
now
have
pluggable
backends,
and
so
you
can
plug
in
the
the
storage
mechanism
of
your
choice
to
deal
with
the
encryption.
That's
probably
the
more
appropriate
way,
because
it's
something
you
should
have
kubernetes
wise
and
it's
a
choice
for
you,
whether
you're
in
a
public
cloud
and
they
use
something
like
a
key
vault
or
if
you're
on
premise
and
you
fit
your
own
system
in
there.
D
So
I'm
just
wondering,
is
what
what's
the
support?
Is
there
like
a
for
the
stable
charts
or
the
charts
that
are
out
there
now
like?
How
do
we
tell
which
ones
are
helm
three,
or
is
there
a
separate
repo
for
the
github
or
something
for
the
helm,
three
charts
or
is
there
something
we
need
to
do
to
convert
them
or
most.
A
A
B
B
To
manage
secrets
with
helm
is
that
your
question
yeah
so
here
this
question
a
lot.
The
recommendation
I
usually
have
is
to
use
a
tool
called
Mozilla
stops
sop
s
and
you
can
store
secrets
encrypted
at
rest.
In
your
repository
and
your
installation
process
then
becomes
cube.
Ctl
like
or
decrypt
the
the
stops
encrypted
files
apply
them
as
raw
kubernetes
animal
and
then
in
your
helm,
chart
just
use
a
secret
ref.
So
you're
not
like
baking
secrets
into
your
into
your
helm,
values
like
the
actual
key
values
for
the
secrets.
E
B
B
A
So
I
think
some
of
this
is
going
to
depend
on
your
I.
Don't
know
if
you've
any
policies
or
processes
a
some
people
want
to
store
their
secrets
out-of-band
from
their
applications.
Others
might
use
a
tool
like
shoppes,
where
you're
going
to
when
you
have
the
chart
at
rest,
maybe
you're,
storing
it
and
get
or
something
like
that.
It
will
encrypt
all
of
the
metadata
on
it.
So
you
can't
actually
read
the
values
of
the
secrets.
A
There
are
a
number
of
strategies
and
plugins
on
how
you
can
do
this,
and
a
lot
of
it
depends
on
the
kind
of
process
you
or
your
organization
wants,
because
there's
a
few
different
ways
to
do
it.
If
you
want
to
chat
afterwards,
we
can
probably
talk
about
it,
or
this
is
something
you
can
do
in
a
search
engine.
Thank.
E
D
I
was
just
gonna,
add
two
as
an
answer
to
his
question.
What
one
of
the
things
we're
looking
at
its
external
secrets
by
GoDaddy
introduces
a
new
kubernetes
object,
and
so
on
that
you
can
use
a
config
map
to
infer
your
secrets,
then
use
a
different
system
to
fetch
those
secrets.
So
you're
not
putting
anything
in
the
helm.
Chart
that's
going
to
reveal
your
secrets,
and
so
we
use
AWS
for
example.
D
F
A
I
think
it's
going
to
depend
on
your
style
and
how
you
want
to
do
things
in
your
internal
processes.
If
you
have
things
like
you're
gonna
pass
it
around
where
lots
of
people
aren't
going
to
know,
but
they
need
to
install
it
easily.
Helm
is
great
for
that.
If
everybody's
gonna
know
all
of
the
intricate
gamal
and
it's
pretty
much
gonna
be
the
same,
it
may
not
be
such
a
thing
right.
A
In
Grenda,
you're
gonna
have
a
lot
of
intricate
yamo
I
mean
I
generated
the
yamo
for
deploying
WordPress
with
the
database
at
one
point,
just
the
simple
one
but
production-ready
it
was
like
13
manifests
now
if
I
knew
those
manifests
and
I
could
go
hack
on
them,
and
it
was
always
going
to
deploy
the
same
things
yeah.
If
I
was
going
to
have
differences
in
different
environments,
if
other
people
were
gonna,
go
install,
those
and
I
didn't
want
them
to
have
to
deal
with
the
complexities
of
installing
them
and
modifying
them.
A
Then
I
find
it
to
be
really
useful
to
package
it
up
as
a
Helmut
art,
because
that
makes
it
easy
to
install
but
I
think
some
of
this
is
going
to
depend
on
your
taste
and
your
workflow
and
what
you're
trying
to
accomplish
and
who
the
people
are
involved
in
this
you've
got
to
deal
with
it
and,
in
my
case,
I,
like
simplicity,
I
want
to
simplify
their
lives
and
what's
the
best
way
to
do
that,
and
that
depends
I
don't
know.
Josh
do.
B
Yeah
I
mean
I.
One
thing
I'd
like
to
point
out
a
lot.
Is
that
there's
not
really
any
one
way
to
use
helm?
It
just
depends
on
your.
Your
workflow
in
your
use
case,
like
I,
was
recently
working
on
a
project
and
there
was
nine
micro
services
as
part
of
this
big
deployment
and
so
I
started
out.
I
was
building
a
helmet
for
each
of
these
services
at
about
number
two
or
three
I
realized
like
they
were
all
pretty
much
the
same
HTTP
listening
service.
B
So
instead
talk
with
the
team
and
decided,
let's
build
a
single
helm
chart
that
works
exactly
the
same
for
every
service,
and
so
we
use
a
single
helm
chart
to
deploy
nine
different
services,
doing
a
different
configuration
for
each
just
setting
a
new
docker
tag
for
that
specific
for
that
specific
service.
But
that's
that's
usually
not
the
way
you
see
how
charts,
but
just
trying
to
get
at
that.
There's
several
different
ways
and.