►
From YouTube: Diagnostics WG meeting December 01 2021
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).
B
A
A
Getting
a
right
balance
of
time
is
always
challenging.
For
example,
the
current
time
is
too
too
early
for
mary,
and
it's
already
too
late
for
stephen
stephen
said:
he's
he's
able
to
adjust
because
he
has
other
meetings.
But
if
we
move
on
either
side,
then
it's
it's
a
tricky
tricky
stuff
to
do
now.
One
of
the
proposals
that
we
have
on
the
table
is
to
rotate.
A
Which,
I
believe
is
to
have
two
two
meetings
at
two
different
times,
so
that
at
at
least
everybody
can
attend
at
least
one
one
meeting.
A
Yeah,
that's
a
valid
question
for
the
the
meeting.
The
the
diagnostic
working
group
itself
is
four
or
five
members
for
regularly
attending
when
it
is
a
single
session,
when
we
split
it
as
we
split
into
half,
then
it
would
be
like
two
or
three
in
either
of
the
meetings
that
that
I
don't
know
how
that
would
affect
the
quality
and
the
and
the
focus
on
the
stated
items
in
the
working
group.
B
C
E
A
A
E
B
D
B
A
A
A
I
will
try
that
michael
okay
and
contact
you
in
case,
if
I'm
unable
to
do
that.
Okay,
thanks.
E
E
C
I
think
we're
about
the
only
people
who
do
it
at
datadog,
apm
vendors.
Do
it
so.
C
In
part,
because
back
porting
stuff
is
so
problematic
so
like
in
order
for
us
to
actually
use
async
hooks
as
we
do
today,
a
bunch
of
stuff
goes
wrong
because
of
performance
performance
gets
really
bad,
and
if
we
try
to
fix
it,
we're
not
allowed
to
backboard
it,
because
almost
everything
you
do
in
async
hooks
is
considered
major.
E
C
C
E
C
E
I
think
that
you
know
the
whole.
The
whole
thing
here
was
like
it's
too
low
level,
which
probably
leads
to
some
of
the
things
you're
talking
about
so
like
what
other
apis
can
we
have
that
are
fill
the
bill,
but
you
know
one
the
we're
comfortable
as
a
as
a
project
making
stable
and
would
also
probably
be
like
you
know,
get
around
some
of
those
issues.
You're
talking
about
so
like
async,
local
storage
is
one,
but
this
was
to
gather
use,
k,
other
use
cases,
key
use
cases.
C
Yeah,
so
the
big
one
is
context
propagation
for
us
what
we
call
scope
management
internally,
so
we
have
some
slides,
I'm
working
on
to
kind
of
present
this
internally
at
our
company
there's
multiple
contexts,
especially
whenever
you
do
an
a
weight
and
you
need
to
gather
not
just
like
async
local
storage,
which
propagates
the
current
execution
stack
onward.
But
after
it
returns
from
the
await,
you
need
to
know
basically
what
it
did
inside
of
the
await.
C
What
async
tasks
were
still
running,
so
you
can
propagate
stuff
further.
The
default
example
there
that
breaks
really
bad
when
you
try
to
use
async
local
storage
style
propagation
is
a
message
queue.
So
aws
has
the
sqs
library.
It's
basically
message
queue.
You
can
use
it
and
get
a
message
out
and
when
you
get
a
message
out
using
async
local
storage,
you
don't
get
anything
if
aws
were
to
attach
something
during
async
local
storage
context
propagation.
You
won't
see
it
outside
of
that
and
that's
bad.
C
So
we
have
an
idea
of
a
different
concept.
This
kind
of
was
hinted
up
at
by
stephen
when
there
was
a
recent
pr,
reverting
some
stuff
about
resource
ownership
in
asynch,
hooks,
there's,
multiple
kinds
of
ownership
and
context
propagation,
and
we
don't
have
a
clear
one
for
what
we're
calling
follows
from
we
do
for
async
local
storage,
which
is
basically
not
follows
from,
but
child
of
so
right
in
global
storage
is
always
the
child
of
a
task
inherits.
C
We
do
it
today,
poorly
with
asus
hooks.
That's
part
of
the
introduction
of
the
promise
hooks
because
it
makes
it
right
more
correct.
E
C
E
So
yeah,
so
I
guess
that's
another
use
case
to
make
sure
that's
called
out
in
terms
of
like
async
hooks
replacement
right.
C
B
E
Yeah,
I
think
this
is
why
I
think
you
know
we
we
should
have
the
discussion
when
we
have
stephen
but
like
it
seems,
like
we've
gone
out,
we've
asked
for
for
feedback
what
we
now
need
to
sort
of
like
spend
a
session
talking
about.
Okay,
what
are
the
next
steps
on
this
and
how
do
we?
A
C
C
So
we,
like,
I
said
earlier
in
the
call
we
are
starting
to
change
our
instrumentation
to
use
diagnostics
channel.
Is
there
like
a
user
success
tracking
going
on
in
this
working
group,
or
is
it
just
kind
of
we
report
back
to
you
if
we
have
problems,
I
join.
E
D
D
E
C
E
Right,
so
I
think
this
group
should
definitely
try
and
help
support
getting
that
message
out.
If
there's
a
you
know,
if
there's
a
success
story,
we
can
point
to
promote,
and
even
just
like
you
know,
if
it's
we
need
to
the
next
part
on
diagnostic
channels,
some
blog
posts
sort
of
explaining
what
it
is
and
how
to
use
it
and
encouraging
people.
That's
also
something
this.
B
A
A
F
A
F
C
Yep
steven-
I
don't
know
if
you
saw
the
pr
on
the
datadog
stuff,
that
we
are
doing
that
with
monkey
patching
in
a
new
thing.
F
C
F
It
would
be
reasonable
to
add.
I
think.
C
F
Yeah
but
yeah
I
mean
just
like
having
we
could
have
like
another
channel.
That
is
just
like
same
thing
as
forget.
What
the
name
of
the
thing
is,
but.
C
A
A
If
no
other
items
on
the
agenda,
if
everybody
has
any
other
items,
otherwise
we
are
done
for
today.
B
B
I
think,
for
I
don't
know
doing
some
lack
of
traction
due
the
fact
that
the
document
it's
easily
it's
easy
to
do.
I
think
that
we
could
promote
it
or
we
could
share
it
publicly
with
the
community.
If
someone
is
interested
in
create
some
tutorial
in
one
of
our
current
list
of
of
missing
tutorials,
I
think
that
would
be
great
because
this
this
work
we
are-
we
have
been
doing
the
past
past
year,
but
since
it
has
finished
in
the
document
we
haven't
exposed
it
to
the
community.
E
B
Yeah,
I
think
so,
because
the
tutorials
is
some
of
them
is
easy,
or,
for
instance,
for
folks
that
who
have
used
at
race.
Events
for
something
would
be
more
easier
as
well,
and
also
there
is
some
some
tutorials
missing
for
specifically
operational
systems,
so,
for
instance,
the
when
you
try
to
use
native
tools
like
the
trace,
it
usually
works
when
you
use
solaris
as
operational
system,
and
I
could
make
it
so.
B
A
A
A
A
E
B
Yeah,
actually,
I've
shared
here
in
the
chat,
an
issue
that
I
have
creating
september
right
with
a
list
of
missing
tutorials.
I
think
that
it
could
be
a
a
good
start.