►
From YouTube: Layer5 Community Meeting (August 21st, 2020)
Description
Welcome all new contributors!
A
We
record
these
calls,
so
you
know,
congratulations.
You
are
youtube
famous
well
or
at
least
as
of
86
subscribers.
I
guess
if
you
consider
that
that's
famous.
A
Hey
for
those
of
you
who
I
haven't
met
before
my
name
is
lee,
I'm
the
I
get
to
be
the
host
of
this
particular
call
in
the
layer,
five
community.
We
have
a
few
different
calls
each
week
and
I'm
pleased
to
not
be
the
host
of
a
couple
of
those,
and
so
should
you
two
be
pleased,
because
you'll
only
have
to
put
up
with
me
so
much
so
on
this
call.
A
A
Please
do
jump
into
the
meeting
minutes.
If
you
click
the
link
and
you
don't
have
access
just
request
it
I'm
going
to
share
the
meeting
minutes
momentarily.
A
And
say
that
these
meeting
minutes
are
like
everything
else
here,
a
collaborative
effort,
so
please
toss
your
name
into
the
list.
Please
help
take
the
meeting
notes
the
minutes
to.
A
A
A
No,
yes,
oh
okay,
all
right!
Well,
I
had
the
fortune
of
working
with
some
good
folks
in
oslo
as
cisco
had
acquired
a
telepresence
company,
forget
the
name,
but
they
had
some
great
telepresence
equipment
about
six
seven
years
ago,
so
random,
random,
all
right!
So
back
back
to
the
meeting
at
hand,
there's
there's
a
number
of
you
who,
I
think
haven't
been
on
a
call
before.
A
If
you
would
there's
a
rite
of
passage-
and
I
think
it's
a
good
one,
if
you
would
to
take
a
quick
moment
to
introduce
yourself
I'll
call
on
a
couple
of
names-
apologies
if
you've
introduced
before,
but
but
it
doesn't
hurt
to
say
hi
and
so
just
randomly
sumeet
nayak.
A
Nice
submit
very
good.
It's
also
nice
to
hear
people
enunciate
their
own
name.
So
I
can
stop
mispronouncing
george.
You
might
have
just
joined
the
call
right
in
the
moment
in
which
we're
asking
everyone
to
introduce
themselves
so
george,
if
you're
in
a
place
where
you
can
say
hi
and
just
kind
of
introduce
yourself,
please
please
do
lots
of
new
folks
on
the
call
so
nice
to
get
to
know
people.
D
A
A
A
Good
sorry,
more
folks
flowing
in
and
so
well,
sun
fumi.
A
Shooting
for
me
how
beautiful
they
are
great
good
to
have
you,
nice
bala,
mr
krishna.
E
Yes
lee:
what
are
you
doing
here?
I'm
calling
from
india
I'm
a
kubernetes
developer
and
I'm
an
aws
developer
also,
and
I
want
to
learn
about
something
like
or
in
a
messed
up
audio.
Like
is
the
one
mystery
for
that
purpose.
When
I
have
joined
here,
awesome.
A
Awesome
very
good
whereabouts
in
india.
A
All
right,
hey,
bala,
educate
me
briefly:
what's
the
deal
with
sometimes
bangalore
and
sometimes
belgaru.
E
A
Okay,
okay,
all
right,
fair
enough
good
to
have
you
bala
nice.
Thank.
A
A
B
B
Okay,
so
you
were
talking
to
me
right,
yep,
yep,
yeah,
oh
yeah,
so
in
the
last
meeting
I
got
to
introduce
myself
but
yeah.
I
will
introduce
again
fine,
so
hi
everyone.
I
am
because
there
are
so
many
new
guys,
I'm
saying
here,
hi
everyone.
I
am
dave
and
I
am
from
india-
and
I
am
here
to
learn
something
new
and
to
network
with
people
yeah.
So
that
is.
A
That
awesome,
yeah,
very
good
you
know
what's
what's
interesting-
is
that
this
is
the
dev
set
a
keyword
and
I'm
going
to
delay
the
start
of
the
meeting
for
even
a
little
bit
longer
just
to
to
speak
to
this,
and
because
because
it's
something
that
I'm
I'm
pretty
pumped
about
and
something
that
I
think
is
a
little
bit
unique
to
this
community,
and
that
is
that
so
so
kubecon
happened
this
last
week
and
kubecon
is
the
you
know,
is
a
significant
event
in
the
cloud
native
ecosystem.
A
It's
hosted
by
the
cloud
native
computing
foundation,
the
cncf
it's
sort
of
the
home
of
kubernetes,
and
then
excuse
me,
like
49
other
projects
by
the
way
measuring
one
of
our
bigger
projects
very
likely
to
be
the
second
project.
If
you
will
like
like
there's
a
lot,
we
engage
in
that
community
a
lot
anyway
on
the
slack
they
have
turned
on
a
bot
that
you
know
anytime,
that
someone
says
says
like
hey:
do
you
guys
know
where
to
go,
find
something
anytime
that
that
phrase
guys
is
used?
A
The
bot
comes
on
and
says:
hey,
you
know,
check
your
check.
Your
language
try
to
be
try
to
use
something
more
gender,
neutral
or
and-
and
I
gotta
say,
like
you
know,
of
the
many
talks
that
I
give
on
stage
or
and
having
my
you
know,
having
been
raised
by
my
mom
who
just
uses
that
euphemism
all
the
time.
Just
like
hey,
you
guys
do
this.
You
got
you
know,
for
it's
been
really
hard
for
me
to
try
to
check
my
language
and
and
anyway,
living
in
texas
right
now.
A
You'd
think
that
that
I'd
have
a
nice,
y'all
I'd
be
saying
y'all
every
every
which
way
I've
yet
to
kind
of
pick
up
on
it,
but
but
I'm
I'm
trying
to
because
because
of
what
is
actually
happening
in
this
community,
which
is
really
nice,
we've
got
kaleche
and
ruth
and
anita,
and
I'm
trying
to
pick
out
other
names.
But
but
just
we've
got
female
nerds
female
geeks
showing
up
here
and
actually
showing
up
some
of
the
rest
of
us.
A
So
to
speak,
and
I
recognize
it
ain't
easy
being
a
young
at
one
point
being
a
young
freckle-faced
pale.
You
know
nerd
myself
in
school
having
been
through
a
lot
of
trials
and
tribulations
over
the
last
couple
of
decades.
I
recognize
like
it's
a
word.
A
Not
all
communities
and
not
all
subject
areas
are
all
that
inviting
so
just
adeola
is
here
with
us
as
well
anyway.
I'm
just
I'm
super
encouraged
that
that
people
have
found
it
a
warm
environment.
So
it's
nice,
so
just
just
a
word
on
that
moving
on
quickly,
if
we
can,
who
else
is
here
new
that
didn't
get
a
chance
to
introduce
themselves.
D
G
A
Very
good
any
anyone
else
we'll
get
going,
get
going
on
the
meeting
minutes
here.
Some
some
good
things
to
share
busy
week
who
all
did
who
all
attended
kubecon
like
who
all
got
to
catch
the
like
a
keynote
or
or
get
in
on
the
free
hall
pass.
A
A
David
g,
now,
no
no
okay,
yeah
boy
anyway.
Coupon
is
kind
of
the
focus
of
our
initial
set
of
discussions.
Part
of
our
announcements.
It's
a
you
know
it's
an
important
event
for
the
technologies
that
we
focus
on
and,
as
such,
you
know
happens
twice
a
year,
and
you
know
we
try
to.
A
You
know
be
a
bit
forward-leaning
in
the
technologies
that
we
focus
on
here,
and
so
we
try
to
share.
For
my
part,
I
think
I've
spoken
at
the
last
eight
q
cons.
I
guess-
and
I've
been
really
pleased
to
be
able
to
share
the
stage
with
pratik
this
year.
Pratik
sahu
he's
a
phd
student
here
at
the
university
of
austin
texas,
he's
been
engaged
in
the
community
for
about
a
year
and
to
finally
be
able
to
get
some
of
his
research
out.
A
There
he'd
been
doing
some
studies
around
well
performance
analysis,
which
is
you
know,
part
of
what
the
mesh
ray
project
makes
its
name
on,
and
we
were.
A
You
know
acknowledging
that
there's
a
lot
of
intelligence
that
you
can
build
into
the
data
plane
of
a
service
mesh,
and
you
can
do
that
in
different
ways
and
so
understanding
that
yeah
and
so
understanding
that
there's
a
little
bit
of
a
cost
to
wherever
you
put
intelligence
wherever
you
put
logic
and
code
it'll
cost
you
anywhere
that
you
place
that
he's
been
trying
to
help
us
quantify
the
overhead,
the
latency.
A
You
know
the
overhead
of
logic,
that's
placed
into
the
data
plane
and
there's
different
ways
that
you
can
place
it
there
and
so
and
so
yeah,
so
he's
been
able
to
so
he's
got
a
great
talk.
I
feel
really
poorly
of
the
talk
that
he
had
given
the
these
are
pre-recorded
talks.
A
I
feel
I
feel
bad
because,
because
it
was
his
first
time
presenting
it
but
great
presentation,
there's
a
lot
of
data
and
info
that
he
and
a
number
of
the
rest
of
you
here
have
been
working
on
in
this
regard.
A
So,
while
there's
a
little
bit
of
that
data,
that's
published
on
that
page
you'll
also
find
I
think,
down
in
the
the
notes
from
prior
meetings:
a
debt,
a
full
deck
and
some
additional
info
that
might
be
of
interest
to
you
so
so
kudos
to
pratik.
On
his
first
kubecon
presentation,
we
had
had
a
couple
of
other
presentations
at
kubecon.
One
of
them
was
about
very
much
so
about
the
the
initiatives
that
are
happening
within
the
layer,
5
community.
There's
a
working
group
there's
a
special
interest
group
in
the
cncf.
A
It's
called
sig
network.
There's
a
a
working
group,
the
service
mesh
working
group-
that's
it's
a
young
working
group,
we're
just
formally
forming
it,
but
I'm
under
its
umbrella
or
in
its
field
of
focus,
is
a
few
initiatives
that
we
focus
on
here.
One
of
them
is
the
smi
service
mesh
interface.
A
Another
one
is
the
spec,
the
standard
that
this
community
has
been
working
on
for
a
while
at
service
mesh
performance,
and
so
we
we
kind
of,
went
through
at
cubecon
some
of
the
stuff
that
that
you
guys
that
y'all
geez,
I
just
said
it
there
right.
I
just
said
you
guys
have
been
working
on
and
we
got
to
tell
people
about
it.
So
it's
really
nice.
So
there's
there's
a
talk
out
there.
A
I
think
if
you
didn't
attend
to
qcon,
and
you
think
that
some
of
these,
some
of
the
things
that
are
presented
are
of
interest,
I
think
we're
all
fortunate
in
that
those
recorded
talks
will
be
posted
on
youtube
to
consume.
At
your
your
leisure,
there
were
two
other
individuals
in
our
community
who
have
been
kind
of
prominent
ant
weiss
and
natish
malhotra.
A
They
both
gave
talks.
Neither
the
two
of
them
are
on
right
now,
and
so
I
don't
want
to
misconstrue
their
their
talk.
The
links
to
their
talks
are
down
below.
A
There's
also,
you
know
at
cubecon,
there's
also
a
number
of
other
side
attractions,
other
smaller
conferences
that
excuse
me
that
go
on
one
of
those
is
called
service
mesh
con,
the
community,
the
layer
five
community
was
called
out
at
the
end
of
the
day
as
being
a
really
interesting.
A
I
think
you
know
a
community
and
projects
that
people
should
really
pay
attention
to,
and
that's
great
that
was
really
great,
also
called
out.
Pretty
prominently
was
smp.
The
service
mesh
performance
specification
that
we're
working
on
that
we
just
talked
about
it
was
called
out
in
context
of
smi
highlighted
as
a
real
area
of
interest
with
you,
which
is
nice.
A
I
know
that,
probably
for
the
majority
of
us
that
are
on
this
call
right
now,
smp
service
mesh
performance
that
specification
and
what
it
means
and
what
we're
doing
there
is
probably
entirely
foreign
which,
if
that's
the
case
for
you,
there's
two
things
to
do.
One
is
to
peruse
this
site
to
get
a
little
bit
of
familiarity.
A
A
Sometimes,
if
I
pause
long
enough
that
just
elicits
someone
to
say
forces
someone
to
say
something:
okay,
good
well,
the
event's
over
I
gotta
say
it
was
the
first
time
it
was
run
as
a
virtual
event.
A
Wait
it
just
wasn't
the
same
at
all.
They
you
know
great.
They
did
as
well
as
they
can,
but
just
virtual
conferences
are
a
different
beast
than
being
able
to
catch
up
with
people
in
person,
so
cool
all
right.
Well,
we've
got
some
other
topics
to
go
through
today,
anirud.
I
think
you
are
you're.
First.
H
Up
it
will
take
me
around
five
minutes
to
set
up
so
can
you
have
something
for
me.
A
Sure
yeah,
let's
let's
come
on
back
the
next
topic-
is
who
wants
to
talk
about
this
one.
I
Okay,
so
we
are
planning
to
add
a
good
first
issue
board
to
the
airfield
which
basically
will
tweet
out
about
any
issues
that
have
the
good
first
issue
tag
on
them
and
that's
good
publicity,
because
it
attracts
any
initial
or
new
contributors
to
the
f5
repo
and
its
projects
and
is
good
for
both
us
and
the
contributors,
because
it
gives
us
a
chance
to
help
people
get
started
and
gives
the
contributors
a
chance
to
get
involved
with
the
community.
I
We
have
an
open
issue.
A
link
is
provided
in
the
meeting
minute
setting,
so
any
thoughts.
J
Anybody
yeah
we
have
discussed.
I
think
we
have
discussed
this
good
first
issue
bought
on
development
call
yeah,
and
I
think
this
is
good
enough
to
add,
like
it
is
good
for
publicity
and
also
good
for
like
new
contributors.
I
Yep,
so
we
have
an
open
issue.
I
Nothing,
I
was
just
expanding
on
my
point,
so
we
have
an
open
issue
for
this.
If
anybody
wants
to
take
this
up,
you're
very,
very
welcome
to
again
we
have
a
link
in
the
minutes.
A
Nice
good
yeah,
it
would
be
even
better
if
that
twitter
account
had
more
followers,
but
either
way
so
long
as
we're
able
to
welcome
a
couple
of
first-time
contributors.
That's
fantastic!
All
right
kush
did.
Are
you
in
a
position
to
talk
about
the
smi
conformance
tool
today,
maybe
show
people
how
that
works.
K
A
Okay!
Well,
let's
I
know:
we've
talked
about
this
tool
a
time
or
two
on
other
calls,
maybe
we'll
chris
I'll,
if,
if
you
are
able
to
by
the
end
of
the
call
that
would
be
great
now,
if
you
can
delete
the
namespaces
or
just
or
just
walk
people
through
what
it's
supposed
to
be.
K
A
H
Yeah,
I
think
I
can
start
a
demo,
so.
H
It
is
okay,
yeah
right,
so
let
me
just
get
the
plan
on
what
we
are
trying
to
do
first,
so
you
have
a
bit
more
context.
So
let's
drive.
J
K
E
H
So
I
guess
the
doc's
visible
now
it
is
yeah,
so
what
we're
planning
to
do
is
add.
Currently
we
support.
We
support
system
commands,
that's
maintaining
measuring
life
cycle
and
configuring,
it
and
perf
to
create
call
performance,
benchmark,
cool
benchmarking
tool,
and
that's
that's
that
are
the
that's.
The
total
cli
support
right
now
we
have
and
we
are
planning
to
add
a
mesh
light
service,
mesh
life
cycle
management
and
for
that
we
have
the
mesh
command
so
with
what
currently
we've
had.
H
We
have
in
mind
is
being
able
to
deploy
a
sub
deploy
service
mesh
through
the
cli.
Currently
this
this
is
the
ui
for
that
and
you
can
actually
go
like
osm.
You
could
simply
go
and
deploy
a
service
mesh
to
this
like
deploy
open
service
mesh.
I
could
simply
do
this
and
that
would
deploy
a
service
map
for
me.
H
So
this
is
the
ui
part,
but
we're
planning
on
adding
cli
support
for
it
and
that's
currently
a
work
in
progress
and
the
other
sub
command
would
be
mesh
validate
where
we
could
actually
run
smi
conformance
to
test.
If
it's,
if
it's,
if
the
service
mesh
confirms
with
the
smi
spec,
so
these
these
two
are
a
work
in
progress.
As
of
now,
and
currently
I
have
a,
I
have
a
template
for
measuring
ctl
match
validate,
so
I'm
on
mexico.
H
So
it's
metroidl
mesh
validate
so
currently
it
it
doesn't
call
the
test.
H
So
there's
a
template
data
right
now
we
are
in
progress
of
making
apis
support,
output
responses
perfectly
right
now,
and
so
currently
we
we
are
using
a
template
here,
so
that
is
the
number
of
success
tests
and
the
number
of
field
tests
and
it
actually
checks
if
the
mesh,
if
measure
is
running
or
not,
and
so
this
is
more
on
a
mesh
validate
and
the
part
for
creating
osm.
So
this
this
you
can
you
can
this
postman
and
I
can
actually
send
an
host
request.
H
Currently,
this
is
how
I'm
deploying
deploying
a
service
mesh
on
a
mesh
tree,
so
I've
actually
called
in
for
I've
called
in
for
osm
driver
and
you
can
see
it
being
deployed.
So
that's
something
we're
working
on
integrating
this
into
the
service
measuring
ctl,
mesh,
create
and
yeah.
So
here
this
is
most
of
what
we
are
trying
to
complete
now
and
hopefully
in
I
guess,
one
to
two
weeks
we
might
be.
H
We
might
have
better
demo
than
this,
and
so,
if
like
it
doesn't
even
have
experience
with
this,
like
has
anyone
worked
with
any
api
client
in.
H
A
A
I
mean
there's
a
number
that
aren't
on
the
call
that
have
as
well
any
reading
part
well
you're,
making
a
call
for
like
for
input
and
participation
on
the
design
of.
H
So
we
do
need
help
with
improving
the
doc
for
measuring
skill,
as
well
as
planning
on
what
different
life
cycle
commands.
Do
we
add
currently
we've
thought
of
measuring
start
and
probably
mesh
delete
too
and
some
sort
of
some
sum
of
functions.
So,
if
there's
any,
I
like,
if
you,
if
you
guys,
have
any
idea
on
what
do
you
think
we
should
add
to
the
mesh
service
mesh
lifecycle
management?
A
Yep,
oh
very
good
and
then
you're
on
a
rude
anything
else,
any
other
anything
else
to
add.
There.
H
Yeah,
so
the
last
thing
here,
I
would
say,
is
like
if
anyone
wants
to
work
with
me
on
mercury,
ctl
or
measuring
ctl
mesh,
validate
or
perhaps
improve
the
dock,
you
could,
let
me
know
I'm
on
slack,
obviously
at
the
layer,
layer,
five
channel
and
just
let
me
know
if
you
want
to
help,
contribute
or
help
me
make
it
better
yep.
I
guess
that's
it.
A
Nice
very
good
yeah,
it's
it's
le
I'll,
reflect
for
just
a
moment
and
maybe
give
you
honored
something
to
noodle
on
as
rest
as
sort
of
the
the
rest
of
us.
I'll
say
this.
That
is,
as
you
familiar
well
as
many
of
us
familiarize
with
meshary
and
its
abilities,
and
we
think
about
the
word
validate
like
like
it's
interesting.
A
To
think
that
I
would
bet
that
there's
a
number
of
you
that
are
on
this
call
right
now
that
are
whoa,
and
maybe
I
should
just
speak
for
myself
but
sort
of,
but
scratching
your
head
and
saying
I
don't
know
what
the
hell
any
rude
is
talking
about.
Much
less.
Could
I
like
offer
up
any
meaningful
feedback,
and
that's
where
you
would
be
wrong.
A
That's
why
I
would
be
wrong
he
in
in
part,
because
I'll
say
this,
that
that
measuring
ctl
is
a
command
line
tool
and
it's
a
tool
that
all
of
you,
if
you're,
if
you're,
going
to
use
mesherie
that
you'll
interact
with
like
you'll,
do
a
measuring
ctl
start
or
something,
and
since
you
touch
it
just
like
you
touch
any
any
ui
like
this
one.
A
This
this
simple
site
that
we
have
that
talks
about
one
of
the
the
projects
in
the
community
that
since
you're
touching
it
in
this
case
with
your
mouse,
you
have
an
opinion
as
to
whether
or
not
you
like
the
feel
of
that
button
that
felt
good
or
it
didn't
it
took
too
darn
long
for
that
bootstrap
effect
to
like
take
a
play.
You
know,
whatever
your
opinion
is
whether
you're
educated
on
user
experience,
design
or
not
turns
out.
Your
opinion
is
valid
because
you're,
a
user
of
this
interface,
so
maybe
it
shouldn't,
be
yellow.
A
It
should
be
something
anyway.
The
point
is
with
meshrectl
you're,
a
user
of
mystery
ctl.
Maybe
you
don't
like
the
fact
that
it's
called
mesh
validate?
Maybe
that
doesn't
really
make
any
sense
to
you
and
here's
how
you
can
as
you
familiarize
with
measuring.
I
would
say
that
I
I
actually
ask
for
your
feedback
on
this,
for
those
that
do
want
to
get
or
are
already
into
meshri.
A
Do
you
are
you
running?
Is
your
service
mesh
configured
in
the
right
ways
so
today
that
validate
command?
Does
this
users
will
see
this
back
in
their
ui
they'll,
see
a
list
of
different
best
practice
checks,
saying
hey,
looks
like
you're
whoops,
you
hit
the
wrong,
you
know,
it'll
run
individual
checks
and
they'll,
say
yep,
yeah,
you're
doing
best
practices
stuff
or
not
you're
doing
it
wrong.
Do
it
this
way,
or
you
know,
it'll
make
a
suggestion,
and
today
we
use
the
term
validate
to
in
context
of
like
analyzing
config.
A
Well,
what
onyroot
is
proposing
is
that
there's,
a
new
form
of
validation
operation
that
that
new
validation,
operation
in
measuring
is
well,
is
actually
what
kush
was
considering
I'm
demoing,
and
it
is
that
mesherie
will
run
a
bunch
of
tests
against
a
service
mesh
that
is
set
up
to
be
smi
compliant
service,
mesh
interface
compliant.
A
This
is
the
service
mesh
interface,
it's
just
a
set
of
apis
if
a
service
mesh
says
that
it
conforms
to
those
apis
which
these
meshes
do
or
they
they
proclaim
to
well,
meshary
is
the
to
the
official
tool
that
will
be
used
by
the
smi
project
to
verify
that
oh
good,
and
so
as
these
projects
and
as
users
go
to
use
meshri
to
verify
smi
compliance.
A
Well,
some
people
are
going
to
use
the
ui
to
invoke
that
operation
and
some
people
are
going
to
use
the
cli
mastery
ctl
to
invoke
that
operation
and
and
the
reason
that
I'm
saying
that
anyone
any
individual
all
of
you
on
this
call
have
valid
feedback
is
as
a
user.
You
can
go
and
get
a
sense
of
how
this
feels
if
whether
or
not
this
is
intuitive
to
you
like
is
this.
A
Now
we
actually
haven't
exposed
that
in
the
cli
we
actually
don't
have
a
command
to
expo.
Invoke
that,
but-
and
so
since
these
two
things
these
concepts
are,
you
know
we
use
this.
Some
of
the
same
words
validate
we're
thinking.
Maybe
they
should
be
kind
of
co-joined
and
kind
of
together,
both
in
the
cli
and
in
the
ui,
but
so
so
an
open
question
and
a
call
for
comments
is
how
does
this
strike
people?
A
For
my
part,
one
of
the
things
that
it's
kind
of
missing
here
is
something
like
I
don't.
You
know
again
like
it
is
like
tight
type
console
or
something
like
that,
like
hey
generally,
people
are
going
to
want
to
run
like
they
will
execute
this
sort
of
command
in
context
of
the
mesh
that
they
have
deployed,
that
might
be
implicit
or
it
might
not
be
something
that
they
have
to
explicitly
define
the
shorter
the
commands,
the
better.
A
If
measuring
ctl,
already
has
the
concept
of
a
context
which
you
know,
those
that
are
anyone
who
uses
cube,
ctl
or
the
docker
cli
you're,
beginning
to
become
sort
of
familiar
with
with
context
or
how
you
might
switch
between
clusters.
So,
anyway,
all
you
don't
have
to
know
any
code
to
to
bring
up
these.
These
considerations.
A
I'm
totally
kidding
he's
not
there
but
anyway,
so
there's
links
to
this
stuff
that
the
link
to
the
messagery
architecture
is
here
kind
of
interesting
as
anyone
goes
to
get
familiar
with
meshi
it's
like
well,
we
were
talking
about
apis
and
clies,
and
so,
where
does
all
that
fit
in?
Well
again,
anyone
like
me
needs
pictures
with
big
round
objects
to
really
understand
anything,
and
so
those
are
there
for
your
reference,
I'll
pause
and
say
having
described
that
any
feedback
for
any
road
any
feedback.
A
Effort,
oh
yeah,
so
david
brings
up
a
good
question.
You
know
sort
of
in
context
of
what
we're
talking
about
here
about
you
know
in
ensuring
best
practices
and
kind
of
checking
on
things
david's.
You
know.
If
I
can
paraphrase
david's
question,
is
you
know
he
does?
Will
mesherie
highlight
the
fact
that
maybe
you're
experiencing
some
traffic
issues,
some
packet
loss
latency
that
you
didn't
really
want?
Maybe
you've
defined
an
slo
for
or
you
know,
a
particular
response
time
in
milliseconds
against
the
service
request,
make
sure
we
let
you
know
about
that.
A
And
the
answer
is
that
it's
that
it's
roadmap,
that
meshery
is
capable
of
measuring
like
more
than
capable
of
measuring
latency.
It.
A
Doesn't
it
doesn't
allow
you
to
specify
an
slo
or
a
threshold
by
which
you
would
want
to
know
about
a
slow
set
of
services
or
slow
response?
It
doesn't
do
that.
Yet
it
does
let
you
if
you
have
prometheus
or
grafana
deployed
in
your
environment
meshery
will
you
can
connect
measuring
to
to
prometheus
and
grafana
or
one
or
the
other
if
you've
invested
time
into
prometheus?
I'm
sorry,
grafana
dashboards,
grafana
panels,
mescheri
will
read
all
those
in
and
let
you
use
the
exact,
let
you
take
advantage
of
your
investment
into
specific
panels.
A
A
What
is
special
to
measuring
today
in
this
regard
is
that
it
will
it'll.
Let
you
easily
test
the
performance
of
or
the
response
time
or
the
throughput
of
a
particular
set
of
your
services
under
whatever
config
you're,
applying
to
the
service
mesh
it'll
measure,
it'll,
generate
load
and
measure
that
which
is
good
and
helpful
in
you
know,
probably
in
non-production
environments
or
on
a
you
know
or
on
a
separate
path.
A
A
It's
related
and
it's
a
good
it's
good
for
preparing
for
a
change
to
testing
out
what
what
a
configuration
change
to
your
mesh,
how
that
might
impact
the
performance
of
your
services
or-
and
that's
all
like
specific
to
your
environment,
specific
to
your
workload
specific
to
your
cluster
sizes,
specific
to
the
the
config
that
you
would
apply
to
the
mesh
and
what
version
of
the
mesh
you're
running
and
all
these
things
and
meshri
is
on
its
journey
to
trying
to
be
a
tool
that
provides
easily
repeatable
set
of
you
know,
tools
around
answering
those
questions.
A
Soon.
Mestre
will
have
performance
test
profiles,
so
you
can
go
out
and
describe
the
type
of
test
that
you
want
to
run
and
recall
it
very
quickly
and
have
that
run
and
have
that
programmatically
invokable.
Maybe
you
want
to
call
that
within
from
your
release
pipeline
and
have
a
certain
set
of
soak
tests
or
capacity
tests
or
whatever
I'm
invoked
from
there?
A
That's
good.
It's
part
of
the
answer.
The
other
part
of
the
answer
is
more
like
hey,
no
sort
of
real-time
assessing
of
the
environment
and
whether
or
not
things
are
going
haywire.
So
we
are
that's
like
the
that
type
of
use
cases
directly
in
our
crosshairs.
It's
actually
to
david's
question.
It's
exactly
what
our
this
presentation
at
cubecon
was
really
like.
A
The
underpinnings
of
it
are
about
is
that
at
some
point
we
will
be
will
provide
people
with
the
ability
to
deploy
an
envoy
filter,
probably
through
wassum,
that
does
these
measurements,
like
actually
watches
for
tracks
latency,
and
it's
not
that
you
can't
track
that
elsewhere.
It's
that
you
might
want
to
do
something
about
it
like
from
the
data
plane
right
then,
and
there
and
maybe
do
something
about
it,
based
on
who's
logged
in
or
or
like
whether
or
not
that
that
request
has
an
open
shopping.
Cart.
A
Maybe
if
that's
what's
important
to
you
or
maybe
those
those
individuals
are
signed
up
for
the
premiere
plan,
and
so
you
know
you
need
to
have
a
much
much
more
intelligence
in
your
qos
or
in
your
handling
of
retries
or
any
number
of
things,
and
so
like
anyway.
I
love
david's
question
that
there's
a
ton
of
things
that
we
want
to
do
in
and
around
that
question,
and
we
want
to
do
it
in
a
little
bit
of
a
novel
way
or
you
know
so.
A
So
any
rude
that
was
all
the
feedback
you
were
looking
for,
so
the
real
time
and
so
david's.
Another
good
question,
which
is
hey,
will
any
of
the
real-time
assessment
stuff
that
we're
just
alluding
to
make
it
to
the
cli
because
david's
totally
right,
like
there's
a
ton
of
values.
Yes,
is
the
short
answer
and
then
I'll
pontificate
about
that
for
a
minute
and
say
that,
like
in
part,
what
any
rude
is
was
speaking
to
about
the
programmatic
invocation
of
smi
conformance
tests.
A
Yeah
there'll
be
some
you
I
mean
this
is
my
estimation
that,
like
there'll,
be
some
users
that,
like
you
say,
have
said
hey
we
want
to
make.
We
want
to
use
that
abstraction.
We
want
to
use
smi,
we
want
to
make
and
we've
chosen
console
or
whatever
mesh,
and
we
want
to
just
make
sure
that
as
we
upgrade
console
or
as
we
upgrade
the
specs
on
smi,
that
we
that
things
are,
you
know
all
good
that
we're
still
that
infrastructure
is
still
in
compliance.
Great,
sometimes
they're
going
to
want
to
do
that.
A
The
service
meshes
themselves,
they're,
absolutely
going
to
want
to
do
that
all
the
time.
Every
time
they
make
a
release,
which
means
that
they're
going
to
be
baking
measuring
into
their
pipelines,
which
means
to
david's
point
they're
going
to
want
to
be
able
to
invoke
that
programmatically.
They
can
either
hit
measures
rest
api
to
do
it
or,
I
suspect,
in
a
ci
pipeline,
they're,
probably
going
to
feel
just
as
comfortable
scripting
around
mastery
ctl,
the
cli.
A
So
to
that
same
use,
case
david's
question
about
some
of
the
real-time
assessment
stuff
making
into
the
cli
like
yeah.
I
guess
I
already
said
it
yes,
that
some
of
that
is
because
people
are
going
to
want
to
do
they're
going
to
want
to
do
that,
they're
going
to
have
that
type
of
control
at
their
fingertips,
both
in
production
environments
and
production,
use
cases.
It's
like
right.
A
How
am
I
doing
right
now,
like
I
want
to
tap
into
this
thread,
maybe
specific
to
this
custom?
This
token,
this
jwt
or
whatever
it
is,
and
get
that
assessment,
and
that's
going
to
be,
I
think,
a
set
of
fantastic
tools,
not
only
that
but
they're
going
to
want
to
do
the
thing
we
talked
about
before,
which
is
well
we're
they're
planning
a
change.
They
want
to
do
a
new
release
of
just
just.
A
They
want
to
change
the
configure
the
mesh
or
find
all
their
infrastructure
staying
the
same,
but
they're
going
to
roll
out
a
new
version
of
their
workload,
they're
going
to
automatically
invoke
those
real-time
assessments
you
know
to,
and
the
thing
is,
is
like
that
real-time
assessment
of
a
plan
to
change
well,
hey,
if
you're,
not
if
traffic
isn't
going
to
those
new
services,
then
it's
kind
of
hard
to
measure
take
a
real-time
assessment
of
the
impact.
Well,
guess
what
mescheri
is
a
load
generator?
A
It
has
three
different
types
of
load
generators
in
it
and
it
can
do
that
so
david,
don't
don't
get
me
started
man,
we're
you're,
not
good
good!
Thank
you
for
that
anirud
kush.
A
Understood
very
good
all
right.
We
have
two
other
topics,
one
that
I'm
really
excited
to
tell
you
guys
about.
If
I
don't
tell
you
today,
it's
fine,
it
can
wait.
There's
some
nice
analysis
being
done
on
mesheri,
but
before
that,
ediola
has
has
a
topic
for
us.
L
Can
you
hear
me?
Okay,
yeah
yeah.
Thank
you
lee.
Well,
it's
not
really
a
topic
per
se,
so
I
wanted
to
say
thank
you
to
leah5,
first
of
all,
because
I
started
my
open
source
contribution
journey
with
them
and
he's
been
bringing
a
lot
of
opportunities,
and
I
know
that
we
recently
completed
an
open
source
challenge
from
chico
africa
and
I
think
I
myself
myself
ruled,
I
think,
precious
as
well
so
yeah.
L
L
We
could
introduce
layer,
5
project
to
first
time
contributors,
and
I
know
that
the
community
is
very
welcoming
and
I
was
like
okay,
that's
a
good
opportunity,
for
you
know
the
community
to
grow
and
also
opportunities
for
people
who,
like
myself,
want
to
start
contributing
to
open
source
project,
and
I
think
that
layer,
5
is
a
very
good
place
for
them
to
start,
and
I
know
that
most
of
them
will
continue
and
you
know
be
committed
to.
So
we
have
a
lot
of
interesting
projects
here.
L
So
the
reason
why
I'm
saying
it
is
because
I
won't
be
able
to
do
it
alone
right,
so
I
wanted
to
see
or
ask
the
community
if
anybody
would
be
willing
to.
You
know,
support
on
that
day,
because
we
have
to
you
know,
introduce
the
project
to
them.
Those
who
select
layer,
5
project,
introduce
it
to
them.
You
know
just
walk
them
through
how
they
can
contribute.
L
You
know
so
the
main
aim
of
that
day
is
to
just
be
amazon,
so
there
will
be
like
a
is
a
virtual
conference,
so
there'll
be
like
a
slack
group
or
a
project
platform
where
people
who
are
contributing.
We
ask
questions
who
guide
them
will
check
if
they're
contributing
well
so
majorly
what
most
of
what
we
do
at
layer,
5.
L
A
A
Okay,
nice
anyway
idiot
like
no
doubt
that's
exactly
what
ruth
and
kaleche
and
anita
and
yourself
and
like,
and
a
lot
of
you
here
have
been
doing
what
I
think
is
a
really
impressive
job
of
both
paying
it
forward
for
others
and
learning
yourself
at
the
same
time,
and
just
you
know
doing
doing
the
right
doing
the
thing
that
is
so
often
not
done
in
other
communities,
which
is
which
is
to
to
highlight
folks
like
usama
and
george
and
many
of
the
other
people
who
introduce
themselves
today
and
and
just
and
be
encouraging
so
yeah
boy.
A
You
know
with
how
frustrating
it
can
be,
which
means
that,
like
if
people
are
raising
their
hand
to
volunteer
it'll,
be
really
impactful.
I
think
we'll
have
less
dead
computers
around
the
world
because
of
that
and
more
people
being
successful
with
open
source.
So
please
reach
out
to
idiola
and.
A
A
Yeah
minute,
before
we're
we're
wrapped
for
the
day,
anyone
have
any
items
that
we
we
missed.
A
Nice
ruth,
I
think
you
I
couldn't
really
hear
you,
but
I
think
you
were
maybe
pointing
out,
is
that
anisat
that
hasn't
had
a
chance
to
introduce.
I
A
L
A
It's
always
the
network.
Anyway,
it's
it's
awesome
and
it's
not
great
to
have
you.
Thank
you
so
much
for
having
an
easily
and
uns
annunciated
name.
We
will
get
along
very
well.
A
Cool
we're
one
minute
over,
I'm
gonna,
I'm
gonna
do
this
for
30
seconds,
because
I'm
just
I'm
so
tickled
about
it.
I
just
can't
I'm
gonna
pee
myself.
I
think,
and
that
is
that
you
know.
One
of
the
things
we
were
able
to
to
accomplish
this
week
is
get
a
little
bit
of
the
word
out
about
the
the
works
that
that
all
of
you
are
doing.
A
There's
a
there's,
a
relatively
small
analyst
firm,
called
amalgam
insights,
it's
they're
kind
of
like
a
gartner
they're
like
a
451
or
a
forester
and
they're
doing
they're
doing
a
bit
of
research,
the
the
particular
researcher
I'm
here,
his
name's
tom.
A
He
had
done
some
research
on
service
mesh
space
in
the
past
and
he's
kind
of
writing
up
one
about
kubernetes
and
the
evolution
of
kubernetes,
and
I
wanted
to
point
out
that,
like
four
times
in
his
this
is
a
draft
of
his
report,
and
so
it's
he's
asking
for
feedback
and
review,
but
like
four
times
in
his
report,
he's
calling
out
in
this
case
he's
calling
out
mescheri.
It's
like
you
know.
The
open
source
project
measuring
is
all
yeah
anyway
I'll
next
week.
A
We'll
kind
of
go
through
some
of
the
other
comments
and
things
I
just
wanted
to
say
that
people
are
really
paying
attention.
There
was,
if
you're,
in
the
cncf
slack,
which
is
open
to
anyone.
If
you
don't
have
the
link
nikhil,
anyone
can
send
it
to
you,
but
you're
welcome
to
join
you'll,
see
people
talking
about
your
projects
as
it's
super,
it's
something
to
be
proud
of.
So,
okay!
A
Well,
that's
it
cubecon's!
A
wrap!
This
meeting's
a
wrap
same
time
next
week,
really
encouraged
to
see
all
the
new
faces.
A
Please
stick
around
and
put
put
anita
and
ruth
and
kaleche
and,
and
everyone
here
to
work
they're
here
to
help.
So
all
right
same
time.
Next
week,
we'll
see
you.