►
From YouTube: Layer5 Community Meeting (July 2nd, 2021)
Description
Layer5 Community Meeting - July 2nd, 2021
Join the community at https://layer5.io/community
Find Layer5 on:
GitHub: https://github.com/layer5io
Twitter: https://twitter.com/layer5
LinkedIn: https://www.linkedin.com/company/layer5
Docker Hub: https://hub.docker.com/u/layer5/
A
Welcome
everybody
to
the
left
community
meeting
today
is
june
july
2nd,
and
we
have
a
couple
of
items
lined
up
today
and
we
have
a
lot
of
newcomers
here
as
well.
So
we
have
a
tradition
to
to
let
everyone
who
is
on
the
call
introduce
themselves
if
they
are
here
for
the
first
time
so.
A
Hungli
tan,
am
I
pronouncing
it
correctly?
Would
you
like
to
introduce.
A
Yourself,
if
you
have
microphone
problems,
then
you
can.
B
Yeah
yeah
yeah,
hi
hi.
Everyone
would.
B
Yeah
my
name's,
I'm
from
hanoi,
vietnam,
I'm
a
freelancer
developer
and
a
community
builder
yeah
nice
to
me
meet
all
of
you
here.
D
Very
good,
very
good
riddle
me
this.
If
you
would,
is
it
or
sorry,
it's
a
bad
phrase
at
what?
What
time
is
it
for
you
is
it?
Is
it
still
friday?
Are
we
on
saturday
yeah?
It's.
B
Still
friday,
it's
10
10
p.m.
D
All
right
any
other
newcomers
on
today's
call.
E
D
So
honey
we
were
just
on
a
call
like
an
hour
ago
and
we
really
could
have
used
you,
but
we
were
oh,
my.
E
D
We
were
having
trouble
with
mini
cube
and
using
the
docker
driver
and
the
networking
that
comes
with
that,
and
so
it's
refreshing
that
you're
here,
it's
nice
nice
to
have
you
good
to
hear
devops
is
it's
kind
of
been
your
focus?
That's
great.
A
You
any
others
who
any
other
newcomers
on
the
call.
Would
you
like
to
introduce
yourself.
A
G
Yeah
yeah
hello,
I'm
a
third
year
engineer.
I'm
currently
third
year
of
my
engineering,
I'm
pursuing
computers,
degree
I'm
freelancing
currently
and
I'm
looking
forward
to
learning
new
things
in
that
by
community.
D
Nice,
a
news
jubilee
gibril,
is
one
of
the
community
community
members
here
and
he
was
noting
that
it
can
occasionally
be
well
embarrassing
to
have
to
introduce
and
sometimes
it's
not
and
and
the
thing
is
is
like
when
it's
not
that's
when
we
need
to
change
things
and
figure
out
how
to
embarrass
you,
and
so
I
just
didn't
hear
anything
embarrassing
just
now.
So
I
like,
could
you
tell
us
maybe
what
your
favorite
color
is
or
who
your
favorite?
A
You
anyone
else.
C
Was
that
just
me
or
I
couldn't
hear
him
yeah
we
I
couldn't.
D
Boy,
I
totally
missed
it.
Who
is
just
introducing,
if
you
don't
mind,
can
you
do
it
again.
I
Hi
guys
krishna
here
I
have
come
up
here
by
winning
and
hackathon
event,
and
I
usually
dwell
with
machine
learning
and
few
computer
vision
data
sets,
and
I
am
very
glad
that
I
am
able
to
learn
so
many
concepts
here
and
would
be
great
to
learn
few
things
here.
Thank
you.
D
I
J
Yeah,
hello,
this
is
bharat
and
adding
to
what
krishna
said.
I
was
in
the
same
team.
J
D
D
D
Cool
great,
I
use
your
voice
was
really
low,
we'll
say
hi
in
chat.
If
we
can
yeah
nice
all
right
anybody
else,
I
feel
like
there's
one
more.
D
K
K
Okay,
I'm
taking
the
student
from
sdit
and
I
came.
I
called
a
chance
to
join
this
community
because
I
won
my
team
won
second
prize
in
jit
hack
and
I'm
here
to
donate.
A
D
It
as
36,
let's,
let's
leave
it
at
36,
that's
that's
great,
and
by
the
way
it
isn't
really
about
the
number.
That's
probably
the
only
way
of
kind
of
measuring
it,
it's
about
well
for
me,
and
it
could
be
different
for
all
the
rest
of
you
and
just
for
my
part.
Well,
it's
about
getting
things.
Learning
things
getting
things
done,
sharing
what
you've
learned
taking
on
new
challenges,
challenging
others
challenging
yourself
getting
your
name
put
in
lights.
D
You
know
it's
just
it's
actually
about
a
lot
of
things.
The
community
itself
is
about
a
lot
of
things.
One
of
the
ways
in
which
to
describe
the
community
is
well
as
a
platform
for
shared
success,
and
so
my
one
piece
of
advice
with
respect
to
how
to
be
successful,
at
least
in
this
community,
is
to
go.
Get
it
like
be
tenacious,
go
grab,
something
it's
an
extremely
rare
situation
that
the
word
the
word
no
is
used.
When
someone
says
I
I
would
like
to
do
something
so
don't
hold
yourself
back
good
go.
A
Yes,
okay,
I
I'll
move
that
next
agenda.
So
first
I
will
start
with
the
updates
on
measuring
and
smp
onboarding
into
the
ctf.
So
we
have
created
a
new
rapport
new
organization
for
measuring
and
moved
all
of
the
measuring
projects
into
this
repository.
So
recently,
mastery
and
space
match
performance
have
been
adopted
into
cncf
cloud
native
computing
foundation.
So
we
have
a
new
repo
now
and
also
for
smp
and
the
onboarding
is
going
well
and
we
have.
A
We
still
have
to
check
a
couple
of
items
in
this
list
for
both
the
projects
and
yeah.
Once
once,
we
have
done
that
we
will
be
officially
in
the
cncf
and
we
have
also
updated
the
cncs
landscape,
so
mushri
and
smp
would
be
under
service
meshes
in
the
service
category
and
yeah
vmware,
sandbox
budget
and
yeah.
L
Tank
in
a
window,
so
hello,
everyone,
we
have
a
thing
called
as
block
kitchen.
So
what
we?
What
I
just
tell
you,
what
it
it
is
for
and
what
we
usually
do
in
it.
So
basically
you
might
have
seen
okay.
Let
me
just
share
my
screen.
D
To
interrupt,
I
didn't
realize
this
happened.
We
skipped
right
over
one
of
today's
announcements
and
right
into
the
topics,
and
so
I'm
obligated
to
interrupt,
and
that
is
to
say
well
adi-
is
our
latest
mesh
mate,
so
congrats?
Thank
you.
Thank
you.
Yeah
thanks
lee,
it
had
been
a
while
coming
adi.
I
think
you've
had
over
the
last
week
week
and
a
half
you've
had
any
number
of
like
I.
I
think
you
might
have
peaked
out
at
a
young
age.
D
Like
I
don't
know,
I
don't
know
if
you
can
have
a
bigger
week
than
what
you've
had.
So
if
you
have
so,
if
you
have
other
news
to
share
as
well,
please
do
but
but
congrats
on
becoming
a
mesh
mate.
That's
fantastic!
It's
just
it's!
It's
a
it's
a
role
and
a
badge
that
you've
earned
that
you've
been
doing
so.
L
Kudos,
thank
you.
Thank
you.
So
yeah
this
week
has
been,
I
would
say,
pretty
pretty
happy
for
me
because
I
got
into
red
hat
and
I
became
a
messmate
and
then
I
got
a
lyft
scholarship,
also
so
yeah
it's.
I
would
say
that
it's
a
hat-trick
thing
for
me.
D
D
On
your
way
to
well
from
the
from
near,
as
I
can
tell
like,
potentially
on
your
way
to
into
a
maintainer
role,
the
stuff
that
that
you
and
and
the
have
been
working
on
on.
D
L
Yeah,
so
speaking
of
blogs,
we
have
a
thing
called
as
a
blog
kitchen,
if
you
guys
have
seen
it
in
slack.
So,
okay
yeah,
we
don't
have
any
updates
before
june
14th
anyways
yeah.
What
we
usually
do
is,
as
you
can
see
on
my
screen.
L
We
have
layer,
5.,
ios
blog,
where
we
usually
deal
with
new
blogs
regarding
any
updates
of
layer,
5
projects,
missouri
the
new
interns
and
if
there's
a
new
maintainer
and
all
we
also
deal
with
new
blocks
when
a
new
release
of
misery
comes
probably
like
a
major
release.
For
example,
another
vlog
might
come
up
soon
for
version
6.0
when
it's
released
so
yeah
and
other
some
of
the
other
stuffs
where
which
happens
is
regarding
kubecon
or
docker
car
or
anything,
and
to
deal
with
everything
and
mapping
of
all
the
blocks.
L
We
have
something
called
as
content
calendar,
so
what
it
basically
does
is.
It
is
like
a
place
where
we
can
track
who's
writing
what
blog,
whether
he
needs
any
he
or
she
needs
any
help.
What
and
all
topics
he's
going
to
write
about
and
where
he
can
find
resources
to
make
his
blog
amazing
and
so
yeah?
That's
about
block
kitchen,
and
we
have
some
pretty
good
vlogs
coming
coming
up
soon
by
the
contributors
of
layer,
5,
so
yeah,
I
would
say
if
you
guys,
are
very
much
interested
in
writing
blogs.
L
You
are
very
much
welcome
to
take
any
of
the
topics
which
is
in
green
and
if
you
have
any
new
idea
that
you
need
to
write
a
blog
about
on
some
of
your
personal
platforms
or
anything.
We
welcome
that
also,
and
if
you
have
anything
in
mind
that
you
need
to
write
and
submit
to
layer,
5
blogs,
you
can
do
that
as
well.
You
can
ping
any
of
us
in
the
blog
kitchen
channel,
so
yeah,
I
would
say,
welcome
to
blog
kitchen
and
hope
to
see
you
guys
cook
some
new
blogs.
D
H
D
Yeah,
that's
that's
good!
There's
some
prior
examples
up
there
as
well
of
what
some
people
have
written
about,
there's
actually
a
lot
of
introductory
material
to
how
to
use
measuring
or
how
to
use
or
other
aspects
of
the
projects.
D
Here
we
really
haven't
written
up
like
like
just
some
of
the
nuances
of
measuring
ctl
systems
start
like
almost
all
of
you
have
run
that
command
or
you
will
soon,
and
some
of
you
had
a
really
pleasant
experience,
and
some
of
you
probably
have
had
stumbled
into
things
and
so
there's
a
lot
of
other
people
who
stumble
into
those
same
things.
D
So
writing
up
a
post
about
that
is
like
like
there's
no
post
to
there
are
blog
posts
too
small,
too
short,
but
not
too
small
in
terms
of
their
technical
content,
not
everyone
can
write,
something
like
michael
wrote,
up
on
the
the
mesh
kit
and
mystery
adapter
library
post,
but
like
this,
this
is
antiquated.
D
In
fact,
it's
like
it's
great.
I
updated
part
of
it
recently,
just
because,
but
it's
from
a
long
time
ago
and
like
the
commands,
have
all
shifted
around,
so
we
could
use
a
cheat
sheet.
That
would
be
great,
but
there
are
a
number
of
people,
as
a
matter
of
fact
in
intel
has
been
hounding
recently
for
like
for
the
for
an
answer
to
this
question,
which
is
how
do
you
do
performance
management
in
measuring,
like
how
do
you
run
a
performance
test
and
they're?
D
Looking
for
just
like
a
simple
video
that
walks
them
through
that
and
there's
around
20
engineers
in
a
team,
that's
focused
on
service
mesh
at
intel
and
they're
they're
waiting.
They
asked
a
couple
of
weeks
ago
so
like
it's,
that
sort
of
a
thing
that
can
be
helpful
as
well,
especially
if
you
go
to
record
that
you'll
learn
a
lot
in
the
process,
so.
D
All
right
well
did
next,
what
what
are
we
talking
about.
L
Yeah,
hello,
everyone,
hello
again,
so
we
have
some
great
participants
and,
of
course,
the
winners
from
jit
hack,
that's
bharat,
krishna,
gahan,
kanis,
kusal,
shruti,
satish,
raya
city
and
a
couple
of
others.
Sorry,
if
I
missed
your
name
and
see
coming
to
your
hackathon,
sorry,
your
internship
specifically,
so
we
will
be
dealing
with
mainly
misery
ui,
so
you
guys
will
be
working
on
a
cnc
of
project
itself
now,
which
is
pretty
amazing
and
yeah.
L
The
internship
will
be
for
a
period
of
three
months
starting
from
today
itself,
or
let's
do
one
thing
it
would
be
from
monday
itself
and
your
internship
manager
will
be
navindo
and
mentors
me
gavrao
and
ayus,
and
some
of
the
other
things
which
you
will
be
contributing
is
one
thing.
I'm
pretty
sure
that
I
will
say
now.
L
You
guys
have
to
contribute
to
missouri
ui,
that's
the
flagship
thing
which
we
have
going
on
and
you
guys
will
be
writing
blogs
before
the
start
of
an
internship
middle
of
an
internship
as
well
as
after
your
internship,
you
will
be
working
with
some
of
the
latest
technologies
like
nexjs
react.js,
and
if
you
are
interested
in
designing,
you
guys
will
work
on
figma
and
yeah.
L
So
I
guess
that's
all
and
we
will
be
sharing
you
a
document
regarding
your
internship
and
all.
So
I
guess
that's
all.
I
would
say:
welcome
to
layer,
5.
D
That's
great
so
adi.
It
sounds
like
that.
Like
over
the
weekend,
you
and
potentially
a
couple
of
others
might
be
working
on
some
specifics
for.
D
The
collection
of
individuals
whose
names
are
in
the
meeting
minutes
for
them
to
work
on,
and
so
I'm
here
to
help
as
well
I'll
try
to
get
a
few
things
lined
up
did.
Is
there
anyone
who's
participated
in
the
read
that
recent
hackathon,
that
is,
does
not
feel
comfortable
with
react.
D
D
L
L
D
D
I've
got
I'm
an
endless
amount.
Okay,
oh
well!
Sayatan
did
you
want
to
talk
about
some
work
that
you've
been
doing
here
recently
and
is?
Is
this
also
a
call
to
action
for
some
volunteers.
M
So
there
are
some
links,
breakdown
like
we
have
to
change
the
layer,
5
links
to
the
hostel,
and
that
needs
to
be
done
in
all
the
repositories.
So
I
think
that's
a
huge
lot
of
tasks.
If
there
is
anyone
who
would
like
to
access,
you
can
just
ping
me
up
with
the
snap,
and
I
will
tell
you
like
what
it
needs
to
do.
M
Not
only
that
there
is
another
one
like
the
health
wanted
level.
That's
the
link
should
be
changed
to
them
all
the
four
organizers,
so
that
is
also
a
big
task.
If
anyone
is
interested,
you
can
just
assist
with
the
springs.
D
Perfect,
oh
thanks
for
that.
So
let
me
give
a
little
more
context
as
to
what
cyan
was
talking
about
and
I'm
still
working
on,
the
enunciation
of
mr
bose's
first
name.
D
It's
been
weeks
of
me
messing
it
up.
So
let
me
share
a
different
screen
and
say
this.
So
if
for
all
of
you
that
are
in
the
community
slack,
there
was
this
announcement
about
sort
of
the
slow
process
of
getting
meshary
and
service
mesh
performance.
D
Those
two
projects
into
the
cncf
there's
a
number
of
you
here
that
have
been
working
on
the
administrative
tasks
for
doing
that,
and
so
some
of
those
tasks
are
well
they're
a
little
bit
disruptive
because
we
end
up
taking
repositories
and
moving
them
around
in
github
and
and
so
those
like
it's
somewhat
pervasive
how
that
each
repo
almost
gets
impacted
by
these
changes
because
things
get
moved
around,
and
so
so
it
is
worth
so
that's
what
mr
bose
was
talking
about
is
making
some
of
some
updates
across
each
repository.
D
Each
repository
needs
a
couple
of
things
tweaked,
and
these
are
a
number
of
these.
Tweaks
are
just
that
they're,
relatively
small
changes
that
need
to
be
made,
and
so,
if
you
want,
if
you're
looking
to
get
your
your
commit
count
up,
your
contributor
count
up
go
see,
mr
buzz,
because
because
this
is
an
easy
way
to
do
that.
The
context
here,
though,
is
that
there
are
four
github
orgs
that
the
the
projects
that
you
all
steward
and
you
all
contribute
to
span.
D
So
there's
the
the
layer,
five
io
org,
that
is
the
initial
organization,
is
the
one
where,
if
there
isn't
a
better
place
for
the
repo
it
it's
there,
and
so
it
does
have
some
of
the
upcoming
projects
like
get
nighthawk,
which
I
was
talking
to
some
of
you
earlier
today.
If
c
plus
is
your
thing,
then,
let's
talk
about
get
nighthawk,
pink
ping
me
and
let's
talk
about
what's
going
on
there?
D
Okay,
good!
This
is
sort
of
the
catch-all.
D
This
is
kind
of
the
a
lot
of
community
things
go
on
here,
some
some
ad-hoc,
repos
still
there's
the
service
mesh
performance
org
right
now
it
just
has
one
repository,
and
that
is
for
the
specification
itself
and
for
the
website,
smp
hyphen,
spec,
dot,
io
and
then
yep,
there's
the
the
meshri
org,
and
so
it
probably
has
20
or
so
repos
for
the
various
measuring
components,
measuring
its
components
and
then
there's
another
one
that
we
well
we've
kind
of
talked
around,
but
we
haven't
talked
about
the
fact
that
there's
a
separate
org.
D
What
is
a
service
mesh
pattern
anyway?
I
see
that
it
is
a
green
windmill.
Twirly
thing:
okay,
that's
good,
but
the
rest
really
isn't
explained
on
this
org.
So
this
really,
this
organization
is
lightly
used
at
the
moment.
If
all
goes
well,
then,
maybe
in
about
a
year
maybe
a
year
and
a
half
cert,
this
concept
service
mesh
patterns
would
be
something
of
a
household
name
or
like
for
those
doing
cloud
native
infrastructure,
the
the
well.
D
D
I
don't
know
how
to
describe
this
succinctly.
This
is
I'm
digressing
into
explaining
what
this
is
in
part,
because
I
think
this
is
just
as
good
of
a
topic
as
any
other
that
we
would
have
talked
about
that.
We've
got
on
the
agenda.
D
So
there's
there's
an
upcoming
book
service
mesh
patterns.
D
It's
a
book
that
talks
about
best
practices
of
how
to
use
different
features
in
a
service
mesh
and
it's
the
first,
that's
the
first
of
what
will
be
probably
a
two-part
series
covering
not
specifically
a
cookbook
on
one
service
mesh
like
hey,
here's
link
or
d,
and
here's
how
you
configure
linker
dn.
Here's
what
you
do.
It
will
talk
a
lot
about
linker
d,
but
it'll
talk
a
lot
about
all
the
meshes
and
when
it
does
it,
you
know
in
each
chapter,
there'll
be
about
30
chapters
in
this
first
book.
D
D
Well,
that
the
book
itself
is
under
early
release,
so
there's
a
couple
of
chapters
that
are
out
there
and
wait
a
second.
This
looks
pretty
familiar
so
chapter,
one
mesh
sync:
how
to
get
started
with
any
service
match
chapter.
One
talks
about
this
thing
that
you
all
know
this
thing
that
you
all
are
contributing
to
it
talks
about
measuring
as
a
matter
of
fact,
the
entire
book
kind
of
talks
about
measuring.
D
That's
because
mesherie
is
capable
of
taking
each
of
the
service
mesh
pattern.
Examples
the
examples
that
are
in
each
of
the
chapters
and.
D
Realizing
that
or
orchestrating
them
making
them
come
to
life
like
applying
them
to
the
infrastructure
and
and
letting
people
take
those
examples
that
will
be.
D
D
I
think
there's
a
couple
of
you
so
as
if
we
haven't
talked
about
adi
enough
today.
Audi
is
oh
wait
a
second.
Maybe
we
can
talk
about
rudraksha
as
well,
so
rude
rocks
just
got
done,
posting
an
example
pattern
and
there's
a
couple
of
others.
So
these
are
example
patterns.
These
are
things
that
meshrie
can
take
and
apply.
This
is
still
early
days
for
measures,
capabilities
around
patterns,
but
they
become
quite
core
to
the
architecture
of
meshi
and
and
how
it
manages
the
infrastructure,
and
it
is
well.
D
I've
said
this
to
a
couple
of
you
and
but
I'll
say
it
to
the
other
20
of
you
that
haven't
heard
this,
and
that
is,
I
was
talking
to
a
messerie
user
yesterday
and
it
wasn't
a
private
conversation.
D
So
they
were
attracted
to
mesherie
in
part
because
they
wanted
to
understand
the
difference
between
the
performance
of
different
types
of
service
measures,
and
so
we
were
talking
we're
kind
of
chatting
about
things
that
meshery
does
today
and
things
that
hopefully
measury
will
do
tomorrow
and
when
we
talked
about
the
patterns
like
he
almost
he
almost
fell
out
of
his
chair.
D
Almost
he
didn't
quite
fall
on
the
ground,
but
he
sort
of
like
slipped
off
to
the
side.
He
was
he
was
taken
aback
and
just
you
know,
really
excited
about
the
work
that's
going
on
there
on
patterns,
so
now
that
I've
talked
about
it.
This
much.
This
is
another
example
of
well
like
we
just
don't
have
a
lot
written
about
what
the
heck
a
surface
mesh
pattern
is
and
how
they
get
used.
There's
a
single
page
right
here
and
it's
caveated
with
the
fact
that
this
these
things
are
still
in
flight.
D
It
does
highlight
the
notion
that
there
would
be
this
future
service
mesh
pattern,
repo
or
org,
which
is
what
got
us
started
on
this
conversation
in
the
first
place,
and
so
the
the
thought
is
that
these
will
be
like
named
patterns
named
and
numbered
patterns
that
people
can
refer
to.
They
can
use
as
a
starting
point,
for
you
know
they
can
augment
their
patterns
to
to
match
what
they
want
to
do,
and
it's
not
just
about
the
patterns
themselves.
D
I
mentioned
a
couple
of
other
projects
that
are
taking
something
of
a
similar
approach
to
describing
your
infrastructure
in
this
way
and
working
with
your
infrastructure.
In
this
way,
for
some
of
you
you're
saying
well,
yeah,
it
doesn't
that's
what
kubernetes
does
right
and
that's
what
it's
like
in
part
like
largely
yes,
they're.
D
I
don't
know
how
to
make
this
concise
and
give
you
concise
examples
to
say
that
this
is
slightly
different.
One
of
the
ways
in
which
meshri
will
bring
value
on
top
of
service
meshes
is
like
this
example
here
is
to
orchestrate
a
canary
rollout.
D
So
some
of
you
are
familiar
with,
and
some
of
you
aren't
like
the
notion
of
like
a
of
different
types
of
ways
to
roll
out
new
versions
of
your
applications
and
to
do
blue
green
deployments
and
et
cetera
service
meshes
are
all
capable
of
letting
you
control
a
rollout
of
a
new
version
of
your
application
so
to
split
traffic
from
your
current
version
of
the
app
to
the
new
version,
maybe
integra
in
a
granular
way
in
a
slow
way.
That's
comfortable.
D
So
in
this
case
that's
what
this
is
describing
you
route
a
certain
percentage
of
traffic
to
one
version
of
the
service
for
so
long
and
then
incrementally
step
it
up
and
do
so
dependent
upon
some
measurement,
some
like
health
measurement,
that
that
you're
watching
these
are
things
that
measure
would
bring
on
top.
This
is
what
I'm
getting
too
so
so
there's
a
lot
more
to
come
with
respect
to
patterns
there.
D
The
the
short
of
the
story
is:
there
are
about
four
github
organizations
that
you
all
we
all
work
across
and
one
of
those
upcoming
ones
is
service
mesh
patterns.
If
there's
anyone
on
the
call
that
fancies
themselves
a
visual
designer
or
is
oriented
towards
user
experience
design,
there
are
some
things
that
we
want
to
do
with
service
mesh
patterns.
D
So,
to
the
extent
that
those
will
be
named
and
numbered
patterns,
it
would
be
really
nice
to
be
able
to
say
so
to
be
able
to
say,
like
hey
the
the
circuit
breaker
pattern
like
that,
one
specifically
here's
it's.
This
is
made
up
at
the
moment.
We
haven't
really
gotten
into
numbering
them,
but
it
the
one
that's
for
circuit
breaking.
D
We
want
to
provide
everyone,
a
visual
depiction
of
what
it
what
circuit
breaking
is
as
they
go
to
learn
and
understand
that
concept,
and
so,
if
you're,
into
visual
design,
there's
a
number
of
these
to
be
made.
D
Topic,
samir
has
a
question
in
chat
link
yeah,
so
you
know.
Is
there
a
specific
measuring
operator
that
takes
care
of
patterns,
or
is
this
a
collaborative
task
of
multiple
measuring
operators
like
sort
of
what
are
the
mechanics
behind
how
it
is
that
mesherie?
You
know.
D
Understands
and
orchestrates
patterns,
so
so
I'm
bringing
up
a
slide
deck.
That's
in
our
when
I
say
our
I
mean
your
because
you
all
should
have
access
to
it
and
if,
if
you
don't
please
let
anita
know,
but
this
architecture
deck.
D
Is
what
I
want
to
point
you
to,
because
if
I
begin
to
describe
it,
you'll,
listen
to
me
for
another
20
minutes,
and
that's
not
good.
So
this
is
probably
the
this.
Is
the
meat
and
potatoes
slide
of
what
mescheri
is
doing
with
patterns
and
where
they're
coming
from,
and
so
the
the
most
juicy
part
of
the
meat
here
so
to
speak.
Is
this.
D
World
view,
but
if
you,
if
you
digest
this
one
and
grock
this
one,
then
you're
doing
good
and
we
can
talk
about
the
other
ones.
So
we'll
probably
talk
more
about
one
of
the
other
ones.
On
wednesday's
mesh
redevelopment
call
there's
a
collection
of
you
that
have
been
working
on
a
visual
topology
in
which
you
can
visually
design
these
patterns.
D
A
So
andrew,
do
you
want
to
talk
about
grafana
charts,
forester.
N
So,
for
those
who
are
not
familiar
with
mesh
map,
mesh
map
is
an
extension
to
the
imagery
project
which
lets
people
visualize
and
design
their
service
missions.
So
we
have
a
designed
pane
over
here.
What
I'm
going
to
demo
today
is
the
design
and
paint
functionality
that
I
have
developed
in
the
past
week.
So
what
we
have
over
here
is
some
components
for
some
service
measures
and
we
have
a
canvas
and
what
the
designer
thing
lets
you
do
is
it
lets.
N
So
last
week
what
we
saw
was
we
were
able
to
make
a
call
to
an
sd
adapter,
that's
running
in
my
local
machine,
which
then
returns
a
set
of
workloads
which
are
defined
in
the
open
application,
and
these
workloads
can
be
then
dropped
into
the
canvas
which
then
converted
into
patterns.
As
we
explained
earlier,
what
patterns
and
cotton
fields
are
and
then
these
patterns
are
then
used
to
create,
can
be
used
to
create
actual
service
mesh
so
yeah,
let
me
just
give
you
a
quick
number.
N
So
what
we'll
do
right
now
is
what
I'll
do
is
I'll,
drag
and
drop
an
sql
mesh
over
here
and
what
measure
does
is
it
requires
some
sort
of
a
configuration
before
you
can
actually
design
a
mesh,
so
what
the
service
mesh
does
is
it
opens
a
pop-up
that
hey,
if
you
want
to
drop
this
onto
the
canvas
you
need
to
first
configure
it
so
I'll,
configure
it
and
then
set
up
a
version
for
seo
and
click
submit
what
this
does
this?
N
It
saves
the
configuration
on
to
the
current
state
of
the
front
end
and
as
soon
as
I
click
apply,
what
it
would
do
is
it
would
send
a
request
to
the
local
adapter.
That's
running
that,
please
start
please,
like.
I
please
look
at
this
request
that
I'm
making
and
look
at
the
pattern
that
I'm
sending
you
and
create
this
spin
up
this
workload.
So
if
you,
if
you
look
at
the
terminal,
wait,
I'm
not
sharing
my
terminal,
I
guess
let
me
just
share.
F
It
it
should.
N
Be
yeah,
so
if
you
look
at
the
terminal
right
now,
this
is
the
window
where
the
steer
adapter
was
running
and,
as
you
can
see,
there
was
a
request
that
was
created
where
a
request
was
sent
from
the
front
end,
which
told
the
adapter
to
create
an
instance,
and
then
that
instance
basically
is
created
in
mini
cube
cluster.
That's
running
on
my
local
machine.
N
O
So
hey
one
second,
and
did
we
take
the
consideration
of
dependencies
like
let's
say
I
am
doing
the
grafana
thing,
but
grafana
doesn't
work
without
prometheus,
so
does
my
mass
map
detects
that
the
dependencies
has
been
there.
N
As
of
now,
we
don't
check
for
any
kind
of
dependencies
per
se,
but
it's
it's
in
like
in
future,
we
will
have
some
sort
of
metadata
with
which
we
would
be
able
to
configure
all
of
that
and
a
lot
of
stuff
like
this.
If
we
can
actually
drop
this
and.
D
Yeah
yeah,
to
add
to
what
soham
had
said
is
well.
Is
that
so
so?
Hopefully,
all
most
of
you
are
familiar
with
grafana
and
prometheus
and
one
is
a
visualizer
of
metrics.
The
other
one
is
a
metrics
collector
and
so
the
visualizer
of
metrics
grafana
to
samir's
point.
Well,
it
can't
really
visualize
anything
if
it
can't
talk
to
prometheus
or
some
metrics
collector,
and
so
the
two
kind
of
go
hand
in
hand,
or
at
least
there's
a
dependency
there
for
grafana
being
dependent
on
prometheus
and
it
gets
nuanced.
D
D
It's
hard
people
make
careers
out
of
this
stuff,
so
visually
depicting
it
is
important,
but
but
also
people
want
to
run
their
infrastructure
in
a
bunch
of
different
ways.
So
I
guess
my
point
here
is
that
we're
trying
to
enable
service
mesh
for
the
rest
of
us,
and
so
we
want
to
make
be
opinionated
and
make
you
know,
make
the
the
happy
path
so
to
speak
or,
like
the
the
eighty
percent
use
case,
path,
really
easy
to
do
so
like.
D
You
want
to
interface
when
you
play
with
laid
down
prometheus
and
grafana,
both
connected
to
them
and
just
started,
reading,
metrics
and
and
actually
started
displaying
them
like
like
right
away
like
right
when
you
bring
up
the
ui
like
great
and
we're
really
close
to
doing
just
that
the
issue
becomes
or
like,
where
the
logic
in
measuring
needs
to
what
it
needs
to
account
for
is
that.
D
Well,
there
are
any
number
of
environments
that
where
people
are
already
running
a
prometheus
operator
across
you
know
like
a
daemon
set
across
all
their
nodes
and
they've
got
their
own
prometheus
set
up,
they've
got
thanos
connected
to
it
or
a
cortex
connected
to
you
know:
they've
got
like
they're
doing
their
own
thing
already,
and
they
don't
really
want
the
they
don't
want
to
use
the
prometheus
add-on.
D
This
concept
of
well
of
measuring,
discovering
an
environment
and
the
component
that
does
that
is
called
mesh
sync,
so
mesh
sync
talks
to
kubernetes
and
it
says
kubernetes
tell
me
about
all
that
all
the
things
all
the
things
that
are
running
in
your
environment,
so
that
mesherie
can
digest
that
you
know
internalize
it
understand
it
visualize
it
and
manage
it
well.
Some
of
what
it
will
discover
is
that
oh,
there
was
an
existing
prometheus
that
was
already
deployed
and
maybe
two
of
them
and
they're
actually
for
different
purposes
or
they're.
D
They're
have
different
metrics
or
whatever
or
they're
they're
set
up
in
an
h,
a
configuration
and
as
mesherie
performs
this
and
by
the
way.
This
is
this:
these
aren't
newfangled
concepts.
These
are
like
tried
and
true
concepts
of
management.
Software
of
manage
of
like
network
management
systems,
which
is
that
infrastructure
and
managed
elements
so
like
a
prometheus,
would
be
a
generic
term
for
prometheus
and
grafana
would
be
an
element
and,
as
they
get
discovered
by
measuring,
they
become
eventually
a
managed
element,
but
part
of
that
discovery.
D
There's
a
few
things
to
track.
There's
kind
of
a
process,
a
finite
state
machine
to
probably
build
in
define
here
and
walk
through
the
finite
state
machine
would
in
this
case
for
discovery,
would
would
be
about
facilitating
a
process
of
registration,
discovery
and
registration
and
ultimately,
affiliation
or
part
of
so
so.
To
give
an
example,
you
take
measuring
you've,
already
got
a
kubernetes
cluster.
You've
already
got
a
prometheus
operator
running,
you
run,
mescheri
meshsync,
interrogates,
kubernetes
discovers
all
these
elements
like
prometheus
brings
back
that
information
and,
in
some
cases,
mesh
hole.
That's
cool!
D
Can
you
okay
anyway,
next
time
you
demo
so
we'll
have
to
do
you
have
to
show
the
visualizer
with
the
layouts?
Those
are
cool.
You
in
some
cases
will
be
able
to
measure
we'll
just
say:
well,
hey
I
found
a
hey.
I
found
a
prometheus.
Obviously,
that's
the
one
that
you
want
to
use,
I'm
just
going
to
connect
to
that
and
start
showing
you
some
charts
and
in
other
cases
it's
like.
Well,
it's
actually
not
really
clear.
D
If
you
want
to
use
that
or
you
want
to
do
something
else
or
if
that's
appropriate
and
like-
and
so
there
needs
to
be
this
regis,
this
process
of
registration,
where
the
the
user
needs
to
come
in
and
just
answer
a
couple
of
questions,
they
need
to
say,
yeah
measuring.
Why
don't
you
go
ahead
and
connect
to
that
prometheus
by
the
way
I've
got
a
couple
of
grafanas
because
of
whatever.
D
So
I
want
you
to
connect
that
grafana
and
that
prometheus
together-
and
I
want
you
to-
I
want
you
to
I'm
going
to
connect
you
to
eight
different
kubernetes
clusters
or
400
kubernetes
clusters
with
you
know:
50
nodes
apiece
like
there
are
absolutely
environments
like
this
and
by
the
way,
the
longer
that
we
go
on
as
a
community
as
a
cloud
native
community.
The
trend
is
generally
to
have
more
clusters,
more
smaller
size
clusters
than
their
than
the
trend
is
to
have
larger,
bigger
clusters
by
clusters.
D
I
mean
kubernetes
clusters
and
the
number
of
nodes
that
they
have
so
inevitably
measure
you.
There
will
be
any
number
of
mesh
sinks,
one
per
kubernetes
cluster
running
sitting
there,
interrogating
that
kubernetes
cluster,
helping
keep
mesherie
in
sync
to
understand
the
the
topology
and
all
of
the
managed
elements.
So
it
will
absolutely
be
the
case
that
there
will
be
multiple
instances
of
prometheus
and
grafana,
and
so
the
point
is
that
we
would
do
well
to
begin
to
articulate
and
define
a
finite
state
machine
in
which
people
walk
through
this
process
of
discovery
and
registration.
D
Part
of
registration
is
that
we
would
define
an
environment
and
that's
kind
of
on
the
roadmap
for
measuring.
So
if
you
navigate
to
the
settings
yeah-
and
here
mesherie
today
is
capable
of
actively
connecting
to
one
kubernetes
cluster.
D
Oh
that's
good!
That's
a
good
starting
point
like
that's,
but
but
it's
scoffable,
it's
kind
of
laughable
like
any
environment
worth
doing.
Anything
materially
interesting
is
running
more
than
one
cluster
measuring
needs
to.
We
need
to
expand,
measuring,
create
this
construct
of
an
environment.
Let
people
create
environments,
load
up,
multiple
connections
to
different
kubernetes
clusters.
Multiple
connections
to
different
prometheus
and
grafanas,
and
so
samir
you
asked
a
great
question
about-
is
measuring
intelligent
enough
to
track
this
dependency,
like,
for
example,
between
prometheus
and
grafana,
and
it's
like
yes,
but.
D
D
But
it's
something
that
we
all
are
going
to
have
to
think
about
and
help
try
to
auto
like
automatically
register
elements
where
it
makes
sense,
and
so
today
today,
if
you
download
the
latest
copy
of
meshri
and
you
deploy
istio
and
then
you
turn
on
yeah.
Thank
you
very
much
and
then
you
turn
on
the
prometheus
and
the
grafana
add-ons
meshsync
will
discover
like
meshri
will
provision
those
things.
Mesh
sync
will
discover
them
and
it
will
auto-connect
or
in
most
cases
it
will
auto-connect.
D
So
if
that's
a
great
thing
for
all
of
you
to
try
out
as
a
matter
of
fact
like
like
it,
there
will
be
certain
cases
where
it
may
not
auto
connect
so
tune
in
next
time.
For
what
are
those
cases
about
like
why
it
wouldn't
auto
connect
and
like
also
what
the
hell
is,
a
finite
state
machine,
but
that's
a
great
topic
for
wednesday
in
which
michael
will
be
presenting
or
anyway
it's
something.
We
should
start
to
think
about.
D
I
won't
digress
into
troubleshooting
your
environment,
samir,
I
get
well,
you
know
what
okay
fine
do:
samir
in
your
environment.
Do
you
get
two
connections
to
prometheus
or
two
connections
to
grafana,
or
is
there
just
a
single
url
in
the
dropdown.
O
O
D
A
A
Thankfully,
thank
you
so
so
we
have
a
couple
of
items
to
discuss
on
the
next
week's
measure.
Development
call,
and
we
also
have
some
items
left
from
today's
call.
So.