►
From YouTube: Meshery CI Meeting (Aug 26th, 2021)
Description
Meshery CI Meeting - Aug 26th, 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
A
A
B
Another
one
depending
upon
who
all
comes
today
is
well
is
the
the
service
mesh
labs
in
part,
because
we
sort
of
bring
together
ci
and
devops
center.
You
know
office-centric
things
so
surya
shankar
since
you're
on,
maybe
and
since
anshu
is
here
or
whoever
is
interested,
like.
Maybe
a
discussion
on
service
mesh
labs,
I'll
drop
that
in.
A
It
is
the
the
issue
I
think
appears
is
not
integral,
but
he
is
recommending
to
enable
this
line
by
line
check
in
order
to
know.
Where
is
the
problem
in
in
the
in
the
linting,
for
example?
A
So
I
think
we
can
add
that
in
the
code
cop
ui,
I
will
read
this
in
order
to
to
enable
this.
I
think
I
have
the
enough
permission
to
do
so.
The
action
team,
the
action
item
to
do
here
is
for
me,
and
I
will
enable
this
in
the
code
code.
A
A
B
Oh
very
good
yep
it
is,
I
mean
it
is
the
answer.
It
is
the
one
on
you
good
rudolph
briefly.
You
might
not
know
this
just
yet,
maybe
in
advance
of
having
dug
in
here,
but
I
guess
one
of
the
questions
for
me
is.
B
Well,
well,
yeah,
I
think
part
of
the
value
of
having
this
enabled
is
that
it
becomes
one
is
you
can
anyone
can
go,
look
and
see
which
lines
are
not
covered
so
like
which
lines
you
know,
if
you're
going
to
try
to
get
it
from
30
to
50,
like
you
know,
what's
already
covered
so
that
you
you
know.
That's
that's
helpful.
That
that's
I
mean
that's
really
helpful,
especially
as
we
get
to
like
higher
percentages.
It's
like
okay!
B
Well,
I
don't
know
if
it
would
randomly
suggest
that
someone
write
a
unit
test
for
something
because
it's
like,
I
think,
some
of
the
other
apps.
They
will
like.
Let's
say
that
we've
got
a
bunch
of
lint
issues.
B
I
it
felt
like
in
the
past
to
me
like
if
there
were
25
lint
issues
that
you
might
get
a
randomly
selected
set
of
five
on
a
pr
that
just
sort
of
tell
you
like,
there's
all
25
that
are
reported,
but
you
might
get
like
a
set
of
five
like
down
at
the
bottom
of
the
pr
saying.
Oh
check
this
out
check
this
out,
and
I
wonder
if
that's
in
part
what
would
happen
here
or
not,
but
just.
A
B
Maybe
the
third
benefit,
actually
it's
kind
of
in
that
screenshot
right
there.
So
I
think
this
is
saying:
there's
an
individual
that
created
a
pr
they've
contributed
some
code
and
it
looks
like
it's
saying
that
the
the
lines
that
were
added
in
that
pr
here's
the
lines
that
you
have
yet
to
cover
so
like
that's
to
be
really
encouraging
to
you
know
to
those
that
are
adding
new
yeah.
So
that's
good!
That's
cool.
A
A
So,
yes,
is,
is
a
huge
amount.
B
It's
yeah,
it's
like
it's
embarrassingly
huge.
Actually
and
it's
a
you
know
it's
a
prohibitor
to
people
potentially
using
either
using
measuring
or
contributing
to
measuring
that
as
a
user
by
default.
If
you
just
down,
if
you
just
install
the
default
configuration
of
meshrie
it
downloads
each
of
the
mesh
readapters,
the
adapters
themselves
are
pretty
they're,
pretty
small.
I
mean,
I
think
it's
like
between
you
know,
maybe
on
average
70
megabytes
or
something
pretty
small,
but
if
that
other
one
is
taking,
you
know
five
minutes
or
hour
long
to
download.
B
B
B
Sorry,
I
got
distracted
by
a
different
note
that
so
when
mescheri,
the
docker
file
from
ashtray
has
different
build
stages
or
I'm
sorry,
it
has
different
stages,
a
build
stage
and
like
a
production-
and
my
ignorance
comes
in
about
here
when
we
go
to
that
second
step.
B
We've
done
when
we
built
some
c
plus
c
and
c,
plus
plus
things
in
phase
one
and
we're
going
to
intentionally
like
the
point
of
the
second
stage,
is
to
like
slim
that
sucker
down
like
trim
the
fat
and-
and
I
can't
believe
that
you
know
we
don't
you
know.
I
can't
believe
that
those
two
c
plus
programs
require
an
image
size,
an
image
of
that
size
like
they
don't
need
all
of
those
kernel,
headers
or
or
they
don't
need
all
that
additional
bloat.
B
So
there
is,
I'm
convinced
that
we
can
reduce
that
one
of
okay
and
so
that
that's
been
a
problem
for
a
while
we've
yet
to
have
someone
raise
their
hand
and
say
hey,
let
me
let
me
really
give
that
a
try
and
what's
sort
of
happened
in
the
time
that
we've
been
discussing.
B
This
is
that
an
acknowledgment
of
that
that
nighthawk
in
specific
is
that
the
way
that
we're
integrating
nighthawk
we
would
like
for
that
to
change
or
then
measure
it
and
that
being,
we
would
like
to
go
ahead
and
pull
nighthawk
out
of
the
mesherie
server
container
image.
Put
it
into
its
own
container.
B
B
So
that
would
be
fantastic
because
you
know
we're
talking
like
dropping
it
to
like
a
hundred
megs
two
hundred,
because
something
you
know:
okay,
wonderful
now
there
is
the
psychological
effect
that
it'll
have
on
people
where
they
go
to
install
measuring
and
again
by
default.
It
would
be
like
instead
of
just
10
containers
like
it's
11
containers
and
you
know
and
they're
like
holy.
B
You
know
what
is
this,
but
if
they're
all
pretty
tight,
you
know
if
they're,
all
pretty
slim
and
it's
kind
of
like
a
you
know
what
app,
what
you
have
an
application.
That
is
this
capable
has
this
many
features
like
what
application
do
you
not
use
today
that
isn't
comprised
of
a
couple
of
you
know
where
it's
a
distributed,
app
you
like,
like
others,
and
so
I
I
don't
think
people
I
feel
like
that
psychological
effect
surmountable,
like
I
don't
feel
like
people
will
push
back
that
much
now.
B
B
This
is
equally
as
well
as
I
am,
but
it's
it's
to
when
someone
runs
a
performance
test,
not
just
generate
load
from
one
load
generator
and
have
that
you
know
and
analyze
that,
like
that's,
that's
great,
but
it's
also
to
instantiate
any
number
of
those
nighthawk
containers
and
then
and
then
invoke
a
performance,
analysis
and
load
generation.
That's
coming
from
this
multi
factor,
so
a
distributed
performance
analysis.
In
order
to
do
that
good.
Well
then,
let's,
if
we
have
nighthawk
separate,
we
can
you
know
spin
up
different
instances
and
spread
share.
B
The
load
share
the
load
generation
anyway.
So
that's
the
yeah
and
to
rudolfo's
point
like
I
am
dumbfounded
that
I
couldn't
find
an
open
issue
talking
about
the
blood.
A
Yeah,
so
for
that
reason
I
growed
this
in
order
to
create
a
mission
and
follow
all
the
track
and
on
this
issue
I
think
we
have
two
issues.
One
is
to
separate
the
get
nighthawk
code
and
the
other
one,
these
c
and
c,
plus
plus
that
you
mentioned.
But
I
I
I
I
don't
too
familiar
with
that,
but
once
I
created
get
knighthood,
could
you
li
create
the
other
one
for
yeah
great.
B
Yeah,
well,
here's
the
I'll
I'll,
be
honest
and
say
that
second
issue
is
it's
the
easier
of
the
two
and
in
part
it's
the
easier
of
the
two,
because
rudraksha
is
sitting
right
here
and
I
know
he's
gonna
he's
gonna
help
on
that.
Second,
one
yeah
in
part
like
like
the
yeah,
so
the
thing
is
like
those
artifacts
the
c
program
and
the
c
plus
plus
program.
Those
artifacts
are
pretty
they're
pretty
you
know
small
binaries.
So,
like
the
container
image
size
that
would
run
one
of
those.
A
A
D
B
And
so
good,
to
give
us
a
leading
or
something
that
will
give
us
an
indicator
as
to
how
small
we
might
be
able
to
get
the
nighthawk
container
is
the
envoy
project.
B
A
If
not,
we
can
continue
with
the
next
one
and
the
next
one
is
propagation
of
issue:
remind
label
and
label.
Commenting
workflow,
I'm
not
familiar
with
this
one.
Is
this
label
new.
B
B
If
I,
if
that's
okay,
I'll
share
my
screen
to
show
okay.
B
Yeah,
so
all
right,
so
we
have
as
a
general
phenomenon
not
specific
to
this
project,
but
your
contributors
come
by
and
find
open
issues
raise
their
hand
and
instead
like
to
work
on
something
and
and
either
in
that
scenario
or
the
scenario
where
contributors
have
been
here.
A
real
long
time
been
working
on
all
kinds
of
things,
but
not
all
of
the
things
that
they
sign
up
for
land
or,
like
you
know,
stuff
gets
stalled
and
people
forget
about
them.
B
You
know
stuff
happens,
and
so
you
know,
we've
got
172
open
issues
just
in
this
repo
and
there's
a
lot
of
repos.
B
So
you
know
occasionally
going
by
and
checking
in
on
people
making
sure
that
they're
not
stuck
or
you're
trying
to
support
them
or
trying
to
figure
out
if
it's
time
for
someone
else
to
work
on
the
issue.
That's
a
good
cure.
You
know
curation
activity,
good
kind
of
project
management
activity
and
it
can
get
a
little
old
like
writing
down
the
same
thing.
Every
time
or
like
you
know,
hey
you're,
just
checking
in
and
how
are
you
doing
you
know?
Can
we
help
you
and
it's
like?
Well?
B
So
there's
in
a
few
of
the
repos
like,
I
think
that
this
is
broadly
propagated
already.
I
believe
that
if
you
go
to
the
vast
majority
of
their
repositories
across
the
four
orgs
by
the
way,
because
when
we
donated
mescheri
when
we
donated
surface
mesh
performance
when
we
created
service
mesh
patterns,
so
there's
the
layer,
you
know
there's
a
few
orgs
and
more
consistent
that
they
are
the
better.
Because
a
lot
of
us
work
across
you
know
all
of
them
and.
B
Excuse
me,
excuse
me,
and
so
we
had
written
a
label,
so
we
created
a
label
that
there's
a
category
of
labels
called
they
kind
of
start
with.
Actually,
actually,
I
think
most
including
rodolfo
knows
this,
but
just
there
is
we
try
to
document
the
crap
out
of
a
lot
of
things,
and
one
of
those
things
is
like
is
process
so
there's
a
process
about
github
get
there
for
our
github
process,
and
this
is
slightly
dated,
but
you
know
pretty
close.
B
I
think,
and
that
is
like
hey
what
labels
do
we
have
and
what
was
their
purpose
and
like
how
are
they
in
anyway?
There's
a
collection
of
labels
issue
invite
okay
yeah,
that's
a
that's
another
one!
That's
pretty
interesting!
We
could
talk
about
another
time,
inbound,
stale,
we'll
fix
so
basically
whatever.
Whenever
there's
an
issue,
whenever
labels
are
added
to
issues
or
pull
requests,
there's
a
workflow
that
will
kick
off
that.
B
The
configuration
of
that
it
says
that
if
you
apply
a
label
to
with
the
name
issue,
slash
remind
then
when
an
issue
is
labeled.
So
if
you
apply
that
label
to
an
issue,
then
it'll
say
this
should
check
in
it's
been
a
while,
since
you
know,
we've
heard
from
you
on
this
issue
or
you
will
go
and
and
please
let
us
you
know,
please
let
us
help
you,
and
so
I
was
the
the
note
here
is
just
to
do.
B
B
Repo
template
that
has
all
these
70-something
labels
defined
and
we
generate
new.
We
just
got
done
creating
two
new
repos
last
week
and
we
did
those
from
template,
and
so
they
get
they
they
benefit
from
having
all
those
those
the
predefined
labels
associated.
When
we
generate
from
template
that
template
also
includes
you
know
a
number
of
predefined
workflows,
but
it's
not
retroactively
applied.
So
there's
no
doubt
that
there's
some
repos
that
need.
B
B
A
E
I
want
to
tell
something
solely
actually.
I
was
looking
into
this
issue
remind
part,
but
I
want
to
show
you
something
like
to
get
more
clearance.
Can
I
share
my
screen
once
sure?
E
Okay,
so
I
think
the
you
are
asking
this
like
whenever
we
will
label
like
this.
I
add
a
level
of
issue
remind
then
a
github
action
will
be
like
this,
and
here
it
will
say
that
whatever
the
statement
is
like
this
way
right:
okay,
okay,
okay,.
B
A
I
just
I
see
that
a
rudraksha
is
interested
in
interesting
on
working
on
this
one.
So
please
could
you
create
an
issue
on
the
repo
that
you
will
add
this
label
in
order
to
open,
appear
and
link
that
issue
through
directions,
but
thanks
for
the
help.
A
A
Okay
thanks
everyone,
so
we
can
continue
with
the
next
topic.
A
A
The
next
topic
is
error.
Codes,
utility
workflow
doesn't
create
comics
in
measure
repo
for
some
adapters.
I
remember
I
remember
something
like
this.
Is
that
a
in
goal
we
are
creating
for
graphql.
If
I
am
not
wrong
some
code
errors-
or
I
I
don't
remember
if
if
it
is
graphql
or
api
in
general,
but
we
are
generating
our
goal,
but
we
are
generating
their
codes
right.
A
D
D
D
D
D
A
So
can
you
assign
to
to
suryas
in
order
to
start
working
on
that
yep.
D
And
this
will
be
a
good
thing,
like
you'll
also
need
to
take
care
of
spreading
it
into
other
repositories
where
we
are
still
not
able
to
do.
This
like
sq,
is
an
example
other
than
this.
If
you
go
to
engine
x,
you'll
see
the
same
issue
there
as
well,
and
basically
every
adapter
that
you
don't
see
here
has
this
issue
and
you
have
to
help
you
can
also
bring
in
other
contributors
and
take
them
into
this.
A
So
here
I
assigned
you:
if
you
have
a
any
question,
please
let
us
know
yeah
great
so
before
continuing
with
the
next
topic.
There
is
another,
a
question
about
you
that
a
is
the
code
called
workflow
supposed
to
run
for
all
purchasing
forks.
It
usually
ends
up
in
error.
A
C
A
A
Is
all
is
service
mesh
labs?
The
thing
that
you
mentioned
before
so.
B
Yeah
so
serious
anchor,
do
you
wanna.
C
So
like
this
is
one
of
the
errors
I
get.
I
end
up
with
this
first
like
no
such
file
or
directory,
but
this
is
like
only
one
of
them
like
this
is
not
persistent
like.
Sometimes
I
get
it
sometimes
I
don't,
but
even
if
I
don't
get
it
and
it's
like
works,
then
when
I
select
the
provider
like
nothing
happens
it
just
like
reloads
the
same
select
provider
page
and
let
me
see
if
I
can.
I
should
I'll
just
refresh
the
page.
B
Catacota
has
four
to
six
or
eight
different
images
to
choose
from
and
they
they
start
you
at
different.
They
have
different
starting
points
like
there's
one
image
for
ubuntu
and
it's
just
ubuntu
to
have
another
image.
That's
ubuntu
plus
mini
cube,
and
I
think
this
is
probably
that
one
they
have,
but
they
have
different.
They
have
different
versions
and
like
the
one
that
we
ideally
use,
is
like
one
that
already
has
a
running
kubernetes,
that's
already
up
and
ready
to
go,
and
then
we
just
start
installing.
B
B
B
B
Okay,
if
you
click
on
the
the
now
so
that
that,
like
that's
an
issue
right
and
we
have
to
figure
out
you're,
running
brave
right
now,
right
yeah.
So
that
seems
like
a
browser.
C
None
of
these
issues
are
like
consistent,
but,
like
I
always
face
some
issue
other
like
regardless
of
the
browser.
B
Okay,
yeah
that
that's
bothersome,
that's
for
us!
That's
in
part
why
we
have
a
project
to
create
the
world
service
mesh
playground
like
we
do
want
to
do
this,
and
do
it
really
well,
but
until
we
get
there
we'll
have
to
leverage
catacota.
So
this
particular
issue
right
here
seems
like
a
brave
slash,
catacota
challenge
before
you
nav
anywhere
too
far.
Can
you
click
on
the
cube?
Ctl
get
nodes
again
and
if
it
doesn't.
C
B
Cool
okay,
so
this
is
almost
got
nothing
to
do
with
measuring
in
our
labs.
Specifically,
it's
just
got
much
more
to
do
with
like
does
your
browser
like
catacota,
or
does
your
browser
allow
a
catacota
to
be
embedded
in
another
environment?
Like
you
know,
there's
some.
C
B
D
C
B
I
think
one
one
thing
that
might
be
happening
is
that
so
I
just
try
I'm
trying
right
now
with
met
the
mastery
and
istio
lab,
and
I
get
the
same
experience
that
you
just
had,
which
is
like
it
just
kind
of
leaves
me
sitting
there
with
nothing
going
on
conceptually,
and
so
I
click
on
cube,
ctl
get
notes.
It's
like.
I
don't
know
what
you're
talking
about
okay.
So
then
I
click
on
the
next
step
and
it's
process.
It's
doing
you
know
it's
downloading,
measuring
like
this,
so
the
question
becomes
is
like.
Well
so
did.
B
B
D
Depends
on
the
service
mesh
that
we
are
trying
to
access
it's
foreign.
B
B
With
them,
okay,
we
should
I
mean
we
should
raise
that
but
like
even
if
it's
just
hey,
you
in
brave
brave,
doesn't
like
this
or
like,
even
if
it's
just
like,
we
should
be
raising
those
up
as
issues
on
github
and
like
I
I
know,
suhani
has
called
for.
B
C
C
B
Well,
I
think
I
just
lost
my
environment
because
I
maximized
it,
but
it
was
this
is
this:
is
my
environment
like
this?
Is
the
catacota
and
and
this
loaded
for
me,
I
just
chose
the
nun
provider
like
you
did,
and
this
loaded
again
like
that.
That's
an
inconsistent
behavior,
which
you
know
is
bothersome
because
it's
like
well
well,
what's
the
issue,
so
so
this
one
right
here
like
to
me
this.
This
is
wrong
already,
like
okay,
so
the
session
started
and
the
user
is
just
sitting
here
going.
B
Okay,
now,
by
the
way,
I've
seen
this
on
other
people's
labs,
not
just
ours.
So
it's
like
okay!
Well,
what
am
I
supposed
to
do
something?
So
I
guess
I'm
supposed
to
do
cube,
ctl
get
nodes,
or
it
does
say
this
session
will
auto
install
minicube
instead
of
other
environment
variables,
necessary,
run
measure.
This
may.
C
That
a
shell
dot
sh
usually
set
stuff
up.
So
if
that's
not
like
nothing
will
be
working,
I
guess.
B
B
So
definitely
raise
an
issue.
Please
please,
please.
If
I
will
event
eventually,
I
I'll
I'll
say
this
like
so
we've
been,
I
think,
you're
hot
on
the
trail.
This
would
be
great.
Please
do
it
if
you
don't
I'm
eventually
just
going
to
close
this
down
for
everyone
and
stop
asking
people
to
help
I'm
trying
to
go
we're
going
to
I'll,
just
go
fix
it
and
then-
and
that's
got
nothing
to
do
with
you
as
much
as
like.
B
It's
just
a
perpetual
frustration
for
me
that
this
seems
like
such
a
great
thing
for
so
many
of
the
contributors
to
come
and
work
on
and
learn
what
devops
are
and
like
and
service
meshes
and
play
with
them
and
learn
them,
and
but
no
one
has
really
sunk
their
teeth
in
and
so
out
of
nothing
more
than
just
frustration.
I
think
I
would
stop
asking
yeah.
Definitely
like
the
brave
you
should
definitely
file
the
issue
about
brave,
like
not.
D
So
this
little
message
that
I
just
sent
in
the
chat
is
about
bash,
not
working
like
this.
In
some
of
the
environments,
I'm
not
exactly
sure
what
is
this,
but
it
is
a
caveat
with
bash,
sometimes
that
it
doesn't
work
like
this
in
all
the
environments.
So
for
this
we
tend
to
use
env
cache
in
bash
script.
B
So
these
are
the
labs.
These
are
the
scenarios.
If
we
go
into
like
the
one
that
we
were
just
doing,
the
istio,
I'm
a
total
ignoramus
here
like
I
have
no
idea
how
this
you
know.
I
haven't
done
any
of
this,
but
so
I
don't
know
if
the
image.
B
B
Maybe
we
should
be
switching
to
kubernetes
cluster
running.
That
sounds
about
right
to
me.
B
Do
anyone
like
to
try
this
out
like
by
the
way
whenever
a
get
a
commit
is
merged
into
the
here?
It
auto
is
a
web
hook
that
fires
and
then
just
automatically.
I
don't
know-
and
this
is
just
my
ignorance
like
I
don't
know
what
the
best
practices
are
for
testing
before
you
actually
publish
that
kind
of
thing,
but
you
know-
and
that's
just
if,
if
anyone
I'm
willing
to
give
like
if
anyone
wants
to
get
is
gonna
del,
you
know
dig
into
this.
B
Let's
get
you
and
get
you
access
to
the
behind
the
scenes
here.
So
in
case
that's
helpful
in
terms
of.
C
B
Because
that
that
may,
like
the
suggestions
that
rude
rocks
was
just
saying,
does
make
a
lot
of
sense
to
me,
but
also,
if
we
can
leapfrog
past
having
to
wait
for
mini
cube
to
install
and
like
it's
just
kubernetes
is
right
there
at
the
command
prompt,
oh
great.
Let's
just
start
with
measuring,
let's
not
have
to
dink
around
with
kubernetes.
A
So
I
think
we
have
a
enough
action
items
in
order
to
start
working
on.
So
please
thank
you
for
your
attending
for
this
meeting.
We
we
will
have
this
meeting
again
in
two
weeks.
B
You
know
as
I
apologize,
and
that
is
that
anshu
has
been
you
know,
kind
of
waiting
patiently
for
a
little
while
to
talk
about
this.
I
made
a
comment
not
too
long
ago
saying
that
the
release
the
auto
generated
release
notes
that
that
they
categorize
as
they
auto,
generate
the
release,
notes
as
release
really
as
release
drafter
auto
generates
those
notes
that
it
will
categorize
pull
requests
the
updates
under
like
the
maintenance
category
or
the
docs
category,
or
the
new
features
category
based
on
labels,
pr
labels
labels
on
the
pr.
B
Okay.
Well,
us
humans
can
try
to
make
sure
that
most
of
the
prs
are
labeled
before
we
merge.
But
you
know
it's
another
step.
Another
thing
to
remember:
it's
like
well,
so
too
can
we're.
Already,
anyway,
we
can
auto
label
pull
requests
based
on
where
files
have
been
changed
in
what
file
paths
and
then,
consequently,
when
the
release
notes
are
generated,
yeah
so
anshu,
that's
the
that's
it
that's.
The
description.
A
I
see
okay,
so
it's
something
that
we
can
change
in
order
to
take
the
labels
instead
of
the
paths.