►
From YouTube: Meshery Development Meeting (May 25th, 2022)
Description
Meshery Development Meeting (May 25th, 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
Yeah
so
welcome
again
are
in
the
developers.
Call
here
is
the
meeting
minute.
So
please
add
your.
You
know.
Attendee
add
your
name
so
that
yeah,
you
are
in
attendance
right,
so
I'm
pasting
it
in
the
chat
too,
so
that
it's
handy
to
youtube
yeah.
A
So,
let's
without
without
wasting
any
time.
Let's
move
on
to
the
first
topic
that
has
which
is
a
modal
on
machine
database.
We
said
so
it's
been
a
long
wait.
So
do
you
have
any
updates
there.
C
B
Okay,
so
I
have
here,
I
have
added
two
buttons
here.
We
are
the
reset
database
and
the
flash
messaging
button,
and
if
we
press
the
reset
database,
it
actually
asks
us
to
confirm
this,
and
if
we
press
no,
it
will
start
to
not
accept
that
request
and
if
we
just
press
this,
it
reset
the
database
same
for
the
flash
messaging.
We
have
here
our
model
of
flush
missing
and
if
you
plus
yes,
it
is
flush
the
meshing.
A
A
A
So
so
the
backend
has
that
functionality
only
what
I've
done
is
I've
initially
thought
of
making
at
a
making
up.
One
go
like
there
is
a
reset
button.
This
is
a
database
button
and
in
the
model
there
will
be
a
check
mark
for
the.
If,
if
user
want
to
do
hard
reset,
if
the
check
mark
is
not
checked,
then
it
will
only
flash
missing
and
if
the
check
mark
is
checked,
it
will
do
the
hard
disk.
So
backhand
has
that
functionality
you
just
have
to
in
the.
A
If
in
the
ui
you
will
see
when
the
request
is
sent
for
request
is
sent
to
back
end.
There
are
three
options.
This
thing
reset
clear
db
and
if
you
want
to
do,
flash
messaging
only
do
clear
db
and.
B
A
First,
I
thought
you
would
do
like
this,
but
this
also
looks
good.
B
B
A
Yeah
there
is
one
doubt
for
me,
is
like
a
user
may
not
know
a
lot
about
meshing.
I
just
you
know.
I
I
think
it.
I
like
it's
not
very
important,
it's
true,
so
a
flush
messaging.
So
someone
like
for
me
as
a
user.
How
would
I
you
know
perceive
it
so
that's
one
thing
like
do:
we
have
something
user
friendly
information
instead
of
machine,
you
are
providing
some
user
friendly
like
flush.
Only
these
these
things
that
are
related
to
machine.
A
Yeah,
that's,
of
course,
not
a
strong
consideration
here,
but
yeah,
maybe
future
constitution.
E
E
A
As
you
will
see,
I
have
based
on
the
ad
for
hydration
also,
so
I
will
tell
about
what
happens
in
both
the
case.
If
you
will
only
flush
messing
data
that
is
equivalent
to
soft
reset,
only
the
tables
are
dropped
and
recreated
means
all
the
data
is
truncated,
and
when
you
do
a
hard
reset,
then
what
happen
is
the
machete
measuring
sql
file
in
config
dot
messaging
config
is
a
move
to
dot
archive
folder
and
a
new
sql
file
is
created.
E
Got
you
okay,
yeah.
You
probably
just
said
that
too.
So
I
probably
missed
that.
Okay,
then
that
makes
sense
in
the
case,
in
the
event
of
a
hard
reset,
how
does
meshri
server
and
its
process
recover
from
that.
E
F
When
meshi
server
started,
it
does
re-initialize
it
if
the
file
is
not
the
tree
initializes
it,
but
if
the,
if
machine
server
is
running
and
so
you're
asking
if
measure
server
is
running
and
then,
while
the
messages
are
running,
we
delete
the
sql
file.
What
happens
in
that
case
in
in
that
case,
as
of
now,
we
do
not
have
a
way
to
region
reinitialize
it
yeah.
We
don't
have
a
way
to
re-initialize
it
right
now.
E
Cool
yeah
good,
so
this
particular
hard
reset
would
have
you
know
in
some
respect
it
would
have
the
undesired
effect
of
basically
just
breaking
your
deployment,
and
so
what's
the
answer
here,
can
we
and
can
we
the
function
that
creates
the
database
on
startup?
F
I
think
in
your
refactoring,
you
are
calling
the
same
function
right
this
the
function
that
is.
A
E
Okay,
great
I'm
not
really
it's
like
your
mic
is
being
bumped
the
whole
time
or
wind
is
on
it
or
something,
but
I
don't
know
where
you
kind
of
oh
yeah,
your
video
is
up,
so
you
know
the
mic
is
not
being
bumped.
I
don't
know
what
it
is.
It's
like.
E
I
don't
know
it's
not
there
anymore,
oh
yeah,
it's
the
fan,
yeah
you're
gonna
have
to
sweat
it
out
man,
because
I
can't
hear
you
cool
good,
okay
yeah,
so
you
took
care
of
the
full
consideration
and
so
user.
Do
you
know
in
your
kind
of
testing
of
the
hard
reset
does
mesh
reserver?
Is
it
successful?
It
moves
the
files
over
re-initializes
the
database
and
then
continues
along
its
merry
way.
D
E
E
No,
I
trust
you,
I
don't
know
if
it
would
be
helpful.
I
think
you
you've
proven
of
the
contributions
you've
done.
You've
proven
that
I
totally
believe
you
that's
wonderful.
As
a
matter
of
fact,
I
will
use
this
function,
it's
kind
of
sad
to
say
that
I'll
use
the
function,
because
that
means
that
occasionally
measuring
gets
into
a
corrupt
state,
and
so
we
need
to
go
fix
those
issues,
but
at
least
we
have
two
ways
to
reset
the
database.
So
this
is
great.
E
The
other
thing
about
the
ui
here
is
everybody,
giving
some
good
feedback.
I
think
one
thing
to
consider
is
you
know
to
obvi's
point
of
the
user
may
or
may
not
know
what
mesh
sync
is
like
okay,
but
in
that
same
vein,
what
might
be,
especially
since
we'll
have
a
modal,
that's
an
opportunity
to
really
explain
to
the
user.
E
What
to
expect
when
to
invoke
this
sort
of
what
to
expect
like
it's
not
like.
We
want
to
have
a
full
paragraph
of
text
sitting
there
to.
It
can
be
very,
you
know,
concise
things,
but
it's
sort
of
like.
What's
the
difference
between
soft
and
hard
and
kind
of,
what
do
you
mean
and
for
anyone
who's
you
know,
for
the
most
part,
we
can
expect
a
relatively
educated
user
so
terms
like
flush
and
terms
like
reset.
E
E
But
if
there's
additional
information
that
we
can,
you
know
let
them
know
like
you
know,
if
you're
not
seeing
any,
if
you're
seeing
stale
data
in
your
dashboards,
you
may
want
to
flush
your
database
if,
if
you're
not
seeing
any
new
data
in
your
database
at
all,
you
may
want
to
hard
reset
kind
of
a
thing
like
some
amount,
and
then
I
wouldn't
worry
it's
too
much
about
the
particulars
of
the
check
boxes
versus
buttons
versus
the
versus,
because.
E
F
I
I
have
one
thing
to
interject
here,
because
the
topic
of
multicultural
came
in
or
in
reference
to
this
particular
operator,
configuration
that
we
are
seeing
here
right
now.
What
I
see
is
the
reset
database
and
plus
mesh
sync
button
are
part
of
the
the
part
of
the
that
box
of
operator
configuration
so,
but
this
operator
configuration
will
be
per
context,
so
each
cluster
will
have
its
own
operator
configuration
so
like
for
maybe
for
three
clusters.
F
The
mystery
operator
button
is
clicked
on,
and
maybe
user
has
clicked
off
it
for
the
rest
of
two
clusters
and
right
now
and
the
recent
databases
plus
flush
mesh
sync
operations
are
global
operations
and
they're,
not
cluster
specific,
so
it
kind
of
in
my
head,
because
I'm
looking
in
the
context
of
multi-cluster
approval,
multi-cluster
pr
so
buttons
that
are
specific
to
some
global
operation
are
inside
of
something
that
is
cluster-specific.
F
So
would
it
be
good
to
take
these
two
buttons
out
of
that
thing
and
I'm
not
good
at
ui,
so
other
folks
will
have
a
better
position
than
I
will
and
one
more
thing.
The
end
note
would
be
that
it
would
be
good
to
have
this
thing
at
the
granularity
of
each
cluster
so
that
if
user
wants
to
flush
data
for
three
clusters,
they
can
and
it
would
be
in
the
back
end.
We
would
need
like
three
lines
of
code
user.
F
We
would
just
get
the
context
id
from
the
query
parameters
and,
while
dropping
the
table
basically
you'll
be
deleting
rows
where
the
cluster
id
matches
the
cluster
id
that
you're
getting
from
the
query
parameters
so
yeah.
We
can
also
go
ahead
and
do
fine
grained
like
wipe
out
data
of
x
amount
of
clusters
and
leave
data
for
y
amount
of
clusters,
and
that
will
be
a
cluster
specific
operation
and
that
can
be
left
inside
of
that
operator.
Configuration
box
yeah
any
any
remarks,
criticism
or
supporting
views
in
that
direction.
B
F
Yeah
good
question
because
right
now
that's
kind
of
an
open
question
and
because
I
think
we
would
need
a
wireframe
or
something
to
think
to
put
in
place
before
we
can
go
ahead
and
talk
about
this
thing,
because
I
don't
think
we
have
any
such
thing.
Last
time
we
had
a
conversation.
F
That
was
the
end
note
of
the
conversation
that
we
need
some
kind
of
a
wireframe
to
have
a
common
dictionary
or
ground
on
which
we
can
have
a
conversation
and
right
now
we
don't
so
one
of
the
one
of
the
suggestions
that
he
gave
was
kind
of
a
table.
So
if
you
have
five
clusters,
you
can
have
some
kind
of
a
table
at
right
side.
When
you
click
on
one
of
these
rows.
F
Maybe
this
operator
configuration
box
pops
up
and
then
you
can
apply
operations
or
maybe
some
other
ui
suggestions
are
welcomed
there
as
well,
but
yeah.
We
would
have
to
have
this
operator
configuration
for
each
of
the
connected
cluster
and
leave
what
you
said
about
flushing
meshing
for
each
for
each
cluster
with
that
that
also
can
be
done.
Yeah
yeah
that
can
be
done.
That
can
that
is
completely
doable.
F
So
when
user
clicks
on
flush
flush
mesh
sync,
he
clicks
that
button
in
context
of
one
cluster
and
the
context
id
of
that
cluster
can
be
passed
to
the
back
end
and
then
the
back
end
will
do
the
fine
grained
deletion.
All
the
reset
database
is
a
very
general
operation,
so
it
would
be
good
to
have
that
at
least
outside
of
the
operator
configuration
box
because
it
it
it
does
the
entire.
A
So
what
I
think
is,
if
the
setting
page
is
going
to
be
revamped
or
redesigned
so
for
as
of
now,
we
can
keep
this
reset
database
of
flash
missing
button,
the
way
it
is
and
when
it
is
redesigned
we
will
consider
all
the
things
like
for
each
specific
cluster,
specific
buttons
of
reset
and
fresh
messing
and
all
those
things
just
okay,.
E
Yeah,
nice
yeah
so
just
immediately
pulling
the
buttons
outside
of
the
specific
operator
paper
and
then
and
then
from
there
it's
kind
of
like
just
this
is
just.
This
is
a
soft
opinion.
That
is,
you
could
just
leave
them
as
two
buttons.
You
could
consolidate
them
into
one
with
a
modal
that
has
checked.
E
You
know,
like
it
kind
of
it's
good,
to
get
this
functionality
in
here
now
as
a
global
piece
of
functionality,
and
just
let
that
land
and
it'll
be
helpful
and
then
yep
and
then
what
we
need
to
do
is
spend
time
doing
ux,
which
is
yep,
which
is
which
is
great
like
having
the
knowledge
of
what
ashish
just
said,
which
is
hey.
It
wouldn't
be
too
difficult
to
just
add
the
ability
to
individually
flush
or
re
flush
data
per
operator
per
mesh
sync
per
context.
E
It's
like
that's
all
the
same
statement
so
great
having
that
understanding,
as
you
think
about
the
user
interface,
by
which
a
user
would
I
don't
know
that,
certainly
like
would
manage
multiple
contexts
in
each
context
having
its
own
operator
there's
a
bunch
of
data
to
show
about
the
operator
and
its
custom
components:
custom
controllers,
there's
a
number
of
operations
like
provision,
the
operator
unprovision
it
reset
meshery's
database,
but
only
like
globally
or
just
flush.
Data
specific
to
one
context,
remove
a
context,
add
a
context
like
upload,
a
cube,
config
parse
through
that
fig
config.
E
Maybe
there's
ten
context
in
there
only
register
three
of
them
like
there's
a
lot
of
flows,
user
flows
to
what
the
user
might
want
to
do,
and
all
of
that
is
completely
achievable
in
you
know,
in
a
single
interface,
it's
just
a
bit
of
forethought
needs
to
be
put
in
so
that
conversations
like
this
become
pretty
short.
E
A
Yeah
yeah,
so
we
have
some
residue
here,
so
we
will
talk
back.
Let's
move
on
to
some.
F
Other
topic,
which,
if
we
are
half
past
time
so
yeah,
let's
so
let
me
show
this.
A
So
our
next
item
is
from.
We
have
already
discussed
a
bit
of
it,
so
yeah
hard
reset
of
database.
A
Yeah
support
machine
learning.
You
can
go
on.
A
So
this
is
the
pf
that
adds
support
for
this,
and
this
is
related
to
ms
singh.
I
I
in
the
messing
I
am
adding
a
environment
variable
which
is
where
and
that
I
am
using
here
in
machine
server
to
get
the
version
of
it.
A
F
A
A
F
The
operator
started
and
then
it
I
saw
that
back
off
thing,
so
it
started
and
then
it
exited
and
then
it
started
again
so.
F
E
E
A
Yes,
remember,
you
said
there
is
a
mesquite
mesh
function
for
the
describe,
but
the
thing
is
the
describe
doesn't
hold
environment
variables
that
we
give
in
in
the
docker
file.
Like
I
checked,
I
was
giving
environment
variable
and
it
was
not
appearing
in
the
describe.
But
if
I
am
providing
environment
variable
in
the
deployment
of
what
template,
then
it
is
available.
So
that
function
was
not
helpful.
E
Here
gotcha,
we
also
like
in
general,
there's
got
to
be
a
better
way
like
in
general.
We
don't
want
to
be
parsing
through
text
at
all.
That's
so
fragile!
That's
so
fragile.
Are
you
you're,
including
cube
ctl's
golang
package,.
A
E
As
opposed
to
like
in
general,
we
don't
want
to
like,
I
know
it's
kind
of
deceiving,
but
clies
are
designed
for
humans,
largely
which
is
exactly
why
we
don't,
which
is
exactly
why
you're
having
a
parse
text,
some
people
script
around
them.
A
lot
of
people
do,
but
that's
because
they
don't
know
how
to
code.
I
would
say
code
very
well
and
I'm
not
saying
that.
That's
true
of
you.
What
I'm
saying
is
like
it's
deceiving.
E
We
don't
want
to
use
a
cli
programmatically
if
we
don't
have
to.
What
we
want
to
do
is
talk
to
cube
api
and
invoke
other
endpoints
that
will
retrieve
data
about
a
deployed
pod,
which
is
in
some
respects
to
say
the
underlying
api
that
that
will
invoke
a
get
pod
or
a
that
invokes.
Basically
the
same.
The
same
api
that
cube
ctl
is
invoking
is
like
the
one
that
we
would
because
qctl
is
not
an
api.
A
F
Also,
we
have,
we
already
have
few
functions
which
which
give
us
the
give
us.
Oh,
I
guess,
broker
and
mesh
sync
objects
like
those,
and
I
think
there
we
do
have
labels
for
in
metadata.
So
if
you
will
okay
I'll
I'll
comment
on
this
pr
with
the
and
I'll
refer
to
that
function
that
I'm
referring
to,
I
think
whether
it's
in
mesh
kit
or
in
measuring
in
one
of
those
things
and
in
one
of
those
places
most
probably
in
meshi.
I
guess
so.
F
A
Yeah,
so
let's
move
on
to
our
next
topic,
which
is
compatibility
matrix
by
adita
yeah.
So
please.
G
Okay,
so
we
have
been
doing
a
multiple
e
to
e
tests
and
we
have
been
storing
the
results
for
different
tests
in
this
docs
and
with
that
all
the
results
we,
you
know,
I
made
a
summary
kind
of
a
matrix
in
which
we
show
different
measuring
components
that
are
compatible
like
how
much
they
are
compatible
with
kubernetes
version.
So
currently
we
only
check
for
kubernetes
version
1.20,
so
these
are
all
the
like.
G
The
we
show
the
percentage
of
how
much
like
how
much
passing
test
are
there
for
individual
measuring
component.
So
this
is
a
basic.
Like
summary
of
all
the
integration
tests
that
we
stored
in
docs.
G
G
What's
next
with
this
siberi,
I
think
now
we
have
to
check
like
this
thing.
The
service
mesh
version
we
need
to
you
know
enable
like
compare
like.
Is
it
compatible
with
the
latest
service
service
mess
version
or
not?
I
think
that
should
be
the
next
thing.
Yeah.
E
E
As
a
matter
of
fact,
I
would
just
you
know
I
would
just
ping
him
to
let
him
know
that,
hey
that
this
thing
has
just
come
forth
there
is
he
and
are
giving
consideration
to
all
and
I'm
sure
others,
but
are
giving
consideration
to
all
the
data
that's
being
collected
or
generated
from
end-to-end
tests
in
the
adapters
data
like
how
many
components
that
we
support
per
adapter
or
per
service
mesh,
pretty
pretty
interesting
to
know
also
something
to
eventually
not
just
expose
inside
of
a
compatibility
matrix,
but
to
expose
on
meshri.io
to
be
able
to
say
on
meshwi.io
ashri
supports
x
number
of
measure
goes
deep
right
and
it
supports
all
the
specific
configurations
for
all
these
meshes
and
here's.
E
This
big
fat
number
click
to
see
more
go
right
over
to
the
compatibility
matrix
in
which
people
can
see
a
nice
high
level.
Summary
of
the
you
know
what
meshview
supports,
what
it's
compatible
with
and
then
I
think,
based
on
what
you've
done,
the
if
you
hover
again
over
the
icon
and
it'll,
say
100,
or
is
there
any
if
a
user
clicks
there,
where
you
take
them
to
the
integration
tests?
For
that
results,
like
all
the
passwords,
I
mean
I
mean.
That's,
that's
just
great.
I
mean,
like
you
know.
E
If
a
user
is
interested
generally,
their
interest
will
wane
as
soon
as
they
hit
that
that
third
page,
if
they're,
if
they
start
from
measuring
io
wow,
5
000
component
or
a
thousand
components,
support
wow.
What
does
that
even
mean?
They
click
it?
They
see
that
there's
10
10
of
these
things.
They
see
the
compatibility
of
kubernetes,
the
different
types
of
kubernetes,
so
there's
more
data.
So
it's
really
impressive
and
people
like
they
mostly
want
to
digest
this.
E
They
might
be
curious
as
to
why
it's
yellow
click
it
and
when
they
see
that
there's
test
after
test
after
test,
and
then
they
like,
if
you
click
that
right
now,
like
wow,
okay,
well,
what
happened,
and
they
click
on
one
of
these
to
see
the
details
inside
like
wow,
you
know
like
they
can
actually
believe
like
for
jared
byers,
who's
who's,
given
some
fantastic
input
on
these
calls,
one
of
he
he
focuses
in
qa
and
like
mario,
who
also
focuses
in
qa.
E
It's
kind
of.
I
would
find
it
hard
to
believe
that
mescheri
is
really
doing
all
of
this,
and
really
you
know
I
that
I
would
confidently
use
the
software
to
administer
these
and
it's
like
well,
no,
actually,
it
is
and
here's
the
proof
when
you
can
drill
down
that
far.
I
would
think
most
people,
their
eyeballs,
would
glaze
over
this
third
page
like
okay,
fine,
I
believe
you
find
they
back
out
and
then
they
go.
E
Try
it
where
we
have
another
opportunity
is:
if
you
go
over
to
arc
adapters,
the
adapters
page,
there's
info
about
adapters,
it's
another
place
to
potentially
expose
number
of
supported
components
or
to
to
dynamically.
Have
the
early
the
the
numbers
of
versions
updated
like
it's.
It's
the
same.
The
same
data
file
should
is,
and
should
be
driving
this
page,
the
compatibility
matrix
page
as
well
as
when
you
go
to
supported
platforms
when
someone
wants
to
know.
What's
the
latest
version
of
the
supported.
E
Well,
eventually,
the
compatibility
matrix,
probably
like
this
page
supported
platforms,
probably
needs
to
be
renamed
into
compatibility
matrix
and
then
what
you've
done.
I
ideally
somehow
you're
figuring
out
how
to
intermingle
the
table.
That's
here
which
has
static
info
with,
like
the
real,
the
you
know,
the
real
test
results
that
you're
having
over
there
so.
G
Yeah
so
like
probably
when
you
figure
out
how
to
like
show
compatibility
between
like
different
platforms
like
docker,
we
have
for
kubernetes
now,
but
for
also
docker
and
everything.
F
Also,
I
had
a
doubt:
can
you
go
back
to
that
page,
the
yeah?
So
what
what
formula
are
you
using
to
calculate
the
percentage
that
100
or
31
points.
F
Okay,
so
I
think
that
would
have
worked
in
a
lot
of
cases,
but
I
think
here
it's
kind
of
deceiving,
because
the
way
these
tests
are
is
like
you
can
have
20
tests
which
are
failing,
and
then
someone
comes
up
with
a
patch
fixes.
Everything
21st
test
passes
everything.
So
now
we
should
say
that
we
100
percent
supported
and
support
everything
and
everything
is
working,
but
the
real,
but
the
answer
that
would
be
calculated
by
this
thing.
It
would
be
1
up
1
divided
by
21,
which
would
be
a
deceiving
number.
F
All
the
past
results
are
good
enough,
but
you
know
someone
might
have
fixed
something
and
now
that
everything
is
working,
but
we
are
using
the
previous
tests
who
have
no
significance
with
the
current
build
to
to
conceptualize
how
performant
our
system
is
so
instead
of
that
like,
for
example,
if
you
click
on
the
recent
the
most
recent
result
and
you
get
like
three
parts
of
working
passing
and
two
or
no
basically,
three
parts
are
running.
F
Two
are
not
running,
just
use
that
for
ex
so,
for
example,
this
you
can
use
for
the
latest
test
result
number
of
not
the
spots,
so
ruth
raj
has
made
this
pr,
which
we
are
propagating
through
all
the
adapters.
Now
not
just
pods.
We
can
have
any
kind
of
resource,
so
number
of
free,
successfully
deployed
resources
divided
by
total
number
of
free
expected
resources.
F
So
in
this
case
it
would
be
for
stu,
it
would
be
hundred
percent,
because
we
have
one
two
three,
four,
five,
six,
seven
ports
and
we
expected
seven
parts
so
seven
by
seven,
it's
100.
So
if
you
can
just
go
ahead
and
change
the
formula
to
reflect
that
it
would
be
nice.
G
Yeah,
I
have
one
question,
though:
so:
will
we
be
comparing
for
the
edge
version
or
the
stable
version?
Then
if
we
do.
F
Good
question,
I
think
I
think
I
think,
okay,
what
do
I
think?
F
Let's,
whichever
is
the
recent
I
think
edge
should
work,
but
again,
maybe
so
we
broke
something
off
in
edge
and
the
latest
is
working,
fine,
so
yeah,
I
I
will
tell
you
after
I
think
right
now.
I
don't
have
a
better
suggestion.
To
this
thing.
Yeah
edge
versus
release
is
like,
whichever
is
the
reason
for
now.
Maybe
that
would
work,
whichever
has
the
most
recent
time
stamp
or
yes,
as
lee
suggested
stable,
of
course
think
users,
so
you
can
use
the
stable
one.
F
G
G
That
can
be
done.
Okay,
so
just
like
the
final
question
that
just
to
be
sure
yeah,
I
need
to
calculate
percent
of
total
number
of
like
test
runs
divided
by
passing
it
right,
like
yeah.
F
Yeah,
I
think
the
reason
for
that
was
not
having
sodium
cni
and
raj
modified
the
core
workflow
to
reflect
the.
I
will
not
go
into
details
for
people
interested
in
this
thing.
We
will
talk
more
about
compatibility
matrix
and
that
of
red,
failing
serium
test
and
tomorrow's
building
release,
call
so
make
sure
to
attend
that
and
yeah.
That's
it.
A
A
Okay,
okay,
so
my
workers,
mike's
work
so
far,
is
to
promote
the
bash
commands
and
the
mystery
smp
action
file
on
this
script,
and
my
objective
is
to
make
them
to
make
measure
measuring
ctrl
data
in
the
sense
like
if
we
use
a
cube
ctl
to
apply
yaml
files
to
deploy
an
app
instead
of
it,
we
will
be
aiming
to
up
the
aiming
to
deploy
the
application
from
the
using
the
power
of
the
mystery
ctl
itself.
A
So
that's
what
the
task
is
all
about,
and
so
far
I've
been
running
the
test,
multiple
tests
on
the
github
action
on
my
branch
and
so
far
so
good,
and
what
the
issue
is
that
the
parent
file,
I'm
also
I've,
also
checked
with
you
using
the
service
mesh
patterns
from
the.
What
do
you
say?
The
book
info.
A
And
still
it's
returning
the
error,
500
code
yeah
when
I
tested
in
local.
It
worked
slightly
pretty
fine,
but
the
this.
This
is
the
only
blocker
I
haven't
right
now
and
for
the
service
measures
so
far
been
work.
Only
three
measures
have
been
worked:
hto,
linker
d
and
or
some
open
service
mesh
rest
of
the
service
meshes
zin
is
working
right
now
to
follow
the
same
procedure
as
well.
So
yeah,
that's
the
update
from
my
side.
So
any
suggestions
to
bypass
this
line
alone.
A
F
I
remember
this
500
I
used
to
get
this,
but
that
got
fixed
is
spews
on
the
call
I
think
used
to
you.
Can
you
remind
do
you?
Are
you
aware
of
the
condition
when
the
ct
used
to
return
500
when
you
used
to
do
message,
little
pattern
apply
yeah.
I
don't
remember
right
now:
okay,
I'll
I'll,
look
into
that
and
I'll
connect
with
you
later.
Okay,.
A
Just
one
minus
suggestion:
what
do
you
think
is.
F
A
Yeah,
so
you
just
said
said
that
you
are
not
able
to
reproduce
the
500
error
on
your
local.
So
probably
because
your
machine
is
already
configured
with
you
know
all
the
setup
you
have
done
like
config
database
and
all
so
probably
try,
try
running
it
on
an
empty
container
or
some
virtual
machine
if
you
have
any
like
yeah
okay,
so
you
are
asking
me
to
run
the
machine
cell
machine
as
a
new
product
and
then
try
to
test
these
commands.
F
In
your
local
system,
but
a
clean
system,
so
you
can
get
an
aws
container
for.
A
The
thing
or
I
just
use
virtual
machine
you
have
to
just
you
know,
execute
these
two
deploy.sh
file
right.
So,
okay,
I
think.
A
You
will
better
be
able
to
you,
know
debug
when
you
are
running
it
on.
You
know
a
machine
rather
than
on
github
action.
Other
than
that
I
have
found
a
you
know
article
that
art
you
know
gives
you
context
of
how
you
can
debug
the
github
actions
and
how
you
can
search
into
the
you
know,
containers
while
the
github
action
is
running.
So,
yes,.
D
A
Yes,
I
get
it
again.
That's.
I
have
also
implemented
the
teammate
session
right
now,
thanks
to
a
suggestion
given
by
shiva-
and
it's
been
quite
full
and
well,
but
still
that
the
debugging
of
the
error
code
500,
is
alone
a
blocker
for
me
other
than
that
no
issues
found
actually.
C
A
A
So
my
topic
is
about
enhanced
user
profiles,
so
I
don't
have
a
for
demo
right
now,
but
I
can
like
talk
about
what
I
have
been
working
on
right
now.
So,
for
example,
whenever
so
like
we
sign
into
mesheri,
then
which
first
choose
a
provider
and,
for
example,
if
we
use
remove
provider,
we
get
this
screen
now.
The
thing
is
that
this
is
coming
from
michigan
cloud
that
takes
care
of
logging
in
sign
up
and
all
those
authentication
stuff
which
we
do
for
each
user.
A
So
whenever
you
click
on
sign
in
with
github,
then,
for
example,
if
I
use,
if
I
use
a
profile
like
any
other
user,
then
I'll
see,
I
will
get
authenticated
with
mystic
cloud
and
then
I'll
get
access
to
this
thing.
Basically,
then,
the
problem,
then
my
work
is
like:
when,
whenever
we
go
to
settings
we
should
we
should
have
more
stuff
related
to
user,
for
example,
the
name
of
the
user
first
name
last
name,
and
apart
from
just
using
those
third
party
authentication
provider,
we
should
also
have
a
sign
in
sign
of
form.
A
D
A
This
is
the
thing
which
I
was
working
on
so
that
we
can
retrieve
those
information
from
the
remote
provider
and
get
those
things
on
to
the
user
in
mystery,
ui
and
more
than
that
also,
for
example,
later
on.
After
my
changes,
we
would
also
be
like
any
user,
can
directly
go
to
measure.layer5.io
and
see
their
profile
here
as
well.
Currently,
we
cannot
do
it
because
the
support
for
it
is
not
here.
A
Actually
like
it
depends
upon
the
role
as
well,
but
if
you
are
a
normal
user,
you
won't
be
able
to
see
this.
So
that's
my
next
thing,
and
this
will
also
help
us
to
keep
track
of
users
in
different
websites
like,
for
example,
layer
5..
We
also
have
those
learning
paths
and
we
would
be
also
able
to
store
their
information
of
each
user,
their
preferences
and
then
they
can
manage
them
based
on
their
use
cases.
A
A
Okay,
thanks
for
now
so
next
thing
is
for
ashish
and
me
our
multi-context
features.
I.
F
A
Yeah
so
yeah,
let's
see
if
the
chance
to
I'll
find,
mario,
so
a
blog
on
evolution
of
mashri
ctl
reference
page
alpha.
That's
that's
just
a
minor
update
like
since,
after
the
pr
on
the
what
you
can
see,
the
author
generation
of
the
message,
ctl
documentation
has
been
merged,
so
I've
started
to
write
the
blog
right
now
on
how
the
mystery
ctl
has
been
evolved
and
how
its
car
design
has
been
revamped,
and
it's
been
how
it's
helpful
to
the
newcomers
today.
So
currently
it's
in
progress
right
now
and
yeah
yeah.
A
Okay,
great
those
are
good
updates,
so
yeah,
mario
cyprus,
end-to-end
tests
year
number
five
five
to
six.
Let's
see
it
yeah
by
the
way
you
can
share
your
screen.
H
Okay
sure
yeah,
I
was
spending
on
sharing
it,
so
just
I'll
just
share
it.
That's
okay!.
H
Awesome,
thank
you.
So
I
was
finally
able
to
send
apple
requests
for
contributing
instructions
for
end-to-end
testing
using
cyprus.
So
the
pr
is
up.
We
posted
it
on
slack
feedback
is
welcome
and
I've
got
the
changes
over
here.
So
basically,
it's
a
new
page
over
here
and
here
the
idea
is
to
describe
a
the
the
framework,
the
test
framework.
So
basically
it's
the
foldering
structure
of
our
ui
cypress,
a
folder
and
located
in
a
measuring
repository.
H
So
here
each
folder
is
described
as
to
a
what's
that
intent.
What
are
the
files
located
there
supposed
to
be
used
for,
and
so
we
have
action
helpers?
What
are
the
fixtures
fixtures
and
there
are
also
some
useful
links
to
the
actual
cypress
docks.
H
So
you
see
this
one
would
redirect
to
the
core
concepts
of
writing
and
organizing
test,
specifically
to
the
what's
a
fixture
file
and
and
also
a
like
high
level
description
of
what
what
it,
what
what's
the
purpose
of
these
fixture
files
for
both
integration
and
end-to-end
tests?
H
There
could
be
more
usage
like
use
cases
for
fixture
files,
but
the
most
common
is
the
stubbing
response,
data
for
integration,
test
and
reusing
data,
as
could
be
test
input,
but
also
like
things
that
tests
are
supposed
to
to
use
to
input
on
the
the
ui
or
also
like
expected
a
expected
conditions
or
data,
so
that
that's
kind
of
like
trying
to
enable
like
data
driven
testing
and
yeah,
so
also
the
other
different
folders
like
it's
at
least
to
me.
H
It
was
a
bit
confusing
why
there's
an
integration,
integration,
folder
and
then
an
integration
and
end-to-end
folder,
but
like
it's
here,
it's
clearly
described
when
an
integration
test
would
be
used.
What
are
the
benefits
versus
an
end-to-end
test?
H
So
basically
it's
testing
both
the
server
and
the
ui
so
on,
and
some
links
to
other
cypress
docs
are
here
on
sitting
data,
bypassing
ui.
What
that
means,
actual
server
responses
and
different
things
like
cypress
routes,
weights
and
assertions.
So
since
anyone
this
if
anyone's
writing
an
end-to-end
test
like
these
links,
would
be
very
helpful
in
that
context.
So
it's
like
a
reference
guide
to
actual
examples
and
like
very
easy
to
understand.
H
Also
the
plugins
like
what
are
they
how
to
use
them
all
these
links,
reference
like
they
would
redirect
someone
to
the
to
the
actual
github
file,
like
the
the
one
in
master
and
some
other
links
for
plugins
guide
and,
finally,
the
support
the
support
files
like
how
whatever
what's
a
custom
command,
how
to
like?
What's
your
purpose
and
even
some
like
a
link
to
our
own
custom
command
script,
and
also,
of
course,
how
to
open
the
cypress.
H
So
I
added
a
custom
like
an
npm
script,
a
in
the
in
the
ui
project.
So
the
idea
is
to
just
use
this
command
and
open
the
this
actually
runs
the
the
dev
server
for
us.
So
it's
actually
easier
than
having
to
open
multiple
consoles
to
open
to
run
the
server
and
then
another
one
to
open
cyprus.
So
this
will
open
in
interactive
mode
and
it'll.
It
it'll
help
like
for
anyone
to
rerun
the
tests
and
a
oh
yeah.
Basically,
that's
it
and
there's
some
useful
links
over
here.
H
The
one
I
want
to
highlight
is
this:
learn
cyprus.
So
they
have
really
good
courses
over
here,
so
that'll
that'll.
I
think
that'll
help
a
lot
and
yeah
yep
and
just
one
more
thing
in
the
measuring
ui
there
is
this
running
cyprus,
integration
test.
This
was
lee's
idea.
I
think
it's
very
good
so
actually
having
a
place
here
to
explain
the
simple
make
command
to
to
run
the
this.
This
is
our
only
for
the
integration
test.
H
Sorry,
so
this
would
run
only
the
front
end
and
some
tests
that
that
stop
the
the
server
responses.
So
that's
all
any
questions
or
comments,
complaints.
H
Yep
all
right
yeah,
I
hope
it
helps.
I
mean
I
think
there.
There
is
a
separate
issue
in
layer,
five
ripple
to
create
a
blog
like
actually
creating
a
test
so
I'll
work
on
that
next,
but
I'll
still
wait
for
some
feedback
on
this
pr
case.
There's
any
and
that's
all
for
right
now.
A
Okay
yeah,
so
so
we
have
handful
of
minutes.
Can
we
discuss
about
the
backstage
and
the
plugin.
E
E
E
E
That,
okay,
all
right
yeah,
I
can
see
that
so
backstage
there's
one
of
the
users
of
mashari
samir
had
at
one
point
requested
that
there'd
be
a
plug-in
for
backstage,
and
so
in
a
similar
like
in
a
similar
vein
as
to
how
someone
else
had
requested
a
well
a
helm,
chart
or
a
plug-in
for
measuring
into
cube
sphere.
Someone
else
has
requested
a
messagery
be
available
in
backstage,
and
so
you
came
to
my
mind
as
someone
that
might
want
to
pick
this
up.
E
This
issue's
gone
stale
a
couple
of
times
and
are
you
familiar
with
so
backstage
is
another
cncf
project.
It's
a
service,
catalog,
there's
a
link
here
to
how
to
create
a
plugin.
E
E
Yeah
and
you
don't
so
so
as
you
do,
that
one
know
that
you
don't
need
to
know
anything
about
backstage
potentially
like
you
know,
because,
but
rather
I
mean
it's
good
that
you
do,
but,
as
you
look
to
like,
hopefully
they've
made
it
fairly
straightforward
about
what
changes.
We
need
to
be
included
in
backstage
to
be
able
to
include.
E
I
don't
know
but
yeah
good,
okay
yeah.
If
that's
the
case,
then
you
know,
maybe
actually
let
me
check
that.
E
Yeah
sounds
good
yeah
as
you
investigate
just
drop
a
comment
on
the
issue.
If
it's
of
interest
else,
I'm
we'll
either
we'll
fight
we'll
grab
another
contributor
and
and
then
maybe
you
can
tag
team
with
them,
if
that
makes
it
a
little
more
tangible,
a
little
more
palatable,
rather
so.
E
Nice,
okay,
yep
good.
So
by
the
way
for
those
that
don't
aren't
familiar
with
siva,
he
had
recently
created
a
github
action
to
publish
mescheri's
helm
chart
to
cubesphere,
which
is
great.
I
mean
he's
going
to
take
the
contents
of
that
action
and
integrate
that
into
mesherie's
workflow
for
really
well
actually,
the
yeah
for
releasing
the
helm
chart
to
mesher's
helm,
chart
to
cube
sphere
just
before
he
does
that
he's
just
gonna
go
manually.
E
Do
that
integration
see
how
much
people
use
it
and
if
it's
worth
kind
of
following
up
to
with
automation
but
yeah
so
jeez
see,
but
you
haven't
even
been
on
the
call
before,
and
it
already
takes
me
five
minutes
to
explain
what
you've
already
done.
That's
great.
E
You
would
think
that,
like
green
is
my
favorite
color,
it's
I
didn't
choose
the
colors,
but
anyway
I
white,
okay,
I
get
it
white,
okay,
I'll,
remember,.
A
So
yeah,
so
we
are
already
passed
a
minute,
so
we
have
one
item
left,
which
is
we
can
discuss
it
in
future
development
calls
so
yeah.
It's
time
to
see
you
all
goodbye
and
thanks
once
again,
you
know
you
joined
it
and
gave
your
precious.
F
And
people
who
have
stayed
till
the
end,
just
one
interesting
thing
see
you
on
the
next
dev
call
because
that
item
that
we
skipped
most
probably
we're
going
to
give
a
live
demo
of
the
multi-context
thing.
We're
pretty
enticing
to
see
so
see
you
in
the
next
call
yeah.