►
From YouTube: Meshery Development Meeting (April 27th, 2022)
Description
Meshery Development Meeting (April 27th, 2022)
Join the community at https://layer5.io/community
Find Meshery on:
GitHub: https://github.com/meshery
Twitter: https://twitter.com/mesheryio
LinkedIn: https://www.linkedin.com/showcase/meshery/
Docker Hub: https://hub.docker.com/u/layer5/
A
D
C
C
C
C
C
D
D
It's
a
nice
overview
of
how
measury
measuring
system
or
the
measuring
deployment
is
how
it
is
a
high
level
if
it's
healthy
and
but
it
is
essentially
all
you
know
more
or
less
inappropriate
for
us
to
have
that
dashboard
or
the
the
connection
status
section
on
the
main
dashboard
like
right
now
for
the
status
of
where
the
project
is
at.
It's.
D
Commonly
referenced
by
anyone
who's
deploying
measuring
they
want
to
make
sure
that
their
measuring
deployment
is
okay,
that
each
of
the
adapters
that
they
have
up
are
in
communication
with
mastery
server.
Mastery
server
is
talking
to
these
different
kubernetes
contexts
that
they're
connected
to
prometheus
and
grafana,
and
the
list
goes
on
right
and
that's
good,
although
that
really
needs
the
whole
thing
like
needs
to
move
to
the
settings
area
and
in
some
respects
it's
a
duplication
of
what
is
in
settings
inside
of
settings.
D
You
get
the
same
info
plus
a
lot
more,
so
you
know
it's
not
that
there
isn't
utility
in
having
a
concise
dashboard
of
your
measuring
deployment
and
in
all
of
its
connections.
There
is,
but
it's
not
the
core
reason
why
people
are
using
measuring
they're,
not
using
measuring
to
make
sure
that
measuring
is
running
well
necessarily
anyway.
D
So
I'm
saying
that
as
background
for
part
of
this
discussion,
so
in
this
discussion,
there's
there's
a
new
kubernetes
context:
switcher
in
the
upper
right
hand,
corner
that
is
primarily
for
purposes
of
quick
access
to
something
that
people
will
quite
commonly
reference.
As
kubernetes
as
measuring
manages
multiple
kubernetes
clusters,
multiple
kubernetes
contexts,
people
are
going
to
want
to
view
to
slice
their
dashboard
views
based
on
on
a
context,
division
or
based
on
one
or
more
contexts,
one
or
more
clusters.
D
D
D
D
So
having
given
you
that
full
context,
both
of
where
we've
been
where
we
are,
where
we'd
like
to
go,
hopefully,
is
again
to
concisely.
Answer
your
question.
Yes,
you
should
show
all
the
chips
here.
Yes,
as
people
choose,
if
you
go
into
settings,
we're
showing
both
chips
and
so
in
the
same
way.
Now
that
you
understand
that
connection
status
is
really
just
a
summarization
of
what
you
find
in
settings,
we
would
want.
D
We
would
desire
for
the
same
behavior
and
we
actually
in
the
same
behavior
being
if
you've
got
a
cluster
connected
or
had
established.
A
connection
at
one
point
then
show
the
chip
here
here
and
on
the
status
page.
The
status
page,
if
you'll
note
it,
doesn't
allow
you
to
remove
a
chip,
it
doesn't
have
the
x
or
in
in
indian
nomenclature.
It
doesn't
have
the
cross,
which
is
totally
confusing
to
me,
but
yes,
the
you
yep.
D
We
don't
have
it
here,
because
the
function
is
really
just
about
connectivity
status
connection
status,
whereas
in
settings
it's
more
about
configuration,
mutating
those
connections,
meaning
creating
or
deleting
them
and
or
updating
them.
One
of
the
things
that's
a
digression
from
the
uniformity
and
consistency
that
we
have
in
terms
of
user
experience.
So
if
you
look
at
the
operator
chip,
meshsync
and
nats,
it's
there's
no
x
on
the
operator.
D
C
C
D
B
C
B
Yeah,
so
why,
while
you
are
selecting
or
deselecting
the
context
from
context
feature
or
from
the
cluster
configuration
settings,
did
you
notice
that
there
are
some
network
requests
that
is
going
to
the
server
to
actually
switch
the
context?
So
you
can
open
your
console
and
see
if,
if
that
thing
is
actually
functional
or
not
so.
B
Okay,
okay,
you
can
open
the
network
panels
right
now,
ctrl
shift,
I
okay
so
go
to
network
yeah,
so
just
click
on
the
cluster
one
at
the
top
switch
the
context,
and
there
is
no
network
request.
That
means
this
is
just
a
kind
of
it's
not
doing
anything
right
now
right
so
you
have
to
like.
There
is
already
an
api
that
handles
the
switching
of
context,
so
you
have
to
bind
that
api
call
with
the
with
this.
Our
selection
and
description.
E
B
B
C
B
You
you
don't
require
to
refresh
the
page,
so
it's
that's
why
this
functionality
is
breaking
right
now.
So
this
whole,
are
you
wearing
the
case
for
switching
context
and
managing?
It
is
kind
of
broken
right
now
and
there
is
a
effort
required
to
fix
those
things.
So
what
you
have
to
do
is:
are
we
already
using
redux?
So
you
have
to
maintain
the
state
of
active
and
inactive
clusters
in
the
redux
store.
B
A
C
B
G
F
D
F
A
F
Meshkit
that
that,
like
that
works,
equivalently
to
cube
ctl
describe
command.
If
you
have
ever
used
tube
ctl
describe
deployment
and
like
getting
details
about
those
resources,
it
is
very
efficient
to
that,
and
this
is
right
to
graph
your
endpoint.
How
it
works
would
be.
It
requires
three
parameters:
name
namespace
and
the
type
of
like
what
kind
of
resource
it
is,
that's
how
it
works,
and
it
will
be
added
to
measure
ui
and
mesh
map
program
so
that
people
can
know
about
the
details
about
the
resources
that
they
get
from
the
cli.
B
F
D
F
D
D
D
F
Okay,
yeah,
I
did
thought
about
that.
The
thing
was
that
cube
ctl
like
package
explicitly
like
just
like
the
command
we
like
in
command.
We
specifically
write
deployment
cube
serial
described
deployment,
qc
descriptor.
The
package
also
requires
to
identify
ourselves
what
kind
of
resource
we
are
trying
to
get
like.
F
Is
it
a
deployment
or
not,
because
it
was
a
query
I
kind
of
had
to
maintain
a
map
so
that,
like
we
get
a
just
a
string
from
our
parry
that
okay,
it
is
a
part
and
we
pass
the
exact
like
kind
of
struct
to
cuba
package.
This
is
the
kind
of
resource
it
is
like.
It
is
kind
of
a
maintenance
thing,
but
it's
just
to
make
sure
that
kind
of
find
a
balance
between
querying
it
and
like
passing
that
passing
that
kind.
D
Yeah
it'd
be
really
nice
if
that
was
dynamically
interpreted,
where
you
don't
have
to
spell
like,
where
you're
still
doing
this
but
you're
doing
it.
Dynamically
you're,
not
hand
coding
the
names,
because
within
the
package
itself,
it
has
the
it
has
this.
How
do
I
use
it?
It
has
descriptions
of
each
of
these
kinds.
D
F
Okay,
cube
skill
is
dynamic
in
that
respect
like
if
you
pass
some
random
kind
and
I'd
say
ramesh,
because
that
kind
doesn't
exist.
It
will
just
say
that
okay,
this
kind
of
resource
doesn't
exist
and
why
I
implemented
this
map
was
just
to
like
make
sure
that
okay,
this
is
a
valid
kind.
It's
not
something.
We
are
yeah.
D
D
This
is
like
hey.
Why
is
lee
asking
me
about
this
stuff?
I
thought
I
was
doing
a
good
job
of
like
taking
care
that,
like
this
is
going
to
work
and
the
people
are
going
to
have
a
good
user
experience.
We're
going
to
be
able
to
say.
Oh
did
you
want
to
look
at
a
role,
no
problem?
I
know
what
a
role
is.
Let
me
get
that
for
you.
You
know
et
cetera
and-
and
here
here
in
the
hearing
lie.
D
The
issue
is
like
one:
yes,
that's
good,
the
the
for
ux
and
for
validation
that
someone's
giving
you
some
junk
or
they
misspelled
it.
You
know
the
the
problem
fault,
the
problem,
the
the
crux
of
the
issue
is
more
or
less
this
that,
like
someone
may
someone
you're
we're
going
to
find
that
someone
asks
to
describe
a
role
for
example,
and
either
that
specific
instance
doesn't
exist
or
roles
as
a
as
a
concept
as
an
object
were
never
present
in
that
kubernetes
deployment,
and
so
which
would
be
weird
right.
D
I
mean
just
like.
Maybe
it
wouldn't
be
as
weird
for
like
an
endpoint
slice
or
something
else,
but
but
suffice
to
say,
if
you're
being
diligent
in
your
code,
like
you,
are
we're
going
to
have
to
address
the
use
case
where
the
instance
where
oh,
that
object,
isn't
present,
and
so
I
need
to
have
error
handling
around
that
that
error
handling
becomes
the
answer
for
well
hey.
I
can
just
remove
all
this
static
listing
this
I'm
going
to
have
to
do
that
logic
anyway,
I'm
gonna
have
to
either
validate
or
handle
the
post.
D
F
H
Okay,
so
we
are
doing
multiple
e2e
end-to-end
testings
across
different
adapters,
and
we
are
documenting
it
in
our
measure,
documentation
which
we
can
see
here,
like
all
the
integration
tests
of
different
adapters
in
stable,
as
well
as
edge
versions.
We
are
storing
them
and
showing
the
users
about
the
test
results.
H
So
so
now
we
have.
We
need
a
summary
of
these
tables
so
that
that's
what
I've
initialized
and
that
started
to
work
on
so
right
now.
The
summary
table
shows
the
overall
compatibility
of
the
stable
version
of
this
particular
measuring
version,
but
in
future
we
would
be
requiring
a
multiple
like
we
need
to.
We
need.
We
need
to
be
able
to
select
multiple
messaging
versions
and
according
to
it,
we
need
to
find
the
overall
compatibility.
H
So
and
one
more
thing
right
now,
the
overall
compatriot
is
not
calculated
accurately
as
of
say,
because
I
have
not
considered
the
edge
version
that
I
tested
for
this
measuring
version.
This
is
just
for
the
stable
version,
I
think
so
yeah
after
calculating
the
compat
like
over
a
compatibility
for
edge
versions.
Also,
then
this
will
show
a
proper
cobra
compatibility.
D
Yeah,
I
think
yeah
it
is
to
perhaps
just
summarize
more
actually
the
compatibility
matrix.
It's
it's
probably
confusing
how
it's
been
described
in
most
like
you
can
probably
drop
the
hard,
the
complexity
of
trying
to
figure
out
the
overall
compatibility
and
just
say
that
we're
stating
our
intention
and
and
our
what
should
be
happening,
which
is
measure
istio
adapter,
should
be
compliant
with
or
compatible
with
istio
version,
112
6.
D
and
the
actual
stat
state
of
it
is
contained
within
the
integration
tests
below
like
like
what
we
would
do
is
like
like
in
the
future.
You
can
work
toward
overall
compatibility
and
get
very
detailed
and
stuff,
but
to
help
move
things
along
quickly.
The
thing
that
we're
trying
to
give
to
users
is
that
the
istio
version,
these
very
top
level
things
which
is
like
istio
version
6.0,
is
compatible
with
this
kubernetes
version.
D
This
docker
version
this
windows
version
this
linux
version
this,
like
these,
these
high-level
system
requirements
and
from
there
like
yeah
the
ability
potentially
for
people
to
drill
into
like
oh
the
specific
versions
of
each
adapter
and
then
the
specific
versions
of
like
so
so.
It's
this
high
level
system,
compatibility
of
just
all
the
service
meshes.
It's
like
it's
like
the
mesh
reverb.
D
It's
this
it's
a
matrix
which
is
not
exactly
what
we
have
right.
Now,
it's
more
like
measuring
on
one
axis,
measuring
all
of
its
components
on
one
axis
and
then
all
the
third-party
components
on
a
different
axis,
and
then
us
just
saying
check,
mark
check,
checkmark,
checkmark
and
whatever
version
that
is,
for
those
components,
is
kind
of
the.
D
There
may
be
the
devil
in
the
details
of
that
check,
mark
which,
at
some
point
you
could
make
a
clickable
thing
and
that
either
redirects
them
to
integration
tests
specific
to
that
those
crosshairs
or
not,
because
yeah,
I
would
say
not
for
a
while,
because
it's
just
too
too
hard
for
you,
like
not
for
you,
but
I
mean
too
hard
for
we're
going
to
get
into
this
situation
where
it's
like.
Yes,
measuring
v060
is
compatible
with
kubernetes
122.
D
D
Yeah,
the
having
the
version
number
is
is
the
important
bit
because
the
people
can
see
from
the
supported
platforms
page
that
these
platforms
are
supported
and
in
concept.
I
think,
if
you
click
on
supported
platforms,
it
will,
I
think,
you'll
have
a
list
of
the
version
number
of
those
platforms
which
is
antiquated
and
so
in
a
lot
of
respects
like
this
is
kind
of
what
we're
looking
for.
Except
this
doesn't
say
what
mesh
reversion
is
just
sort
of.
D
D
D
D
H
Okay,
yeah,
that's
yeah,
that
that's
what
I
was
also
going
to
do,
but
there
was
one
issue
regarding
that,
like
suppose,
for
mercury
version,
5
l
they're
only
test
for
h2
linked
at
osm,
like
we
don't
have
dash
for
like
say,
puma
or
ngnx
or
traffic
mesh,
so
like
should
be
like
what
should
we
do
for
that
values
that
should
be
just
and
a
like?
Should
we
just
put
nh
and
state
that
it's
not
tested.
D
A
E
So
when
we
do
a
performance
test,
then
we
get
this
kind
of
a
graph.
This
basically
includes
the
dynamic
results
from
our
testing,
but
it
doesn't
give
static
results.
Dynamic
means
the
results
which
you
get
out
of
our
after
doing
the
test
and
the
starting
result
is
the
data
for
our
kubernetes
cluster
and
the
state
of
it
which
it
undergoes
when
we're
doing
the
test.
E
So
basically,
this
thing
is
already
there
in
our
system,
but
it's
just
that
it
was
removed
from
the
ui
and
and-
and
what
I
would
be
doing
is
that
I
would
be
adding
that
back
to
this
performance
crowd.
So
by
the
end
of
this
movie,
I
would
be
able
to
do
it.
So
that's
the
starting
board,
and
if
you
go
to
the
console
log,
you
would
be
able
to
see
all
those
details.
E
H
B
I
Hi
hi
so
yeah
I
couldn't.
I
could
make
out
some
some
stuff
what
you
were
going
through
earlier.
I
think
you
were
working
on
adding
new
features
to
your
react
application.
If
that
would
make
sense,
it's
very
new
to
me.
I
think
that
was
what
happening
so.
Could
you
guys
explain
me
what
was
going
on
in
this
meeting
like
the
basic
overview.
B
Yeah,
so
in
this
meeting
we
just
us,
you
know
revisit
items
we
have
done
the
by
the
week
or
or
the
future
items
we
need
to
do.
You
know,
while
developing
any
kind
of
thing
like
a
server,
graphql
iui,
it's
not
related
to
this
diy,
and
just
like
that,
we
welcome
our
new
commerce.
You
know
we,
we
love
to
know
more
about
them,
how
where
they
come
from
and
what's
yeah.
So
would
you
like
to
tell
a
bit
about
yourself.
I
Yeah,
so
I'm
actually
a
student
right
now
enrolled
in
a
college.
So
all
this
is,
I
I
really
want
to
you
know,
be
diligent
in
open
source
and
contributing
to
open
source,
because
it's
a
really
nice
way
to
you
know,
share
your
work
and
gain
knowledge
and
meet
amazing
people.
So,
yes,
I
wanted
to
be
diligent
in
contributing
to
open
source
so
mystery.
I
just
came
across.
It
came
across
misery
recently,
so
yeah
I
received
a
very
warm
welcome
from
lee
and
all
of
you
guys
as
well.
I
I
B
D
D
J
D
Part
of
what
will
happen
in
those
in
those
those
continuous
integration
workflows
is
a
deployment
of
measuring,
will
happen
and
certain
functional
tests
will
be
invoked.
So
we
can
verify
that
mesherie's
features
and
its
functions
still
work
and
there's
there's
this
many
features
that
meshri
has
and
there's
this
much
of
them
that
have
been
that
are
covered
through
this
testing,
and
so
mario
is
not
only
calling
out
for
participation
from
those
that
are
interested
to
like,
as
a
matter
of
fact,
there's
a
mesh
rebuild
and
release
meeting
tomorrow.
D
You
know,
need
improvement,
so
I'm
hoping
that
people
engage
and
ask
questions
get
get
I
shoot.
I
might.
I
might
go,
get
involved
as
well
and
make
make
a
bit
of
a
mess
just
because,
like
the
the
tests
here
are
so
important
if,
as
we
go
to
make
a
release
or
go
to
merge
a
pr,
it's
so
much
more
comfortable
when
you
can
see
that
all
the
features
the
functions
are
still
working.
J
Yeah
I
just
wanted
to
share
the
right.
Now,
I'm
going
over
the
existing
document,
I
was
just
trying
to
get
my
like
put
myself
under
like
someone
contributing
the
mentoring
ui
right,
so
I
was
comparing
the
instructions
that
are
there
today
and
the
make
file
our
ui
contributing
ui
was
referring
to
run
fast,
like
the
runs
run,
a
yeah
the
run
fast
make,
but
it
looks
like
it's
deprecated.
Is
that
correct.
E
J
Correct,
so
that's
I'm
changing
that
right
now
and
also
I'm
modifying
the
a
like
the
headings
just
because
they're
not
showing
up
in
the
they're
in
the
content
to
the
right.
So
that's
like
it's
not
easy
to
see.
What's
the
what
content
is
in
there
right,
I've
done
I'm
working
on
those
changes
right
now,
just
to
you
know,
make
it
more
usable,
clear
information-
and
I
was
just
thinking
that
can
I
share
the
what
I
have
so
far.
A
J
Okay,
I
need
to
reopen
zoom
to
share
my
screen,
so
that's
the
summer
but
anyway.
So
I'm
working
on
that-
and
I
was
just
just
a
quick
question
to
everyone.
So
when
we're
running
cypress,
so
we
have
some
integration
tests,
so
they
do
not
really
they
mock
the
server
response.
So
would
that
be
enough
for
now
to
share
how
to
run
those
steps
just
as
the
as
it
is,
or
should
we
also
add,
like
a
make,
commands
to
a
little
bit
east
more
easily.