►
From YouTube: Layer5 Newcomers Meeting (Oct 22nd, 2020)
Description
@Hacktoberfest AMA with @lcalcote.
B
Good
morning
I
was
just
taking
the
opportunity
to
hang
out
with
jeanette
jeanette
and
I
are
early
risers.
Apparently
we
we
got,
we
got
the
meeting
warmed
up
like
half
an
hour
early,
so
we're.
B
A
Gillette,
have
you
been
on
a
meeting
before
no?
No,
this
is
my
first
making
all
right.
Welcome
about
to
harass
you,
so
please
do
go
ahead
and
introduce
yourself.
C
Okay,
hi
everyone.
My
name
is
jeanette,
I'm
a
cs
major
in
my.
C
A
Great
welcome
to
the
community
we'll
be
starting
up,
and
then
we
start
helping.
You
start
out,
hopefully
all
right.
Well,
let's
start
properly.
A
I
usually
start
these
out
with
a
small
intro,
so
welcome
to
video
meeting
everybody.
We
are
taking
this
month
to
transfer
these
meetings
to
the
hack
to
the
best
initiative,
so
a
special
welcome
to
any
and
all
have
to
profess
participants
who
have
been
contributing
over
the
last
month
or,
if
you're,
just
starting
out.
Well,
you
still
have
10
days
to
go
the
speeding,
especially
important
for
you,
then,
because
we'll
help
you
find
open
issues
will
help.
You
find
any
domain
that
you
want
to
get
into.
A
All
right,
there
are
a
lot
of
links
in
the
meeting
when
it
stay
so
before
we
get
to
bamboozled
I'll
point
them
out
to
you.
A
Right
all
right
so,
for
the
third
point
go
check
out
the
hacktoberfest
page.
You
have
everything
you
could
possibly
need
in
there
for
either
the
hack
to
the
first
or
for
starting
out
as
a
newcomer
in
the
community.
A
You'll
find
links
so
you'll
find
resources.
Here
you
find
people
to
contact.
If
you
have
any
issues,
you
can
basically
contact
anybody
in
the
community,
really,
I'm
pretty
sure
they'll
help
you
out
or
point
to
somebody
who
can
help
you
out
there's
a
list
of
tutorials
here
which
are
specifically
for
how
to
profess.
If
you
want
a
list
of
more
detailed
list
of
tutorials
that
we've
been
doing
again,
yep
go
ahead,
have
a
look
at
jesus
should
really
be
making
small
artists
all
right.
Go.
Have
a
look
at
the
newcomers.
A
Playlist
you'll
find
tutorials
you'll
find
basic
walkthroughs
of
our
projects
about
to
get
started
of
how
to
set
up
even
your
local
systems
to
get
started
with
our
projects
or
with
git
or
with
any
of
the
basic
resources.
A
Lee
will
be
pointing
them
out
in
more
detail.
I
suppose,
if
you
want
to
come
up
with
other
meetings,
other
life
community
meetings
go.
Have
a
look
at
this
link.
A
A
That's
all
for
the
housekeeping
task!
Suppose!
Oh
yeah,
if
you
are,
I
know
a
newcomer
to
open
source,
even
if
you
aren't
a
newcomer
to
open
source
and
just
need
some
help
to
get
started
or
just
need
some
help
to
find
a
domain
that
suits
you
better,
go
ahead.
Take
our
open
source
questionnaire,
you'll,
find
it
on
the
hack
to
the
first
page
as
well,
so
go
either
way
go
fill
it
up,
find
a
tutorial
that
you
want
to
have.
A
We
usually
do
this
on
a
most
popular
basis,
so
vote
definitely
counts
and
will
happen
we'll
be
having
each
of
these
in
future
meetings
even
after
they
have
to
profess.
So
please
do
stay
with
us.
Join
us
for
meetings,
we'll
get
you
set
up
one
way
or
the
other
and
yep
you'll
find
a
lot
of
details
in
this
form
as
well.
You'll
find
a
lot
of
links
which
are
mostly
added
to
this
meeting
minute
as
well.
A
All
right,
that's
it
for
me,
that's
all
for
me
all
right!
So
today
we
have
well.
We
have
an
open
session
with
d
carcut.
The
only
person
with
this
video
on
so
you'll
be
pretty
easy
to
spot
all
right.
This
is
an
open
session,
so
feel
free
to
ask
us
anything.
You
feel
free
to
ask
us
anything
on
any
meeting.
We're
not
that
strict
about
the
agenda.
A
You
can
add
any
topic
that
you
want
to
ask
or
discuss,
but
today
is
well
an
intro
to
life
eyes
by
its
founder,
who
we
are
all
very
proud
of,
and
I'll
just
stop
talking
and
definitely
talking.
B
Yeah,
so
hey
welcome
everyone
to
to
add
to
what
srito
is
saying
is
that
for
this
month
we
had
been
using
the
newcomers,
call
as
a
hacktoberfest
meetup
kind
of
doubling
up,
because
a
large
part
of
the
newcomers
agenda
is
getting
people
introduced
to
the
projects,
helping
them
find
an
interesting
area,
something
that
they're
passionate
about
something
that
aligns
with
their
technological
focus
or
just
an
area
of
growth
for
them
or
an
area
of
interest.
B
There
are
a
number
of
projects,
I'm
going
to
pull
up
a
deck
that
walks
through
a
couple
just
to
highlight
them
so
that
so
that
you've
seen
them,
because
there
might
be
a
few
that
you're
interested
in
that
you
haven't
seen
just
yet.
There
are
a
couple
of
new
projects
or
more
than
a
couple
of
new
projects
that
have
yet
to
be
well
documented.
B
B
The
list
of
tutorials
are
in
the
meeting
minutes
that
sridi
is
sharing,
they've
been
recorded
and
you
can
go
back
and
look
at
the
slides,
go
back
and
and
play
the
recordings
what
street
was
pointing
out
earlier
with
on
the
questionnaire
the
the
questionnaire
that's
on
the
hacktoberfest
page,
that
asks
for
your
input
on
other
tutorials
that
you
would
like
to
see,
and
so
I'm
in
some
respects
here
today
to
pinch
hit,
I'm
here
today
for
you,
this
is
probably
the
right
way
to
think
about
it,
and
so
you
should
consider
yourself
very
sp
lucky.
B
No.
What
I'm
trying
to
say
is
I'm
here
for
you
and
so
it's
your
agenda.
I
can
talk
to
you
all
day
about
different
projects
and
various
things
that
are
going
on,
introduce
you
to
them
and
for
lack
of
anyone
asking
questions
or
suggesting
something.
That's
what
I'll
do-
and
I
suspect
will
generate
discussion
through
that.
B
But
I
want
you
to
internalize
that
for
a
moment
that
I'm,
for
my
part-
and
I
know
for
sriti
and
for
others
that
are
here,
there's
a
number
of
others
that
are
on
the
call
that
that
are
representatives
of
the
community
and
who've
been
instrumental
in
a
number
of
the
projects
that
you
see.
B
They're
they're
here
for
you
as
well,
so
internalize
that
be
selfish.
Take
the
next
50
minutes
and
be
selfish.
Ask
a
silly
question:
amia
does
a
great
job
of
oh!
No,
I'm
just
of
doing
that
just
kidding
it
does
a
great.
He
does
a
great
job
of
bringing
up
of
putting
his
thinking
cap
on
and
asking
questions.
So
one
of
the
questions
that
was
just
asked
was
to
get
clarified.
B
B
This
is
this
isn't
related
to
service
meshes
per
se,
but
it
is
related
to
well
a
lot
of
other
projects,
yeah
strappy,
dot,
io,
so
you're
all
familiar
with
wordpress.
I'm
sure
wordpress
is
a
very
popular
example
of
a
content
management
system.
B
If
you
want
to
give
it
a
classify
it
in
a
you
know
as
to
what
type
of
tool
it
is,
it's
a
blogging
platform
right,
but
it
helps
you
manage
content,
manage
posts
and
images
and
video
and
other
things
well,
there's
a
newfangled
content
management
system
or
sort
of
a
next-gen
of
wordpress.
If
you
will,
there
are
many
different
content
management
systems,
but
one
of
those
is
called
strappy,
and
so
here
in
this
community
we're
establishing
an
instance
of
strappy
the
the
the
link
to
it
is
in
the
chat
strappy.io.
B
And
emil
has
been
standing
up
an
instance
for
us,
and
actually
I
don't
know
that
I
could
directly
answer
your
question
because
I
I
haven't
looked.
There
was
another
contributor
here,
dhruv
he's
still
here,
he
had
looked
and
sort
of
determined
that
our
probably
our
best
approach
was
to
fork
a
strappy
and
do
deployments
from
there,
and
so
we
have
already
forked
strappy.
B
B
So
part
of
that
is
is
to
consider
why
it
is
that
you're
here
or
like
hey.
What
are
you
interested
in?
What
can
I
tell
you.
B
So
all
of
you
that
are
on
the
call
you
should
all
have
access
to
the
community
shared
drive
and
I'll
show
you
that
if
you
haven't
seen
it,
you
should
there's
a
layer.
5
community
shared
drive,
a
google
drive,
it's
got
a
ton
of
info.
B
It's
got
a
lot
of
assets,
images
pictures
posts,
a
lot
of
a
lot
of
work,
that's
been
done
by
other
you,
some
of
you
names
very
much.
So
your
names
are
on
some
of
these
things.
We
just
gave
a
workshop
for.
B
Ccoss
earlier
this
week
there
were
a
number
of
spanish-speaking
individuals
who
had
joined,
which
is
awesome,
we've
done
a
similar
workshop
last
year,
and
so
and
so
there's
the
measuring
documentation
is
currently
being
translated
to
spanish.
B
Thanks
to
those
contributors
today,
we're
giving
two
kubecon
talks,
so
we're
recording
two
cubecon
talks,
we'll
be
hosting
a
third
booth
at
kubecon,
not
a
layer,
five
booth,
but
we
maintain
service,
mesh
interface
and
open
source
project
smi,
so
we'll
be
ma
manning
or
personing
that
booth
you're
all
so
the
the
two
cubecon
talks
that
we're
giving
I'm
delivering
and
have
invited
different
individuals
from
the
community
to
join
me
on
stage
for
a
couple
of
reasons,
one
because
it's
a
way
of
rewarding
those
that
spend
time
here
too,
because
the
world
doesn't
need
to
really
hear
me
speak
that
much
more.
B
Rather,
I
would
much.
Rather
they
hear
you
speak.
I
know
that
sounds
entirely
uncomfortable,
probably
for
most
of
you
but
I'll
be
there
with
you,
we'll
rehearse
at
least
30
minutes
before
we
record
it
since
I'm
mentioning
some
of
these
activities
is
like
these
are
there's
this
this
open
source
community
here?
Isn't
it
isn't
like
it's
like
some
other
ones
you'll
be
in,
but
but
it's
also
intentionally
not
like
other
ones.
B
I,
if
anyone
wants
to
speak
I'd,
love
for
you
to
speak.
I've
got
people
asking
me
more
times
than
I
can
to
go,
speak
and
again
it
doesn't
need
to
be
me,
I'd
love
for
it
to
be
someone
else.
I've
got
more
two
books
being
written
at
the
moment.
That's
ridiculous!
I've
got
there
are
so
many
other
things
to
go:
focus
on
I'd
love
for
one
of
the
books,
that's
being
written.
It's
called
service
mesh
patterns,
there's
60
patterns,
we've
written
them
down
just
to
just
the
outline
of
what
they
are.
B
B
They
talk
about
some
of
the
like
some
of
the
projects.
There's
a
service
mesh
landscape
out
there,
okay,
I'm
gonna
start
going
and
I'm
gonna
monologue
you
through
the
next
40
minutes.
If
you
don't
interrupt
with
a
question,
so
do
it
and
don't
apologize.
B
But
one
of
the
first
projects
that
the
community
started
on-
and
this
really
is
a
community
project.
It's
it's
called
the
service
mesh
landscape
that
we
have
it
is
in
in
its
current
form.
This
is
what
it
looks
like
it
helps.
People
understand
the
fact
that
there
are
a
bunch
of
meshes
out
there
and
the
fact
that
it's
a
multi-mesh
world,
some
of
the
meshes
that
you
see
here
some
have
already
died.
B
Some
have
already
been
acquired
and
some
you
have
yet
to
see.
There's
one
that's
being
added
right
now.
The
product
manager
at
f5
at
nginx
for
nginx
service
mesh,
has
an
open
pr
where,
as
a
community
contributor,
he's
adding
and
updating
nginx
service
mesh
to
this
page.
As
a
matter
of
fact,
it's
an
open
pr
right
now,
so
I
am
willing
to
make
someone
a
bet
with
a
fair
amount
of
money
that
the
nginx
service
mesh
won't
be
the
last
one
that
we
see
as
a
new
mesh
added.
B
So
this
has
been
a
curation
of
information.
It's
one
of
our
projects.
Now
this
project
gets
referenced
by
the
cncf.
It's
been
a
few
times
now
that
the
executive
director
of
the
cncf
has
pointed
people
to
this
landscape,
which
is
fantastic,
considering
how
many
people
have
come
to
bear
on
the
cncf's
landscape.
B
I
was
part
of
the
team
that
originally
created
this
particular
landscape
and
helped
create
the
structure
or
the
the
groups
that
you
see
here.
This
isn't
the
only
landscape
at
the
cncf.
There
are
sub
landscapes.
B
I
chair
the
cncf
network
special
interest
group,
which
is
the
group
that
presides
over
see
the
the
green,
the
first
green
line
here.
It's
kind
of
small.
B
But
but
everything
from
this
first
screen
line
to
the
right,
the
the
big
projects
here,
the
ones
that
are
actually
in
the
cncf,
these
all
fall
into.
Oh
this
hasn't
been
updated.
We
just
accepted
bfe
into
the
cncf
a
little
while
ago,
so
that
should
be
highlighted,
but
these
all
fall
into
our
domain
of
they'll
report
into
that
special
interest
group.
So
mesherie
is
here.
B
It
is
our
intention
that
measury
go
into
the
cncf
by
the
way
it's
been
a
long
time
coming,
but
the
project
just
passed,
500
stars
yesterday,
which
which
also
by
the
way
the
cost
of
admittance
to
this
call,
is
a
github
star.
So
if
you
haven't
started
the
project,
you
you
must,
you
have
to.
B
B
B
The
part
of
the
the
reason
that
things
like
this
book
is
pertinent
to
you
all
is.
I
was
just
seeing
usana's
face
and
thinking
about
the
pr
that
she
got
merged
yesterday
for
mastery
ctl
and
thinking
yeah.
As
this
book
is
written
out,
her
work
is
going
to
be
used.
I
don't
know
how
many
times
I'm
hoping
thousands
and
thousands
of
times,
because
here's
what
we're
going
to
do.
B
There
are
a
few
that
have
been
written,
there's
a
second
edition
of
this
coming
out
in
two
weeks.
It's
why
I
only
have
four
hours
of
sleep,
because
I
was
providing
feedback
on
the
last
copy
edits
on
the
new
one.
Coming
out
this
this
first
one
right
here
is
free
to
you,
there's
a
link
to
it
in
it
just
mention
it
in
slack
and
you'll
get
a
link
to
it.
It's
a
great
introduction
to
the
space
it's
two
years
old,
now
still
quite
relevant,
still
pretty
good,
but
the
second
edition
is
coming
out
shortly.
B
This
is
the
other
book
that
I'm
referring
to
mescheri
will
be
the
tool
that
is
referenced
in
every
single
pattern
as
the
tool
to
use
to
stand
up
and
play
with
and
get
familiar
with
each
of
these
patterns.
B
You
can
see
that
happening
inside
of
mesh
area
already,
if
you
consider
something
like
data
plane,
extensibility,
there's
a
lot
of
excitement
and
buzz
around
webassembly
wassum
and
how
that
works.
We
demonstrated
one
of
the
community
members.
Abhishek
had
demonstrated
a
week
ago,
last
friday's
community
meeting
how
meshrie
can
be
used
to
dynamically
load,
a
web
assembly
filter,
and
if
you
want,
you
can
watch
that
recording
or
you
can
go
to
look
at
that
project
so
in
a
roundabout
way,
we're
kind
of
getting
around
to
talking
about
the
projects.
B
That
project
is
called
image
hub.
It's
a
sample
application
for
helping
people,
learn
about
webassembly
and
webassembly's
use
inside
of
service
meshes,
so
image
hub
is
listed
here.
It's
a
repository
inside
of
the
layer,
five
io
org
oops.
I
didn't
realize
it
goes
there
anyway.
You
can
also
watch.
This
was
a
talk
given
at
dockercon
2020
just
a
few
months
ago.
B
B
B
B
B
And
so
so,
for
some
of
you,
this
might
be
a
project
that
you
want
to
jump
into
is
to
cut
your
teeth
on
some
gatsby
on
some
react,
some
css
html
and
some
javascript.
B
Here's
the
image,
here's
the
new
image
hub
page.
For
my
part,
I
kind
of
consider
that
these
pages
that
there
should
be
something
that
offsets
the
page
here,
the
content
here
from
the
rest
of
the
page,
because
it
it
just
feels
like
things-
are
free-floating.
I
don't
know
if
anyone
else
gets
that
feeling.
B
Okay:
okay,
other
things
we
get
noticed
forrester,
is
writing
about
us.
I'm
gonna
go
speak
with
gartner
and
451
they've
all
asked
to
be
briefed
on
the
works
that
you're
doing
and
we'll
there's
another
research.
An
analyst
firm,
that's
written
about
us
recently,
amalgam,
a
lesser-known
firm
but
they've
published
their
research
and
your
works
are
included,
and
so
we'll
get
a
copy
and
post
that
as
well
we're
creating
a
new
standard
for
the
industry
called
smp,
which
is
a
service
mesh
performance.
B
B
It's
lonely!
Being
an
industry.
Leader
takes
a
lot
of
explaining
and
explaining
and
re-explaining
we're
going
to
create
mesh
mark
as
part
of
that
effort,
a
whole
new
way
of
measuring
the
performance
of
your
mesh
and
articulating
it
succinctly
if
anyone's
familiar
with
app
decks,
that's
kind
of
what
we're
going
to
do,
but
for
service
meshes,
so
aptx
you're.
B
B
This
is
a
bit
of
an
explanation
of
it.
The
other
way
of
understanding
meshmark
this
new
way
of
measuring
the
performance
of
a
mesh,
is
to
reflect
on
app
decks.
Appdex
is
a
toss
back
to
the
early
2000s.
I
think
I
don't
know
to
to
a
different
internet
yeah,
there's
a
formula
that's
used
to
to
measure
and
determine
the
whether
or
not
someone
is
happy
how
well
your
application
is
doing
the
quality
of
experience,
and
so
it's
a
unit
of
measure
for
measuring
whether
or
not
people
are
happy
with
an
app.
B
B
And
when
we
do
it,
we
really
want
to
do
it
in
in
coordination
with
other
with
other
nerds
other
academic
nerds.
B
We're
fortunate
enough
to
to
know
a
few,
and
some
of
you
might
know
a
few
or
some
of
you
might
want
to
some
of
you
are
students
might
want
to
pick
this
up
as
a
final
project
or
you're
going
on
to
a
master's
or
a
phd
studies.
Great
there's
a
lot
to
dig
into
here.
It's
a
it's
a
good
way
to
to
make
your
mark
to
leave
your
name
on
something
that
will
be
a.
E
Probably,
can
I
say
so
like
miss
mark
is
built
upon
which
takes
tech.
B
Yep,
it
isn't
it's,
it's
there's,
no
implementation
of
it
today,
the
formula
for
it,
the
formula
by
which
you
would
calculate.
B
Yeah
yeah,
you
know
how,
when,
if
you're,
if
you're
in,
if
you're
an
operator
or
if
you're
responsible
for
an
application,
you're,
probably
going
to
look
at
some
signals,
some
metrics
to
determine
whether
or
not
the
app
is
running?
Well,
hey,
what's
the
what's,
our
you'll
you'll
go
over.
Maybe
look
at
the
four
golden
signals
coming
out
of
the
sre
book
or
or
you'll
use
red
you'll
use.
You'll.
Look
at
the
number
of
errors
that
your
app
is
producing
the
amount
of
the
how
much
how
responsive
it
is.
B
You'll,
look
at
a
bunch
of
different
things,
but,
and
so
what
aptx
was
trying
to
do
is
to
incorporate
a
bit
of
the
softer
side
of
whether
or
not
people
like
your
app
might
have
zero
errors,
be
extraordinarily
responsive
people,
click
on
it,
and
it's
there
like
that,
but
that
doesn't
mean
they
actually
like
it
or
enjoy
it.
They
might
be
rage
clicking.
Unlike
this,
the
section
that
looks
like
it's
supposed
to
be
clicked,
but
it
doesn't
do
anything
and
then
there
you
know
there
are
many
javascript
offerings
out
there.
B
That
will
try
to
help
you
identify
if
people
are
rage.
Clicking
or
not,
well
guess
what
those
clicks
are
going
over
the
network.
Guess
what
controls
the
network?
Guess?
What
can
measure
that
yeah
intentionally,
we
haven't
defined
the
formula
for
mesh
decks.
B
For
my
part,
or
I
mean
you
know
that
that
can
we
can
go,
do
that,
I'm
waiting
for
I'm
waiting
for
someone
for
someone
at
a
university
to
step
up
and
say
that
they
would
like
to
because
I'd
love
for
someone
else
to
own
it.
I'm
trying
to
give
space
to
others
to
let
them
their
mark
in
some
respects.
It
doesn't
really
matter
what
the
formula
is.
B
Meshri
is
the
tool
that
will
is
implementing
smp
today
and
hence
will
implement
meshmark
all
of
the
service
meshes
that
any
of
you
can
name
on
the
landscape,
any
of
them
that
are
still
around.
They
have
agreed
at
one
point
in
time
to
run,
measuring
in
their
build
process
their
release
process
and
produce
their
mesh
mark
and
publish
those
results.
Just
like
they've
agreed
to
publish
the
results
of
the
smi
conformance
tests
that
meshrie
will
run.
B
So
let
me
send
you
the
link
to
this
deck,
so
you
so
folks
can
kind
of
walk
through
it
and
then
let
me
stop
and
say
what
questions
do
you
guys
have.
F
B
Yeah
they
are
so
let
me
they
are
the
way
that
they
would
produce
that
is
by
taking
measuring
and
embedding
it
in
their
release
process.
So,
if
they're,
using
whatever
ci
system
that
they're
using
circle,
ci
or
travis
or
jenkins,
or
jenkins,
x
or
github
actions
that
they
would
make
a
call
out,
they
might
leverage
measuring,
ctl
and
I'll
they'll
invoke
a
performance
test
in
mesherie
that
performance
test
will
they
can
run
a
bunch
of
different
tests
and
eventually
it
will
produce
a
mesh
mark.
B
B
B
We
have
only
recently
gotten
meshary's
tooling
to
the
point
where
you
can
pro
easily
programmatically
invoke
a
performance
test.
I
expect
that
most
of
those
service
mesh
groups
they'll
end
up
using.
B
To
do
measure
ctl,
perf
and,
and
it
expects
a
few
different
parameters
about
what
you
want
to
test.
What
we'll
do
is
each
of
those
individual
service,
mesh
groups
istio
or
what
have
you
they
will
they'll
they'll
identify
a
given
account
that
they
use
to
authenticate
to
measure
we'll
flag
that
account
and
save
their
results
separately.
B
Massaged
the
data
which
we
do
because
when
measure
is
released,
a
random
but
there's
a
secret
key
included
inside
of
meshri,
that
is
used
to
verify
the
integrity
of
the
measuring
binary
so
that
nobody
is
screwed
with
it
to
like
bump
up
their
score
same
thing
for
smi
conformance
tests
so
that
nobody
passes
their
tests
will
only
allow
certain
accounts
to
count
in
terms
of
having
come
having
being
representative
of
that
project.
So
it's
a
lot
of
long
long
term
strategy.
B
A
lot
of
planning
like
to
get
us
to
this
point
and
we're
actually
almost
there
for
both
of
those
things
when
measuring,
particularly
as
we
like
push
mastery
into
the
cncf
it'll,
be
the
tool
that
determines
smi
conformance,
whether
or
not
a
mesh
is
conforming
to
smi
it'll,
be
the
tool
that
says
hey
in
accordance
with
this.
This
new
service
mesh
performance
standard,
here's
how
these
meshes
line
up,
we'll,
go
and
borrow
20,
something
or
however
many
nodes.
B
We
want
from
the
cncf's
lab
to
run
some
performance
tests
on
bare
metal
systems
and
then
publish
those
results
and
there's
just
there's
just
a
lot
of
stuff
to
do.
Well.
Does
that
is
that
what
you're
asking.
G
Lee
I
had
a
question
so
the
performance
test
right
some
could
run
on
a
different
hardware.
Some
could
run
in
a
different
environment.
So
how
are
we
accounting
for
that?
So
they
might
get
a
so
that
is
where
this
service
mark
kind
of
thing
is
coming
into
picture.
Where
we
define
what
should
you
measure
basically
or
something
of
the?
Maybe
I
I
did
not
understand
the
mishma
part
completely.
B
Right
yep,
the
great
questions,
yeah
it's
and
I'm
being
brief
or
I'm
being
very
quick
to
like
say,
say
a
bunch
of
concepts
in
the
same
sentence
and
you're
right
that,
like
smp
as
a
standard
and
as
a
tool
to
collect
and
capture
and
characterize
the
performance
of
a
mesh
in
a
consistent
way
is
one
is,
is
a
thing
very
closely
related
to
that.
B
Is
this
notion
that
we
would
create
a
new
unit
of
measure
of
a
new
yardstick
called
meshmark
that
would
that
would
use
a
formula
to
take
into
account
like
a
bunch
of
variables,
including
variables,
like
you
just
said
exactly
like
you
just
said,
which
is
wait.
Your
your
environment
was
50
clusters
with
40
nodes,
a
piece
each
of
them
had
different
amounts
of
cpu
and
memory,
and,
like
oh,
my
gosh
that
that's
so
many
variables
right
there
and
then,
and
then
you
say,
yeah
we're
running
this
mesh
okay
at
this
version.
B
Okay
with
this
configuration,
which
is
like
there's
a
ton
of
different
ways
to
configure
a
mesh
under
this
workload-
oh
geez
on
this
version
under
this
amount
of
requests
and
except
like
there's
a
reason
why,
if
you
go,
google
for,
like
you,
you'll
go
find
posts
that
talk
about
the
performance
comparative
between
service
meshes.
You
won't
find
one
from
us
because
it's
really
hard.
I
commend
those
that
have
tried
to
publish
their
results
and
things
and
they
are
they're
somewhat
helpful,
the
workloads
that
they
use
in
those
tests.
B
I
think
exactly
to
the
point
that
that
you
were
making
is
like
well,
they
were
running
some
database
intensive
workload
that
did
such
and
such,
which
was
great
for
them
because
that's
what
they
do,
but
has
nothing
to
do
with
the
the
image
resizing
workload
that
we
use
over
here,
which
has
a
totally
different
signature.
B
If
you
will
it's,
maybe
and
so,
and
so
that's
in
part
why
mesherry
was
built
is
to
like
to
like
pull
together
a
lot
of
the
tools
that
people
normally
have
to
write
glue
code
around
and
scripts
around
to
do
load
generation,
statistical
analysis,
capture,
all
the
details
of
their
environment,
calculate
it
all
up,
have
a
bunch
of
charts
that
show
all
the
the
various
ways
in
which
you're
measuring
the
performance
and
characterizing
the
environments
and
it's
like,
and
you
dedicate
a
few
humans
to
it
for
a
couple
years.
B
We're
trying
to
give
people
an
easy
to
use
tool.
That's
repeatable
that
they
can
bring
into
their
own
environment,
to
test
against
their
their
environment
and
then
to
sit
there
and
like
manipulate
to
turn
on
load
balancing
around
robin
as
an
algorithm
and
then
switch
over
to
weighted
round
robin
and
just
change
that
one
variable
and
see
the
difference
for
themselves
and
their
own
environment
in
accordance
with
their
own
systems.
B
G
C
G
Like
a
couple
of
projects
like
yeah,
like
http,
request
responses
and
then
a
project
like
an
app
that
you
run
on
google's
app
right,
the
hipster
app,
you
would
run
and
then
measure
the
performance
so
similar
to
that.
I
think
maybe
we
will
add
more
so
as
and
when
someone
comes
up
with
a
specific
workload.
Types
right.
B
Yeah,
that's
a
yeah
totally
we
will,
and
we've
that's
been.
A
point
of
discussion
of
any
number
of
times
has
been
like
what
well
what,
if
you're
just
going
to
do
a
generic
study.
B
What
type
of
workloads
are
representative
of
distributed
systems
microservices
that
people
are
running
and
there
isn't
a
single
answer
for
it,
but
there
are
some
fairly
decent
representatives,
some
of
the
units,
some
of
the
professors
that
we've
engaged
with
well
one.
They
very
much
so
have
that
question,
because
they're
somewhat
removed
from
the
from
enter
from
they're,
often
somewhat
removed
from
people
who
are
actually
running
workloads,
and
so
a
lot
of
times.
They'll
they'll
ask
us
this
question
and
we'll
characterize
some
of
those.
B
Whenever
we've
gotten
our
hands
on
a
sample
app
that
we
think
that
that
works,
we've
tried
to
include
it
and
make
it
so
that
you
can
deploy
that
sample
app
on
as
many
of
the
meshes
as
possible,
because
it's
one
thing,
because
that
just
that
unto
its
own
leaving
performance
aside,
is
just
interesting
and
helpful
to
people
to
be
able
to
take
an
application,
run
it
on
linker
d.
Take
that
same
app,
run
it
on
console
and
just
experience.
The
difference
in
terms
of
functionality
between
the
meshes.
B
B
She
she
and
her
students
had
been
working
on
two
applications
I'll
have
to
these
are
like
these
are
in
the
notes
in
the
there's,
a
social
network
application
and
another
one
another
one
that
I
forget
what,
but
we
were
trying
to
work
with
her
and
her
students
to
incorporate
that
sample
app
because
they
they
went
through
the
pain
of
like
trying
to
create
this
sample
app
to
make
it
be
a
good
representative
they're
trying
to
instrument
the
app
all
along
the
way
so
that
they
can
measure
some
things.
B
They've
got
some
these
some
hypotheses
that
they'd
like
to
prove
and
disprove.
It
was
great,
very
relevant
study,
part
of
our
initial.
This
was
like
two
years
ago,
so
part
of
our
initial
challenge
was
that,
while
christina
is
a
very
bright
academic
and
so
her
students,
they
are
like
football
players
doing
ballerina
when
it
comes
to
code
and
so
which
couldn't
it
was
a
lot
of
work
to
to
rejigger
their
app
to
run
on
kubernetes.
B
It
eventually
it
does
now
and
we
did
help
them
get
there
and-
and
we
haven't
incorporated
that
it's
not
we've
done
work
to
do
it,
but
it
hasn't
landed.
G
B
B
B
Yeah,
okay,
let
me
bring
this
up
so
yeah,
so
there's
definitely
notes
inside
of
inside
of
here
I'm
going
to
try
to
do
two
things
at
once.
I'm
going
to
try
to
answer
part
of
the
question
that
emil
was
asking
about.
B
B
We've
worked
a
lot
with
this
team
for
performance
benchmarking.
They
publish
a
list
of
tools
that
you
might
want
to
use
to
to
do
these
things
and
guess
what
it's
like,
auspiciously,
missing,
meshary
and
the
last
interaction
I
I
had
with
them
was
they'll
accept,
pull
requests
to
list
meshri
great.
We
should
go.
Do
that
so
that
more
people
see
it.
B
Did
there's
an
a
github
repository.
B
B
And
then,
by
the
way,
is
it
hosanna
or
husseina.
G
B
Very
good,
so
within
yeah,
if
anyone
else
has
questions
feel
free
free
to
bring
them
up.
I'm
just
gonna
dig
up
the
name
of
that.
B
B
G
So
whenever
another
service
mesh
wants
to
join
missionary,
so
what
would
be
the
work
involved
from
their
side?
So
what
is
that
they
have
to.
B
Yeah,
that's
a
that's
an
awesome
question
thanks
for
asking
so
so
it's
a
very
relevant
question
because
the
nginx
service
mesh
has
so
by
the
way.
Just
I
I
don't
know.
B
Yes,
yeah
who's,
saying
that
I'm
gonna
put
the
link
to
this.
I
don't
know
if
this
is.
I
suspect
this
isn't
really
what
you
were
going
after
or
what
you're
looking
for,
but
the
the
link
to
this
github
that
I'm
posting
it'll
have
a
few
sample
apps
that
it's
been
a
couple
years
since
we've
looked
at
them
and
helped
over
here,
and
we
might
want
to
look
at
incorporating
like
these
three
that
have
been
released
into
as
sample
apps,
that
meshrie
can
deploy,
could
could
be
here
here.
B
They
are
right
here
at
the
time
this
was
kind
of
silly
because
they
like
had
embed.
Oh
it's
funny
because
they're
using
wrk
too
anyway.
Obviously
they're
studying
some
similar
things.
I'll
say
that
christina
is
less
than
friendly
as
part
of
our
issue
she's
just
I
just
was
sort
of
awkward
about
many
other
professors,
weren't.
B
B
It's
relatively
the
adapters
themselves
are
relatively
lightweight
like
there's,
not
a
whole
ton
that
goes
into
them.
We
have
had
three
or
four
vmware
is
actually
one
of
them
groups
that
have
come
over
and
written
their
adapters.
B
We
don't
have
a
vmware
tanzu
service
mesh
adapter,
mostly
because
that,
because
we've
been
working
with
that
team
for
a
year
in
advance
of
them
finally
releasing
the
product
and
now
that
it's
released,
we
haven't
gone
back
to
go
to
go
look
in
part,
because
I
think
you
need
to
go
probably
download
like
sign
in
and
download
the
bits
they're
not
just
generally
available,
but
the
process
is
fairly
lightweight.
We,
I
think,
we've
yeah.
B
B
B
I
believe
there's
a
version
of
this
deck
on
this
page
that
highlights
who
wrote
theirs
and
who
didn't
so
so
citrix
wrote
their
own
octorine
wrote
their
own.
B
B
Yeah
and
so
hussein-
actually
that's
a
great
we
should
you
we're
going
to
talk
about
measuring
adapters.
I
think
on
friday,
because
it's
it's
a
really
good
conversation
to
have
the
adapters
themselves,
they're
kind
of
in
a
v1
architecture
and
and
what
ended
up
happening
to
us.
Is
we
repeated
ourselves
a
bunch
of
times
over
like
there's?
B
Some
of
the
same
code
is
in
each
of
the
individual
adapters
and-
and
we
shouldn't
do
that,
and
so
we've
known
for
a
while
that
we
wanted
to
pull
out
common
functionality
like
hey,
to
deploy
a
service
mesh
using
helm.
As
an
example,
that's
pretty
common
functionality,
a
lot
of
service
meshes
have
helm
templates,
helm
charts
that
you
can
go
use
so
we're
refactoring
to
pull
some
of
the
common
code
base
into
an
adapter
library
that
can
be
imported
into
each
of
the
adapters.
B
So
we
can
stop
repeating
ourselves
good,
that's
an
excellent
step
toward
being
having
a
higher
quality
project
easier
to
sustain
project.
We're
going
to
talk
about
this
on
friday.
I
think
there's
or
I
know
for
a
long
time
when
we
before
we
first
started.
Writing
the
first
adapter
the
open
question
was
well.
Should
these
be
individual
kubernetes
custom
controllers?
B
Should
we
create
custom
resources
that
represent
these
different
types
of
service
meshes,
and
so
we
may
end
up
with
a
year
from
now,
six
months
or
nine
months
from
now,
a
third
rendition
of
the
what
the
adapters
actually
look
like
it
might
be
that
they
are
less
about
running
as
individual
containers
and
more
about
running
as
individual
kubernetes
controllers.
B
If
we
go
that
route,
part
of
us
thinking
that
way
has
been
the
creation
of
the
measuring
operator,
the
term
the
kubernetes
concepts
in
term.
I
don't
know
if
it's
confusing
for
everyone
else
here
or
not,
but
it
can
get
confusing.
Certainly
for
me,
the
difference
between
an
operator
and
a
controller
and
suffice
to
say
that
there
is
a
collection
of
individuals
that
meet
on
tuesday
that
work
on
kind
of
the
mystery
operator,
a
component
of
it
and
the
the
first
component
that
they're
really
focusing
on
is
mesh
sync.
B
So
now
I'm
bringing
up
all
kinds
of
topics
and
projects
like
I
was
telling
you
everyone
here
earlier-
is
like
there's
a
lot
of
projects
as
good.
It's
a
lot
to
do.
Mesh
sync
is
very,
very
simply
it's
just
the
fact
that
mesh
re
needs
to
stay
synchronized
with
what's
going
on
on
a
mesh,
and
what's
going
on
inside
of
kubernetes,
pretty
simple
in
concept.
B
There's
a
v1,
an
early
version
of
what
mesh
sync
is
in
in
it's
it's
inside
of
meshre
server.
Today,
it's
very
simple:
it's
basically
measury
server
reaching
out
and
interrogating
kubernetes
saying
give
me
a
list
of
of
all
of
your
deployments,
I'll
look
through
there
and
see
if
any
of
those
look
like
a
mesh
if
they
do
I'll
I'll,
pull
that
out
and
I'll
I'll
get
more
information
about
them,
I'll
pull
out
the
services
and
the
deployments.
B
B
B
B
Some
material
that
you
can
peruse
on
this
on
mesh
sync
like
like
hey,
like
here's,
a
good
folder
to
just
digest
a
few
things
within
some
of
this
needs
cleaned
up
a
bit
to
help
clarify
I'll,
be
working
on
that.
So.
B
B
B
Anybody
I
know
that
I
know
it
probably
sounds
like
I'm
placating
hussannah,
but
I'm
not
the
code
that
you're
you've
changed
usana
the
mastery
ctl
system
start
well,
charles
who's,
on
the
call
can
tell
you
anyone
that
comes
to
use,
mesherie
they'll
be
invoking.
Hussain
has
code
like
that's
how
you
start
measuring
today,
so
so
already
like
anyway.
I'm
excited
because
it's
like
already,
you
can't
use
mastery
without
touching
her
santa's
code.
A
Right
thanks
very
sessionally,
thanks
for
spending
15
minutes
with
this,
you
find
all
the
links
that
lee
has
referred
to,
that
I
have
represented
the
meeting.
Let's
go
look
and
that's
all
that's
all
for
today.
B
Nice
see
you
see
you
guys
tomorrow
on
the
community,
call
bye
all
right,
see
you
guys.