►
From YouTube: What’s new in Gloo Mesh 2.0 & Future Roadmap
Description
SoloCon 2022:
What's new in Gloo Mesh 2.0 and Future Roadmap
Speakers:
Chris Gaun
Director of Product Management, Solo.io
Neeraj Poddar
Head of Engineering, Solo.io
Abstract:
Join Chris Gaun and Neeraj Poddar to learn more about what's coming up in the Gloo Mesh product roadmap.
Track:
Service Mesh and Application Networking
A
First,
I
want
to
give
you
a
little
bit
of
a
business
update
over
the
past
year.
In
the
last
solo
con,
we
introduced
glue
mesh
1.0
since
that
time,
we've
added
a
whole
slew
of
leadership
from
the
istio
community,
naraj
lynn,
that
have
come
on
board.
These
are
people
with
that
have
been
on
the
technical
operating
committee,
the
steering
committee
of
istio
and
have
really
guided
istio
and
our
key
leadership
they're
now
here
at
solo,
telling
us
and
communicating
to
us
and
building
our
products,
and
it's
just
great
to
have
them
here.
A
We've
also
made
a
whole
bunch
of
huge
contributions
to
istio
the
upstream
community,
the
release
cycles,
and
then
we
also
have
seen
the
latest
and
first
analyst
report
come
out
around
service
mesh
solo.
I
o
has
been
a
leader
in
this
space
since
the
very
beginning,
and
now
we
have
the
recognition
from
this
analyst
report
from
gigaom.
We
are
right
in
the
middle
in
the
leadership
position
ahead
of
everybody
else.
I
want
to
give
you
a
general
update
on
product
here
at
solo.
A
First,
we
have
that
slack
community.
We
have
over
4.7
000
users
now
in
there
to
help
you
our
users,
we,
the
product
managers,
are
in
there,
I'm
in
there.
I
love
hearing
from
you.
I
love
getting
dms
on
how
you're
doing
with
the
product
or
feedback
on
the
product.
It's
great.
We
take
that
very
seriously
and
make
sure
that
all
the
feedback
that
you're
giving
us
is
considered
when
we
actually
build
the
new
versions
of
the
product.
A
Secondly,
we've
done
hundreds
of
releases
this
year
since
the
last
solocon
we're
talking
about.
If
you
add
up
the
gas,
the
patch
releases,
the
beta
releases
across
all
of
our
product
lines.
It's
a
dizzying
array.
I
have
to
actually
go
and
look
up
what
version
of
the
product
we're
on
70
to
80
percent
of
the
features
that
we're
delivering
the
issues
that
we
address
come
directly
from
the
feedback
of
you,
our
users
and
customers.
We
take
this
very
seriously.
A
A
I
o
we
have
two
main
product
lines:
one
is
glue
edge.
This
is
one
of
our
api
gateways.
It's
based
on
envoy,
it's
one
of
our
most
mature
and
older
products,
and
then
we
have
glue
mesh,
which
is
now
in
the
2.0
phase.
This
is
our
enterprise
sdo,
offering
in
that
we
have
a
bunch
of
modules,
including
developer
portal,
graphql
wasom
assembly
hub
and
new
ones.
Coming
out,
all
the
time
we
as
a
product
team
like
to
deal
directly
with
the
customers,
we're
in
meetings
with
customers
all
the
time
every
week.
A
This
is
a
classical
model
of
the
networking
layers
from
layer
1
to
layer.
7.
you've
heard
a
bunch
of
announcements
here
at
solo
con
where
we've
announced
ebpf
features.
We've
announced
graphql
we've
announced
a
new
version
of
glue
mesh,
we've
announced
new
api
gateway
features
and
you're
curious
on
how
that
all
fits
together.
A
A
This
is
something
that
ebpf
adds
unique
features:
layer,
4
and
above
this
is
something
that
you
have
uniqueness
and
feature
rich
features
in
istio,
and
then
we
even
added
a
layer,
eight,
which
is
the
graphql
layer
which
we
think
actually
you
know,
adds
another
layer
above
this
and
sort
of
aggregates
this
together
and
makes
it
easy
to
for
developers
to
use
all
the
other
layers
below.
So
before
we
go
into
glue
mesh,
2.1
and
beyond.
I
wanted
to
also
cover
blue
mesh
2.0,
I'm
going
to
invite
narage
on
to
talk
about
the
features.
A
I
know
we
didn't
in
the
keynote
we
wanted
to
have
it
in
this
session
as
well.
Glue
mesh
2.0
has
deep
foundational
changes
that
allow
us
to
address
things
that
we
heard
from
our
users
from
the
one.x
you
wanted
multi-tenancy
support.
You
wanted
it
to
be
easier
to
add
new
clusters
without
any
reconfiguration,
and
you
wanted
to
have
fine-grained
policy
applied
to
this
multi-tenancy,
and
this
is
something
that
we
delivered
in
2.0
and
niraj.
Will
you
please
come
on
board
and
introduce
us
to
all
the
great
features
thanks.
B
So,
as
chris
mentioned,
we
are
making
foundational
changes
in
2.0,
both
in
our
architecture
and
apis
to
meet
a
great
number
of
requirements
for
our
customers.
We
are
also
focused
on
delivering
tactical
use
cases
so
that
we
can
say
that
glue.
Mesh
2.0
is
the
best
service
mesh
platform
out
there
in
the
market.
So
I'm
very
excited
to
show
you
all
the
amazing
enhancements
we
have
made,
along
with
the
redesign
of
our
api
and
architecture.
So,
first
off
we
have
added
support
for
hto
lifecycle
management.
B
What
I
mean
by
that
is
now
you
have
high
level
blue
mesh
apis
available
that
you
can
use
to
automatically
install
and
upgrade
stu
for
you
in
your
cluster.
This
was
a
major
requirement
for
many
of
our
customers,
who
did
not
want
to
take
the
burden
and
complexity
of
managing
a
steer
themselves
so
going
forward.
You
don't
have
to
deal
with
whether
you
have
to
install
the
steer
using
helm.
B
Is
your
ctl
still
operator
it's
all
hidden
away,
and
you
can
just
use
the
blue
mesh
apis
to
do
the
job
for
you
moving
on
in
2.1
and
2.2,
we
will
even
provide
support
to
take
over
your
existing
s2
installation,
so
many
of
our
customers
already
have
installed
stu,
but
I'm
sure
they
would
want
to
get
simplicity
in
the
environment
and
use
the
glue
mesh
apis.
So
we'll
allow
you
to
move
from
existing
s2
installation
to
using
the
hto
lifecycle
management
apis
from
glue
mesh.
B
We
have
made
a
lot
of
new
api
changes
in
2.0
to
ensure
we
add
multi-tendency,
and
one
of
the
core
principles
of
providing
multi
in
our
system
is
through
the
concept
of
workspaces.
So
I'm
going
to
show
you
next
written
very
briefly
what
workspaces
are
and
how
they
enable
you
to
get
multitenancy
in
gloomesh
2.0.
B
We
have
been
adding
and
continuously
improving
high
availability
and
scalability
in
our
in
our
system
in
2.0,
I'm
very
excited
to
say
we
have
support
so
that
you
can
scale
out
your
glue,
mesh
agents
running
in
your
workflow
cluster.
So
as
your
services
scale
in
one
cluster,
you
can
scale
out
your
agents,
so
that
will
make
sure
that
the
metrics
and
the
locks
that
they
are
collecting
from
your
services,
they
all
can
be
send
reliably
to
the
management
plan.
Moving
on.
B
We
have
also
added
the
capability
of
scaling
out
the
glue
mesh
management
plane
that
allows
you
to
go
beyond
multi
to
go
at
a
scale
of
multiple
clusters
from
tens
to
hundreds
of
clusters,
and
you
can
make
sure
if
you
scale
the
glue
mesh
control
print
correctly.
You
have
the
right
reliability,
you
still
have
the
same
performance
and
you
get
the
centralized
configuration
security
and
visibility
all
with
a
single
pane
of
glass.
We
are
going
to
continue
our
work
for
high
availability
in
future
releases.
B
We
are
going
to
even
add
support
for
beyond
one
region,
so
you
have
more
reliability
in
the
critical
blue
mesh
management
plane
cluster.
We
have
made
huge
enhancements
in
the
ui,
I'm
very
excited
to
share.
We
have
a
brand
new
ui
which
is
oriented
around
our
2.0
apis
and
concepts
like
workspaces.
I'll
briefly
show
you
some
of
the
designs
and
pages
we
have,
and
I'm
very
excited
to
share
that
work
with
you
all.
B
B
Many
of
our
customers
have
legacy
applications
running
nvms
that
they
are
migrating
to
kubernetes
or
they
might
never
migrate
to
kubernetes,
depending
on
the
complexity
of
the
legacy
app
itself.
So
what
we
provide
now
is
a
simplified
one-stop
command,
which
you
can
use
to
onboard
your
virtual
machine
to
the
glue
mesh
in
this
to
control
plane.
Basically,
after
using
your
command,
your
virtual
machine
will
have
the
right
files.
It
has
the
right
configuration
and
the
right
binaries
so
that
it
can
be
part
of
your
mesh
you're,
going
to
continue
our
support
here.
B
Virtual
machine
integration
in
future
release,
you
will
get
things
like
adding
policies
for
virtual
machine
from
the
glue
mesh
apis
and
being
able
to
observe,
what's
happening
in
terms
of
traffic
management
and
in
terms
of
requests
coming
in
on
your
virtual
machines
and
the
metrics
all
through
the
blue,
mesh,
dashboard
and
ui,
and
we
continue
support
to
make
sure
the
glue
mesh
gateways
are
having
the
same
filters
and
parity
with
the
glue
edge
will
continue
to
do
so
in
future.
Releases
too,
like
I
was
saying,
we
have
added
the
concept
of
workspaces
in
blue
mesh.
B
2.0
workspaces
are
the
key
logical
boundary
for
a
t1
application.
They
are
the
building
blocks
for
multi-tenancy
in
the
glue
mesh
product.
Basically,
in
a
workspace,
you
get
strong
isolation,
guarantees
for
configuration
and
service
and
traffic
management.
All
the
configuration
that
you
apply
in
workspace
cannot
affect
another
workspace
at
the
same
time,
by
default,
you
can
configure
it
so
that
only
services
within
one
workspace
can
talk
to
each
other.
B
So
with
this
concept,
where
a
workspace
spans
multiple
name
spaces
across
one
or
different
clusters,
you
can
draw
a
boundary
for
your
applications,
team
and
easily
and
safely
and
securely
onboard
them
the
glue
mesh
platform
at
the
end
of
the
day.
This
is
simplifying
your
user
experience
for
a
single
cluster
or
a
multicluster
setup,
because
you
have
added
very
flexible
label
based
selections.
B
B
We
have
optimized
the
data
plane
con
resource
consumption,
particularly
this
is
important
for
customers,
where
the
clusters
have
large
number
of
services
and
deployments.
What
that
leads
to
is
istio
automatically
configuring
the
sidecars
for
for
communication
throughout
the
cluster,
causing
a
lot
of
cpu
and
memory
overhead
with
the
concept
of
workspaces,
because
we
know
your
inputs
and
exports
and
your
service
connectivity.
We
automatically
create
the
sidecar
resources.
A
B
So,
like
I
was
saying,
we
have
a
brand
new
ui,
and
this
is
what
the
login
page
looks
like
you
can
sign
into
the
ui
through
and
integrate
it
with
a
single
sign
on
through
your
ldap
or
your
oidc
systems.
Once
you
log
in
you
will
see
the
screen,
which
gives
you
an
overview
of
the
different
workspaces
and
clusters.
B
You
can
see
your
destinations,
you
can
see
the
routes
you
can
see
the
policies.
Basically,
all
the
blue
mesh
resources
that
you
would
like
to
have,
that
is
attached
to
one
workspace
is
visible
right
here
in
one
spot
and
we
have
significantly
made
changes
in
our
graph
to
make
it
more
usable.
At
the
same
time,
we
organize
that
around
workspaces,
which
makes
it
more
scalable
so
in
here
you
can
see
that
the
bookstore
workspace
is
communicating
with
the
pastoral
workspace
and
all
the
services
are
logically
divided
in
that
neat
workspace
boundary.
A
Thanks
niraj
now,
let's
get
into
the
2.1
roadmap,
there's
a
lot
of
stuff
that
we
delivered
in
the
2.0
and
you're
going
to
see
a
lot
of
stuff
on
this
slide
in
the
2.1.
We
have
a
team
of
engineers,
an
army
of
engineers,
ready
to
take
this
on.
First
and
foremost,
our
customer
requests
rfes
that
we're
getting
from
you
are
users.
This
is
again
extremely
important.
We
triage
hundreds
of
issues
per
week
that
are
coming
from
you.
A
Secondly,
is
the
sdo
lifecycle
management?
We
did
part
of
this
in
the
2.0.
We
have
the
ability
to
upgrade
istio
with
glue
mesh.
We
have
the
ability
to
install
istio
with
glue
mesh.
We
want
the
ability
to
also
take
over
brownfield
istio
when
we're
walking
into
a
new
user.
You
often
have
istio
set
up
and
we
wanted
to
be
able
to
take
over
the
life
cycle.
A
Management
of
that
brownfield
istio
instance
within
glue
mesh,
replace
it
with
our
images
so
that
you
can
get
all
the
goodness
all
the
long-term
support
of
sdo
right
there
in
in
your
cluster,
we've
also
wanted
to
add
graphql
to
glue
mesh.
We
did
an
excellent
job
of
adding
that
to
glue
edge
and
we're
not
stopping
there.
We're
going
to
now
add
that
to
glue
mesh
and
glue
mesh
gateway,
there's
going
to
be
not
major
api
changes,
but
since
we
made
a
large
change
in
the
api
there's
some
cleanup
work.
A
There's
some
additional
refinement
that
we're
going
to
do
in
the
2.1
as
well
we're
going
to
get
feedback
from
you
we're
going
to
address
that
as
we
move
forward
and
then
one
other
thing
that
we're
doing
is,
we
did
high
availability,
partly
in
the
2.0
right.
We
added
scaling
ability
for
high
availability,
where
you
could
actually
have
more
instances
of
the
management
plane
like
classic
high
availability,
that
you
would,
you
know,
have
an
xcd
or
or
the
kubernetes
control
plane
or
the
istio
control
plane.
A
A
Next,
we've
added
a
bunch
of
observability
improvements
in
the
2.0,
as
you
saw
the
animation.
Some
of
this
is
cosmetic,
like
the
animations,
the
ability
to
see
the
latency
with
those
animations
the
crawling
ants.
I
love
it-
the
mtls
logo,
the
ability
to
move
nodes
and
clusters.
It's
all
wonderful.
We
want
to
not
stop
there.
We
want
to
add
more
features:
more
capabilities
to
observability,
including
seeing
your
virtual
machines
in
the
in
the
graph
tcp
traffic,
the
ability
to
see
ebpf
within
that
graph
we're
just
going
to
keep
on
improving
upon
it.
A
I
mean
it's
a
world-class
feature
right
now
and
we're
not
going
to
stop
next
we're
going
to
talk
about
breaking
glass.
This
is
a
common
request
that
we
got
from
you,
our
user.
A
lot
of
you
are
using
istio
today
and
you
like
our
api.
You
like
the
value
that
we're
bringing
with
glue
mesh,
but
our
api
is
not
going
to
ever
replicate
everything
that
you
would
do
in
istio
and,
furthermore,
you
were
sometimes
have
set
up
a
process
or
you're
using
the
istio
resources
directly,
and
you
want
to
continue
to
do
that.
A
Next,
let's
talk
about
virtual
machines.
You
saw
some
virtual
machine
features
in
the
2.0,
namely
we're
able
to
configure
the
virtual
machine.
This
is
important
to
you,
our
user,
because
you
don't
want
to
go
through
the
20
steps
to
do
this
in
a
manual
method.
You
just
want
a
single
command
line
where
you
get
the
configuration
so
that
you
could
put
the
sidecar
in
so
that
you
could
get
discovery.
A
A
Next,
what
we
want
to
do
is
take
that
virtual
machine
and
be
able
to
apply
the
glue
mesh
policies
and
that
the
ability
and
have
the
ability
to
see
that
virtual
machine
in
the
observability
platform
that
we
have
built
next,
I
want
to
talk
about
the
ui,
the
ui
man.
That
is
great.
I
love
it.
It
looks
so
fresh.
It
looks
so
enterprisey
if
that's
a
word.
It
has
a
left
hand
side
menu.
It's
very
easy
to
find
things.
I
love
it.
A
It's
just
so
great
a
big
shout
out
to
the
team
of
all
the
work,
hard
work.
They
did
there,
but
we're
not
going
to
stop
there.
We
want
to
add
the
ability-
and
we've
heard
this
from
a
lot
of
users-
to
have
write
capabilities
right
now.
The
ui
is
read
only
we
want
to
be
able
for
you
to
actually
change
the
configuration
directly
in
the
ui.
A
Now
I
want
to
talk
about
glue
mesh
gateway,
we've
delivered
a
whole
bunch
of
features
to
get
parity
between
glue
edge
and
glue
mesh
gateway.
If
you're
using
istio
glue,
mesh
gateway
is
often
the
first
choice
for
using
an
api
gateway,
we're
going
to
add
egress
to
the
platform
as
well.
This
is
a
common
request
from
ue
user.
This
is
very
straightforward.
It's
just
the
adding
egress
very
important
to
our
use
cases.
A
Next,
we
want
to
add
the
portal
capabilities.
We
built
a
excellent
developer
portal,
that's
highly
utilized
in
glue
edge.
We
want
to
build
that
in
glue
mesh
as
well.
Now
this
is
a
little
bit
nuanced.
We
have
a
lot
of
the
api
management
features
that
are
going
to
come
to
workspaces
the
ability
to
put
your
apis
in
there
and
manage
those
and
share
those
with
different
teams,
but
there's
another
level
right.
A
A
A
Also,
a
lot
of
users
wanted
to
have
what
you
would
probably
call
a
istio
native
api
gateway.
So
all
those
api
gateway
features
built
into
the
istio
ingress
itself,
and
also
we
had
users
that
you
know
like
the
api
gateway
features
and
thought
they
were
getting
all
of
that
with
istio,
and
there
is
some
overlap
you
could
see
in
this
slide.
Mtls
and
some
of
the
other
features
you
do
get
that
with
istio
out
of
the
box,
but
taking
those
basic
building
blocks
that
are
in
istio
and
building
an
enterprise.
A
Api
gateway
is
a
monumental
task.
There's
a
large
mounting
that
you
would
need
to
climb
in
order
to
you
know,
get
the
federation
to
get
the
external
authorization
and
authentication
and
build
a
complete
product,
and
so
this
is
what
we
wanted
to
do,
and
this
is
what
we've
done
with
glue
mesh
gateway
in
in
glue
mesh
and
we're
going
to
continue.
Adding
features
here
as
well.
A
A
This
includes
the
lambda
filters
that
we
built
that
have
enabled
it
enabled
users
to
move
easily
from
aws
api
gateway
and
just
put
in
glue
edge
gateway
as
a
very
simple
and
easy
replacement.
We
have
customers
that
are
completely
wholesale,
moving
50
or
more
teams
from
aws
api
gateway
to
our
api
gateway,
because
we
have
the
ec2
support
because
we
have
the
lambda
support
and
obviously
we
have
support
for
services
on
kubernetes.
A
We
have
customers
where
60
to
80
percent
of
the
services
in
their
organization
are
in
lambda
and
we
need
to
build
that
support
in
glue
mesh
gateway,
so
you're
going
to
see
that
in
the
2.1
and
the
2.2
and
versions
going
forward
again,
thanks
for
joining
me
on
this
roadmap
session,
thanks
for
being
with
us
since
the
1.0
and
this
whole
journey,
it's
been
excellent.
I
love
being
the
product
manager
here.
I
love
working
with
the
team
here,
the
engineers,
the
deep
istio
expertise
that
we
have
here.