►
From YouTube: CNCF Service Mesh Interfact Project Meeting 2020-09-02
Description
CNCF Service Mesh Interfact Project Meeting 2020-09-02
A
B
A
B
The
yeah,
the
planning
earning
service,
meshes
that
we
we
have
for
a
couple
of
years,
a
kind
of
in-house
service
mesh,
and
we
are
looking
at
on
the
the
market
to
maybe
move
to
an
off-the-shelf
solution,
and
and
in
this
evaluation
process
I
have
discovered
smi.
I
I
would
say
I
like
it.
The
the
support
by
the
vendor
is
not
up
to
the
level
I
would
like
it
to
be
so
I
can
easily
compare,
but
that's
that's
the
whole
story.
I
guess.
A
D
F
A
E
E
A
Yeah,
that's
fair
cool,
so
next
one
the
first
community
blog
post,
it's
hopefully
just
the
first
and
not
the
last.
So
let's
congratulations,
that's
already
great
and
hopefully
many
more
to
follow
anyone
here.
Who
has
an
idea
like,
for
example,
patrice
if
you
want
to,
I
don't
know
how
much
you
can
share,
but
kind
of
like
us
like
we
would
kind
of
like
be
all
over
you
and
ask
you
about.
A
B
C
B
E
B
A
It's
important
because
so
many
folks
at
the
current
point
in
time
are
evaluating
and
are
kicking
tires
and
trying
to
figure
out
coming
up
with
a
framework
how
to
actually
look
at
stuff.
So
it's
super
important
if,
if
you
can
share
I'm
aware
of
that,
you
know
very
often
this
cannot
be
shared,
but
if
you
can
share
that
thing
like
this
is
the
way
how
we
look
at
it
and
there's
no
right
and
wrong.
It's
like
it's
your
way
right.
You.
A
E
B
On
on
on
this
one
specifically,
I
have-
I
have
indeed
commented
that
yeah,
it's
it
says
access
everywhere
and
then
suddenly
you
have
this
target
prefix
and
I
think
it
would
it's
definitely
a
good
idea.
Now,
of
course,
I
don't
know
what
are
the
implications
of
that,
but
but
yes,
that
that
would
sound
terribly
more
consistent
to
call
it
traffic
access
instead
of
traffic
targets
and
same
for
traffic
split.
A
Cool
yeah,
I
think,
all
the
usability
ux.
You
know
what
what
it's
not
necessary,
that
that
people
who
write
specs
have
a
wrong
perception
or
whatever
it's
just
is
that
what
most
people
who
are
in
that
area
consider
as
the
most
natural,
the
most
self-explanatory
term
or
not?
So
you
know
all
that
input
is
super
super
useful
right.
A
E
I
don't
think
we
have
anyone
present
who
is
actively
working
on
that
which
is
okay.
I
just
want
again
I
I
was
looking
at
the
last
couple
of
meetings
and
I
was
like
we
probably
don't
want
to
forget
that
somebody
is
asking
about
that.
Maybe
kind
of
an
action
item
that
I
can
take
going
forward
is
try
to
figure
out
who
is
going
to
work
on
that
going
forward.
E
D
Yeah
I
mean
I'm
interested
in
that
because
I'd
like
to
have
maybe
like
an
uber
demo
of
all
the
providers.
If
that
makes
sense,
and
then
I
was
going
to
ask,
I
noticed
that
that
istio
adapted
it's
in
a
separate
like
repo,
is
that
due
to
is
there's
like
some
ci
testing
or
something
is
there
a
reason?
It's
like
spawned
off.
D
D
B
I
was
just
saying
I
do
have
experience
with
with
console
not
with
istio
but
for
concert.
That's
the
same
as,
as
you
said,
bridget
it's
you
have
the
console
repo
and
then
they
you
have
a
another
smi
console
adapter
or
something
like
that,
which
kind
of
translates
the
the
smi
api
into
the
native
console.
One.
A
B
A
For
for
kind
of
like
how
we
would
structure
the
I
I
don't
know,
maybe
we
do
and
I'm
not
aware
of
it.
That
could
perfectly
be
the
case
in
terms
of
how
we
structure
the
the
overall
github
organization
right,
if
we
have
one
per
whatever,
but
do
we
want
something?
Do
we
like
from
from
working
backwards
from
a
user
coming
to
it
like?
What
would
you
expect?
Would
you
like?
D
No,
you
know
what
I
mean
I'll
give
my
two
cents,
I've
been
kind
of
talking
to
a
lot
of
filled,
people
etc,
and
I
think
the
you
know
when
we
show
off
some
of
this.
Some
of
this.
You
know
people
who
are
kind
of
working
in
more
of
like
a
homogeneous
environment.
They
don't
understand
the
value
of
smi
and
I
think
if
we
can
have
some
examples
right
next
to
each
other,
showcasing
the
different
providers
and
saying
hey
here's
the
common
api.
D
I
think
the
light
bulb
comes
on,
I'm
just
finding
that
a
lot
of
people
who
like
dedicated
their
life
to
like
istio
et
cetera
they
when
you
show
them
this
they're
like
well,
I
don't
get
it
you
know,
but
then
you
say:
hey
look!
Here's
a
way!
If
you
have
a
common
tool
chain
with
an
api
across
you
know
multiple
measures,
but
but
right
now
I
think
what's
happening
it's
it's
hard
to
do
a
search
and
get
that
kind
of
consolidated
view
of
kind
of
what
smi
is
touching
and
I
think
that's
missing
right.
A
I
used
to
be
active
in
w3c
many
years
ago,
15
years
or
whatever,
and
there
essentially
in
order
to
graduate
to
tr
essentially
was
the
requirement
to
have
three
interpretable
implementations
right,
so
you
could
not
be
an
attack.
Tr,
I
believe,
was
the
highest
one
without
showing
you
have
three
interval
implementations
and
that's
something
that
I
I
think
makes
it
it's
not
about
a
shootout
right,
but
it's
not
about
oh,
like
I'm,
covering
more.
It's
showing
that
actually
look.
This
spec
does
something
for
you
exactly.
You
can
move
stuff
around.
You
get.
A
That
shows
that
value,
I'm
absolutely,
I
would
also
be
you
know,
I'm
more
than
happy
not
in
the
driving
seat,
but
if
you
have,
if
you
would
own
that,
I
would
be
more
than
happy
to
do
my
share
in
terms
of
like
you
know,
whatever
is
necessary.
Implementing
writing
dogs.
Whatever
is
necessary,
just
kind
of
really.
E
C
A
E
I
have
a
suggestion.
My
suggestion
is,
someone
should
decide
if
they
want
to
moderate
on
september.
16Th
michael
has
done
a
very
good
job
of
showing
us
what
that
looks
like
and
someone
else
could
decide
if
they
want
to
take
notes
on
september
16th
and
I've
been
doing
that
and
the
notes
are,
I
will
drop
them
into
the
chat.