►
From YouTube: Layer5 Community Meeting (March 26th, 2021)
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
B
C
C
B
See
I
would
I
don't
it's
a
it's
a
little-known
fact
around
these
parts
that
I
am
more
or
less
incapable
of
saying
good
evening
at
this
early
an
hour.
But
since
I
already
got
to
say
good
morning
to
jesus,
I
the
good
evening
was
was.
B
Well,
hey,
while
we're
hanging
out
aditya
and
everybody
else
slap
your
name
into
the
meeting
minutes.
This
will
help
us
get
organized
a
little
bit.
B
B
B
While
I'm
we'll
spend
one
more
minute
before
we
get
started
and
I'll
say,
hussein
is
nice
to
see.
I
was
in
the
kubernetes
sig
network
slack
yesterday
and
saw
a
familiar
name,
so
it
was
nice
to
bump
in
pump
into
you
over
there.
I
forgot
I
was
about
to
say
hi.
I
got
shy.
E
Yeah,
actually
we
had
a
meeting
in
vmware,
so
they
were
asking
us
to
contribute,
so
they
have
given
these
channels
to
follow
and
joined.
So
I
joined.
B
D
B
I'm
sorry
good,
that's
that's
something.
That's
pretty
cool
did
and
I
suspect
you
know,
given
your
focus
around
networking
that
you
know
like
hey,
that's
a
natural
place
for
you
to
was
the
direction.
B
Was
it
general
encouragement
toward
open
source
projects
that
are
relevant
and
interesting
and
or
was
it
like
explicit
direction?
Hey
there's
a
couple
of
goals,
particularly
within
networking
within
kubernetes
that
or
what.
A
B
Well,
who's
saying:
let's
make
that
a
topic
if
we
could
there's
a
couple
of
other
things
I
want
to
tell
you
are
offered
to
you:
okay,
all
right
we're
about
five
after
folks.
B
We
are
about
five
after
I'm,
just
getting
a
smile
on
my
face,
seeing
seeing
some
some
names
that,
like
anirud,
that
just
tickles
me.
B
Good,
okay,
all
right,
hey,
so
somebody
else
should
host
this
call.
Clearly,
I'm
disorganized.
It's
the
26th
of
march.
We
record
our
community
meetings.
This
is
the
layer
5
community
meeting.
We
have
a
meeting
every
day,
some
of
you
some
of
those
meetings.
B
Almost
all
of
them
are
on
the
community
calendar
which
is
linked
somewhere
in
here
here
which,
by
the
way,
if
you
want
to
oh,
so
this
will
let
you
subscribe
to
it,
but
as
a
random
piece
of
information,
if
you
go
to
meet.layer5.io
you'll
get
a
different
version
of
the
calendar
here,
you
can
subscribe
to
it
here
in
your
mail.
Your
calendar
client
also
there's
a
nice
quick
reference,
and
I've
highlighted
this
fact
a
few
different
times.
B
The
reason
I
feel
compelled
to
is
that
that
url
meet.layer5.io,
I
don't
think
we
really
have
it
written
down
like
published
anywhere,
and
so
it's
easy
to,
but
it's
such
an
easy
thing
to
remember,
especially
for
many
of
you
who
join
multiple
calls
like
jubril
who
who's
who
was
just
on
yesterday's
ci
call
so
nice
to
see
you
brill
on
today's
call.
B
There's
a
number
of
people
on
this.
Call
that
hadn't
haven't
gotten
a
chance
to
meet
you.
Can
you
take
a
moment
to
introduce,
say
hi
and.
F
B
Nice
ow
nice
did
you
bro
all
right?
Well,
I
have
yet
to
figure
out
the
thing
to
make
you
know
to
like
harass
you
about
to
to
just
you
know,
to
make
you
feel
warmly
welcomed,
but
I'll
find
it
it's
anyway,
nice
to
have
you
bill
thanks
for
introducing
good,
to
have
you're,
frankly
like
good
to
have
another
african
there's.
Quite
a
few
nigerians
who've
joined
and
quite
a
few,
a
number
of
other
countries
in
africa.
B
That
are
more
toward
the
western
part
of
the
country,
whose
names
I
cannot
remember
anyway,
yeah.
B
B
Okay,
all
right
good.
All
right,
then,
we
also
have
aditya,
who
hasn't
had
a
chance
to
introduce
on
this
call
before.
D
C
H
B
Hey,
what's
the
what's
the
deal
when
when
when
do
you
say
banglore
and
when
do
you
say
bengaluru
they
they
they
both
are
the
same.
But.
C
B
Nope
nope
all
right
well,
just
before
I
forget
we
actually
bengaluru
bangalore
bengal,
is
was
top
of
mind
because
of
kubernetes
community
days.
We
should
talk
about
that.
So
hey
welcome
everybody.
That's
that's
new!
Welcome
back
everybody
else.
We
we
will
oftentimes
list
out
folks
individually
here,
there's
an
item,
a
couple
of
announcements.
So
there's
an
item
that
happened
this
last
week.
Abhishek.
Do
you
want
to
talk
about
this?
First?
One.
J
Yep
all
right,
okay,
so
we
just
recently
released
a
fiverr
version
of
mashri
which,
which
has
got
a
significant
changes,
or
we
can
say
enhancements.
J
B
Have
yeah
you
know
sorry,
I
I
got
you
all
mixed
up,
actually
just
just
just
the
announcement.
I
guess
about
the
fact
that
it,
but
but
your
topic
to
go
in
depth
below
yeah,
but
so
it's
so
yeah,
I'm
a
little
disorganized
today
and
I
apologize
there's
there's
one
other
item,
though,
that
you
abhishek
you
and
asuko,
maybe
michael,
had
been
working
on.
Let
me
let
me
share
it.
It's
a
blog
post,
there's
a
draft
of
that.
B
J
So
we've
introduced
a
new
adapter
for
the
traffic
service
mesh
which
was
earlier
called.
I
don't
remember,
can
you
remind
me
what
was
it
called
earlier.
J
M-A-E-S-H-
and
it
has
been
so-
an
adapter
has
been
created
for
the
traffic
traffic
mesh,
which
is
a
part
of
the
fiber
release,
and
there
is
a
beta.
So
it
is,
it
is
an
s
beta
version.
J
There
is
another
adapter
that
got
released,
which
is
for
nginx
service
mesh,
which
is
all
also
in
the
beta
version
and
yeah.
We
are
kind
of
accepting
prs
in
there.
So
if
anyone
is
interested,
they
can
start
with
adapters
as
because
they
are
relatively
a
very
simple
code
base
and
dynamic
detection
of
oh
yeah.
J
We've
we've
introduced
a
meshi
plugin
system,
which
is
basically
to
dynamically
load,
any
remote
provider
plugin,
which
has
a
specific
functionality
that
measurey
doesn't
do
and
on
under
the
wood
we've
used
different
technologies
like
go,
plugins
and
dynamic,
loading
of
react,
components,
etc.
So
that's
another
place
that
that
can
be
explored.
J
J
Apart
from
that,
this
mesh
kit,
which
is
already
there,
I
don't
think,
there's
anything
new
in
there.
There
is
certainly
mesh
sync,
which
basically
helps
us
to
help
measure
you
to
sync
the
cluster
state
with
sync
with
the
cluster
state.
J
I
will
come
back
to
that
when
my
turn
comes,
I
see
and
machine
operator
which
manages
different
controllers
or
different
custom
controllers
inside
kubernetes,
we'll
come
back
to
that
as
well,
and
meshes
got
a
new
graphql
endpoint
which,
which
is
being
rigorously
regressively,
used
in
a
couple
of
components
in
the
dashboard,
and
apart
from
that,
meshri's
state
management
is
done
with
a
very,
very
small
scale,
database
called
sqlite,
which
is
a
part
of
the
machine
container
itself.
J
B
That's
it!
That's
all
that,
that's
all
you
guys
did
sorry.
B
A
it's
a
major
release,
wow
okay,
some
of
you
might
have
questions
on
all
that
big
mouthful,
so
I'll
check.
You
you'll
speak
in
depth
a
little
bit
more
on
some
of
those,
but
all
of
the
links
that
I
was
just
pulling
up.
B
Those
are
all
those
are
all
available
in
the
community
drive
so
which
means
they're
all
available
to
all
of
you.
So
that's
the
so
we
need
to.
We
need
to
get
the
word
out
a
little
bit,
we'll
finish
up,
that
blog
post
have
that
go
out.
We'll
certainly
share
that.
I
think
asuko
was
considering
writing
a
blog
post
on
dev.2,
but
the
more
blog
posts,
the
merrier
or
the
more
noise.
B
The
better
that
noise
will
be
hopefully
will
like,
hopefully
coincide
with
the
consideration
of
measuring
and
service
mesh
performance
for
adoption
in
the
cncf.
B
So
both
of
those
projects
have
been
put
up
for
or
are
being
proposed
for,
donation
in
the
cncf,
and
so
this
coming
tuesday
there'll
be
an
evaluation
of
sandbox
projects,
which
is
the
level
at
which
we've
proposed
mescheri
and
smp,
and
so
so
next
week.
We'll
know
if
it'll
be
a
big
it'll,
be
a
big
moment
for
the
projects.
B
It
won't
really
or
won't
at
all
change
what
we're
doing
here
or
how,
how
we
do
things
or
the
roadmap
for
the
projects
or
any
of
that
as
a
matter
of
fact,
though,
it
might
actually
change
the
roadmaps
a
little
bit
only
because
there
will
be
more
activity.
One
of
the
things
that's
that's
happened
for
something
like
well,
both
measuring
and
service
mesh
performance.
B
B
I
have
my
doubts
about
yeah
anyway,
about
google
participating,
we'll
have
to
press
it
upon
them,
but
that
brings
us
back
to
the
topic
husseina
about
sigs,
so
who's
saying
that
there's
the
I
you
might
have
been
on
this
call
before
have
you
been
on
a
cncf
sig
network
call.
B
Okay,
not
that
just
so
that
you're
aware
that
it's
there
just
and
it's
the
reason
I'm
mentioning
it
one
one
is
because
you're
all
invited
to
go
to
the
see
the
cncfc
network
call
there's
a
lot
of
times.
We
end
up
discussing
the
well.
We
host
the
service
mesh
working
group
in
this
in
the
cncf
sig
network,
and
in
there
we
actually
talk
about
all
your
projects
a
lot.
Actually
we
talk
about
smp.
B
We
talk
about
knight,
get
nighthawk.
B
We
talk
about
service,
mesh,
conformance,
smi
conformance,
so
it's
a
for
not
only
a
place
to
discuss
the
things
that
you're
you
all
are
working
on,
but
it's
also
a
very,
very
friendly
place
because,
because
they're
your
projects,
so
so
who's
saying,
I
thought
I'd
just
point
it
out
because,
like
hey
you're
like
in
a
lot
of
respects
like
if
someone
were,
if
I
was
or
in
your
shoes
and
someone
were
to
suggest
that
hey,
maybe
that
would
be
a
new
thing
we
should
go
do
is
go
contribute
to
some
of
these
projects.
B
But
the
just
yeah,
I
guess,
there's
a
there's
a
lot
of
things
to
do
inside
of
the
sig
network.
More
than
I
can
handle,
which
means
that
there's
an
open
role
for
a
tech
lead-
or
maybe
two
tech
leads
so
who's.
Saying
that
just
word
of
word
of
warning
as
you
join
and
poke
around
like,
I
will.
B
I
will
I
will
make
some
things
available
to
you.
Okay,
next
topic
is
so
by
the
way.
If,
if
anybody
wants
to
know
like
hey,
if
you
want
to
know
where
those
that
meeting
is
it's
actually
just
go
to
meet.layer5.io,
there's
a
link
to
the
meeting
minutes
for
the
cncfc
network
and
the
service
mesh
working
group,
and
so
yeah
we've
got
a
few
kubecon
talks
coming
up.
I
think
two,
I
don't
know,
maybe
three
and
so
again
this
is
a
potential
place
to
participate.
B
I
think
dhruv
jumped
in
last
time
and
participated
in
one
of
them,
but
that
we
will
be
talking
about
our
we'll
be
talking
about
what
we're
doing
inside
the
cncf
sig
network.
So
we'll
be
talking
about
these
projects,
your
projects
again,
so
that's
great,
there's
kubernetes
community
days
has
been
kind
of
a
small.
B
It's
been
a
small,
it's
not
brand.
These
aren't
brand
new,
but
they've
are
now
being
taken
much
more
broadly
and
and
many
new
ones
popping
up,
and
so
there's
a
kcd
kubernetes
community
days,
africa,
whose
call
for
papers,
I
think,
ends
in
five
days.
I
think
and
another
one
for
bengaluru,
both
of
which
I
would
love
to
see
people
in
this
community
submitting
to,
let
me
know
if
I
can
help
or
let
me
know,
if
you're
going
to
do
it,
and
I
can
certainly
help
there's
a
lot
of
stuff
to
talk
about.
B
There,
while
we're
on
that
subject,
well
I'll
try
to
bring
up
the
I'll
try
to
bring
up
where
it
is
that
the
other
kcd
days
are
coming
to
and
where
some
are
proposed,
for
I
just
had
a
yeah
I'll
bring
that
up
later.
B
Okay,
let's
switch
these
topics
around
a
little
bit
so
that
it's
not
just
me
talking
this
whole
time.
One
is
perhaps
ruth
totally
putting
you
on
the
spot
on
this.
If
ruth.
If,
if
I
bring
up,
I
figured
you
might
be
able
to
introduce
the
newcomer
questionnaire
and
then
we
might
be
able
to
just
reflect
upon
some
of
the
responses.
G
Okay,
hi
hi
everybody,
so
I've
got
some
music
around
me.
I
hope
I'm
quite
audible.
G
G
G
It
has
not
started
yet
yeah
so
yeah,
so
we
have
for
those
that
are
not
aware
of
the
call
every
thursday
which
we
had
one
yesterday
with
like,
we
have
a
full
there's
like
a
full.
G
We
had
quite
of
seven
persons
during
the
call
which
the
jubilee
was
part
of
those
presents,
and
we
already
have
these
codes
on
thursdays,
where
it's
just
a
way
to
you
know,
introduce
a
newcomer
city
project
where
people
save
space,
where
you
can
like
ask
questions,
and
you
know
we
also
do
tutorials
in
doing
the
new
commercial,
but
for
a
while
now,
we've
not
had
any
tutorial
come
up
because
we've
not
actually
had
like
persons
to
you
know
giving
those
tutorials
right.
G
We've
had
to
draw
out
the
past
and
also
on
the
on
the
page.
You
see
like
a
list
of
tutorials
if
you
could
show
that
table
for
tutorials
on
the
on
the
site
itself,
so
we
have
a
couple
of
them,
but
looking
at
these
responses
here,
you
see
that
there
are
still
some
which
you
have
not
like
really
covered
up
on.
So
we,
I
love
volunteers.
G
How
gl
websites
work
introduction
to
gatsby
and
all
of
them
right?
So
if
you
you
want
to
take
up
a
tutorial,
don't
worry!
It's
not
something
scary
right,
we'll
take
you
through
the
process
of
creating
this
slide
and
you
know
giving
a
tutorial
just
a
30
minutes
tutorial.
G
So
if
you
want
to
do
this,
which
is
like
a
call
for
volunteer,
so
you
can
hit
me
up
on
slack
or
you
can
just
put
it
up
in
the
newcomers
channel
and
yeah
and
anybody's
welcome
to
you,
know,
modulates
the
meeting
and
part
of
the
newcomers
meeting
right.
So
you
can
also
contribute,
as
far
posting
the
call
or
showing
up
or
answering
questions
for
newcomers
in
newcomers
channel.
So
yeah,
that's
like
physically!
I'm
sorry
for
music
here.
B
Yeah,
I'm
jive,
I'm
sounds
like
you're
having
a
good
time.
I'm
I've
got
a
bit
of
fomo
going
on
nice
thanks
thanks
for
that.
So
some
of
you
are
already
aware
of
the
newcomers
meeting
on
thursdays.
It's
not
a
one-time
thing.
You
can
hang
out
in
newcomer's
meeting
as
long
as
you
like,
there's
also
the
mesh
mates
program.
So
ruth
is
a
mesh
mate
and
so
in,
amongst
a
number
of
other
things
that
she
does
in
the
community.
B
B
Good,
so
I'll
just
reinforce
the
fact
that
ruth
was
calling
for
volunteers
for
tutorials.
So
in
the
past,
she
and
others
have
helped
organize
a
steady
flow
of
tutorials
on
technologies
in
that
thursday
meeting
and
most
of
them
not
all
of
them,
but
most
of
them
are
listed
in
this
table.
B
We're
missing
one
and
actually
it's
an
opportunity
for
any
of
you
that
want
to
contribute
to
this
site.
There's
an
open
issue
to
list
out
another
tutorial
that
was
given
this
site
if
you're
unaware
are,
is
open
source
and
was
created
by
this
community
and
so
as
such,
it's
still
maintained
by
this
community.
So
so
there's
an
opportunity
to
contribute
right.
There.
B
Okay,
hey
as
a
note
speaking
of
the
community
and
speaking
of
these
projects
being
evaluated
by
the
cncf
one
of
the
ways
in
which
people
and
the
cncf
measure,
the
popularity
of
a
project
of
an
open
source
project
is
by
count
of
stars
on
github.
B
I
mean
slap,
your
stars
across
all
the
repos.
That
would
be
great,
but
I
guess
I
won't
be
shy
about
asking
you
to
do
that
if
you're
here,
hopefully
you're
learning
something
you're
being
built
up
you're
giving
back
to
others.
Hopefully
you
end
up
enjoying
the
community
in
their
projects,
and
so
asking
for
a
star
is
the
least
you
can
do,
but
yeah
put
put
your
star
in
there.
It
helps
it's
not
a
great
metric,
but
it's
the
it's.
B
Certainly
a
metric
that
everyone
uses
so
it'll
help
it'll
help
the
projects
get
into
the
cncf.
All
right
next
topic
up
utkarsh
databases
in
mesri,
server.
K
Yeah,
so
this
is
just
a
quick
update,
so
we
so
not
not
that
the
machine
server
is
entirely
using
sqlite
now
so
mystery
server
earlier
used
to
use
bitcast
for
storing
everything,
just
like
as
the
sharing
on
the
screen.
But
now
recently
we
have
started
migrating
that
storage
from
the
cast
sqlite.
K
A
lot
of
things
are
being
stored
in
sqlites,
for
example,
performance
profiles.
Mesh
sync
is
something
that
was
the
main
like.
That
is
something
which
brought
in
sqlite
so
that
began
store
mixing
data
into
into
database
and
then
can
do
smart
queries
on
it,
but
now
machine
servers,
other
components
are
also
are
also
starting
to
leverage
the
sqli
database.
K
So
recently
we
moved
to
performance
profiles
and
the
services
performance
results
into
sqli
databases,
a
few
things
like
user
preferences,
those
are
still
being
stored
in
bitcast,
but
they'd
be
definitely
moving
to
sqli
database.
So
why
are
we
moving?
Actually,
the
task
is
a
good
database,
a
good
key
value
for
database,
but
it
doesn't
offers
the
basically
good
querying
capabilities
which
we
actually
need.
K
So
we
can
do
happy
stuff,
which
we
would
which
we
were
doing
so
that
bitcast
can
do
some
kind
of
queries,
but
it's
a
very,
very
minor
abstraction
on
top
of
file
system
just
do
the
serialization
and
those
kind
of
stuff.
So
that
is
a
main
motivation
due
to
which
we
are
moving
towards
the
skill
right
so
that
we
can
do
good
queries
using
sql,
so
basically
leveraging
issue
as
much
as
we
can
in
mystery
server,
just
like
a
few
like
other
cloud
providers
so
yeah.
L
So
if
you
don't
mind
me
asking
questions
why
exactly
sqlite
like
there
are
the
que
okay,
do
you
perform
any
complex
querying
or
is
that
a
necessity
for
it.
K
So
so
the
queries
are
not
necessarily
extremely
complicated,
but
they
are
certainly
complicated
for
bitcast.
So
if
you
are
performing
joins
or
something
like
that,
then
so,
if
you're
using
bitcast,
then
first
the
bitcast
also
has
trouble,
while
so
because
it's
just
a
very
minor
abstraction
of
our
file
system.
So
if
you
want
to
read
something
from
it,
so
it
will
acquire
a
lock
and
then
won't.
K
Let
anything
else
read
through
the
same
file
unless
until
the
log
so
yeah,
basically
we
can
do
same
kind
of
queries
using
bitcoins,
but
it's
extremely
hacky,
because
it's
at
the
end,
it's
just
a
key
value.
Third
database,
you
can't
provide
a
key.
It
will
give
you
the
value,
but
then
you
will
have
to
go
and
basically
do
stuff
and
go
so
that
you
get
the
data
that
you
were
asking
for,
but
sqlite
that
offers
sql
and
we
do
certain
kind
of
queries
which
require
multiple
tables
to
so.
K
Basically,
we
need
access
to
multiple
tables
at
the
same
time
and
measuring
mesh
sync
uses
that
and
the
performance
profiles
are
also
using
that.
So
a
very
simple
example
is
apple
used
to
order
the
results
you
can
order
results
by
like
when
they
were
created
or
when
they
were
updated.
Doing
this
in
bitcast
again
is
possible,
but
you
retrieve
all
of
the
data
into
the
memory
and
then
you
actually
sort
it
in
memory
and
then
push
down
the
results.
K
L
Use
so
yeah.
I
know
that
part
was
fine
between
bitcast
and
sqlite.
That
part
I
sort
of
understood,
but
what
I
didn't
get
was
sqlite
itself.
L
Okay,
I'm
not
sure
if
you're
gonna
run
into
this
issue,
but
when
you
handle
multiple
users
and
multiple
access
access
rules
right,
like
you,
have
a
lot
of
issues
at
least
I
personally
face
quite
a
few
issues.
So
would
you
face
something
like
that,
while
like
down
the
line?
L
So
if
you
want
a
particular
user
to
you
know,
if
you
want
a
particular
authentication
method
for
a
particular
user
and
then
there
is
one
a
sort
of
super
user
or
you
know
a
measuring
a
user
for
measuring
to
interface
with
the
system.
So
when
you're
sort
of
dealing
with
these
multiple,
what
is
access
points
or
people
or
systems
accessing
it?
You
might
run
into
issues
with
sqlite
itself.
I'm
not
sure
if
I
was
clear
enough,
I
don't
know
if
it's
going
to
be
an
issue
also
so.
K
I
think
I
understand,
but,
but
I
I
don't
know
if
we
I
mean
I
that
would
be
a
certain,
it
would
be
a
use
case
that
is
providing
access
based
on
the
users,
but
that
is
definitely
going
to
be
a
use
case.
I
believe,
but
I'm
not
sure
where
exactly
we'll
be
getting
stuff.
Using
this
kid.
L
Okay,
I'll
do
one
thing
I'll
look
into
that
aspect
where
I
got
stuck
a
little
more,
hopefully
I'll,
be
able
to
frame
it
together
and
I'll
be
able
to
get
you
that
information.
Then
we
can
see
if
it's
truly
an
issue.
K
Yeah
yeah
and,
of
course,
yeah,
even
even
if
yeah
so
I
hope
I
hope
we
don't
have
to
migrate
from
sqlite
to
some
other
database,
but
even
if
we
do
currently
we
are
using
com,
so
that
is
kind
of
database
neutral
yeah.
Hopefully
we
don't.
B
And-
and
I
pronounce
I
don't
I
to
be
honest-
I
wasn't
listening
to
the
conversation
I
apologized
but
was
postgresql
the
suggestion,
I'm
just
guessing.
L
Postgres
does
solve
that
issue
from
what
I
know
I
had
to
in
my
project.
I
had
to
switch
over
to
postgres
because
sqlite
wasn't
cutting
it,
but
then
again
just
shifting
the
entire
thing
to
a
perspective
would
be
another
hell
in
in
on
itself.
So
I'll,
we'll
I'll
see
if
you
know
that
issue
true
is
truly
going
to
be
an
issue.
So
if
it
is,
then
we
might
have
to
support
stress
or
something
equivalent.
If
it
isn't,
then
it's
peaceful,
we
can
still
continue
using
sqlite.
L
Okay,
so
it
doesn't
behave
well
in
multi-user
environments
for
some
reason
the
access
can.
I
I
mean
this
was
an
issue
about
I
think
a
year
and
a
half
ago.
I
don't
know
if
they
fixed
it.
The
exact
issue.
Again,
I
don't
remember,
but
that's
that's
the
rough
justified.
H
Hey
hi,
so
the
update
is,
we
recently
updated
our
specs
for
the
smi
conformance
tool
to
the
latest
smi
spec,
which
is
0.6.0,
and
we
have
now
changed
our
site
to
load
the
dynamic
data
which
we
have.
So
if
you
go
to
the
near
fire
side,
you
can
see
the
results
for
each
of
the
services
for
the
latest
spec
of
sma,
and
if
you
hover
over
them,
you
can
also
see
the
detail
of
where
they
failed.
Why
they
feel.
B
Without
context
that
this
is
a
new
spec,
the
average
user
or
visitor
might
think
that
this
spec
isn't
doing
so
well
because
there's
a
bunch
of
red
x's
across
but
given
context
that
this
is
new,
then
that
there
was
an
update
and
a
shift
in
and
how
it
works.
B
So
osm
open
service
smash.
What
do
you
know?
H
Yeah
so
like,
as
you
said,
the
problem
is
still:
we
haven't
implemented
our
versioning
for
the
smith,
which
we
are
running.
We
are
basically
updating
into
the
latest
fake,
which
is
there,
but
probably
in
future.
We
will
have
a
way
in
which
you
could
go
back
and
see
whether
it
is
compliant
with
0.5.0,
0.4.0
and
stuff.
Like
that,
because
a
lot
of
measure
here,
I
have
not
upgraded
to
the
latest
0.6.0
so
like
for,
as
you
can
hover
to
traffic
miss
itself,
you
can
see
that
it
has
failed
at
step.
Four.
H
That
means
that
it
has
capability
for
traffic
access,
but
it
has
not
capability
for
the
v1
alpha
3.
they
most
likely
have
capability
for
the
previous
version
like
v1.
For
now
we
are
ready.
Some,
I
add
traffic
speed,
so
the
current
version
is
p
one
alpha
four,
but
since
they
are
failing
over
there,
they
have
capability
for
d1
alpha
three.
In
that
case,
and.
B
B
So
here,
okay,
so
as
a
as
a
visitor
as
a
someone
reading
the
report
somewhat
intuitively,
I
begin
to
understand
that,
like
hey,
there's
that
all
tests
have
passed,
there's
a
full
circle
here,
they
are
fully.
You
know
capable
that
oh
okay
they're,
you
know
they've
got
partial
okay,
you
know
they're
partially
passing,
but
there's
a
failure
along
the
way
and
so
drew.
This
is
mostly
commentary.
B
I
don't
expect
that
we
actually
do
anything
about
this,
but
just
so
so
traffic
mesh
made
it
so
far
with
traffic
access
kind
of
up
to
a
certain
point
and
they
all
it
also
made
it
like
up
to
a
certain
point
with
traffic
split.
H
Yeah,
that's
because,
like
traffic
split,
has
more
number
of
tests
than
traffic
accidents
so
like
if
they
haven't,
haven't
lived
up
to
the
half
of
the
test,
then
they
won't
show
half
of
them.
If,
if
it's
half
complaint
or
not
like
it's
a
lose
way
to
describe
this,
I'm
like
in
the
next
series.
I
would
suggest
rather
that
if
they
don't
fail
the
first
step
itself,
which
is
to
see
whether
they
have
the
traffic
access
resource
itself,
then
they
should
give
a
half
a
partially
compatible
like
let's
discuss
this
later.
H
B
Makes
sense
so.
L
Right,
so
do
you
think
it
will
be
useful
to
have
instead
of
a
half
check
or
something
you
can
have
like
a
sort
of
sub
list
or
something
showing
what
tests
have
passed?
What
have
failed
because
some
people
might
not
want
the
entire?
What
do
you
say
the
compliance
entirely,
and
maybe
just
one
of
the
use
cases?
Perhaps
I'm
not
sure
that
this
might
be
a
total
edge
case,
but
just
for
that
sake
is
it
doesn't
make
sense.
H
Yeah,
it
does
actually
like
the
current
thought
in
our
mind,
was
that
basically
they
will
see
that
which
text
case
is
failing,
and
they
will
come
over
here
and
see
which,
which
particular
test
case
is
trailing,
and
they
could
like
view
in
detail
what
is
happening
but
to
upgrade
the
table
to
show
which,
like
it
makes
sense
if
all
like
we
like.
Currently,
what
is
happening
is
like
if
one
of
the
test
is
failing,
it
is
not
checking
for
the
rest
of
the
test.
H
So,
like
you
know
that
up
to
this
point,
all
these
things
are
passing,
and
after
this
point,
all
these
things
are
not
checked
itself,
because
this
was
failed
over
here.
So
like
doing,
that
would
only
make
sense
if,
like
we
were
continuing
with
the
test,
if
the,
if
one
test
was
failing
itself,
that's
what
I
think.
B
Yeah,
so
to
rephrase
that
one
thing
that
that,
if
some
that
is
not
obvious
to
pranav
to
you
or
to
others,
is
that
is
that
what
dhruv
just
said
is
the
tests
themselves
are
run
sequentially,
and
so,
if
you,
if
you
see
a
tool
tip
that
says,
oh,
you
failed
at
step.
Four,
then
implicitly,
you're
green
from
one
through
three
and
then
it's
it's
at
four
that
you've
now
but
pronounced
to
your
point
like
oh
okay.
B
Well,
it
took
me
explaining
that
to
you
to
actually
understand
that
and
then
so,
like
there's
room
for
improvement
on
and
so
actually
that
room
for
improvement
brings
up.
A
good
brings
up
a
topic,
one
that
I
forgot
to
cover
earlier,
I
believe,
or
that
I
moved
it
down
below.
So
we
may
as
well
just
jump
to
it
real
quick,
and
that
is
that
many
of
you
are
familiar
with
the
fact
that
we
participate
in
internship
programs.
We
do
a
lot
of
internships.
B
Some
of
you
are
interning
right
now
we
do
internships
directly
through
layer
five.
We
also
do
internships
in
combination
of
layer,
five
and
other
these
other
communities,
maybe
or
these
other
organizations
there
are
some
of
those
are
coming
up,
and
some
of
you
have
no
doubt
some
of
you
have
joined
it
much
much
earlier
in
interest
interest
of
those,
and
some
of
you
are
graduates
of
those.
Some
of
you
are
interested
in
them
in
them
now,
and
I
don't
know
that
we've
called
it
out
recently
so
so
now
is
an
opportune
time.
B
I'm
calling
these
out
in
part
because
those
some
of
those
enhancements
that
pranav
is
referring
to
are
listed.
I
believe
one
of
those
is
listed,
maybe
not,
but
it
could
be.
B
B
So
there's
some
of
the
project
ideas
for
this
year
for
gsoc
are
posted
here
we've
I
will
have
more
lfx
projects
coming
up
and
we
just
applied
for
gsod
the
gsat
one
is
really
intriguing
to
me.
Some
of
you
may
be
saying
what
documentation
intriguing.
I
don't.
What
well
actually
like
documentation
is
extraordinarily
important
and
so
so
get
off
your
high
horse.
If
that's
what
you're
thinking,
but
in
this
case
this
particular
project
is
well,
it's
really
cool.
B
Actually,
one
of
one
of
it's,
not
one
of
a
kind
it'll
be
related
to
something
like
this.
If
you've
seen
play
with
docker
or
play
with
kubernetes,
which
is
which
I
know
austin
ends
up
using
a
fair
bit,
which
is
great
we're
going
to
make
a
play
with
service
mesh
play
with
mesherie,
and
this
falls
under
the
documentation
category.
If
you
will
so
so,
that's
actually
what
the
pro
proposed
project
is
about.
It
will
be
running
meshery
as
a
service,
basically,
but
then
wiping
it
clean
every
you
know
hour
or
something
like
that.
B
Something
else,
oh
little
known
fact.
Actually
ruth
was
talking
with
a
newcomer
rabja
earlier
yesterday
and
part
of
the
discussion
was
what
was
what's
the
genesis.
Why
did
measure
get
created
in
the
first
place?
What's
measuring
about
what
was
what
was
it
created
to
do
and
little
known
fact,
so,
the
the
maintainers
of
the
creators
of
this
are
good
good
friends
of
mine.
This
is
how
mesherie
got
started
when
we
first
began
the
first
code
from
sri
started
here
inside
of
play
with
docker.
B
B
So
so
there's
a
fun
factoid.
We
should
write
a
blog
post
about
that.
Okay,
I
remember
now.
The
the
other
thing
I
wanted
to
talk
about
with
respect
to
the
internship
programs
is
that
many
many
people
have
this.
The
question
of
hey,
that's
great,
like
I
see
some
of
you've
got
all
kinds
of
questions
about.
How
do
I
best
position
myself
to
get
an
internship
to
get
one
of
these
or
one
with
layer,
five
or
whatever?
B
I
Yes,
yes,
yes,
hello,
everyone,
so
I'm
quite
excited
to
present
or
discuss
with
the
community
today
regarding
discuss.layer5
and
lan.layer5,
because
on
the
lender
layer
5,
we
are
yet
to
get
started
on
designs,
which
also
I
umly
proposed
discussing
it
to
the
community
fest
to
know
their
opinion
on
what
we'll
be
looking
forward
to
what
we
want
to
look,
what
we
want
most
to
be
implemented
on
both
the
discuss
and
the
land
on
layer
5,
because
they'll
be
interior
parts
of
the
products
we
are
working
on
at
layer
5
and
let
me
quickly
move
don't
mind.
I
I
My
major
interest
right
now
is
knowing
what
the
community
looks
forward
to
on
to
be
implemented,
like
features
we'll
be
looking
out
for,
is
we're
trying
to
build
a
learning
platform
for
layer
5
seo,
or
we
don't
want
it
to
just
look
plain
without
us.
Wanting
to
educate
people
about
service,
mesh
measuring
and
everything
without
them,
then
giving
what
exactly
wants,
because
we
we
might.
I
I
I
want
life.
I
I
design
for
my
perspective
without
doing
the
research
we
should
have
in
this
meeting
we're
going
to
have
a
disconnect
between
the
community
and
layer
5
so,
which
is
why
I'm
trying
to
first
get
everyone's
impulse
regarding
what
we
want,
what
to
be
looking
out
for
features
layouts
and
that
will
spice
up
the
community
and
then
boost
newcomers
coming
in
and
getting
educated
and
also
with
the
old
timers.
Sorry
for
using
that
word
being
very,
very
involved
in
the
community
and
also
for
the
discourse.
I
Also,
since
we
are
making
it
more
interactive,
we
we've
added
something
like
a
search
feature
which
you
can
search.
Not
it's
not
the
standard
of
google
yet,
but
we
want
it
to
be
to
the
aspect
where
you
can
just
come
and
search
about
kubernetes
and
then
docker
play
with
docker
labs
play
with
mystery
and
all
and
you
get
what
you
want
because
before
before
we
got
to
this
iteration,
it
was
a
long
process
almost
like
five
or
six
discussions
went
into
this
meeting
so,
which
is
why
I'm
bringing
up
the
call
right
now
to
know?
I
And
I
don't
know
if
I
could
get
any
impute
right
now,
even
if
he's
in
chat
boss
and
regarding
what
the
comments
will
be
looking
out
for
so
so.
L
Have
you
seen
the
istio
discussed
or
testio
website
like
how
it
looks
and
how.
I
L
Yeah,
no,
no,
so
the
look
of
it
is
not
that
great,
but
definitely
seeing
the
figma
things
that
your
yours
already
looks
much
better,
but
the
markup
support
and
the
certain
autocomplete
features.
When
you
start
typing
and
commenting
that's
quite
good.
I
know
it.
It
will
add
a
lot
more
work,
but
it's
a
very
nice
to
have
thing.
If
you
could
it's
just
a
request.
I
So
if
the,
if
there
are
more
imputes,
if
you
have
anything
features,
you'll
be
looking
out
for
layouts
you'll
be
looking
out
for
things
you
like
also
actually
design
spec.
I
would
really.
I
Appreciate
google
form
and
then
drop
it
in
the
in
the
community
and
the
websites
and
channels.
So
you
drop
your
suggestions.
You
drop
your
inputs
I'll,
create
two
forms,
one
for
the
discourse
and
one
for
the
land
on
layer
5.
So
we
could
make
the
the
projects
all
projects
we
are
working
on,
so
I
would
like
to
carry
everybody
along
and
then
make
them
very
successful.
Just
the
same
way,
we
launched
the
new
layer,
5.
leah5.ng
website
and
your
community
was
excited.
I
B
Nice
balance
great
so
implicit
to
what
austin
was
saying
is
that
the
go
participate
in
the
creation
of
discuss.layer5.io?
The
design
spec
is
in
the
chat,
austin
and
josh
patel
and
there's
some
others
that
are
participating.
B
There's
a
number
of
others
that
are
participating,
but
josh
patel
is
saddled
with
with
the
responsibility
overall
delivering
it.
So
certainly
ping
josh
he'll
be
talking
more
about
it.
In
the
monday
meetings,
the
website's
meetings
on
monday
austin
was
giving
a
prelude
to
a
forthcoming
larger
project
around
creating
a
subdomain
for
expanding
the
learning
materials
that
we
have
so
that'll
be
that'll,
be
exciting,
that'll,
be
that
yep
that
would
augment
play
so
learn.layer5.I
would
augment
play.layer5.io
or
play.mastery.io
that
hosted
instance
of
meshrey.
B
C
So
what
I
kind
of
had
in
mind
was
like
I
just
went
through
dogs
of
measuring
and
what
I
thought
was
that
what
if
we
can
move
from
one
framework
which
is
with
the
measuring,
as
you
had
said,
is
in
jekyll,
so
what
I
thought
was
we
can
move
to
move
completely
to
doc.
Docusaurus.
C
And
so
completely,
if
you
get
started
on
this,
we
can
also
implement
the
algolia
search
for
our
site
and
also,
as
docusaurus
supports
internationalization.
B
Mind
hello,
so
very
good,
so.
B
So
this
isn't
the
first
time
that
that
someone
has
suggested
a
potential
shift
to
docusaurus
a
number
of
reasons
why
we
didn't
do
it
in
the
past
and
why
why
I
support
us
like,
for
my
part,
I
support
us
doing
it
yeah,
although
I
should
clarify
that
I
I
support
the
suggestion,
but
won't
offer
any
assistance
and
then
I'll
explain.
Yeah.
B
Ended
up
staying
with
jekyll
in
the
past
is
because
we
have
like
five
websites
now
for
different
projects,
all
of
which
are
on
the
same
technology.
Jekyll
super
easy
to
get
started.
People
are
able
to
come
and
contribute,
and
then
some
of
those
people
end
up
sticking
around
okay,
it
jekyll
and
depending
upon
what
theme
you're
using
can
be
lacking
in
a
few
areas.
So
if
alonso.
A
B
He
has,
you
know,
has
worked
toward
bringing
like
international
language
support.
Yeah
we've
discussed
how
we
can
bring
versioning
to
this
theme,
and
so
it's
been
extra
work
work
that
you
would
have
otherwise,
probably
gotten
directly
out
of
docusource.
So
it's
clearly
a
better
framework,
and
so
I'm
supportive
of
that
suggestion.
What
I
was
trying
to
clarify
is.
B
So
actually
I'm
gonna-
I'm
gonna,
explain
this
for
just
a
moment,
so
so
supportive
but
we're
all.
But
for
my
part
now
this
is
just
my
part.
There
are
many
others
in
the
community
that
may
have
may
be
able
to
engage
and
assist.
We've
been
doing
a
lot
of
tech,
debt,
elimination
within
the
project,
so
all
of
the
adapters
from
sri
they
were
almost
all
of
them
have
been
rewritten.
B
Recently,
we've
been
working
on
kind
of
moving
away
from
bitcast
getting
over
to
sql
we've
been
bringing
in
a
new
operator,
we've
been
doing
like
a
lot
of
stuff
on
the
back
end.
That
is
the
right
thing
to
do,
but
from
a
user-facing
perspective
nobody
really
knows
or
cares
like
it
doesn't,
doesn't
really
change
eventually,
it'll
significantly
change
and
enhance
the
experience
and
the
user-facing
features
that
people
will.
You
know
benefit
from
in
measuring,
so
I'm
using
that
tech,
debt
elimination
as
an
example
of
like
moving
away
from
jekyll,
is
a
bit
like
that.
B
Like
sort
of
tech
debt
on
hey,
that's
that
was
the
right
choice
at
the
time.
Maybe
we're
going
to
shift
to
something
else,
but
I've
so
for
my
part,
I've
spent
all
of
my
calories.
B
For
my
part,
I
don't
have
any
space
left
for,
and
the
issue
with
so
you're
certainly
welcome
to
fork
this
work
on.
Thank
you
source
separately
or
we
can
create
another
repo
like
you
were,
you
know
suggesting
we
can
create
another
repo
and
you
can
work
on
that
in
context
of
the
community
and
you
can
and
we'll
I'll
we'll
tell
people
about
it.
Help
kind
of
get.
You
know,
help
uplift,
to
the
extent
that
we
can.
B
What
I
was
trying
to
say
is:
there's
a
zero
percent
chance
that
you
will
be
successful
in
delivering
that
without
assistance
from
maintainers
here
and
expanding
any
amount
of
energy
toward
like
at
some
point
we'll
have
to
so
it's
a
challenge
for
us,
because
it's
a
good
suggestion.
We
should
probably
do
that.
We
should
probably
do
it.
Post,
1.0
or
like
after
measuring,
has
reached
sort
of
self-sustaining
mode
or
reached
escape
velocity
in
terms
of
like
lots
of
users
benefiting
because
today
it's
the
project
is
well
done,
but
we
don't.
B
But
what
we
see
like
represented
on
the
community
call
is
30-something
contributors.
This
is
awesome
by
the
way
way
more
people
than
even
the
istio
community
call
gets,
which
really
says
something
about
the
way
that
community
is
run,
but
no
or
very
few
users
represented
on
today's
call,
which
means
we
haven't
reached
pmf
product
mark
a
product
market
fit
meaning
like
it's.
It's
features
aren't
compelling
enough
to
keep
bringing
people
back
now,
don't
get
me
wrong.
B
We've
had
about
a
thousand
people
use
measuring
individual
people,
and
many
have
come
back
to
use
it
again
and
again,
but
we
haven't,
but
so
that's
the
context
in
which
I
want
to
both
be
encouraging
and
then
I'll
also
be
a
realist
and
say
even
the
smallest
feature.
You
can
ask
the
vendor
or
husseina.
B
That,
like
they
were
adding
a
flag
just
this
week
on
dash
dash
silent
flag
or
dash
dash,
yes
or
something
I
forget,
and
I
can't
tell
you
how
many
times
they
went
back
and
forth
not
back
and
forth,
arguing,
but
back
and
forth,
like
refining
the
implementation
they
had
to
write
it
down,
they
had
to
do
peer
review,
reviews
and
etc,
and
they
could
have
been
spending
that
time
creating
better
support
for
kubernetes,
which
is
what
is
on
the
roadmap
they're,
not
doing
the
wrong
thing,
they're
doing
the
right,
they're
doing
the
right
thing.
B
It's
just
like
it's
a
slight
distraction
and
it's
okay.
Okay,
it's
like
so
I
I'm
giving
confusing
guidance
because
on
one
hand,
I'm
like
hey
for
my
part,
like
what
a
great
idea
like
like
if,
if
it's
facilitating
api,
docs
and
some
interactivity
with
like
the
apis
of
measuring
and
like
if
it's
clearly
if
it
has
versioning-
and
you
know
that
international
language,
support
and
stuff
like
actually
there's
only
two
two
just
two
detractions
from
doing
it.
One
is
that
we
lose
the
economy
of
scope.
K
B
So
it's
a
so
hey
so
here
I'll
I'll
I'll
put
I'll
I'll
lay
out
the
gauntlet
it
let
if
you
can
do
it
for
all
the
sites.
C
That
would
be
kind
of
tough,
but
I
would
I
would
first
get
started
with
this
misery
thing
and
then
move
on
to
the
other
sides.
Also
like
once
I
finish
with
this
misery,
then
I
can
move
on
to
the
other
sides
as
well.
B
E
C
C
Like
what
I
had
in
mind
was
like,
if
you
had
gone
through
this
flutter
page,
you
can,
you
can
run,
run
the
app
and
all
right
there
right.
So
if
we
are
planning
to
get
a
playground
for
measuring
the
future,
we
can
just
try
to
implement
some
of
the
functionality
in
the
documentation
itself
like
when
it's
a
small
bit
of
that
or
something.
C
Yeah,
what
I
said
was
like
if
you
had
gone
through,
if
you
have
seen
the
flutter
or
any
other
react
native
or
other
sites,
they
have
this
playground
kind
of
thing
which
is
built
inside
the
documentation
itself.
C
So
I
thought
I
just
had
an
idea
like
we
in
the
in
the
future,
once
we,
if
we
move
the
machine
dogs
to
some
other
framework
or
anything,
we
can
implement
this
playground
kind
of
thing
inside
the
documentation
itself.
B
Gosh
yep
makes
sense,
makes
sense
all
right
we're
just
that
we're
a
few
minutes
after
we
should.
We
should
wrap
up
adika
thanks
for
walking
through
that,
we'll
I'll
follow
up
with
you
and
slack.
B
Everybody
for
the
presentations,
the
discussion
today,
so
nice,
that's
a
wrap,
see
you
next
week
same
time.