►
From YouTube: Migrating 5G Core to the Edge, Wisely and Confidently with Kubernetes - Marco Antonio Gonzalez Ortiz
Description
Don't miss out! Join us at our upcoming event: KubeCon + CloudNativeCon Europe in Amsterdam, The Netherlands from 18 - 21 April, 2023. Learn more at https://kubecon.io The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects.
Migrating 5G Core to the Edge, Wisely and Confidently with Kubernetes - Marco Antonio Gonzalez Ortiz, Ericsson
In recent times, Multi-Access Edge Computing (MEC), a 5G specification, has risen as a game-changing solution for data exchange latency and cybersecurity threats identified among the variety of profitable use cases identified while migrating 5G to the edge. This proposal will describe three simple, yet consistent migration stages of a pre-existing 5G-Core Cloud Native solution to an ETSI-compliant 3GPP 5G EdgeApp (Edge Application) architecture. The first stage will include a review of the different considerations while re-architecting Kubernetes for an EdgeApp. Etcd performance in large Edge environments will be examined, along with pod-scheduling strategies and consistent data-storage for persistent data in a K8s cluster. Subsequently, an integrated MEC-5G architecture will be presented, along with MEC deployment scenarios, including MEC application mobility in a 5G containerized solution. The second stage will involve a thorough review of existing Kubernetes open-source edge projects: k0s, K3S, microK8S and KubeEdge, analysing their deployment costs, Edge-oriented level of maturity, documentation and community support. The final stage will outline some interesting examples of deployed MEC Use-cases, including AI platform for edge devices and 5G-Core over Bare-metal.
A
A
The
contents
for
today
I
want
to
try
to
be
quite
a
simple
of
many
of
the
5D
topics.
One
introduction
then
I'm
going
to
mention
the
problem
for
this
presentation.
What
is
5gh,
what
what
is
Mac
if
you
have
a
head
of
it?
What
is
the
strategies
that
kubernetes
that
can
be
applied
in
kubernetes
to
achieve
this
5D
core
at
the
edge
in
both
hcd
performance,
Sports
scheduling
and
at
the
final
I'm
going
to
do
a
demo
analysis?
It's
not
going
to
be
a
full
demo.
I!
A
A
Oh
great
two,
so
I'm
going
to
do
something
really
simple:
I'm
not
going
to
dig
in
deep
into
5G
for
for
us
for
the
presentations
it's
going
to
be
just
an
application
that
is
going
to
be
run
at
the
edge.
So
basically,
Edge
I
can
Define
as
a
scenario
where
we
have
a
data
producer
trying
to
send
a
bunch
of
data,
and
in
order
to
get
the
response,
we
are
trying
to
put
the
server
close
to
the
to
the
data
producer.
That's
the
main
concept
of
edge
Solutions.
A
Of
course
we
have
different
scenarios,
Computing
of
loading,
dedication
and
data
processing,
and
so
on
so
forth.
But
when
we're
trying
to
do
this,
there
are
common
of
important
questions.
We
we
have
is
this
solutions
it's
actually
going
to
be
profitable,
because
at
the
end
of
the
day,
5G
is
one
of
the
solutions
for
Enterprises.
Therefore,
the
question
is
massive
iot.
Is
it
really
going
to
help
me
this
5D
Edge?
This
is
definitely
going
to
handle
critical
workloads.
A
Is
it
going
to
have
some
redundancy
so
that
those
two
questions
are
appears
many
forums
when,
when
we
talk
about
5G
and
the
edge
and
the
most
important
easy
saving
cost
is
simply
profitable?
Are
we
saving
money
with
this?
I
will
make
him
I
will
be
spending
more
money
and
faster
TTM
or
time
to
Market,
5G,
core
and
5G
Solutions
is
meant
to
be
meant
to
have
a
fast
delivery
to
the
customer,
because
this
day
is
obviously
customer
wants
everything,
yes
of
the
speed
of
light.
So
that's
what
we're
expecting
from
5G.
A
A
A
A
Let's
talk
about
infrastructure,
let's
talk
about
what
makes
5G
at
the
edge
works,
and
if
you
want
to
focus
on
that,
let's
only
talk
about
the
red
boxes,
red
color
boxes.
This
is
one
standard
architecture.
Basically,
it's
defining
how
different
operators
trying
to
apply
these
use
cases,
application
Computing
of
loading,
multi-user
multi-natural
application
b2x,
which
is
a
vehicle
to
everything
and
automobile
automobile
workloads
and
so
on
and
so
forth.
Many
use
cases.
A
So,
let's
dig
into
those
two
red
boxes:
okay,
kubernetes
and
5dh
kubernetes
101,
you
guys
know
what
Etc
is
right.
It's
a
it's
one
of
the
key
topics
of
more
basic
topics.
We
have
it's
really
consistent
distributed
keyboard.
You
store
and
use
the
Wrath
consensus.
Protocol.
I
was
surprised
that
my
colleague
from
well
I
call
it
the
speaker
from
Hitachi
talk
about
this
consensus
protocol
and
what
are
the
challenges?
Indeed,
there
are
some
challenges
we
are
trying
to
solve
with
file
decor.
A
There
are
different
use
cases
in
the
representation
that
talk
about
Financial,
Bank
solution,
right
extra
consistency.
The
data
must
be
replicated
in
both
ends.
That
applies
for
a
bank
solution.
If
we
move
on
to
the
5th
iot,
which
means
devices,
there
is
no
such
priority
or
have
this
strong
consistency,
but
yet
having
this
extra
consistency
makes
a
huge
sacrifice
for
the
hcd.
Let's
go
to
the
this
small
diagram.
A
Every
time
there
is
support
scaling
out
every
time
I'm
trying
to
oh,
the
cluster
is
trying
to
collect
a
new
pot,
there's
a
lot
of
bunch
of
communication
between
the
AP
server,
because
the
controller
and
in
between
each
of
these
steps
we
have
the
famous
data
store
at
CD.
What
happened
every
time?
We
do
one
call
every
time
we
do
one
operation
get
or
to
update.
There
has
to
be
this
famous
strongly
consistency
synchronization.
A
If
we
have
three
data
store,
as
you
can
see
in
a
presentation,
there's
going
to
be
a
synchronization
between
these
three
nodes
before
replying
to
the
AP
API
server,
and
this
go
all
the
way
until
we
schedule
a
port.
We
release
the
port.
If
we
have
a
container
registry,
we
pull
the
image
and
every
time
we
have
this
external
consistency
operation,
you
may
think,
but
Marco.
This
is
just
common
kubernetes.
Why
does
this
affect
us?
A
I
mean
it's
just
a
few
milliseconds
few
seconds
overall,
that's
true,
but
let's
think
it
this
way,
apart
from
5D
at
the
edge
we're
also
trying
to
use
bare
metal,
we
were
trying
to
not
rely
on
one
Bender,
which
we
are
trying
to
open
the
different
options
with
AWS
I'm
sure
we
will
clout.
Today.
I
saw
a
bunch
of
presentation
just
talking
about
those
public
providers
and
you
guys
know
that
they
charge
for
everything
traffic
right
every
in
all
traffic
they
charge.
A
A
What
is
the
strategy
in
this
case?
Well,
this
is
kind
of
the
opposite,
but
my
the
previous
speaker
mentioned
it's
more
strongly,
consistency
to
lazy
thinking
for
that
we
use
the
conflict-free
to
look
at
data
types
which
it's
not.
As
the
name
says,
it
doesn't
need
to
perform
immediate
communication
with
other
notes.
Instead,
it
definitely
answers
the
query
or
replies
the
API
API
server
and
performs
this
of
these
two
replication
State
base
or
operation
based,
of
course,
the
first
one
is.
A
A
Of
course,
I'm.
Focusing
the
solution
on
the
massive
iot
scenario.
You
see
one
of
the
keys
now
that
we
have
so
far
worldwide
in
5G
core
h,
second
strategy,
again
basic
concept:
how
to
do
a
post
strategy,
guys
how
we
Excel
we
do
the
basic
HPA,
BPA
and
cluster
of
scaling,
those
three
options
we
have
been
working.
A
A
It's
meant
to
be
teaches
many
use
cases,
video
gaming,
streaming
tick,
tockers
biggest
tool,
so
one
user
can
do
many
things.
So
it's
not
like
I
want
user
went
traffic.
It's
not
one
using
one
request.
It's
one
user,
many
requests.
So
if
you
think
that
this
is
going
to
solve
your
problem,
you
may
ended
up
having
an
outage.
So
that's
what
that's
why
Telecom
players
don't
want
and
for
that
one
of
the
strategy,
it's
a
develop
and
improve
the
proactive
Auto
scalar.
This
is
obviously
the
research.
A
If
you
can
see
the
diagram
on
my
left,
there
is
a
kind
of
like
a
more
aligned
to
what
is
the
machine
learning
interested
to
the
5G.
For
instance,
we
have
one
control
Loop
in
this
case,
for
the
sake
of
the
presentations
every
20
seconds.
It's
going
to
evaluate
it's
going
to
receive
metrics,
but
not
only
CPU
memory
storage.
Not
only
the
common
ones,
more
data
that
is
aligned
to
the
client
is
aligned
to
the
use
case.
That's
a
big
difference
between
5G
and,
let's
say
any
database
server.
A
It's
not
that
straightforward
So
based
on
that
You
can
predict
the
behavior
and
the
take
upon
action.
It's
not
just
I'm
going
to
increase
when
I
I
found
an
issue
that,
in
in
a
telecom
service,
is
just
service
impact
in
its
customer.
Complaining
it's
impact
to
the
to
the
brand.
So
that's
something
at
least
so
far
is
one
of
the
key
discussions
for
5gh
one
of
the
benefits.
A
Of
course,
it's
a
proactive
workflow
for
auto
scaling,
multimetrics
that
can
be
used,
taking
the
consideration,
of
course,
limitation
and
constraints
of
the
resources.
This
is
important
because,
as
I
mentioned
before,
5gh
is
not
going
to
use
one
simple
brand.
One
simple
flavor:
it's
going
to
use
multi
Edge
for
cloud
providers.
It's
going
to
use
as
I
think
today
was
plenty
bare
metal,
so
different
options.
So
for
this
Randomness
for
these
variety
of
options,
we
can't
use
this
proactive,
both
outskiller
and
we
can
use
the
benefit
of
it.
A
A
As
you
can
see,
the
center
is
focused
on
the
cni,
CRI
and
CSI,
which
I've
discussed
today.
Also
I
I
forgot
to
add
the
the
mesh
what's
also
key
for
5D
edge.
Of
course,
these
four
projects
open
source
I'm
going
to
discuss
about
them
just
as
a
summary,
because
of
the
time
and
above
that,
as
we
can
see,
there's
Argo
amazing,
RCI
CD,
some
of
them
the
numbering
well
now
has
been
anchor.
A
That
I'm
using
nephew
is
an
open
source
project
is
having
a
lot
of
attention
these
days
in
previous
Keep
Calm.
They
have
a
lot
of
use
cases,
which
is
why
quite
interesting
and
on
up
with
just
a
standard
to
said
how
it
is
going
to
be
deployed.
On
top
of
that,
of
course,
we
have
the
application.
As
I
said,
for
us,
we
are
developers,
integrators
5G
is
getting
up
on
just
an
application.
A
Of
course
it's
complex,
but
at
the
end
of
the
day
it's
just
one
application
and
for
we
have
these
two
options
with
the
open
source,
we
are
mainly
used
for
any
for
many
International
Conference,
so
I
try
to
put
them
in
this
presentation
the
table
on
the
left.
It's
a
summary
and
the
comparison
between
these
four
projects
for
cute
H,
as
you
can
see,
slightly
Advantage
is
assigned
to
the
k2s,
and
while
the
cube
Edge
is
having
some
sort
of
like
a
complexity,
some
not
available.
A
Instead,
I
just
simply
use
one
cluster,
Master,
Now,
pre-worker
notes,
public
IP
and
the
reason.
Why
is
that?
Because
I'm
using
our
I
used
AWS
services
and
specifically
one
ec2
notes,
which
just
brutal
machines,
the
iot
devices
in
this
case
are
IP
cameras,
which
is
a
basic
solution
for
just
streaming
or
direct
streaming.
A
Of
course,
we
have
Primitives
for
an
ele
key
for
monitoring
and
data
and
observation
denotes
a
specification
in
this
demo.
I
try
to
use
at
32
cores,
100,
GB,
Ram,
500gb
storage
and
the
233
iops
when
I
start
with
my
colleague
about
this
scenario.
That's
also
that
always
a
question.
Why
do
you
pick
those
values?
Well,
there
are
many
reasons.
Of
course,
one
of
the
key
ones
is
that
as
a
community
Builder
I
have
free
I
have
some
points
to
use,
so
I
didn't
I.
A
In
fact,
this
let's
say
flavor
is
used
for
the
actual
5D
core
that
it
was
offered
to
the
public,
so
I
just
want
to
try.
The
Red
Bull
has
all
the
the
Lotus
the
load
tester
code,
as
I
mentioned
before
in
this
presentation.
For
the
sake
of
your
time,
I'm
not
going
to
do
any
video
demo
guys
just
going
to
go
straight
to
the
summary.
Please
go
through
this
deeper
GitHub
repo
one
and
I
must
mention
this.
A
I
basically
found
one
amazing
load
test
repo
and
Fork
it
and
modify
it.
Some
smart,
real
smart
guys
made
this
load
test.
So
it's
not
for
me
to
rip
in
a
wheel.
It's
just
to
use
and
I
try
to
apply
for
my
use
case
this
six
test
cases.
Kpis
is
what
I
consider
relevant
if
we
focus
on
the
hcd
or
the
database
consumption.
If
you
can
see,
we
have
a
cluster
API.
How
is
that
how
fat
it
is
and
how
fast
is
their
create
deployment?
A
Initially,
if
you
can
see
for
all
these
values
which
are
in
milliseconds,
then
one
who
still
kind
of
like
give
us
some-
let's,
let's
say
some
discussion-
is
how
much
time
it
takes
for
the
pots
to
be
running.
Of
course,
we
can
deploy
easily
deploy
a
cluster-
that's
quite
easy,
but
if
you
have
to
scale
and
considering
the
QR
limitation
we
have
for
the
hcd.
Of
course,
the
four
projects
have
some
struggles.
A
I
still
have
some
struggles,
these
four
projects
that
use
different
cni
CSI
and,
of
course
we
have
a
lot
of
customization
for
a
sake
of
this.
For
this
demo,
I
just
use
default
a
cni
default
configuration
as
you
can
see.
Cube
Edge
has
the
highest
response,
which
doesn't
imply
that
it's
not
the
the
best
option,
one
of
the
the
ones
that
was
kind
of
steady
along
the
the
the
test
which,
by
the
way,
which
is
the
worst
yes
I
scale
out
from
then
replica
to
100
replicas.
A
So
it's
just
one
scale
out
and
just
I
was
trying
to
monitoring.
How
fast
are
for
the
pods
to
be
running,
because
this
implies
I
have
10
000,
IP
cameras.
I
have
many
cameras,
so
they
are
increasing
the
traffic
loads
rapidly.
So
what
I
expect
I'm
expect
my
cluster
to
get
globally
used
to
this
increase
and
adjust
and
also
increase
the
capacity
that's
pretty
much
the
demo
they
perform
in
half
time.
A
Okay,
takeaways,
as
I
said,
5G
core
The
Edge
have
many
advantages
and
many
challenges,
one
of
them
that
I
present
today
are
two
the
hcd
dependency,
the
hcd.
Let's
say:
let's
see
the
limitation
when
we're
trying
to
be
too
strong
consistency
in
that
store.
There
are
many
options
that
or
many
use
cases
that
doesn't
require
actually
to
be
strongly
consistency.
A
Also,
the
post
strategy
is
one
of
the
key
topics
are
still
a
lot
of
research.
A
lot
of
tests,
a
lot
of
failure
and
tests
doing
recently
to
for
the
fourth
strategy,
is
trying
we're
trying
to
get
out
of
the
basic
of
the
scaling
and
yes,
I
use
a
more
predictive
one,
which
was
the
customer
usually
requires
at
the
very
end,
I
talk
about
the
current
executing
projects
design,
and
there
shows
some
speed
when
we
are
starting
up
when
when
we
are
deploying
the
there
is
a
cluster.
A
But
still
there
are
some
observation
or
there
are
still
not
wasting
some
latencies
when
we're
trying
to
actually
get
the
pots
running,
which
is
when
the
application
is
is
already
ready
for
for
taking
traffic.
That's
pretty
much
my
my
demo
for
today
and
with
that
thanks
so
much
for
your
time
and
thank
you.