►
From YouTube: Kubernetes for Product Managers Part 1
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Hi,
my
name
is
victor
arnaud.
I
am
one
of
the
product
managers
here
at
gitlab,
and
this
video
is
to
prepare
product
managers
to
think
about
kubernetes
and
to
be
able
to
consider
it
for
various
products.
They
want
to
build,
perhaps
not
only
product
managers
but
like
product
people
who
are
not
very
technical
and
not
deep
into
cloud-native
technologies.
A
A
A
By
the
end,
I
hope,
with
a
bit
of
imagination,
you
will
understand
why
the
following
snippets
could
be
exciting,
for
example,
for
gitlab
and
how
we
can
reach
it.
What
what
are
the
underlying
building
blocks
within
kubernetes?
That
allow
us
to
to
make
meaning
out
of
that
symbol,
so
it
will.
These
videos
will
have
actually
two
parts
at
least
that's
what
I
planned
today.
A
The
first
part
covers
some
very
basic
things
that
might
be
useful,
even
if
you
want
to
to
start
using
kubernetes
as
a
user.
The
second
focuses
exclusively
on
on
the
the
basic
building
blocks
and
enabling
power
of
kubernetes
still,
I
think,
the
it's
hard
to
understand
the
second
part
without
the
third,
the
first
one.
So,
let's
start.
A
A
Thus
kubata
seems
to
be
complex,
but
if
you
think
about
it,
there
were
virtual
machines,
bare
metal
servers
even
before
kubernetes,
that
system
administrators
had
to
maintain
and
they
managed
to
build
sizable
infrastructure,
huge
infrastructure
with
that,
and
that
was
complex
as
well
extremely
complex.
They
had
to
take
care
of
deployments
of
security,
patching
and
everything
else.
The
same
was
true
for
kubernetes
as
well
and
from
purely
operations.
A
A
The
same
is
true
with
kubernetes.
You
don't
know
how
you
get
the
desired
result.
The
cluster
takes
care
of
that
for
the.
For
this
reason,
you
have
to
understand
that
little
ci
cd
is
imperative.
You
have
to
codify
all
the
steps
you
want.
You
have
to
run
commands
that
will
do
something
that
will
build
your
docker
image.
You
don't
just
say
that.
Give
me
a
docker
image.
You
have
to
say
that
okay,
docker
run
or
docker,
build
whatever
in
kubernetes.
A
Contrary
to
that
the
ml5s
describe
the
desired
state
and
the
cluster
takes
care
about
reaching
that
state.
This
is
the
magic
that
we
should
understand.
We
describe
the
desired
state.
If
you
look
at
the
yaml
here,
for
example,
you
can
probably
figure
out
what
that
desired.
State
is,
and
it's
relatively
easy
to
imagine
an
interactive
code
how
to
reach
the
same
thing,
which
would
say
that
create
me
and
m
space
and
then
provisional
namespace
service,
a
service
account
in
that
namespace.
A
A
So
following
this
decorative
pattern
has
actually
some
very,
very
interesting
benefits.
Besides
making
it
accessible
to
every
user
because
yaml
is
accessible,
it
even
allows
a
programmed
transformation
of
the
ammo,
and
that
makes
it
easy
to
scale
to
different
scenarios
with
all
kinds
of
tools
that
are
that
exist
within
the
ecosystem.
A
A
I
know
that
even
within
gitlab,
not
everybody
thinks
that
it
is
feasible.
I
personally
think
that
we
don't
know
yet
how
to
make
it
feasible,
but
it
is
feasible
and
personal.
I
even
have
an
idea.
You
might
figure
that
out
yourself,
I'm
not
speaking
about
that
in
these
videos.
A
Besides
the
desired
state,
I
would
like
to
speak
a
little
bit
about
labels
and
annotations.
The
ml
files
can
contain
all
kinds
of
metadata
about
the
resources,
and
these
are
typically
in
the
form
of
labels
and
annotations.
It's
very
important
to
understand
that
labels
are
at
the
core
of
filtering
of
resources.
A
This
means
that
when
one
wants,
when
someone
wants
to
be
the
dashboard,
for
example,
of
course,
that
dashboard
can
contain
views
to
list
all
the
pods
or
all
the
namespaces,
but
very
likely
it
will
have
functionality
to
show
the
list
of
pods
with
a
specific
label
or
even
all
the
resources
with
a
specific
label,
not
just
by
the
kind.
The
kind
was
the
namespace
and
the
service
account.
A
A
Just
a
quick
recap:
decorativeness
think
about
it.
What
it
means
to
be
fully
decorative,
what
might
be
the
edge
case?
Is
there
why?
Why
is
it
powerful?
How
could
we
use
that
within
gitlab?
What
where
does
gitlab
face
decorativeness
and
the
other
thing
is
about
filtering?
It's
more
just
to
understand
that
labels
are
there
for
us
to
help
with
filtering
and
building
all
kinds
of
tools
that
can
interact
through
the
filters,
and
that
was
the
first
part.