►
From YouTube: 2021-04-21 meeting
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
A
A
A
A
Yes,
yes,
if
I
got
this,
I
will
try
to
wrong,
but
I,
but
I
didn't
see
these
resources
are
destroyed.
I
mean
the
the
already
exist
results.
B
A
A
B
Of
the
other
steps
fails,
it'll
still
call
destroy
in
the
very
rare
case
that,
for
some
reason
it
didn't
get
called
we
would
have
those
things
left
over
and
that
we
would
just
need
another
job
to
keep
to
try
to
find
old
stuff
to
delete.
I
guess,
but
I
don't
think
that's
gonna
happen
very
often,
so
we
could
follow
up
with
the
cleanup
for
that.
C
A
Oh,
I
think
this
wig
from
aws
side.
We
have
not
too
many
too
much
to
report,
but
we
are
working
on
the
csv
workflow,
so
you
will
see
ourselves
pr
pushed
in
last
week
and
we
also
did
some
change
goal
changing
downstream
repo
aws
of
visibility.
So
we
hope
next
week
we
can.
A
We
can
finish
most
of
the
work
of
csd
and
publish
some
public
there
for
python
node.js
and
java
yeah.
A
So
I
don't
know
if
anyone
is
trying
that
I
think
I
posted
that
issue
one
month
ago,
but
I
see
no
one
reply.
A
B
D
Well,
I
I
actually
have
a
question
regarding
I
guess:
lambdas
and
tags,
and
whether
or
not
there's
any
any
plan
to
support
tags
like
aws
tags
from
lambdas,
or
is
that
like
looking
through
the
layers
it
didn't
look
like
that
was
currently
implemented?
Is
that
something
that's
kind
of
on
the
on
the
plans
at
all.
D
B
D
I'm
curious,
if
there's,
if
there's
any,
if
you
know
you
guys
have
thought
at
all
about
the
about
being
able
to
pull
them
from
pulled
attacks
from
like
the,
I
guess,
the
aws
api
and
loading
them
into
resources.
D
I
mean
I
would
assume
something
like
this
with
assuming
it's
available
through
an
aws
api.
I
guess
it
would
just
have
to
be
loaded
inside
a
resource
detector
or
something
like
that.
B
D
B
D
D
B
I
wanted
to
bring
up
one
discussion
point
if
there
aren't
any
done
with
that
topic,
so
I
put
a
discussion,
the
red
and
the
open
tomato
lambda
repo.
It
is
about
whether
we
want
to
publish
separate
or
combine
the
layers
for
sdk
and
collector
good.
Anyone
have
any
thoughts
on
that.
C
So
yeah,
I
actually
I
have
a
question
about.
I
just
went
through
the
design
doc
and
you
know
so
so
this
is
supposed
to
work
with
the
like.
You
know
the
open,
telemetry
java
agent
or
you
know
other
language
agents
and
does
it
have
to
work
with
the
hotel
collector
as
well
or
you
know,
okay,
so
so
so
it's
the
same
model.
So
there
is
no
exporter
on
the
on
the
sdk.
It's
it's
off.
The
collector.
C
C
Okay,
so
so
so
the
lambda,
so
the
app
is
in
the
lambda,
and
you
have
the
sdk
on
another
layer
is
that
is
that
for
the
okay,
okay
and.
B
C
Oh
so
it
could
be
the
same,
so
the
sdk
with
the
auto
instrumentation
or
you
know
the
that
will
be
in
one
layer
and
and
the
collector
was
in
the
same
layer
or
separate.
B
At
least
that's
what
the
duck
said
surprising
limit
right.
This
is
the
first
time
I
heard
of
that.
Really
at
least
that's
also.
Google
highlights,
if
you
say,
italy
is
lambda
layer
limit
five
layers.
C
B
C
B
B
A
B
A
D
B
A
I
remember
we
had
a
discussion
probably
three
months
ago,
so
it
depends
on
customer
behavior.
If
a
customer
is
a
developer
as
a
hotel
developer,
they
must
be
very
familiar
with
open
parametric
structure.
I
know
that
the
different
data
sdk
and
the
collector
so
for
these
power
users
they
would
like
separate
layers,
but
also
normal
for
normal
users.
We
can
imagine
that
most
of
the
developers,
they
are
not
a
professional
hotel
developer
for
their
site.
A
C
Yeah
and
well
I
you
know
I
actually
haven't
looked
at
this
project
very
closely.
Is
there
a
good
like
an
example
or
a
sample
to
start
with
that
works
for
the
out
of
the
box
collector?
We
have.
We
have.
A
A
And
maybe
if
you
know
that
in
aws
lab.console
you
may
see
some
integrated,
it's
like
official
aws
official
layer
so
for
that
layer,
user
just
need
to
click
some
button
and
go
to
enable
them.
So
I
think,
sooner
or
later,
open
permission
number
will
be
deeply
integrated
into
lambda.
In
such
case,
I
think
one
layer
is
better
than
today.
Let's
see
easier
to
integrate
into
a
good.
A
Console
yeah
so
right
now
customer
can
just
click
a
button
to
enable
lambda
inside
it's.
It's
also
a
monitor
tool
or
click
a
button
to
enable
a
configuration
proxy.
So
we
can
imagine
if
we
have
a
two-step
review.
How
can
customer
use
one
click
button
to
universally
that's
possible,
but
the
value
are
still
simpler
than
player.
C
The
the
reason
I
said
separate
layers
was
I've
seen
you
know
like
we
have
neurally
has
a
hotel
collector.
You
know
just
an
instance
that
you
can
just
start
up
without
configuring.
Anything
right
just
have
environment
variable
as
an
api
key,
and
and
that's
it
you
don't
you
know
you
don't
have
to
set
up
the
exporter.
You
do
not
even
know
that
there's
an
exporter
that
lies
in
that
collector
right,
yep
yep,
so
it.
A
C
Of
transparent
for
for
a
user
of
this
collector
right,
they
don't
have
to
know
for
exporters.
A
B
C
B
B
B
B
D
Might
make
sense
to
capture
this?
The
discussion
inside
the
github
discussion
just
to
make
sure
we're
making.
B
A
So
in
upstream,
open
television
lab
or
we
want
to
publish
it
to
separate
layer
sdk,
but
in
downstream
we
keep
our
previous
decision.
We
will
publish
combined
layers
and
for
the
other
contributor.
If
they
want
to
maintain
their
downstream
level,
they
only
need
to
maintain
their
collector
yeah
yeah.
That
makes
sense
yeah.
B
B
B
B
A
No,
no
that's
a
very,
very
beginning
implementation.
So
before
before
alex
provided
his
extension,
my
invitation
is,
I
separated
to
a
standalone,
but
I
think,
as
a
essential
one,
one
processor
is
always
better
than
two,
so
you
mean
you
want
to
use
the
expression
for
your
shell
to
launch
a
collector
right.
B
A
That
means
that
means
it's
very
hard
to
manage
the
collector,
because
the
character
is
running
in
a
standalone
process.
How
could
you
control
it?
From
extension,
you
know
extension
will
receive
some
event
from
longer
long
time,
so,
probably
in
future
character
become
more
powerful
than
right
now,
if,
if
class
can
receive
some
event,
because
one
process
is
easier.
A
B
A
A
So
give
an
example:
if
you
fashion
crash
language
sandbox,
you
don't
know
that
and
report
the
whole
environment.
But
if
you
separate
the
extension
and
collector
if
collector
crashes,
who
will
know
that
ramadan
would
not
report
that
we
need
to
manage
that.
We
need
to
add
a
health
check.
B
A
But
anyway,
I
believe
we
can
solve
the
law
of
issue
the
adot
idiot
structure.
It
does
apply,
for
example,
pretty
like
character
extension,
so
if
a
dot
collector
can
change
the
log
level,
we
also
can
change
the
level
just
we
need
to
investigate
them.
How
to
do
that.
We
don't
need
to
separate
the
whole
process.