►
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
We'll
start
with
three
kubernetes
clusters:
one
of
them
will
be
the
management
cluster
that
will
use
this
km
Alias
to
access
it
and,
as
you
can
see,
there's
nothing
really
deployed
here
apart
from
metal
OB
that
we'll
use
to
expose
some
Services
externally
and
the
other
two
clusters
are
called
the
workload
clusters
and
we'll
use
K1
and
K2
aliasis
to
access
them
again.
Nothing's
deployed
on
here.
So
let's
start
by
downloading
the
mesh
cuddle
CLI
and
we
can
use
this
CLI
to
deploy
various
glue
components
directly
to
our
clusters.
A
A
So
if
we
look
at
the
diagram,
you'll
see
that
the
the
workload
clusters
on
the
bottom
have
these
agents
which
are
going
to
initiate
a
connection
to
the
management
cluster,
and
you
can
see
it's
like
One
Direction,
so
the
management
Server
doesn't
know
about
how
to
reach
the
workload
clusters.
It's
always
the
agents
that
are
initiating
the
class,
the
the
connection.
A
So,
let's
click
register
these
agents.
There
are
some
security
mechanisms
in
place
to
prevent
someone
from
registering
other
clusters
to
the
management
plane
without
permission.
First,
the
platform
team
needs
to
create
a
kubernetes
cluster
object
in
the
management
plane
for
the
cluster
that
needs
to
be
registered,
and
then,
when
you
were
ready
to
register
that
cluster,
you
would
need
this.
A
Any
of
the
options
we
set
here
are
just
telling
the
agent
how
to
communicate
with
the
management
server.
The
the
and
the
Telemetry
Gateway,
which
is
an
open,
Telemetry,
collector,
that's
running
in
the
management
plane,
that's
going
to
collect
the
metrics
from
all
the
various
workload
clusters.
So
we
do
that
for
the
first
cluster.
A
Give
that
a
second
and
then
we
can
check.
We
can
check
to
see
what
are
the
different
components
that
were
deployed
so
on
the
management
server.
We
have
the
two
like
the
management,
server
pods
and
we
have
the
UI
that's
using
redis
for
caching
and
a
Prometheus
server
to
access
the
metrics
that
have
been
collected
through
the
open,
Telemetry
collector
called
the
Telemetry
Gateway
and
then
on
the
workload
clusters.
A
You'll
see
the
glue
mesh
agent,
that's
that's
reaching
out
to
the
glue
mesh
management
server
and
we
also
see
a
glue
mesh
collect
blue
mesh
Telemetry
collector,
which
is
another
open.
Telemetry
collector
that's
going
to
capture
and
collect
all
the
metrics
from
the
local
cluster,
and
this
is
a
Damon
said,
but
we
only
have
one
node
here
so
and
then
this
pod
will
then
send
these
metrics
to
the
Telemetry
Gateway
on
the
management
cluster.
So
now
we
can
use
mesh
cuddle
to
check
the
connectivity
and
the
health
of
the
different
workload
clusters.
A
We
can
see
all
kinds
of
information
about
about
the
licensing.
We
can
see
the
various
components
that
are
deployed
on
the
management
cluster.
We
can
see
that
they're
healthy.
We
can
see
the
agents
that
are
connected
and
registered,
and
we
can
also
see
which
pod
that
they're
connected
to
on
the
management
server-
and
this
is
important
because
we
can
scale
the
deployment
on
the
management
server
to
multiple
pods,
and
this
allows
us
to
spread
the
connections
of
different
workload
clusters,
allowing
us
to
scale
to
dozens
or
even
hundreds
of
workload.
A
We
can
also
take
a
look
at
the
the
glue
platform
dashboard,
the
UI,
so
we'll
get
the
IP
of
the
of
the
UI
and
then
access
it
in
my
browser
and
I
can
see,
and
I
can
see
the
two
clusters
that
are
registered
the
workload
clusters
right
now.
I
can
just
see
the
version
of
the
kubernetes,
but
istio
is
not
deployed
yet.
So
we
don't
have
any
information
about
that.
A
Yet,
and
I
also
don't
have
any
workspaces
because
I
don't
have
any
workspaces
created
yet
glue
will
actually
emit
metrics
about
the
the
agents
that
are
connected
to
the
management
plan.
So
to
do
that,
we
can
look
at
the
Prometheus
metrics
that
are
Exposed
on
the
slash
metrics
endpoint
on
the
management
server
pod.