►
From YouTube: Exploring Our User-Base Challenges: Getting Insight From StackOverflow - Javier Salmeron, Bitnami
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
Exploring Our User-Base Challenges: Getting Insight From StackOverflow - Javier Salmeron, Bitnami
Kubernetes has become the de-facto container orchestration system that developers and system administrator use to go Cloud-Native. As more users adopt Kubernetes, supporting those users has become a critical need. Stackoverflow has been since the inception the place where to get support, even though other channels like Slack are also heavily used. It should not be surprising that the amount of questions in StackOverflow has increased by more than 800% in only two years. Getting to know our users, understand their pain points and and use cases is the best way to improve the ecosystem. In this talk we will show some deep analysis of Stackoverflow questions. We will show the major trends and describe the major pain points felt by our users. We will also discuss the most common use cases discovered and the most familiar tools that people use in combination with k8s.
To Learn More: https://sched.co/GrTW
A
But,
more
importantly,
this
is
a
call
for
action,
so
in
the
end
which
chlorine,
the
user
base
and
what
what
I
think
is
important
here
is
to
come
up
with
ideas
or
come
up
with
several
questions
on
how
we
could
improve
the
user
experience
in
collectors
in
general.
That's
what
I
think
it's
important!
That's
the
real
value
of
being
here
being
here
in
keep
count.
So
that's
my
aim.
So
that's
what
I
I
wanted
to
tell
you
today.
So
without
further
ado.
Let
me
just
do
a
brief
introduction
about
why
we're
doing
this.
Why
myself?
A
A
I,
don't
know
how
many
of
you
know
the
story
of
of
benami
the
history,
but
we've
been
doing
it
since
the
very
beginning
since
well,
we
had
installers
cloud
images,
multi
tier
deployments
and,
of
course,
they're
covering
this
landscape
is
no
exception.
So
we
would
like
to
improve
the
user
spirit
when
they
work
about
containers
without
health
charts,
with
our
applications
that
BKP
our
cubes
and,
of
course,
in
the
trainer's
that
we
provide.
So
with
this
idea,
we
want
to
improve
the
user
experience
in
this
area.
A
The
question
will
be
how
so
one
of
the
in
on
spirits,
one
of
the
things
that
has
worked
for
us
is
to
basically
listen
to
what
users
have
to
say.
Got
reservation:
try
to
find
out
what
their
needs
are,
what
they
struggle
with,
and
this
is
what
we're
going
to
do
in
the
case
of
ernet.
So
we
want
to
to
see
what
people
struggle
the
most.
We
want
to
see
how
people
use
coordinate
these,
what
applications
they
normally
use
and,
more
importantly,
what
we
can
do,
so
we
want
to
improve
or
use
experience.
A
We
know
that
going
to
user
debate
works.
So,
okay,
where
is
the?
Where
are
we
going
to
go
to
find
this
information?
As
a
first
step,
we
decided
to
go
to
stack
overflow,
I,
suppose
that
this
qi
site
requires
little
introduction,
I,
suppose
that
once
in
our
lives
at
least
stack
overflow
save
the
day
of
our
Teddy
development
right.
A
So
some
bits
of
beta
has
around
10
million
bits
per
day
every
days
around
7,000
questions
are
created
in
different
topics
and
what's
more
interesting,
we
had
a
wide
variety
of
profile,
so
we
have
since
fronting
developers
there
Bob's
she's,
admitting
students,
so
this
variety
of
or
digital
identity
of
users
is
another
value
to
I.
Think
we
have
embedded
information
that
we
can
start
here.
A
So
ok
now
we
know
what
we
want
to
do
where
we're
going
to
do
it.
So
what
exactly?
We
we've
done.
Basically,
we're
going
to
review
this
year.
2018
and
we've
gone
through
around
1,000
questions
that
were
created
when
this
year
order
by
views,
and
we
try
to
instruct
basically,
which
kind
of
isn't
their
user
had
why
they
should
happened
if
possible,
what
platform
they
were
using
on
which
application
they
were
trying
to
deploy.
We
try
to
attempt
it
to
strike
these
four
bits
of
data
circuit.
A
A
This
hybrid,
that
this
interest
and
currently
has
is
it's
also
being
transferred
to
stack
overflow
well
by
the
way
feel
free
to
participate
at
any
moment,
so
we're
here
to
try
to
find
out
what
we
can
do.
So,
although
in
this
case,
what
we
see
if
we
go
to
the
number
of
questions
created
every
year,
we
see
that
it's
doubling
every
year.
A
So
these
years
around
7,000
questions
were
created
among
kubernetes,
so
you
see
there's
a
tendency
of
going
double
and
double
and
double
which
is
remarkable,
and
this
means
that
we
have
more
use
cases
more
experience,
issues,
also
more
concept,
there's
a
lot
of
challenge
here,
but
also
to
it's
interesting
to
see
how
it
compares
to
other
orchestration
platforms.
So
the
nests.
What
I
checked
is
ok,
how
does
it
compare,
for
example,
OpenStack
Norma,
dr.
A
Swan
I,
also
added
Heroku,
and
if
we
see
the
number
of
questions
created,
we
see
that
all
of
them
combined
are
the
more
or
less
the
equivalent
of
the
question
that
we
have
our
kubernetes
even
more.
If
we
add
openshift
so
we
see,
we
will
confirm
that
indeed
color,
that
is
having
an
important
moment.
We
see
Hiroko
here
I
was
hesitant
to
a
lead
or
not.
That
has
a
fair
share.
If
we
see
also
the
tendency
of
question
being
created
every
year,
we
see
that
while
cornutus
is
growing,
this
is
a
percentage
of
Stack.
A
However,
if
we
put
in
two
more
and
we
go
a
bit
further,
how
does
it
compare
to
the
top
categories
in
Stack
Overflow?
So
is
it
going
how
how
important
is
it
becoming
compared
to
the
other?
So
if
we
go
to
the
top
cut
salaries
for
a
prominent
languages
like
JavaScript
or
no
GS,
our
in
Java,
do
you
think
I'll
put
into
the
air?
Do
you
think
the
kubernetes
is
at
the
same
level,
okay,
yeah,
just
to
put
on
a
sample
this
year
around
7,000
communities?
Questions
were
created.
A
How
do
you
think
compared
to,
for
example,
that
JavaScript
questions
created
in
the
stack
overflow
I'll
leave
some
seconds
well
in
a
month
the
manage
a
box
with
questions
is
triple
to
the
mono
communities
questions
in
a
year.
So,
as
we
can
see
here
in
this
graph,
this
small
line
there
at
the
bottom
is
the
tendency
that
kubernetes
is
having.
Of
course,
it
still
has
a
long
way
before
becoming
one
of
the
top
categories.
A
The
truth
is
that
not
yet
so
around
37%
of
the
questions
are
solved,
whereas
around
62%
are
still
not
source
of
there's
an
issue
here.
So
there
are,
we
have
used
as
cumulative
user.
They
want
to
have
the
stack
overflow
to
find
answers,
and
this
there's
a
a
big
number
of
them
that
don't
get
the
solution
to
her
answers.
So
once
again,
we
still
need
to
think
so
how
we
can
do
to
try
to
increase,
and
other
people
have
find
solutions.
A
A
A
In
this
sense,
the
application
developer,
according
to
the
organ
that
CN
CF,
can
design
build,
configure
and
dispose
cloud
native
applications
in
cover
Nettie's.
What
does
it
mean?
It
means
that
it
needs
to
know
the
basic
coverages
objects
like
deployment
services
staple
set
volumes.
It
needs
to
know
how
to
use
the
API.
It
needs
to
know
a
bit
about
poor
patters
because
it
needs
to
deploy
applications
in
kubernetes.
You
need
to
know
how
to
manage
the
application
lifecycle.
We
first
sample
it
upgrades
and
also
how
to
troubleshoot
to
some
basic
troubleshoot
or
applications.
A
Internet
is,
of
course,
it's
cloud
native
application.
They
need
to
know
how
to
develop
cognitive
applications,
containerize
applications,
so
you
can
see
I
think
most
of
us
will
fit
into
this
category.
In
this
sense,
when
we
approach
kubernetes
because
we
are
developers
want
to
deploy
applications
and
to
exploit
the
capabilities
of
cloud
native
applications.
A
Okay,
so
that's
one
other
profiles,
the
other
one
is
the
communities
administrator
people
deploy
applications
in
a
cluster
someone
or
some
entity
has
to
create
the
cluster
has
to
create
the
cluster
to
know
in
order
to
clear
the
clutter
needs
to
know
about
the
coverage
architecture,
the
components
it
has
in
this
to
know
our
how
to
install
controllers
like
ingress
in
external,
my
logging
monitoring
maintenance
of
clusters
about
authorization
Arabic.
These
are
more
so
needs
to
know.
A
This
question
may
be
more
a
bit
simpler
because
Stack
Overflow
is
mainly
a
developer
community
right,
so
it
makes
sense
that
most
of
the
questions
are
for
application
developers
write
it.
At
least
it's
like
that,
so
you
can
see
that
most
of
them
are
question
the
C
key
ad
time,
so
application
developer
questions
around
60%,
but
this
is
also
a
fair
share
of
questions
regarding
aministration
even
more
than
another
page
is
like,
for
example,
server
fault,
which
is
more
aim
to
see.
A
A
When
you
read
through
the
questions
more
or
less
didn't
want
what
I've
done
is
trying
to
classify
in
basically
four
groups.
This
is
what
I
found.
The
first
category
is
still
about
how,
when
it
comes
to
creating
deployments,
that
means
I.
Have
my
application.
I
want
to
create
a
part
safe
instead,
basically,
I
find
nations
with
that
or
I.
Don't
know
how
to
create
it
or
I,
don't
understand
properly.
A
How,
then,
if
I'm
between
container
and
a
pod
and
I
stay
full
set,
I
still
have
questions
about
how,
for
example,
use
configuration
like
environment
variables
how
to
perform
lightning
readiness
checks,
how
to
upgrade
my
application.
So
this
kind
of
questions
that
we
found
for
something
you
Saban's
operation,
every
cron
jobs
or
operators.
A
This
is
one
of
the
main
terrorists
in
this
sense,
I
also
selected.
Some
highlights
we
can
see,
for
example,
lots
of
issues
in
upgrading
stateful
sets.
We
see
one
that
I
find
quite
interesting,
and
maybe
some
of
you
had
spirits,
which
is
the
thought
dependency.
So,
if
I
have
my
application
and
I
have
a
back-end,
how
do
I
make
my
application
feel
for
the
backend
to
be
ready?
For
example,
we
see
a
lot
of
questions
regarding
that.
A
Maybe
it's
not
a
bad
question,
maybe
when
a
different
paradigm
I,
don't
know,
but
the
thing
is
that
we'll
see
questions
like
that
and
also
I
found
for
those
who
are
developing
their
custom
applications
and
in
a
news,
private
registry
I
see
a
lot
of
news
experience
issues
with
that
specific
part.
They
try
to
use
a
private
registry
and
something
doesn't
work.
I,
don't
know
if
you
have
the
chance
to
work
with
that
to
create
the
whole
image
secrets.
People
find
lots
of
problems
with
that.
A
A
A
So,
for
example,
so
we
seized
with
load,
balancer
I
wanna
externally
be,
and
it
doesn't
work
we
see
on
this.
One
is
I.
Think
is
quite
interesting
issues
with
ingress
and
not
maybe
using
English
and
on
its
own
it.
When
you
want
to
use
more
advanced
operations
with
ingress,
you
want
to
exploit
their
capability
for
a
sample
of
the
engine,
s
English
controller,
and
you
want
to
do
more
advanced
operations
and
English
on
its
own.
Doesn't
seem
to
give
what
they
need,
for
example,
word
for
writing
an
URL.
A
A
So
some
highlights,
for
example,
I
see
a
lot
of
cases
of
people
not
using
the
current
selector,
and
it
seems
something
in
principle,
simple,
simple,
but
people
find
issues
with
that,
so
wouldn't
be
a
way
to
maybe
make
that
process
easier,
maybe
some
validation
to
some
something
that
we
could
make.
We
could
make
them
to
avoid
these
kind
of
issues
also
regarding
HTTP
traffic,
using
certificates
or
the
kind
of
Resurrection.
You
see
lots
of
issues
with.
Regarding
that
and
in
the
case
of
ingress
I,
you
can
see
also
questions
regarding
non
HTTP
traffic.
A
They
want
to
use
an
English
willful,
for
example
a
database
and
didn't
know
how
to
deal
with
that
she'll
be
using
recently,
not
the
finished
others
some
spectators
regarding
ingress
and
maybe
we're
qualifying
or
improving.
So
it
provides
what
some
users
demand.
That's
open
to
debate,
ok,
make
it
bit
faster,
so
third,
category
persistence,
so
volume,
so
the
thing
of
adding
external
data
to
the
two
deployments
in
the
form
of
volumes,
but
also
config
map
or
secrets.
So,
for
example,
they
create
modules
and
it
doesn't
work
especially
well,
for
example,
in
cloud
providers.
A
You
see
issues
when
mounting
the
volumes,
especially
because
maybe
they
don't
know
how
to
properly
mountain.
If
it's
a
vine
should
it
and
it's
mounted
like
a
directory,
or
especially
in
the
case
of
Secrets,
on
concrete
maps.
This
mountain
theme
doesn't
seem
to
be
fully
understood
permission
issues,
especially
with
the
case
of
non
food
containers.
Suddenly
they
they
launch
the
application,
they
say
permission
denied,
and
this
is
really
unique
in
each
case
of
off,
depending
on
the
cloud,
for
example,
also
regarding
persistent
body
and
persistent
boiling
claims
433
between
them.
A
How
do
I
do
sit
in
stateful
said
because
it
has
like
a
template.
For
example,
you
see
also
concept
issues
regarding
that.
This
is
also
something
that
is
previous
myself
and
for
something
trainings.
This
is
not
that
easy
display.
This
kind
of
concepts
also
how
to
debug
this
kind
of
mounting
errors
were
to
look
for
this
information.
It's
not
that
clear,
but
where
I
can
find
the
cause
of
these
issues.
A
Okay,
some
highlights,
for
example,
I
sent
debugging
mounting
errors,
and
this
one
regarding,
for
example,
I
see
quite
common
conflict
map
and
secrets
when
I
update
them.
What
happens
so
should
I
restart
my
deployments,
should
I
remove
it
and
recreated
it's
not
clear,
and
so
how
to
deal
with
that
with
this
kind
of
entities
that
sometimes
need
to
be
updated?
A
What's
the
correct
procedure
to
work
with
that,
and
finally,
the
development
tools
with
application
developers
and
want
to
create
cloud
native
applications,
and
normally
you
will
want
to
use
for
sample
people
or
containers
we
could
practices
which
are
those
cool
practices.
How
to
develop
a
culture
was
the
proper
way
to
create
it.
I
have
my
milk
micro
service
application
with
several
elements.
How
should
these
we
translate
into
health
chart
and,
most
importantly,
have
a
CI
CD,
Jenkins
kiddo,
for
example?
A
A
It
seems
to
me,
when
I
look
into
the
cases
that
maybe
it
would
be
interested
to
have
some
kind
of
tool,
some
kind
of
even
visual
tool,
to
help
us
to
create
the
whole
process
from
deployment
to
persist
into
communication
that
validates
that
what
we
created
actually
has
the
effect
that
we
want,
especially
when
you
have
health
chat.
It
will
be
quite
interesting
to
have
this
kind
of
tool
the
helps
to
avoid
these
kind
of
errors.
A
Also,
this
idea
development
tools,
maybe
if
they
have
some
kind
of
of
solution
that
goes
through
all
of
the
process,
but
make
because
maybe
several
developers
don't
want
to
create
the
channel
of
the
whole
chart,
maybe
the
sir
easier
wait
to
create.
Maybe
there
are
patterns
around
that
now
that
I
hear
a
cube,
column
and
I
go
to
the
booth.
A
I
start,
seeing
some
examples
of
tools
that
go
so
that
way
now
I
like
now
that
I'm
dealing
with
all
those
pie,
charts
now
I
wrote
both
and
I
start
thinking
about,
ok,
which
which
percentage
is
this
attacking?
So
you
see
it
ceased
your
help
in
this
the
communication
part
of
this
other
solution,
helping,
for
example,
development
tools.
A
So
how
can
we
help
them
another
way
to
help
them
I?
Think
it's
sorry
is
to
know
which
kind
of
applications
they
commonly
deploy
once
again
stack
overflow
application
developer
community.
So
it
makes
sense
that,
for
example,
one
group
is
custom
applications,
but
there's
also
infrastructure
applications
and
web
applications.
A
This
again
confirms
the
fame
that
there
must
be
a
way
to
make
this
process
easier,
but
I
want
to
go
a
bit
also
for
the
infrastructure
world
so
which
are
the
the
most
popular
in
stack.
Overflow
I.
Think
most
of
us
have
deal
dealt
with.
One
of
those
I
think
that's
leadership.
Press
in
this
case
my
sequel,
mongrel
for
elasticsearch
Revis
Kafka
in
the
end,
is
our
infrastructure
apps
that
help
therapy,
not
only
they
have
an
application
and
they
want
to
use
some
back-end.
A
So
the
common
package
can
be
seen
here,
so
my
stable,
my
really
phosphorus
and
I
think
it's
important
here
to
say
that
we
have
currently
in
stable
health
charts
that
for
most
of
these
applications
right,
so
it's
important
to
make
sure
and
for
example,
how
community
that
org
charts
have
covered
most
of
the
case
that
we
won't
cover
different
typologies.
Have
the
production
requires
that,
but
any
application
will
require
on
the
case
of
helm,
it's
important
indication
happy
to
save
a
sample
in
Minami.
We
work
with.
A
A
Well,
the
top
sister
I
found
Jenkins
the
law
impartial
WordPress
kid,
but
in
this
case
we
see
clear
that
there's
a
huge
percentage
of
people
using,
for
example,
Jenkins.
The
one
seconds
confirms
the
necessity
of
having
a
pipeline
for
developing
applications,
most
people
they
deploy
Jenkins,
but
then
they
want
to
integrate.
The
pilot
doesn't
work.
So
you
can
see
here
the
necessity
clearly
also
we
have
WordPress,
but
we
can
still
stupid
love
and
sooner
queue
artifactory.
They
also
aim
for,
for
example,
developers.
A
What
can
we
do
to
improve
this?
We
have
help
junkies
helped.
We
had
to
take
this
culture
at
the
moment.
Can
are
these
easily?
Do
they
easily
integrate
with
a
kubernetes
plug
for
a
rally
running?
It's
easy
to
configure
others
once
again.
Isis
is
the
whole
solution
that
covers
all
these
pipeline
process.
A
A
First
one
it
has
to
do
with
cluster
installation
so
either
from
scratch.
Reducing
chip
IBM
cops,
some
of
us
which
may
sound
familiar
even
cloud
Menace
applications,
and
you
see
issues
for
example,
component,
like
he
fled
of
cute
controller
manager.
They
they
have
the
cluster
they
want
to
install
and
these
components
were
correct.
A
They
find
it
super
the
container
granting
container
Dior
docker,
especially
with
networking,
so
OneNote,
doesn't
contact
the
other
I
want
to
deploy
container
networking
and
something
goes
wrong
also
issues
with
DNS,
because
we're
in
a
cluster
I,
don't
know
if
any
of
you
have
done.
It
is
a
complex
process
making
on
this
component
work.
Sometimes
it's
not
that
easy.
A
So
you
can
see
several
questions
or
not
in
these
elements,
for
example,
container
networking
you
see
several
issues,
for
example
with
mini
queue,
starting
which
I
think
if
mini
cave
is
one
of
the
platform
that
we
want
to
use
to
onboard
people
in
Korea
at
ease.
This
is
surprised
me
that
there
are
so
many
cases
that
they
do
minion
starts
and
it
doesn't
work,
call
it
because
of
a
operating
system
they
use,
maybe
because
of
the
in
case
of
Mac
OS,
the
container
technology.
Something
goes
wrong
with
Bertil
mostly
have
to
it's
another.
A
A
We
have
a
running
cluster
and
they
want
to
do
operations
like,
for
example,
the
iboga
component
stops
working.
They
want
to
add
monitoring
log
into
that
Prometheus.
How
does
it
work?
How
does
do
I
install
it?
How
do
I
perform
maintenance
operations
in
a
node
I
want
to
remove
all
the
pot?
How
do
I
do
it,
as
you
may
imagine,
doing
the
upgrade
for
cluster
things
go
wrong?
A
How
is
the
upgrade
process
at
the
moment?
Is
there
anything
we
can
do
to
improve
it?
For
example,
in
this
case,
I'd
highlight
cluster
upgrades,
also
how
to
force
the
removal
of
pot,
because
for
some
reason
you
see
several
cases
that
they
want
to
remove
pause
for
a
note
and
they
don't
seem
to
go
away.
They
are
stuck
in
a
terminating
process.
How
do
I
the
bug
that
it
leads
to
me
like
when
I
look
into
all
these
cases?
A
A
A
For
example,
you
get
an
enabler,
for
example,
with
our
bug,
or
do
you
see
that
is
not
accessing
for
some
reason:
it's
not
properly
configured,
it
has
no
proper
configuration
and
it
makes
sense
in
the
case,
for
example,
now
in
a
product
cluster,
but
you
can
see
a
lot
of
issues
recently.
For
example,
in
eks
I
found
several
issues
that
they
want
to
them.
They
configure
it.
We
supposed
to
be
semi
automated,
but
things
don't
work
as
you
should.
A
We
have
to
also
become
some
issues
well
of
the
three
categories.
In
this
case,
you
can
see
that
quite
similar
seven,
you
simmer
that
maintenance,
installation
up
ESS
share
a
more
or
less
the
same
parts
of
the
pie
and
in
this
sense,
I
would
I
think
I
know
you
heard
of
BK
PR,
for
example,
is
a
talk
tomorrow
and,
if
you're
interested
in,
for
example,
getting
attacking
the
cluster
maintenance,
part
or
also
part
of
the
cluster
installation.
A
Part
I
recommend
you
to
take
a
look
because,
as
I
said
before,
hearing
cubicle,
we
start
seeing
several
looting
that
attack
this
case.
It's
recommended
to
take
a
look
also
for
the
English
and
communication
issues.
I
recommend
you
to
go
because
do
you
see
that
how
a
way
to
attack
this
kind
of
issues
with
automation,
for
example,
ok
and
now
the
last
question
anything
this
one
also
has.
A
What
you
see
here,
d
ke
has
a
weak
share,
but
I
also
see
that
mini,
give
and
also
custom
clusters
in
chivay.
Dm
have
an
important
part
of
the
time,
and
this
has
to
do
with
something
that
I
also
found
in
several
of
the
previous
questions,
and
it
has
to
do
with
the
kind
of
user
that
goes
to
stack
overflow.
A
You
see
in
most
of
the
cases,
it's
not
we're
not
talking
about
users
having
communities
in
production
and
having
issues
you
don't
see
in
you,
don't
see
real
production
problems.
Most
of
them
have
to
do
with
people
that
are
learning.
People
are
experimenting
with
kubernetes
people
that
are
evaluating
for
migrating
their
applications
to
kubernetes.
A
A
So
I
was
discussing
this
because
so
maybe
yeah.
This
relates
the
first
question
that
maybe
we're
still
there's
a
lot
of
structure,
but
we
still
need
to
have
more
adoption
in
order
to
have
more
production,
more
production
cases
so
well.
This
helped
us
also
to
see
and
to
think
ok
what
we
can
do
right
to
have
more
adoption
then
start
seeing
more
production,
real
production
cases,
so
maybe
Stack
Overflow
sort
of
platform-
even
I,
don't
know,
but
this
is
also
something
about
the
profit
of
the
user
so
well.
This
is
what
I've
tried
to
find.
A
I
went
to
Stack
Overflow.
Try
to
see
different
questions,
try
to
learn
more
about
abusers.
I
come
up
with
several
questions,
myself
I
share
with
you,
but
to
summarize
we
see
that
kubernetes
popular
disk
only
up,
but
in
a
case
Stack
Overflow
still
there's
a
long
way
to
go,
which
is
interesting.
This
challenge
here,
because
I'm
sure
we
can
do
more
to
have
more
question
answered
to
make
people
find
solution
to
their
issues
and
also
yeah,
because
I'm
assuming
we
see
that
there's
a
lot
of
people
trying
to
learn.
A
So
ok,
for
example,
we
see
that
several
cases
there
in
the
versus
a
location,
Asian
developers
in
creating
deployments
services
how
to
interface
to
use
persistent
and
ensure
that
there's
a
there's
a
lot
of
intersection
between
each
of
the
use
cases
in
so
in
this
case,
it's
more
about
these
kinds
of
issues
or
something
not
well-defined.
Some
issues
are
sometimes
are
trivial,
so,
as
I
said
before,
it's
clear
that
there's
something
we
can
do
to
try
to
simplify
the
process.
A
Also,
there's
a
clear
need
how
to
improve
the
debugging
of
a
tool,
so
maybe
to
add
validation
tools
that
validate
not
only,
for
example,
deployments
are
also
take,
or
at
least
solar
shooting
tools
or
a
clear
table.
Shooting
line,
there's
different
ways
that
automation
there's
also
documentation,
so
we
need
more
tools
within
that,
so
the
blogging
and
simplified
process
and
they're
still
to
finish
in
the
end.
A
This
is
just
an
example:
we've
taken,
Stack
Overflow
is
to
be
interested
in
to,
for
example,
see
github
to
see
the
cumulative
slack
Channel
and
comparing
and
to
see
the
different
news
cases.
But
well,
it's
clear
here
is
that
we
need
to
provide
a
better
user
experience.
I
thought
something
we
should
never
stop
looking
into
not
only
identifying
all
of
them,
but
in
the
case
of
kubernetes
I.
Just
think
about
this,
maybe
will
it
make
sense
to
per
sample
even
have
a
special
interest
group.
A
I,
don't
know
in
the
end
what
I
wanted
to
show
here
either.
This
are
so
many
open
challenges
here
and
so
I
love
to
hear
drink
you've
come
to
take
a
look
to
discuss
to
see
what
we
can
do
in
this
matter.
Maybe
this
in
one
of
the
boots.
We
have
one
of
the
solution
that
we
have,
but
we
require,
or
maybe
there's
something
new-
that
we
need
to
create
so.
A
A
If
you,
google,
you
see
that
there
are
some
kind
of
solutions,
as
well
as
aim
for
developers
that
try
to
simplify
the
whole
process
but
yeah
in
the
end,
we
need
to
have
these
developer
community
also
mature
I,
think
also
the
cloud
native
approach
is
still
new
service,
so
but
there's
any
I
think
it
makes
sense
totally
to
have
this
kind
of
a
community
and
movie
these
aim.
So
this
pattern
that
this
goes
through
very
simple
cases
of
an
application,
so
ham
charts.
A
We
could
have
hunter,
for
example,
death
for
these
kind
of
developments
as
I
know.
There's
there's
an
open
were
in
this
sense
to
have
this
chalice.
Also,
that's
why
I
don't
here
I
just
want
to
try
to
think
about
that.
So
well.
Basically,
that's
what
I
wanted
to
say!
Thank
you.
Yes,
if,
by
the
end
of
these
toggles,
try
to
think
of
all
these
issues
and
they're
all
the
necessity
of
having
to
improve
the
visual
experience
telling
I
happy
hope
that
you
enjoy
the
talk.
So
thank
you
very
much.