►
From YouTube: Keynote: Super-Charge Your Mesh Services With Enhanced Data Access- Idit Levine, Founder & CEO, Solo
Description
Don’t miss out! Join us at our next event: KubeCon + CloudNativeCon Europe 2022 in Valencia, Spain from May 17-20. 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.
Keynote: Super-Charge Your Mesh Services With Enhanced Data Access- Idit Levine, Founder & CEO, Solo.io
It is clear that service mesh is ‘crossing the chasm’ as more and more organizations not only deploy service mesh into production, but deploy it in large scale, mission critical environments. Development and operations teams value the common security model, fine grained traffic management and end-to-end observability provided by the mesh. Organizations are also realizing that the services and APIs that run within the mesh can expose valuable data and functionality, but application developers still need to write a lot of code to harness this information, limiting the adoption and ultimate benefit these services can provide.
In her talk, Idit Levine, CEO of Solo.io, will discuss innovative use cases that can be enabled by extending a service mesh. She will explore how a mesh can be used to simplify complex data and information access to support a wide variety of clients with minimal code to unlock hidden insights of an organization.
A
We
sit,
did
it
okay,
awesome
guys?
Thank
you.
So
much
can
I
start
nothing.
Good
awesome,
hi
guys!
Thank
you
so
much
for
coming
here.
My
name
is
adid
and
I'm
the
founder
and
ceo
of
solo
io,
and
I'm
really
really
excited
to
be
here
for
two
reasons:
number
one
really
really
passionate
about
what
we're
going
to
talk
today
and
number
two.
I
think
it
was
two
years
since
I
talked
to
a
real
audience.
A
So,
let's
start
so.
As
I
said,
solo
is
the
company
that
I
founded
and
we
are
focusing
on
solving
the
application
networking
problem,
so
the
connectivity
and
as
one
we're
basically
leading
this
stl
community
are
stuck
is
basically
sto
and
envoy,
and
therefore
we
have
working
very
very
hard
upstream.
We
have
two
tlc
members
from
the
six
that
exist
in
sto
and
we're
working
very
hard
to
kind
of
like
push
the
project
effort.
A
We
have
basically
a
platform
called
glue
mesh,
and
this
platform
is
basically
a
built
from
four
pieces,
the
first
one,
it's
basically
the
mexico
air
core
and
that's
basically,
what
it's.
Basically,
an
sto
upstream
within
hands
from
us,
like
external
load
and
weight.
Limiting
eastwood
is
a
threat,
rotation
and
a
lot
of
other
stuff
and
big
big
focus
on
multi-clusters.
A
The
second
one
is,
we
basically
took
the
sto
ingress
gateway
and
basically
make
it
a
full-blown
api
gateway.
So
what
we
did
we
extended
with
a
lot
of
envy
filter
like
waff,
like
like
data
loss
provision
like
soap,
the
only
sto
soap
that
exists
in
the
market
and
so
on
and
the
last
and
not
least,
we
have
developer
portal
again.
The
only
developer
portal
that
exe
forest
here
in
the
market
today-
and
it's
supporting
you
know,
open
api
and
as
well
as
jrpc.
A
So
that's
what
we
have
and
the
last
and
not
least,
what
we
have
is
extension
and
extensions,
is
basically
the
ability
to
customize
the
service
mesh
to
your
own
need
and
we're
doing
it
with
wasm.
So
it's
built
into
our
platform,
it's
working
seamlessly
and
you
will
be
able
to
write
your
own
custom
filters
to
android.
A
Okay.
So
that's
great!
So,
as
I
said,
service
mesh
envoy
proxy,
that's
where
we're
focusing
and
service
match
is
solving
a
lot
of
problem.
As
you
all
know,
because
that's
why
you
are
here,
one
of
them
is
networking
the
other
one
is
observability
and
the
last
and
not
least,
is
security
and
policy.
That's
the
stuff
that
they
are
focusing
on.
A
A
So
this
is
where
graphql
is
coming
into
the
place
right
and,
as
you
know,
hopefully,
you
know,
graphql
id
is
basically
to
abstract
that
logic
and
basically
suggested
that,
as
as
as
an
abstraction.
Basically,
you
will
make
one
request
to
the
graphql
service
and
the
graphql
service
that
is
basically
configured
by
configuration
and
declarative
configuration
will
go
and
do
all
those
run
through
for
you,
okay,
that's
great.
So
what
is
graphql
graphql
is
basically
the
ability
to
do
to
to
ask
all
your
services
use
all
your
services
as
a
query.
A
A
Now
it
doesn't
mean
that
it's
need
to
do
it
only
for
one
service
right.
It
can
do
it
from
a
lot
of
service.
It
doesn't
really
care
how
much
service
you
need
to
query
from
and
that's
what
you
can
see.
You
have
three
services
here:
you're
asking
a
query:
you're,
getting
the
response,
he's
going
and
doing
all
the
round
trip
for
you.
You
don't
know
honestly,
it's
abstract
for
me.
A
So
that's
what
graphql
is
and
the
biggest
point
about
graphql-
and
I
think
it's
really
really
important
is
how
easy
to
make
you
to
consume
it,
and
I
think
this
is
something
that
you
will
hear
from
a
lot
of
our
front-end
team
and
ui
team.
They
want
graphql
because
then
it's
one
query
that
they're
putting
in
their
application
and
basically
the
data
is
coming
to
them.
So
it's
a
declarative
query,
it's
quick
and
it's
bringing
you
very
easy
to
consume,
which
one
one
request.
A
So
now
the
question
is:
what's
the
problem,
why
am
I
telling
you
about
it?
This
is
a
service
mesh.
So
we'll
tell
you
why
because
actually
build
this
graphql
service,
it's
actually
taking
a
lot
of
time.
You
need
to
take
some
libraries
right
and
this
library
usually
a
lot
of
the
time
that
at
least
the
leading
one
is
node.js,
because
it's
a
ui
kind
of
like
frontend
team,
and
then
you
need
to
give
them
the
schema.
A
And
then
you
need
to
build
in
the
resolver
which
basically
will
resolve
to
that
graph,
and
then
you
need
to
deploy
it
and
build
it
right.
This
is
a
another
thing
that
you
need
to
take
care
on
your
infrastructure
and,
of
course,
you
know
you
can
have
something
in
your
infrastructure
without
security
and
observability
that
too
risky,
specifically
because
it's
kind
of
like
acting
right
now
as
your
gateway.
So
it's
really
really
dangerous.
A
So,
when
you're
looking
at
this,
we
try
to
figure
out
if
there
is
any
way
for
us
to
make
that
thing
simpler,
we're
working
with
a
lot
of
customer
in
production-
and
we
constantly
heard
about
that
conflict
about
there
is
this
graphql
there
is
a
service
mesh.
I
already
have
a
gateway.
How
is
that
going
to
work?
A
And
that's
exactly
why
we
decided
to
make
it
better
so
envoy
has
this
amazing
ability
to
basically
be
extendable,
and
this
is
why
we're
a
big
fan
of
that
and
what
you
can
do
is
basically,
you
have
the
the
notion
of
fill
filter
chain
and
the
request
is
coming
to
the
filtered
chain
and
going
to
a
filters
that
you
can
put
your
own
custom
filter
inside
and
that's
what
solo
doing
a
lot
right.
So
if
you
see
about
what
our
differentiator
we
know
and
we're
in
and
out
so
we
have
rough
support.
A
We
have
data
loss
prevention,
filter.
We
have
aws
slum
that
the
calling
creating
the
relationship
and
the
the
trust
immediately.
We
have
request
and
response
transformation,
one
of
the
most
honestly
probably
being
used
filter
that
we
have.
We
brought
soap
to
envoy
right
and
therefore
to
sto
as
well,
and
now
we
basically
brought
graphql
to
us
to
envoy.
So
what
does
it
mean?
A
Basically,
what
we
just
did
right
now
is
that
every
service,
slash
gateway
in
your
infrastructure
know
how
to
talk
and
become
a
graphql
server
without
you
need
to
do
much
so
the
question:
what
do
you
need
to
do
and,
as
I
said,
what
you
need
to
do
is
to
give
the
schema
and
the
resolver,
but
we
actually
thought
that
we
don't
even
have
to
let
you
do
this,
so
we
found
a
way
to
auto
generate
the
schema
in
the
resolver
for
new
for
those
services.
So
how
do
we
do
this?
A
Either
we're
using
open
api
or
we're
using
you
know
swagger
over
using
you
know,
wisdom
or
anything
that
we
can
in
order
to
kind
of
like
understand
what
the
schema
should
be
and
they're
basically
converted
as
the
resolver.
So
again,
you
basically
can
make
everything
in
your
infrastructure,
with
a
click
of
a
button
now
basically
serve
graphql.
A
So
what
it
was
very
important
to
us
is
also
to
stitch
it,
because
if
everything
right
now
is
a
little
graphql
to
represent
those
paths,
we
also
need
to
stitch
it
and
that's
what
we'll
do
in
the
gateway.
So
we
actually
build
into
that
envoy,
stitching
and
federation
is
coming
next,
so,
basically
again
click
of
a
button
you're
just
telling
us
the
relationship
between
them
and
boom
everything
declarative.
You
can
actually
have
graphql
everything
in
your
services
and
stitch
it
on
the
gateway.
A
Okay.
So
now
we
have
also
another
piece
on
that
puzzle
right.
This
is
this
is
our
product
right.
This
is
what
we
are
selling
and
now
we're
selling
only
also
that
piece
which
is
again
flip
over
button.
Everything
can
be
queryable,
but
in
glow
it's
actually
even
more
interesting
in
our
platform,
because
in
our
platform
there
is
the
ability
to
actually
create
what's
called
a
workspace.
So
what
does
it
mean?
A
workspace?
A
Your
workspace
is
basically
your
ability
to
take
namespace
cross
cluster
in
same
class
there,
wherever
you
want
and
basically
create
a
workspace
and
delegate
it
to
your
team
and
when
you
delegate
to
the
team,
you
also
tell
me
what
they're
allowed
are
not
allowed
to
do
okay,
so
basically,
this
is
a
very,
very
interesting
way
to
actually
expose
and
delegate
that
in
your
organization,
the
beautiful
of
it
is
now.
You
can
also
export
export
services
in
your
workspace
to
the
other.
A
So,
for
instance,
I'm
the
billing
service-
and
I
wanted
to
be
considerable
by
all
my
company.
I
will
be
able
to
expose
it
and
what
will
happen
in
the
other
services.
They
will
see
a
catalog
and
those
catalog
is
basically
will
allow
them
to
import
that
service.
So
now
they
will
basically
be
able
to
say
I
want
to
take
those.
You
know
this
is
the
services
that
are
exposed
to
me
and
click
on
button.
A
I
can
consume
them
to
my
workspace
and,
if
you're
consuming
it,
let's
do
it
the
right
way
and
what
we
basically
did.
We
said:
let's
embed
a
developer
portal
in
that
process.
So
when
you
actually
try
to
import
a
service
in
the
catalog,
you
will
be
able
to
get
everything
the
developer,
giving
you,
which
is
basically
the
docs,
try
it
get
the
keys
whatever
you
want,
onboarding
your
people
very
easily,
and
until
now
you
could
have
done
it
with
open
api
and
grpc.
But
now
we
also
you
can
expose
anyone.
A
So
that's
what
I
have,
but
I
wanted
to
kind
of.
Like
summarize
what
I
just
talked
about,
because
I
think
it's
extremely
important.
What
we're
solving
is
a
huge,
huge
technology
problem.
It's
not
easy
to
solve
trust
me
anyway.
It's
not
easy
writing
c,
plus
flash
I've
seen
graphql
inside,
and
it's
even
harder,
though,
which
I
think
is
very
very
important
is
the
fact
that
we're
solving
a
conflict
inside
your
team.
We
had
customers
coming
to
us
constantly
and
said
my
team
won
graphql.
How
is
that
working
with
service
mesh?
A
So
here
is
how
it
works
in
service
measure.
It's
part
of
your
service,
it's
taking
advantage
of
your
service
mesh.
You
don't
need
anything
in
your
thing
and
in
a
quicker
button
you
can
basically
enable
your
your
you
know
your
velocity
on
your
organization.
So
that's
why
I
am-
and
I
think
I'm
out
of
time
anyway.
So
thank
you
so
much
and
hopefully
you
enjoy.