►
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
From Zero to Production with Kubernetes - Daniel Lopez Ridruejo & Gus Lees, Bitnami
This session provides a practical intro to deploying Kubernetes applications into production. Kubernetes is becoming the standard platform for deploying applications in the enterprise. The Helm project provides tooling for packaging and an extensive catalog of ready to run applications. However, deploying applications in production requires much more: monitoring, logging, certificate management and so on. The community has adopted a de-facto ‘production stack’ that includes popular projects such as Prometheus, Vault, ElasticSearch, Fluentd and others. Unfortunately, putting together this basic plumbing is not standardized, and results in a lot of wasted, duplicated efforts. This session will introduce the Kubernetes production runtime, an open source stack that can simplify and automate production deployments, and explain how it relates to Knative and Gitlab's Autodevops.
To learn more: https://sched.co/GrYb
A
I
want
to
thank
you,
I
love
you
for
for
coming,
there's
a
lot
of
very
cool
sessions,
these
Keep
Calm.
So
we
appreciate
you
joining
us
today
to
talk
about
moving
applications
to
production
in
communities
and
what
gas
and
I
are
gonna
be
covering
today
is
about
how
to
bridge
that
gap
from
when
you
just
instantiate
a
cluster,
and
you
need
to
stall
all
the
different
pieces
and
to
have
a
production
environment
like
monitoring
and
login
certificate
management,
and
so
on.
So
a
little
bit
about
us.
A
We
both
work
for
for
Vienna
me
and
have
a
background
in
in
open
source.
Then
we
have
been
involved
in
in
different
projects
for
over
20
years.
Each
in
the
case
of
gas,
OpenStack
and
and
Debian
in
my
case
is,
is
is
apache
as
well
as
working
a
number
of
commercial,
open
source
companies
or
or
gas
in
his
case,
he's
paying
over
10
years
at
Google.
So
a
lot
of
what
we're
gonna
be
presenting
today
is,
is
the
evolution
or
is
informed,
but
a
lot
of
for
our
experiences.
A
The
other
part
of
our
experience
that
that
has
conducted
asked
to
work
on
the
Vietnamese
communities
production
run
time,
which
is
the
open
source
project
that
we're
gonna
be
presenting,
is
the
experience
at
benami
of
packaging,
a
bunch
of
other
people's
software.
We
are
one
of
the
main
contributors
to
to
the
home
project
and
in
particular
to
to
helm
charts.
We
also
cooperated
with
happy,
oh
and
in
case
Anand,
and
with
Microsoft
on
the
on
the
recently
announced
Sina
projects.
We
have
a
lot
of
experience,
packaging
third-party
software,
so
people
can
can
consume
it.
A
We
have
a
lot
of
experience
in
producing
our
own
open
source
software
that
runs
on
top
of
communities
that
ranges
from
small
command-line
utilities
to
to
more,
like
you
know,
web
oriented
or
more
news
oriented
type
of
applications
that
you
need
to
deploy
on
on
communities
and
as
well
as
we
ourselves
inside
bitNami.
We
are.
A
We
are
users
and
operators
of
kubernetes
clusters
and
at
least
four
different
clouds
for
multiple
years,
and
you
know
multiple
clustering
English
one
of
these
clouds,
so
we
have
gone
through
through
a
lot
of
a
lot
of,
and
one
of
those
is
a
nice.
A
recurring
theme
is
when
you're
packaging
softwar
for
other
people
to
consume
you're.
A
So
you
see
there
is
we
start
seeing
a
lot
of
patterns
that
repeat
themselves
and
that's
not
good,
there's
something
that
we
wanted
to
fix,
because
it's
a
pain
point
for
us
and
it's
a
pain
point
for
for
a
lot
of
other
people,
and
this
is
the
gap
that
was
talking
before.
Is
you
know
once
you
install
or
you
get
like
instantiate
a
cluster
in
DKA
or
or
aks
or
case,
or
any
of
the
kubernetes
service
providers?
Is
you
need
to
build?
On
top
of
that?
You
can
there's.
A
A
lot
of
things
that
need
to
be
in
place
is
typically
not
clear
for
somebody
who
is
just
new
to
kubernetes
what
is
the
best
monitoring
solution
where
people
are
using
for
for
login?
So
it's
you
know
you
need
to
do
a
little
bit
of
googling
literal.
You
know
asking
or
you
know,
seeing
what
people
are
using
and
then
you
end
up
implementing
all
this.
It
tends
to
be
like
very
manual.
A
It
tends
to
be
different
from
for
for
different
people,
so
you
have
all
these
variations
or
how
you
know
the
basics
and
components
are
installed
in
within
the
same
organization.
Sometimes
you
know,
or
definitely
within
different
different
teams
and
different
companies.
So
what
we
have
come
up
with
is-
and
this
is
solution
that
we
want
to
present
today
and
we'll
also
mention
at
the
end,
how
other
people
or
their
open
source
projects
are
solving.
This
is
what
we
call
the
community
product
run
time,
which
is
a
standard
set
of
manifest.
A
That
gives
you
kind
of
the
basic
building
blocks
that
everybody
aspects
like
monitoring
and
in
English
controller
certificate
manager,
all
the
very
basic
pieces
and
nothing
else,
and
we
do
it
in
a
way
that
it
just
works
out
of
the
box.
Like
you,
if
you
don't
know,
you
know
what
you
should
be
installing
you
just
exit
it
one
command,
and
we
give
you
all
these
different
pieces
working
together.
A
But-
and
this
is
a
very,
very
important,
but
for
us-
and
this
is
where
gas
being
here
a
control
freak
helps
out-
is
we
make
it
super
super
easy
for
the
administrators
to
customize
that
and
specify
that
Delta
to
to
to
make
it
work
exactly
as
as
they
want.
So
it's
kind
of
batteries
included,
but
we
make
it
super
easy,
and
this
was
one
of
the
or
you
know,
design
goals
for
people
to
customize
it.
We
don't
have
a
lot
of
time.
A
I
wanted
to
give
a
live
demo,
but
the
you
know,
the
closest
that
we
can
do
in
the
time
that
we
have
is
I'd
recorded
a
video
in
which
it
shows
in
a
minute
what
this
looks
like
and
what
you're
gonna
see
in
the
video
is
how
how
we
don't
see
in
the
video
is
how
we
can
deploy
a
home
charm.
This
is
cubes.
Is
it's
another
open-source
application,
the
momentum,
but
you
can
deploy
the
same
home
chart
using
the
command
line
and
the
result
is
gonna,
be
the
same.
A
One
we're
gonna
be
deploying
WordPress
using
a
ham
chart
will
give
it
a
name
and
then
specify
in
the
options
that
we
want
the
domain
name
that
we
want
for
them
for
the
WordPress
blog
as
well
as
we
want
an
English
controller.
We
want
SSL
certificate
and-
and
we
want
monitoring-
and
just
with
that,
once
the
word
press
charges
deployed
into
the
cluster
is
gonna,
be
automatically
you
know,
monitors
all
application
is
running.
You
can
see
that
it
has
a
valid
certificate
through
lesson
script
automatically.
Without
having
to
do
anything,
the
application
is
being
monitored.
A
So
you
can
go
to
to
Prometheus
and
start
taking.
A
look
at
at
the
metrics
in
this
case
is,
is
you
know,
access
access,
access
lock
for
the
for
the
sorry,
the
visas
to
the
to
the
website,
as
well
as
with
cabana?
You
can
take
a
look
at
the
access
log
for
four
httpd
or
any
other.
You
know
queries
that
you
may
want
to
do
like
taking
a
look
at
the
different
containers
that
make
the
deployment.
A
So
this
is
an
overview
of
what
you
can
do
once
this
is
things
in
place
and
is
installing
the
cluster
you
can
launch
an
application
and
with
basically
little
or
no
modification,
the
application
is
going
to
be
monitored
and
secure,
and
you
know
certificate
management
and
so
on
now
I'm
gonna.
Let
gas
tell
you
how
we
implemented
this
and
how
you
can
use
it.
B
Hey
good
morning,
everyone
so
I'm,
an
infrastructure
guy
I've
spent
my
whole
career
working
as
a
programmer
and
assisted
min,
and
it's
a
combination
of
both
those
things.
And
so,
when
we
set
up
our
cluster
internally,
we
wanted
to
provide
the
standard
services
to
our
other
developers,
and
so
we
went
on
this
journey
where
we,
you
know
the
same
thing,
we've
all
done.
B
B
B
This
is
all
the
stuff
you
would
set
up
yourself
by
hand
anyway,
unlike
some
of
the
upstream
examples,
we're
defaulting
to
production,
so
we
default
to
n
plus
one
availability
we
default
to
our
back,
enabled
we
default
to
persistent
volumes
all
the
things
you
would
want
for
production,
so
you
should
just
be
able
to
take
this
out
of
the
box
and
run
your
real
production
clusters
using
this
setup.
That's
that's
our
goal.
B
So
the
way
we
go
about
this,
we
we've
effectively
put
together
a
project
that
has
three
parts
its
there's,
the
the
de
facto
specification
of
what
an
application
can
expect
like
how
does
application
interact
with
this
and
that's
all
the
standard
ways.
If
you
want
to
your
Prometheus
to
scrape
your
pod,
you
put
the
standard,
scrape
equals
true
annotation,
the
same
thing.
You
do
normally
so
we've
kind
of
standardized
those
things
if
you
use
BK
PR.
B
If
you're
writing
an
application
at
targets,
peak
APR,
you
can
just
assume
that
it's
going
to
follow
that
that's
standard,
then
we
have
a
bunch
of
manifest
templates.
That's
the
main.
The
real
main
export
from
this
project
is
is
the
manifest
templates
that
describe
what's
going
to
be
installed
and
then
to
make
it
easier
to
consume.
We
have
a
simple
installer
tool
that
just
runs
all
that,
for
you
automatically
with
a
single
sort
of
command
line.
We
follow
the
kubernetes
support
models,
so
we
support
two
consecutive
releases
at
the
moment
with
the
1.0
release.
B
Dubuque
APR
we're
targeting
a
KS
and
GK
clusters,
but
obviously
eks
and
amazon
cops
are
hot
on
our
list.
For
the
next
couple,
and
then
after
that,
we'll
just
follow
demand,
whatever
people
want
to
see
we'll
try
to
support
all
Apache
to
licensed
all
available
on
that
github
project
and
yet
so
those
first
three
parts
of
the
install
of
the
manifests
and
the
de
facto
specification.
B
So
the
install
experience,
here's
the
simple
installer
I'm
going
to
come
back
later
on
and
show
you
the
more
flexible
version,
but
the
simple
experience
start
with
a
fresh
cluster,
whatever
whatever
you
get
off,
GK
or
aks,
or
the
future
supporter
platform
start
with
the
empty
cluster
and
you
download
the
tool
you
run
the
tool
with
some
command
line
flags
and
then
magic,
it
configures
danis
for
you.
It
sets
up
the
service
accounts
for
you
and
you
can
just
use
it.
So
you
can
get!
B
You
can
go
to
the
various
consoles,
cabana
Prometheus
alert
manager,
these
sorts
of
ones
they're
all
provided
over
TLS.
We
configure
the
DNS
names
for
you,
they're
all
protected,
so
you
have
to
sign
in
with
our
credentials
on
whatever
your
relevant
cloud
is
so
aks
OAuth
or
GK
and
Google
Earth
and
it
works,
and
when
you
want
to
do
an
upgrade,
you
come
back.
You
download
the
new
version
of
the
tool
and
run
it
again
on
the
same
cluster
and
it'll.
B
Take
you
to
a
new
version
of
manifest
and
do
a
clean
upgrade
so
just
expanding
on
those.
What
really
happens?
The
Installer
tool
goes
and
does
a
lot
of
non-cuban
ad
setup.
So
it
sets
up
service
account,
set
the
service
principles,
it
generates
passwords
and
then
it
stores
the
result
of
that
in
a
little
JSON
file
and
then
that
JSON
file
are
basically
your
cluster
local
parameters.
So
we
get
those
clusters
specific
parameters
and
we
combine
them
with
the
standard
manifests
and
then
that
generates
the
actual
gamal
objects
for
your
cluster.
B
And
then,
when
you
come
back
into
an
upgrade,
the
upgrade
Riis
is
that
same
JSON
file
and
just
sort
of
like
terraform
it'll
fill
in
any
blanks.
So
if
you've
got
if
the
new
version
needs
some
extra
cloud
resources
that
aren't
in
that
JSON
filed
and
it'll
go
and
create
just
those
missing
pieces,
so
very
similar
to
sort
of
terraform
approach.
Now
is
that
same
thing,
your
diagrams
we've
got
that
JSON
coming
in
and
then
the
templates
coming
in
they
go
through
an
engine
called
J
sonnet,
which
I'll
talk
a
little
bit
about
later.
B
J
sonnet
is
a
small
language.
It's
really
good
for
manipulating
JSON.
It's
really
good
for
getting
JSON
in
and
manipulating
it,
and
then
speeding
out
regular
JSON,
now
I'm
guessing
as
infrastructure
guy
I'm.
Guessing
all
of
you
as
soon
as
I
said
a
magic
installer.
You
all
just
went
yeah
I'm
done
I'm,
not
using
this
tool.
B
So
let
me
tell
you
how
I
expect
you
to
actually
use
this
in
production
that,
in
the
Cupra
installer
is
good
for
demos.
It's
good
for
quick
trying
this
out
in
your
cluster,
but
none
of
us
obviously
are
going
to
run
a
black
box
and
trust
our
infrastructure
too
little
black
box
taller.
So
let
me
come
back
to
this
diagram
again.
B
As
I
said,
the
Installer
does
some
non-cuban
native
stuff
and
it
writes
a
JSON
file
with
the
results.
If
you
have
a
if
you
have
a
a
complicated
setup,
if
you
have
multiple
clusters,
if
you
use
a
stranger
or
setup,
if
you
like
us,
we
we
have
various
internal
projects
and
approvals
and
things
so
there's
no
one
really
who
can
just
run
a
simple
installer.
So
you
can
do
all
those
steps
by
hand.
B
There's
not
a
lot
of
steps
there
just
have
to
create
the
service
accounts
and
get
the
various
IDs
and
store
them
in
distress
and
faults.
So
you
can
create
that
JSON
file.
However,
you
want
you
can
use
terraform,
you
can
use
internal
processes,
you
can
do
it
manually,
whatever
you
need
to
do
for
your
particular
situation,
then
separately
he
get
the
templates.
B
You
can
use
a
tool
cube
C
of
G,
which
is
from
the
caisson
it
project
which
can
consume
those
JSON
templates
that
are
exported
by
our
project
and
smoosh
them
together
with
those
with
JSON
parameters,
and
you
get
all
the
amyl
objects,
and
you
can
look
at
that.
Yeah
Molly,
you
can
differ
versus
your
live
cluster.
You
can
have
a
look
at
what's
going
to
happen
it
you
can
see
exactly
what's
going
to
be
created
and
then
you
can
either
push
that
directly
in
cube
CFG
or
you
can
store
those
that
yam
all
on
disk.
B
You
do
whatever
you
to
do
push
it
through
your
existing
workflow
using
keeps
you
till
apply,
or
you
know,
various
other
get-ups
tools,
whatever
you
currently
use
to
do
that
you
can
do
that
with
this
yeah.
This
is
how
expected
to
actually
be
used
in
production
right.
You
don't
trust
some
little
install
at
all.
You
want
to
see
the
pieces,
you
want
control,
your
invisibility
and
and
you
you
need
to
be
in
power
for
the
whole
process,
so
the
advanced
version
do
up
a
hand
this
little
JSON.
B
Its
overlay
is
really
what
merges
the
the
parameters
and
the
templates
together,
and
it
looks
like
that
little
chunk
down
the
bottom
there,
so
you
can
see
it's
importing
the
templates
with
the
version
and
the
platform
and
in
skating
the
local
parameters
JSON,
and
that
provides
arguments
effectively
to
those
templates
that
then
get
expanded
by
JSON
it.
What
your
sonnet
allows
you
to
do
is
put
any
other
override.
You
like,
in
that
same
little
file
lists
using
the
JSON
at
language.
You
can
see
you
can
change
anything
that
gets
installed.
B
You
can
say
I,
don't
want
elasticsearch,
you
can
say.
I
do
also
want
sto,
and
you
can
configure
these
extra
pieces
in
there.
You
can
remove
those
other
pieces,
you
can
modify
any
parameter
anywhere
and
then
the
result
can
get
generated
into
regular
objects
and
then
gets
pushed
into
cluster.
You
have
full
control
over
what
gets
installed
and
in.
If
you
really
want
to
you
can,
if
you
were
infrastructure
team
for
a
larger
company,
you
could
take
this.
B
You
could
provide
some
some
overrides,
an
overlay
for
your
particular
company
and
your
particular
conventions,
and
then
you
could
re
export
that
you
could
say:
hey
application
teams.
You
can
now
take
the
result
of
this
and
you
can
then
apply
it
again
in
your
own
clusters,
so
you
can
have
chains
of
overlays
effectively
and
customize
this
again
and
again
and
republish
again
and
again,
depending
on
how
it
fits
your
organizational
structure.
B
So
this
is
a
slightly
odd
idea
where
we're
trying
to
do
standardization,
but
we
also
want
to
provide
flexibility
and
and
control.
So
the
general
realization,
I
guess
was
that
in
practice
in
theory
this
is
this
is
unsolvable.
In
theory,
this
is
ridiculously
complex,
but
in
practice
it's
not
in
practice.
People
do
very
simple
over
eyes
like,
like
me:
I
have
a
cluster
at
home.
That
is
mixed
architecture.
I've
got
a
couple
of
raspberry
PI's
I've
got
a
couple
of
old
AMD
laptops.
B
You
mixed
architecture,
I
can't
take
any
standard,
helm,
chart
and
use
it
because
I
need
to
add
a
node
architecture
restriction
that
says
only
run
this
on
an
D.
64
only
run
this
on
on
arm
and
no
one
ever
thinks
of
providing
that
as
a
parameter,
but
with
JSON
it
I
can
override
anything.
I
can
provide
that
one
one
parameter
and
I've
customized
it
in
a
way
that
the
original
author
didn't
intend,
but
it's
a
very
simple
customization,
it's
very
simple
for
me
to
carry
that
customization
forward
over
future
versions.
B
So
in
theory
this
is
a
big,
complex
problem
in
practice.
It's
really
simple.
So
our
kind
of
social
contract
here
is
we're
gonna
test.
The
base
manifests
together
we're
gonna
test
them
on
the
rest
platforms,
we're
going
to
say
these
all
work
together
and
we
verified
that
and
if
you
do
want
to
do
any
overrides,
that's
like--that's
on
you
to
test
those
overrides,
but
in
practice
that's
actually
really
simple.
In
most
in
most
practical
cases,
and
then
you
can
carry
that
that
that
overlay,
you
can
carry
that
forward
across
future
versions.
B
Quite
quite
simply,
and
then
all
you
have
to
do
is
very
expandable
and
make
sure
it
does
what
you
want
and
re
you
know,
verify
the
result
and
then,
when
you
happy
push
it
to
your
cluster,
here's
an
example.
This
is
JSON
at
syntax
looks
a
little
bit
like
JSON.
It's
got
some
extra
pluses.
The
plus
means
merge
it
in
with
the
the
parent
object,
merge
it
in
with
the
template.
B
So
this
is
my
contrived
example
here:
I'm
getting
the
nginx
ingress
controller
and
I'm
multiplying
the
grace
period
by
two
for
some
reason,
contrived
reason
just
to
show
that
I
can
do
mathematics
in
here
and
I
can
refer
to
the
parent
object
if
I
want
to
and
lots
of
other
things,
there's
a
whole.
It's
a
whole
mini
language.
You
can
look
at
the
JSON
at
org
website
to
see
some
pretty
good
examples.
B
It
works
very
well
with
kubernetes,
because
there's
a
lot
of
duplication
in
kuberdas
objects
and
and
this
language
you
find
really
really
powerful
in
working
with
that
and
it's
the
basis
for
the
case
on
it
to
land
the
cube
CFG
tool
they're
both
chosen
to
build
around
the
JSON
a
language
alright.
So
obviously
this
is
just
one
part
of
a
big
workflow
and
we
deliberately
don't
want
to
be
those
other
parts
of
the
workflow
we
just
want
to
handle
the
the
resource
template
part
that
standard
infrastructure
at
the
end
of
the
day
it
just
produces.
B
Regular
kubernetes
manifests
kubernetes
yell.
So
you
can
use
this
with
any
other
tool
in
practice.
You
would
want
to
wrap
it
in
other
tools.
You
want
to
have
some
sort
of
Jenkins
or
gates
or
whatever,
to
store
the
the
configs
that
you've
got
version
them
and
then
eventually
to
push
them
to
your
cluster
for
application
management.
This
works
with
all
the
standard
applications
you
can
use
helm.
You
can
use
case
on
it.
You
can
use
cube
apps.
You
can
use
all
these
other
tools,
they
all
work,
just
fine
with
it.
I
felt
bad.
B
Putting
this
slide
together,
because
I've
only
got
a
few
logos
on
here,
but
this
is
like
everything.
Everything
in
ecosystem
should
be
on
this
slide
here,
because
this
is
just
what
you
would
do
yourself
normally
anyway,
and
so
it
works
with
all
the
other
tools,
just
as
it
normally
would.
If
you
had
built
it
yourself
and
one
that
deserves
a
special
dedicated
slide
is
Kay
native
Kay
native
has
a
lot
of
overlap
with
this
project.
The
green
bits
there,
that's
really
because
everyone
has
these.
B
Everyone
wants
these
same
stacks
and
Kay
native,
just
like
everyone
else
wants
to
have
the
same
stacks,
so
they
they
duplicate
some
of
the
pieces.
We
have
you
could
install
Kennedy,
Van
Beek
APR
into
the
same
cluster
it'll
work.
Fine,
it's
a
little
bit
of
a
waste
because
you'll
be
running
to
Prometheus's
you'll,
be
running
two
elastic
searches,
it'll
work,
fine,
just
wasting
some
resources
so
right
now,
if
you
were
to
do
that,
you
would
want
to
provide
an
overlay
that
turned
off
those
duplicates.
B
No,
it
would
be
okay
if
I,
if
I
had
to
summarize
the
two
projects,
people
ask
me:
what's
the
difference.
So
if
I,
just
summarize
our
I'd
say,
K
native
is
quite
opinionated,
it's
quite
advanced.
It's
bringing
not
just
these,
not
just
monitoring
and
logging,
but
it's
also
recommending
a
way
to
develop
your
applications,
a
way
to
roll
out
your
applications
a
way
to
build
them.
It's
it's
a
lot
more
bolder
in
scope.
B
B
Just
just
the
boring
bits,
so
if
I,
if
I,
had
a
magic
wand,
we
could
merge
these
two
projects
somehow
by
getting
K
native
to
reuse,
the
BK
PR
templates
for
that
common
part,
and
we
might
actually
explore
that
in
the
future,
and
so
then
K
native
would
be
BK
PR,
plus
an
overlay
that
added
the
K
native
special
magic.
That
would
be
a
reasonable
way
to
to
unify
this
to
in
the
future,
so
yeah,
so
K
native
is
much
more
advanced.
Much
more
bolder.
B
If
you
like,
alright,
so
that's
the
end
we
we
have
really.
What
we're
trying
to
do
is
just
provide
basic
infrastructure
for
everyone
that
we
can
all
agree
on
as
a
community,
and
we
can
all
reuse
as
sis
admins.
So
we
don't
have
to
relearn
how
to
do
this
stuff.
We
don't
have
to
learn
what
to
install.
We
have
to
learn
how
to
install
it
as
application
providers.
We
no
longer
have
to
talk
about.
Maybe
you're
not
running
prometheus,
maybe
you're
haven't
got
TLS
enabled.
B
We
could
just
assume
that
there's
always
going
to
be
an
ingress,
there's
always
going
to
be
TLS.
We
can
always
have
security.
We
can
assume
that
our
back
is
always
there,
so
we're
just
trying
to
provide
a
more
a
slightly
higher
level
standard
for
kubernetes.
So,
instead
of
just
deployments
and
daemon
sets,
we
can
say
deployments
and
daemon
sets
and
TLS
and
Prometheus,
and
then
we
can
all
just
kind
of
raise
the
water
level
just
that
slightly
higher
and
start
thinking
about
more
interesting
problems.
Instead
of
how
do
I
deploy
a
LeWitt
manager?
B
This?
Is
it
so
we'll
be
around?
If
you
learn
any
questions
we
rounded
the
bitNami
stands
takes
a
couple
of
days.
You
can
go
to
ku
Prada,
oh,
you
can
download
it.
Try
it
out.
We
don't
have
a
dedicated
slack
channel
right
now,
but
the
hash
key
tops
one
on
the
convenience
like
channel
would
be
a
reasonably
on
topic
channel.
If
you
wanted
to
discuss
any
of
this
and
yeah
any
questions,
you
could
probably
ask
you
look
a
little
bit
of
time.
A
Yeah
and
then
this
hair,
one
of
main
trains.
We
are
at
working
right
now
and
EK
a
support
and
a
field
of
yeah
form.
So
we
just
released
a
project
right
now
we're
looking
for
a
lot
of
feedback,
so
maybe
just
want
to
try
it
out
and
give
us
feedback.
That's
really
welcome,
even
even
more
work
on
our
contributions
and
pull
requests
and.
B
Into
part
two,
some
of
the
other,
if
you
want
to
use
it,
if
you
want
easier
to
different
platforms,
then
let
us
know
and
that'll
help
steer
our
priorities.
More
importantly,
if
you
don't
wanna
use
this
for
some
reason
to
also
come
and
talk
to
us,
and
let
us
know
what
that
reluctance
is
about,
and
an
infrastructure
guy
I
totally
understand
not
willing
to
give
up
control,
which
is
why
that
was
such
a
big
part
of
this.