►
From YouTube: Meshery Development Meeting (Dec 08th, 2021)
Description
Meshery Development Meeting - December 08th, 2021
Join the community at https://layer5.io/community
Find Layer5 on:
GitHub: https://github.com/layer5io
Twitter: https://twitter.com/layer5
LinkedIn: https://www.linkedin.com/company/layer5
Docker Hub: https://hub.docker.com/u/layer5/
A
Welcome
everyone
to
the
mystery
development
meeting
today
is
the
8th
of
december
and
we
are
at
kubecon
week.
So
we
have
kubecon
china
kicking
off
this
week
and
like
every
con,
we
we
do
represent
ourselves
a
lot
there
and
so
like
going
going
over
the
first
first
topic.
For
today
we
have
measuri
and
service
mesh
performance
project
office
servers
at
kubecon,
so
these
officers
are
scheduled
back
to
back
on
friday.
A
We
also
have
a
talk
by
mesh
mate,
anita
on
solving
the
service
mesh
adapters
dilemma.
A
I
think
we
mentioned
this
on
the
committee
called
last
week,
but
to
give
a
quick
update
on
it,
like
anita,
will
be
talking
about
machary,
obviously,
and
yeah
check
check
out
our
talk,
the
link,
all
the
links
are
in
the
minutes
and
we
also
have
the
tag
network
and
the
service
mesh
working
group
event
at
kubecon.
A
So
a
couple
of
our
projects
are
represented
within
the
service
mesh
working
group
in
the
cncf
tag
network,
so
measuring
machinery,
services
performance
and
get
night
talk
being
being
three
of
them,
so
this
is
on
thursday,
so
feel
free
to
jump
in
on
all
of
these
sessions
and
interact
with
interact
with
the
maintainers
and
learn
more
about
the
project
and
yeah.
As
I
mentioned,
all
the
links
are
in
the
meeting
minutes
and
the
link
to
it
is
in
the
chat
all
right.
A
So
let's
get
an
update
from
nitish
on
the
on
the
changes
to
measure
ui.
Oh,
I
guess
nitish
is
not
on
the
call
maybe
aditya
like.
Could
you
give
us
a
quick
update.
B
Sure,
okay,
so
last
last
we
spoke.
Ashanthan
was
working
on
a
couple
of
things
from.
Could
you
go
to
the
progress
tracker.
B
So
shanthan
was
working
on
a
couple
of
items,
so
basically
right
now
for
the
v
0.70
the
roadmap
for
measuring
ui.
We
are
shifting
over
to
redux
state
management.
So
now
we
are
in
need
of
people
to
create
reusable
react
components
to
recreate
measuring
ui.
So
this
is
the
progress
tracker
in
which
we
have
a
list
of
all
the
components
that
need
to
be
created.
B
So
if
anyone's
interested
in
creating
react
components,
they
could
put
their
name
against
one
of
these
components
and
get
started
on
working
with
them
and
yeah.
So
I
think
that's
about
it.
There's
no
other
update.
As
of
now,
I
think
we
are
going
to
redesign
a
couple
of
stuff
so
after
that
we'll
get
to
it.
C
C
D
C
C
Some
of
the
uis
will
just
be
enhanced.
It
can
remain
the
same
or
could
kind
of
be
enhanced,
but
a
lot
of
them
will
probably
get
more
dense,
there's
just
a
lot
more
functionality
forthcoming
and
with
a
little
bit
of
forethought
a
little
bit
of
user
experience,
design,
there's
less
pages
that
users
necessarily
have
to
navigate
to.
So
we
we
one
of
the
things
that
dish
had
worked
on
and
so
had.
C
Well,
there's
a
bunch
of
people.
Actually
I
can't
list
them
off,
but
have
they
had
worked
on
the
connectivity
wizard,
basically
like
an
initial
configuration
wizard,
which
is
currently
hidden
in
mashrey's
ui,
but
it's
a
concept.
That's
like
the
implementation
was,
is
challenged
and
that's
why
it's
hidden.
C
It
doesn't
function
as
it
necessarily
as
it
should,
but
that
concept
is
still,
I
would
suggest,
is
still
very
much
needed
people
they
come
to
a
service
mesh,
their
conf,
the
service
meshes
are
confusing
and
meshery
is
one
of
its
goals
is
to
make
them
much
more
palatable
part
of
getting
mesherie
up
and
connected
and
talking
to
the
various
system
entities
you
know
prometheus
and
jaeger,
and
and
each
of
the
different
service
meshes
and
kubernetes
means
that
you've
got
to
go,
set
up
and
configure
measuring
and
so
a
little
bit
of
a
a
step
through
connectivity
wizard
bringing
that
back
in
the
dot.
C
B
A
All
right,
thank
you.
Aditya
before
we
move
on,
I
think
we
have
a
new
camera
among
us
so
hungry.
Are
you
joining
this
particular
call
for
the
first
time?
If
so,
would
you
like
to
say
hi
to
the
community.
C
A
C
Yeah,
all
right,
so
in
so
that
next
topic
is
the
one.
That's
highlighted
it's
a
design
review
for
well
how
messages
and
message
channels
are
leveraged
inside
of
measuring
there's
an
open
consideration
around
yeah.
Well,
there's
a
yeah
there's
a
few
action
items
in
this
area,
so
today
measure
uses
go
channels
as
part
of
its
communication.
C
It
also
uses
gnats
as
another
part
of
its
communication
and
as
we
look
to
enhance
the
communication
between
the
various
measuring
components,
both
internal
to
both
internal
server
components
as
well
as
external
adapters,
the
performance
load,
generators,
the
measuring
operator
and
mesh
sync,
and
that
kind
of
a
thing
there's
a
there
are
two
design
specs
that
discuss
how
to
how
these
the
messaging
system
and
the
message
format
and
the
use
cases
for
them
unfolds
and
can
be
enhanced.
C
C
C
There's
a
collection
of
and
bomi
since
you're
on
the
call
you
might
be
able
to
do
this
in.
I
think
it's
in
the
measuring
channel
there's
a
collection
of
follow-up
action
items
on
this
topic
that
might
be
good
to
copy
and
paste
into
the
meeting
minutes
so
that
we
can
be
sure
that
they're
persisted
so
before
we
lose
them
in
slack.
C
But
okay,
assuming
no
questions
or
comments
on
that
topic,
it's
kind
of
a
deep
one.
The
next
topic
is,
is
really
like.
The
the
dot
six
release
is
upon
us.
It's
you
know
collectively.
The
project
is
essentially
ready
for
this
release:
they're,
like
its
predecessor,
the
dot
5
release,
which
was
jam-packed
full
of,
like
I'm,
not
sure,
two
or
three
or
more
new
architectural
components.
C
The
dot
6
release
also
includes
a
couple
of
new
architectural
components
and
some
significant
redesigns
in
some
areas.
C
So
it's
kind
of
a
big
thing
to
write
up
and
tell
the
world
about
which
also
means
it's
kind
of
a
big
thing
to
test,
and
so
in
tomorrow's
build
and
release
call
there's
a
bi-weekly
build
and
release
call
on
thursdays.
C
C
Is
it
meeting
criteria
and
is
it
time
for
the
full
release?
So
there's
a
test?
There's
a
spreadsheet
out
there
that
tracks
the
test
cases.
We
need
to
put
put
a
focus
on
it.
You're
on
the
call
you've
recently
brought
forth
a
lot
of
a
significant
amount
of
integration
testing
for
the
adapters.
C
If
you
would
before
tomorrow's
call,
do
you
mind
highlighting
the
the
test
cases
in
this
sheet
that
you
consider
those
integration
tests
cover.
C
E
Hey
lee,
I
was
testing
the
mesh
yesterday,
just
deploying
it
locally
with
the
rc2,
build
and
found
and
found
an
issue
with
the
liquidity.
Adapter
causes
the
container
to
crash
for
that
type
of
issue.
I
did
create
a.
I
did,
create
a
github
issue
for
the
link
or
the
adapter,
but
is
there
another
place
that
we're
tracking
issues
like
that
around?
This
is
about
tracking
system
that
you
use,
or
is
it
all
inside
github
yeah?
It's
a
great
man.
It's
that.
C
Like
yes,
and
no
there
we've,
I
hear
you
it's
github
is
we've
begun
to
use
click
up
a
bit
as
like
a
the
free
version
of
that
tool
to
help
track
across
repos
like
and
help
track
priorities
and
various
features?
And
when
I
say
we
like
there's
just
a
small
collection
of
folks,
who've
been
tracking
a
few
items
in
there
and-
and
so
we
don't.
C
I
figured
I'd
mention
that
it's
not
it's
probably
not
right.
Now,
it's
not
the
real
answer,
like
the
real
answer
is
no
one
of
the
best
places
for
us
to
probably
call
attention
to
like
a
a
release.
C
Pat
there's
hacks
that
we
might
use
or
tricks
that
we
might
use
is
when
you
search
through
when
you
do
a
github
search
on
the
web.
Ui
is
to
do
a
search,
that's
inclusive,
of
either
all
repos
in
a
repository
or
all
repo
or
all
repo.
I'm
sorry
did,
I
just
say
all
repos
and
repo.
I
mean
all
repos
in
an
org
like
the
measuring
org,
to
the
extent
that
measuring
depends
on
service
mesh
performance
that
spec
or
on
service
mesh
patterns.
C
Like
that
org
there's
a
few
orgs,
we
can
have
commonly
referenced
github
searches
for
like
specific
tags,
so
so
jared.
This
is.
This
might
be
a
good
topic
for
tomorrow's
build
and
release.
Call
that
currently
there
are
there's
a
a
github
process
document
in
the
community
drive.
It
talks
about
the
various
ways
in
which
we
have
different
labels
on
github
issues,
different
bots.
C
What
else
does
it
talk
about?
Just
like
it
is
a
process
document
about
what
we
do
with
in
handling
github
and
one
of
those
one
of
the
things
that
it
does.
Is
it
itemizes
what
issue
labels
are
available
and
what
their
purpose
is?
There's
I
think
three
or
four
labels
that
talk
about
severity
of
the
issue,
whether
it's
like
urgent,
high
priority
low
priority.
I
think
jared.
C
If
you
give
it
just
a
little
bit
of
thought,
you
might
suggest
you
might
have
a
scheme
to
suggest
like
either
we
change
the
names
of
those
or
we
designate
one
of
those
as
being
like,
or
we
create
a
new
one
like
one
of
those
as
being
release
blocking
or
you
know,
we've
got
priority
levels
or
something
like
that
like
it,
but
and
then,
and
then
we
can
have
a
commonly
referenced
github
search,
just
a
hyperlink
that
searches
across
repos
for
any
issue
that
has
that
label
and
that
might
help
drive
the
discussions
in
well
in
on
this
call
and
the
rest
of
the
calls.
E
C
Yeah,
no
thanks,
yeah,
don't
yeah,
don't
hesitate
to
suggest
a
process
if
you'd
like
to.
We
can't
really
go
wrong.
We
should
just
start
somewhere
with
issue
labels
or
or
something.
A
So
for
this
issue,
like
jared,
also
mentions
two
possible
solutions
like
is
there
like?
Is
there
one
we
can
choose
right
away,
or
would
we
need
to
look
into
it
like
if
anyone
has
any
opinions
about
that,
we
can
either
revert
the
image
back
to
this
or
disable
cgo?
E
Okay,
yeah
and,
like
I
said,
I'd,
be
happy
to
to
take
this
one
on
and
just
I
wondered
which
which
solution
would
be
preferred,
but
I
can
submit
the
pr
for
this.
C
Jared
that
would
be
fantastic,
so
yeah
just
seeing
the
issue
for
the
first
time,
just
as
one
voice
of
others
that
might
chime
in
here.
So
so
the
the
back
story
here,
part
of
the
back
story
here-
is
we'd
like
to
have
small
image
sizes.
C
That's
nice,
we'd
like
to
have
the
fewer
lines
of
code
and
the
fewer
dependencies
in
those
images,
the
fewer
the
smaller
the
like
attack,
surface
area,
the
smaller
the
smaller
the
number
of
security
issues,
hopefully,
and
so
that's
part
of
what
was
driving
towards
like
a
distrelus
image.
C
In
this
case,
though,
the
current
one
that's
using
to
your
point,
isn't
working,
so
we
can
revert
back
to
the
okay
to
a
different
distro-less
image.
Okay,.
E
Word
so
so
the
one
there
was
that's
the
image
that
was
being
used
before
the
that
pr
I
referenced
up
there.
It's
the
same
image.
It's
I
mean
more
or
less.
The
static
image
is
just.
It
doesn't
include
a
few
of
the
ellipses
packages
that
the
the
base
one
does.
It
does
say
that
the
the
tag
was
bumped,
but
I
I
don't
know
if
that's
exactly
accurate,
because
it
was
changed
to
a
different
to
include
different
packages.
E
It's
not
like
a
version
bump
and
I'm
not
sure
exactly
which
cves
were
included
in
this,
but
the
that
image
was
updated
a
couple
days
ago.
These
are
checking
it
that
in
gcr
I
saw
that
these
images
are
updated.
Pretty
frequently
looks
like
I
saw
one
a
couple
days
ago
and
one
about
a
week
or
two
ago,
so
they're
updated
relatively
regularly
yeah.
F
I
think
in
other
adapters
we
are
using
the
distro-less
node.js
one
and
I
think
we
should
use
that
one,
because
with
that
way,
we
wouldn't
have
any
possible
issues
with
the
the
cube
open.
Api
adjacent
schema
binary
that
we're
using
like
we
had
with
app
mesh
adapter
a
while
ago.
So
I
think
we
should
be
uniform
in
the
way
what
image
we
are
using
across
all
the
adapters.
So
I
think
we
can
go
with
that.
C
Nice,
it
sounds
like
there's
a
suggestion
to
go
with
your
first
option
jared
and
in
doing
so,
yeah
uniformity
across
the
adapters
is
ideal.
G
Actually,
the
binary
that
on
line
23,
if
you
can
actually
see
that
this
binary
cube,
we
open
api
json
schema.
Actually,
that's
it.
That's
something
that
that
adapters
use
to
to
convert
open
api
specifications
to
json
schema,
and
that
is
unfortunately
written
in
note
just
and
it
is
a
binary,
but
it
packs
a
lot
of
it
basically
comes
with
notice
runtime.
So
it
seems
like
it's
a
single
binary,
although
it's
entire
notice
and
that
actually
relies
on
a
lot
of
lipsy
dependency
because
no
just
a
tendency
process.
G
So
it
has
a
lot
of
dependency.
It's
it's
not
a
static
binary.
So
actually
that's
the
reason
we
are
using
notches.
We
have
to.
We
have
to
resort
to
using
nodejs
images
in,
even
if
we
are,
even
if
we
are
completely
statically,
compiling
the
binaries,
so
yeah.
Actually,
that's
so
I
think
that's
very
pointing
with
that.
Why
do
we
need
not
just
in
golden?
Actually
that
was
the
reason.
G
E
So
so
this
was
in
the
linker
d
adapter
in
the
docker
file,
and
I
didn't
see
that
schema
referenced.
There
wait
is
this
the
oh.
This
is
istio
yeah,
but
in
the
linker
d1
I
didn't
see
yeah
that
that
schema.
G
Yeah,
if
it
is
not
here,
it
should
be.
Actually
I
mean
it
is
not
here,
and
actually
it
should
be,
because
every
so
every
adapter
every
adapter
actually
relies
on
that
binary,
at
least
for
now
to
because
it
does
a
lot
of
auto
complete
generation.
So
actually
that
is
why
it
needs
that
particular
binary.
So
actually
it
should
be
here
and
probably
probably
we
missed
one
of
the
adapters.
While
we
were
adding
that.
E
Okay,
so
would
you
like
me
to
go
ahead
and
add
that
schema
and
change
the
image
as
well.
E
Okay,
all
right
cool.
Sorry,
I
think
we're
a
bit
off
topic
for
this
thing,
but
I'll
we
can
sync
up
and
slack,
maybe
after
this
and
and
I'll
I'll
clarify
make
sure
I'm
on
the
right
path.
A
Yeah,
thank
you
jason.
Thank
you
for
bringing
this
up
like
yeah.
It
helps
like
having
a
broken
linkedin
adapter
like
really
hurts
the
v0.6
release,
and
thank
you
for
noting
it
and
taking
time
to
actually
go
and
fix
it.
E
First
thing:
I'm
actually,
I
think
I
think
that
maybe
I'm
just
bad
luck,
because
every
time
I
touch
measuring
like
it's
something's
broken,
because
I
this
actually
seemed
to
block
me
from
even
starting
measuring
and
this
change
was
merged
a
couple
weeks
ago.
So
I
guess
nobody
ran
into
it
yet.
H
H
I
was
just
asking
like
the
whole
attendees
like:
is
there
any
best?
We
could
ask
that
allows
us
to
detect
this
faster,
like
during.
E
It
should
be
I
to
be
honest,
I
haven't
dug
into
the
the
tests
that
are
being
ran,
but
I
would
think
that
just
because
all
I
did
was
pull
the
latest
measuring
and
and
deploy
it
and
I
ran
into
this
and
then
was
also
able
to
repro
it
by
just
doing
docker
run
on
the
on
the
container
and
it
and
it
popped
right
up
very
consistently.
A
Yeah,
I
think,
like
we
can
cover
it
in
the
e2e
test
like
to
make
sure
that
everything
is
functional.
The
adapters
are
operational
and
running
yeah,
so
maybe
probably
we
should.
We
should
iterate
over
this
e2e
workflow
and
include
this
as
well.
A
All
right
so
so
we
will
have
the
build
and
release
meeting
tomorrow.
So
I
think
we
can
come
back
to
this
tomorrow
and
get
the
status
update
on
this
as
well
all
right.
Moving
over
to
aditya
and
other.
I
think
aditya
has
been
working
on
the
service
smash
buttons,
catalog.
B
Screen
yeah,
so
I
had
been
working
on
the
service
mesh
patterns
page,
which
is
now
the
service
mesh
catalog.
B
So
basically,
here
the
we
list
of
a
number
of
patterns
which
you'll
be
able
to
import
into
your
ui
right
now,
there's
just
a
button
that
copies
the
link
to
this
pattern,
and
you
would
be
able
to
paste
that
in
measuring
ui,
one
roadblock
that
I'm
facing
is
filtering
out
these
patterns
based
on
these
filters.
So
that's
the
only
action
item
left
here,
other
than
other
than
that.
This
is
like
the
preliminary
stage
of
how
it
looks
right
now.
C
Yeah
yeah
yeah-
I
will
just
by
the
way
to
follow
up
on
what
jared
is
saying
like
it's
not
just
because
jared
is
a
quality
like
has
a
quality
assurance
background,
because
some
of
the
bugs
that
he's
coming
across
are
blatant
or
others
would
hit
them
as
well.
It's
more
just
that,
there's
that
there's
that
we
might
have
to
well.
We've
we've
been
holding
off
on
the
dot
six
release,
because
we've
been
trying
to
slow
down
to
do
testing,
but
we
haven't
been
doing
it
well.
C
C
That's
going
to
help
in
coverage
in
terms
of
the
adapters,
not
so
much
in
terms
like
some
in
terms
of
mastery
server
push
has
brought
forth
a
lot
and
well
or
helped
champion
a
significant
jump
in
unit
test
coverage
in
mastery.
Ctl
mario
has
been
digging
into
functional
test
coverage
using
cypress
in
from
the
perspective
of
mesh
reui
there's
various
each
of
these
vectors
like
are
necessary
and
unit
test
coverage,
is
necessary.
The
integration
test
coverage
it
makes
for
quick,
wins
or
quick
stroking.
C
You
know,
broad
sweeping
coverage
so
rodolfo
had
been
here
is
currently
a
mastery.
Maintainer
he's
been
inactive
for
longer
than
the
two-month
period
that
we
have
in
our
governance,
docs
and
so
mario,
both
because
you
live
you're
like
geographically
adjacent
to
rodolfo,
also
because
you've
got
something
of
a
similar
focus
in
terms
of
cyprus
and
things
you
might.
C
You
might
strike
up
a
conversation
with
him
and
see
where
he's
at
in
terms
of
his
ability
to
to
participate
his
ability
to
maybe
engage
and
help
advance
some
of
those
some
of
the
cyprus
tests
that
you're
leading,
but
it's
a
it's.
If
people
try
the
project
and
they
they
hit
an
ugly
error
message
or
it
just
falls
over
or
like
there's.
Only
like
it's
it's
okay
for
some
time
we
are,
you
know,
for
a
project
to
be
like
that.
We're
kind
of
you
know
we're.
C
We
also
do
need
there
have
been
two
working
group
leads
in
the
past
for
like
a
build
and
release
working
group,
but
I
think
that
the
folks
that
are
here
are
either
spread
too
thin
or
too
focused
on
new
functionality,
and
so
yeah
I
mean
I
will
have
to
either
identify
someone
new
to
step
up
or
spread
the
love
around
or
slow
down.
Functionality
like
I
was
going
to
talk
about
doing
design,
spec
reviews
and
scheduling
those
4.7.
C
We
should
probably
talk
about
scheduling.
You
know
bug
smash
days
or
like
the
bug,
smash
reviews
instead
and
how
we
can
get
better
coverage
relatively
quickly
with
integration
tests.
C
I
think
we've
got
a
number
we've.
A
number
of
you've
made
significant
progress,
both
in
github
github,
workflows
and
in
cyprus,
and
so
and
then
also
the
some
unit
tests,
but
but
it's
more
the
integration
tests
that
are
going
to
get
us
somewhere
quickly.
C
So
please
please
think
on
this,
and
so
one
of
the
things
that
I
was
encouraged,
that
jared
just
said
was
that
he
might
he
might
submit
a
pr
which,
like
jared,
your
engagement
in
specific,
is
like
is
fantastic,
any
which
way
prs
or
not,
but
like
that
was
the
first
time
I'd
heard
you
say
that,
so
that's
that's
exciting.
E
Yes,
it
took
a
while
to
get
through
some
of
the
the
hoops
and
you
know
yeah.
C
C
Nice,
but
anyway
I
mean
like
jared,
you
know
it's
not
fun
to
hear,
or
I
don't
think
but
but
like
and
jared
wasn't
here
to
make
us
wasn't
necessarily
trying
to
make
a
statement
for
trying
to
like,
but
the
the
reality
is
like
is
really
important.
Whenever
I
mean
it's
like
significantly
important
yeah,
please
please
think
on
this
a
little
bit.
If
you
all,
would
it's
not
okay,
hey
by
the
way
so
of
what
aditya
is
showing?
This
is
something
else
to
think
about.
C
I
think
the
call
to
action
for
most
of
us
is
to
think
about
this.
A
little
bit
so
wrap
your
mind
around
this
concept.
That
measury
is
it's
a
management
plane,
for
you
know
aimed
at
service
meshes.
It
also
orchestrates
or
interacts
quite
a
bit
with
kubernetes.
C
It
lets
you
deploy.
Applications
on
kubernetes
or
on
service
meshes
configure
kubernetes
computer
service
meshes
like
some
of
you
that
have
been
either
working
on
or
watching
as
mesh
map
is
evolving.
I
think
it
becomes
a
lot
more
obvious
that
mescheri
is
capable
of
configuring.
C
The
the
service
meshes
that
we
have
adapters
for
okay,
good
as
people
go
to
make
those
configurations
they
create
designs,
as
people
go
to
use
mesh
re
that
they
it's
nice
to
have
functionality
out
of
the
box
or
more
functionality
out
of
the
box
and
like
as
an
example,
webassembly
filters
and
how
many
of
those
are
included
out
of
the
box
and
the
fact
that
people
can
bring
their
own
like
the
more
that
come
out
of
the
box,
the
more
well
conceptually
the
more
value
that
people
find
users
find
right.
C
C
I
can't
believe
I
have
that
memorized,
but
I
think
that's
what
it
is.
It
is
it's
littered
with
extension
points,
and
so
it's
through
extension
points
and
through
that
meshery
becomes
a
platform.
It's
an
extensible
platform
part
of
the
the
content,
then
that
people
can
bring
is
slide
five,
the
content
that
people
can
bring
it
comes
in
a
couple
of
different
forms.
C
We
actually
start
to
talk
about
content
in
the
dot
six
release,
so
it's
it's
patterns,
applications
filters,
but-
and
that's
probably
not
the
end
of
our
that's,
not
probably
it's
for
sure.
Not
the
end
of
the
content
like
meshri
will
have
a
policy
engine
mesh.
We
also
have
a
workflow
engine
or
more
more
of
a
workflow
engine
than
it
does
today.
C
There
the
users
will
be
able
to
configure
policy
which
becomes
content.
Measuring
will
come
out
of
the
box
with
like
best
practices.
So
you
know
content
in
that
regard.
Policy
in
that
regard,
workflows
in
this
regard,
it
will
come
with
add-ons,
so
pravin
was
just
having
an
open
thought
on
a
call
that
we
were
having
yesterday
about
some
of
the
add-ons
that
measury,
you
know
like
supports
in
a
first-class
way,
so
measuring
supports
prometheus
and
jaeger
and
grafana
and
the
you
know
these
things
in
kind
of
a
first-class
way.
C
These
patterns
that
are
emerging
are,
in
my
mind,
one
form
of
content,
a
pattern.
This
is
what
we
just
said.
You
can
have
a
pattern,
a
filter
of
your
workload
like
an
application
and
then
ongoing.
You
can
have
more,
and
so
I
was
speaking
with
the
artifact
hub,
maintain
or
one
of
the
artifact
hub
maintainers
yesterday
about
like
actually
a
dt.
C
If
you
don't
mind,
if
you
go
to
artifacthub.io
on
the
front
page
you'll
probably
see
like
you
might,
you
might
think
of
artifact
hub
as
being
helm,
centric,
and
certainly
I
think,
that's
probably
the
majority
of
their
packages
are
there.
But
if
you
scroll
down
a
little
bit,
you'll
see
some
other
examples
of
packages.
Package
types
keep
going
package
types
other
than
keep
going
there.
C
The
blue
package
types
other
than
helm
charts,
that's
the
first
in
the
list,
but
there's
falco
rules,
open
policies,
tinkerbell
actions
coordinates
plugins,
captain
plugins,
cada
scalars
of
cada
scalars.
The
main
maintainer
of
cadet
was
just
sent
an
email
to
me
this
morning,
asking
to
jump
into
the
service
mesh
patterns
book.
So
so,
there's
a
lot
of
people
interested
in
what
you
all
are
doing,
and
so
my
question
is
well:
should
there
be
measury
extensions
listed
here
and
should
you
know
maybe
should
we
do
both?
C
Should
we
have
some
of
these,
these
extensions,
these
patterns
to
start
with
out
on
artifact
hub
and
als,
or
it's
either
an
or
or
it's
an
and
or
should
we
have
a
service
mesh
catalog
that
you
know
the
measuring
service
mesh
catalog
where
people
can
bring
their
their
filters
bring
their
patterns
share
their
apps
do
their.
C
Think
about
this,
if
you
would
please
there's
a
mock-up
for
what
this
looks
like.
So
this
is
a
this
effort
is
like
a
line
item
on
the
dot.
Seven
release,
there's
a
mock-up
on
like
slide
28,
oh
I'm
so
close,
yeah,
27
or
no
I'm
sorry
slide.
30.
C
C
I
Yeah
one
question:
so
once
we
have
a
basic
history
installation
right
canary
way,
so
as
you
told
each
day
like
you
just
import
the
url
and
apply
to
your
measuring
and
the
workflow
continues
automatic
or
like.
Is
it
a
controllable
way.
C
Yeah,
that's
it
yeah
good
question.
It
would
be
a
controllable
way
where
okay,
it
right
now,
it's
a
concept
that
you'd
be
able
to
click
to
import
right.
Now.
What
you
do
is
you
click
to
get
the
url
and
then,
when
you
go
when
you
go
over
to
your
measuring
deployment
in
this
case,
since
this
is
a
pattern,
you'd
go
to
the
patterns
section
in
your
measure,
ui
and
you
click
import
and
you
paste
in
the
url
and
then
and
so
from
there.
You
can
then
say
well
yeah.
I
Okay,
the
only
way
I
ask
this
question
is
like
hto
right,
so
I'm
I'm
targeting
hto.
So
basically,
easter
comes
up
two
ways
of
installation
right,
one
is
cni
way
and
the
other
is
a
like.
A
sidecar
injection,
so
people
may
install
like
measuring,
doesn't
support
as
of
now
like
cni
way
of
installation,
as
I
as
I
know,
but
it's
like
it's
like
a
sidecar
sidecar
injection
within
it
container
right.
C
Correct,
although
pu
shortened
event,
that's
correct,
but
preven,
but
just
let
me
clarify,
qualify
push
novendy.
Do
you
guys
want
to
talk
about
mastery
ctl
mesh
is
that
does
that
currently
support
provisioning,
a
service
mesh
through
measuring
ctl.
A
Yes,
so
yeah
like
the
measuring
ui
measuring
ctl
mesh,
deploy,
there's
a
command
called
measure,
gtl
mesh
deploy
which
you
can
use
to
deploy
as
you
mentioned
store.
So
if
you,
the
only
requisite
is
that
you
need
to
have
the
measuring
studio,
adapter
running
and
measuring
ctl
will
handle
deploying
the
service
mesh.
I
I
So
then
normally
it
will
install
through
each
to
your
ctl,
that's
okay,
but
now
what
thing
is
like
if
you
go
for
a
basic
canary
way
of
upgrading
right,
so
the
canary
wave
upgrades
differs
from
like
whether
you
install
the
hto
using
sienna,
ht
or
cni
plug-in
hdrc,
and
I
I
think
yes,
I
think
I'm
just
checking
the
installation
part
of
the
hto
right.
So
when
you
install
hto
so
normally
what
happens?
I
There
are
two
ways
to
install
hto
one
is
it
comes
up
with
like
a
unit
container
like
unit
pod
unit,
pod
comes
and
configures
the
repeatable
rules
and
goes
away
right.
So
that's
the
installation.
C
I
C
I
I
C
Istio
doesn't
install
itself
through
an
init
container,
istio
istio
when
you
deploy
it
like
that's
how
you
onboard
a
workload
site.
I
No,
my
question
is:
how
do
I
install
store
right
today
when
I
go
and
print
a
machinery
and
install
hto,
so
normally
it
install
the
hto?
So
basically
so,
and
and
normally
when
you
yeah,
as
you
said
it's
true,
but
normally
that
web
books
will
be
always
enabled
so
the
way
books
comes
up
with
like
html
installation
right.
So
now,
when
you
deploying
workload,
the
unit
container
comes
up
and
like
and
it
deploys
the
application
right.
I
I
The
canary
way
of
canada,
you
cannot,
you
cannot
do
a
canary
with
the
hto
cni
so
because
cnn,
you
need
to
take
it
separately
in
the
canary,
like.
C
Yeah,
okay,
yeah!
I
got
you
now
yeah.
So
it's
a
great
question.
Actually
it's
a
great
question
for
us
to
do
a
workshop
on
like
a
workshop
on
on
istio,
actually
like
I'll
I'll
to
understanding
your
question.
Now.
Yes,
you're
right,
that's
kind
of
the
most
concise
way
that
I
could
not
tell
you
but
yeah
so
leaving
mesherie
aside,
because
your
question
has
nothing
to
do
with
with
measuring
yup
you're
right,
but
I
wouldn't
say
installing
istio,
I
would
say:
how
does
a
workload
come
onto
the
mesh.
C
Yeah
yeah,
or
rather
yeah
correct
the
way
that
I
would
phrase
that
is
that
measures
should
be
aware
of
applications
that
are
workloads
that
are
running
in
your
kubernetes
environment
and
it
measures
should
because
it's
a
service
mesh
manager.
It
should
also
be
cognizant
of
the
fact
that
that
workload
is
either
currently
running
on
a
service,
mesh
or
off
of
the
service
mesh.
C
What
qualifies
as
running
on
a
mesh
or
off
of
a
mesh
is
the
is
service
mesh
specific
in
istio's
case
that
has
to
do
with
the
presence
of
a
sidecar
in
other
service
mesh
cases.
That's
not
the
same,
but
irrespective
yep
mashri
should
become
should
be
aware
of
that.
Mesh
sync
does
take
a
look
at
that
that
data
is
internally
collected
and
yet
to
really
be
exposed
in
mesherie's
ui
like
in
a
way
in
which
it's
like,
hey,
here's
all
your
apps.
These
are
on
the
mesh.
These
are
off
the
mesh
yeah.
I
C
I
Maybe
what
I'm
thinking
is
like
we
can
add
a
like
ht
installation
is
done
so
after
that
we
enable
the
sidecar
injection
right.
So
that's
the
next
step
like
what
we
do
so
when
we
do
a
sidecar
injection
at
that
time,
we
can
ask
whether
you
want
to
do
a
unique
container
or
survey
of
cycle
injection
or
cmi
fighter
injection.
C
Oh
right,
yeah
sure
yeah,
the
your
yup.
I
think
the
the
thing
that
you're
hitting
on
is
that
by
the
way
nevindy
can
you
take
back
the
hosts
and
share
the
meeting
minutes.
We
only
got
a
few
minutes
left
so
yeah
prevent
you
right.
C
The
current
management
interface,
the
current
ui
in
meshary
for
like
provisioning,
istio
and
then
subsequently,
configuring
it
matt
needs
to
be
put
in
the
trash
bin
and
then
a
new
one
needs
to
come
forth.
That's
like
highly
configurable,
where
you
can
begin
to
see
what
that
might
look
like
is
well
abhishek.
Kumar
is
not
on
the
call,
but
but
it
would
look
like
I'll
say
I'll.
Send
you
a
link
I'll
get
a
look:
okay,
okay,.
I
Thanks,
maybe
I'm
very
way
behind
this
community
yeah?
No,
I
don't
know
I'm
just
experimenting
like
high
level
and
putting
my
thoughts
onto
it,
so
that,
like
I,
will
also
learn
from
it.
So
before
I
make
my
plans
on
this
yeah.
C
You're
in
the
right
you're
in
the
right
place
for
that
yep.
Those
are
good
questions.
I
mean
those
are
good
suggestions
like
some
of
your
questions
become
suggestions
because
mastery
should
be
you
should
there
should
be
an
easy
button
like
I
don't
really
know
like
the
user,
doesn't
really
know.
I
don't
know
I
just.
I
just
want
a
service
message.
I
just
want
that
one
like
do
it
do
intelligent,
like
defaults
for
me
and
also
should
have
the
advanced
button
where
it's
like.
C
Well,
no,
let
me
you
know,
configure
each
item
before
you.
Do
it,
and
and
mesh
should
be
making
that
obvious
to
you
right
now.
It's
like
it's
got
the
easy
button
which
is
helpful
to
kind
of
getting
up
and
going,
but
like
but
yeah
when
you
want
to
get
serious
about
it
when
you,
as
you
start
to
want
to
make
some
of
your
own
choices,
I.
I
A
A
Thank
you,
praveen.
Thank
you.
One
solution
for
the
patents
catalog
was
that
we
could
maybe
add
the
download
button
so
yep.
So
since
we
are
running
out
of
time,
let's
move
on
to
the
next
topic
from
mano
mano.
J
Yeah,
so
it
was,
it
is
just
a
ui
revamp
where
I
change
some
of
the
ui
components,
added
a
better
error
message
and
a
better
statement,
and
the
second
thing
which
I
had
in
the
pr
was
a
check
for
the
json
that
we
are
inputting,
whether
it's
a
valid
json
or
not.
So,
basically,
the
issue
which
was
we
are
facing
that
if,
if
somebody
was
initiating
a
grifana
dashboard
in
a
prometheus,
we
are
in
the
prometheus
term.
We
should
not
be
allowing
grifana
queries
over
here.
J
C
J
C
A
We
have
one
more
topic
for
today's
call
state
from
ashish,
which
would
you
like
to
start
this
discussion.
F
It's
not
very
important
if,
if
you
ask
and
I'll
go
ahead
with
that
or
I'll
talk
about
that
in
the
github,
pr
that
I
created
so
and
should
I
go
ahead
and
talk
about
it
for
a
while.
C
Just
a
final
call
for
as
we
gear
up
for
the
v070-
oh
thank
you,
hussaina
hussaina
and
others
are
identified
as
feature
champions
in
that
roadmap.
So
please,
please
do
take
a
look
at
the
roadmap.
C
Engage
with
your
comments,
your
questions,
your
suggestions,
and
I
think
that's
it
we're
at
time
and
so
navindu,
thanks
for
hosting
very
good
to
see
all
of
you
catch
you
in
discuss
or
on
tomorrow's
building
release,
call.