►
From YouTube: Extending the Docker Compose Experience to Service Mesh with Meshery Layer5 and HashiCorp
Description
Extending the Docker Compose Experience to Service Mesh with Meshery
Presented at DockerCon 2022 with Lee Calcote of Layer5 and Nic Jackson of HashiCorp. More details: https://layer5.io/blog/docker/extending-the-docker-compose-experience-to-service-mesh
A
B
So
for
me,
as
a
kind
of
a
developer,
I
found
docker
in
like
I'd
like
to
say
about
eight
years
ago
and
docker
for
mac.
It
might
be
around.
I
might
be,
you
know,
maybe
I'm
sure
it's
around
there
anyway,
and
what
I
was
enabled
to
do
with
docker
for
mac
was.
I
could
do
things
like
run
my
sql
containers.
I
could
run
postgres
databases
and
I
could
run
them
really
easily
really
quickly
and
I
could
do
it
in
a
kind
of
a
disposable
way.
B
I
didn't
need
to
kind
of
pollute
my
my
local
environment,
and
this
was
like
a
really
big
thing
for
me.
I
kind
of
like
found
this
intriguing
very
early
on
and
and
being
able
to
then
sort
of,
extend
that
and
keep
that
sort
of
clean
room
environment,
as
I
kind
of
extended
it
out
of
testing
was
super
useful.
A
I'm
right
there
with
you
nick
actually
docker
desktop,
has
been
a
staple
of
my
daily
routine
for
about
as
long
as
I
can
remember.
I
think
it
was
at
dockercon
eu.
It
might
have
been
copenhagen
when
the
integration
with
kubernetes
first
came
out
and
I've
been
a
fan
ever
since
it's
just
been
really
exciting
to
develop
an
extension
for
meshri.
That
now
makes
brings
that
same
experience
for
service
mesh
in
the
same
way
in
which
docker
has
brought
that
bundled
experience
for
kubernetes
what's
been
pretty.
A
B
Right
and
and
like
you
know,
I've
been
I've
been
using
doc
composer.
I
think.
As
long
as
there's
been
docker
compose,
and
I
I'm
pretty
still
clued
up,
you
know
I
I
still
pretty
think
today
that,
like
there's,
there's
no
easier
way
to
create
a
sort
of
a
simple
multi-tier
application,
just
specifically
in
in
sort
of
the
development
environment
and
when
it
comes
to
service
mesh,
like
should
developers
care
about
service
mesh
like
do
they
have
to
care
about
it.
B
I
think
the
answer
is
yes
and-
and
I
think
that's
because
you
need
to
think
about
what
I
think's
like
kind
of
three
key
things,
which
is
that
the
way
you
implement
reliability
patterns
is
changing
you're
kind
of
no
longer
codifying
them
into
your
application.
It's
now
externalized
in
the
service
match,
and
you
need
to
kind
of
like
learn,
service,
mesh
and
and
kind
of
understand
how
it
works
as
a
platform
component
being
able
to
kind
of
have
a
tight
feedback
loop
for
developing
and
deploying
workloads
on
console
service
measure.
B
Other
service
meshes
is
really
important
and
we
need
a
great
developer.
Experience
like
the
key
thing
is
that
we
need
to
be
able
to
have
that
developer
experience
be
able
to
run
locally,
potentially
disconnected
that
we're
kind
of
used
to
so
onto
like
the
the
kind
of
the
reliability
just
kind
of
digging
into
that
a
little
bit
deeper.
This
is
how
I
used
to
do
things.
So
if
I
was
writing
something,
let's
say
with
spring,
I
would
kind
of
codify
my
my
reliability
inside
of
my
application.
B
Those
again
is
is
a
really
important
learning
capability,
and
then
I
think
you've
got
to
think
about
the
zero
trust
environment
service
match
gives
you
you
can't
sort
of
take
for
granted
that
a
network
connection
is
going
to
be
available
like
you.
Could
you
could
sort
of
think
about
externalized
connections
in
xero
trust?
Well,
I
might
not
be
able
to
make
a
call
to
like.
I
don't
know,
google,
or
something
like
that.
You
know
that
that
external
network
connection
in
a
zero
trust
environment
is
potentially
not
there.
B
It's
enabled
by
configuration
so
being
able
to
kind
of
test
my
applications
and
being
able
to
kind
of
you
know
just
play
around
with
things
before
I
have
to
push
them
into
a
larger
environment,
really
helps
with
that
fast
feedback,
loop
and
then
there's
performance,
testing
and
behavioral
benchmarking.
I
mean
you
know,
there's
one
of
the
kind
of
the
staples
of
measuring
the
ability
that
I
can
kind
of
run
performance
tests
against
my
application
that
the
fact
that
I
can
kind
of
run
these
behavioral
benchmarks.
What
happens
if
this
particular
service
fails?
B
Does
my
sort
of
configuration
around
reliability
work
correctly
and
then
helping
me
to
kind
of
run?
Some
diagnostics
when,
when
I
need
to
all
of
those
things,
are
really
really
important,
but
you
know
I
could
talk
about
those
things
all
day,
but
maybe
it's
just
better.
If
I
just
show
you
how
it
works.
A
With
docker
desktop
running
and
the
mesherie
extension
installed,
we
can
see
that
meshrey
has
provisioned
and
discovered
console
as
a
service
mesh.
It
has
performed
this
operation
using
the
meshri
adapter
for
console,
which
is
one
of
a
number
of
adapters
that
meshri
has.
It
is
connected
to
the
docker
desktop
instance
of
kubernetes
meshery
is
a
multi-kubernetes
manager
capable
of
performing
lifecycle
management
of
kubernetes,
as
well
as
10
different
service
meshes.
A
In
this
case
console
before
we
deploy,
we
can
also
use
meshri
to
take
that
same
docker,
compose
application
and
go
and
further
explore
all
of
the
capabilities
of
console.
We
can
select
the
version
of
console
that
we'd
like
to
work
with
explore
its
differentiated
capabilities
in
this
case,
like
its
load,
balancing
capabilities
with
hash
policies.
A
B
The
last
points
I
had
on
that
slide
earlier
on
was
about
developer
experience
and
one
of
the
things
that
I've
loved
with
working
with
docker
over
the
last
eight,
eight
or
so
years,
is
that
developer
experience.
It
just
makes
it
really
easy
that
I
can
just
get
out
my
laptop,
and
I
can
just
go,
do
my
job
and
I
think
now,
with
docker
extensions
and
and
the
measuring
extension
to
be
able
to
deliver
service
mesh
into
that
environment.
I
kind
of
continue
that
trend.
A
You
know
more
than
that,
I'm
impressed
by
how
much
you
are
able
to
do
without
leaving
docker
desktop.
That's
really
something!
The
extensions
are
pretty
interesting.
The
mesherie
extension
is
now
available
for
everyone
else.
It's
time
to
go
out
and
try
the
mesher
extension
deploy
a
console
mesh
learn
something
new.