►
From YouTube: 2023-02-15 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
A
C
C
Yeah
Dennis
I
think
on
vacation,
so
yeah
whatever
the
sick
meeting
again.
C
All
right
and
I
think
we
can
get
started
so
hello,
everybody
today,
the
first
bind
on
the
agenda
is
the
same
as
we
had
for
the
last
few
weeks.
That's
this
exponential
histogram
PR
by
Matt,
where
out
there,
where
the
first
part
has
already
been
merged.
There
is
a
second
part
open
and
then
already
reviewed
that
one
would
be
good
to
have
another
review
on
that
to
for
us
to
be
able
to
get
that
one
merged
and
then
move
on
to
part
three.
That's
the
video
out.
C
A
Yeah,
so
the
sandbox
is
moving
forward.
I
got
I,
have
the
API
merged
when
I
try
to
rethink
with
version
1,
13
I
think
it
was.
A
There
was
a
bug
identified
because
we
now
have
CI
integration
and
yes,
we're
failing,
because
I
failed
a
copy
over
the
previous.config.js
so
effectively
we
had
one
part
of
it.
Yes,
let's
say
I
want
double
quotes
and
the
other
part
saying
I
went
single
quotes
and
it
failed.
So
because
the
prettier
con
pick
wasn't
there,
which
says
useful
quotes
So.
A
So
this
particular
PR
is
just
updating
the
automatic
merge
script
to
say,
you
know,
make
sure
you
copy,
like
in
the
case
of
line
142
copy.
The
prettier
config
from
the
JS
repo
auto
merge
for
auto
merge.
Js
is
effective.
The
entire
JS
repo
and
says
you
know,
drop
that
into
the
root
of
the
the
sandbox
ads
into
missing
definitions.
There
I
don't
know
why
I
added
cross
mode
module
references
adds
in
the
prettier
things.
A
As
part
of
the
auto
merge
script,
it
checks
to
see
whether
the
whether
there
is
an
outstanding
PR,
so
on
line
49
of
the
Czech
PR
exists.
Cs,
that's
just
saying
restricted
to
the
current
user.
So
while
the
open
Telemetry
bot
had
a
PR
open
it,
wouldn't
let
me
create
a
test
PR.
So
that's
just
restricting
it
to
the
current
person
and
I
just
rebooked
the
the
lips
scripts
a
little
bit,
so
they
just
call
it
and
then
pass
additional
arguments
to
the
face.
One
there
identical
yeah.
A
I
think
in
the
conversation
I
put
in
for
the
pr
I
put
in
a
link
to
the
PR
that
I
could
I
generated
locally.
It
was
just
now
closed,
so
it
was
like
number
39.
If
you
go
back
to
the
previous,
so
yeah
to
example,
PR
murder
result.
A
So
you
can
see
what
this
PR
changes
versus,
there's
already
an
open
one
there
for
open,
Telemetry
bot,
which
is
failing
because
of
the
yes
length
issues.
So
you
can
like.
C
A
A
But
yeah
moving
forward,
it's
still
on
I,
think
On
Target
to
have
most
of
the
web
packages
done
by
the
end
of
the
month.
So
once
this
get,
this
one
goes
in.
I'll
then
merge
it
get
it
up
to
date,
I'll
then
start
effectively,
I'm
commenting
out
the
the
config
TS
to
start
bringing
in
more
of
the
web
packages.
At
the
moment
it
is
just
the
API
from
Js.
A
That's
all
running
using
Rush
running
web
worker
tests
running
a
test.
You
know
running
node
web
node,
browser
and
web
worker
tests
and
creating
bundles
for
everything,
which
is
what
this
script
does
when
it
brings
it
over.
C
A
And
once
I
get
to
the
bugs
which
do
exist
because
there
are
some,
there
is
some
code
that
there's
some
browser
tests
that
fail
when
I
try
and
run
it
a
web
worker
because
visibility
change
doesn't
exist,
so
I
have
to
create
a
PR
on
the
js
on
that
and
remember
the
JS
or
the
country
repo
to
get
them
pushed
in,
so
that
they
can
then
propagate
their
way
back
to
the
sandbox.
C
Yeah
I
think
there's
a
few
of
those
out
there.
They
keep
popping
up
from
time
to
time,
but
they
don't
think
we
have
issues
for
all
of
them.
A
Yeah
there's
one
of
the
reasons
for
making
the
sandbox
run
both
browser
and
web
worker.
If
I
can
is
because
yeah
we
have
a
number
of
instrumentations
that
don't
work
when
it
runs
in
a
web
worker
because
they
directly
wrap
this
document.
C
C
If
there
are
no
more
topics,
then
I
guess
we
can
move
on
to
the
back
triage.
The
first
one
here
seems
some
container
ID
being
set
wrong.
C
That
is,
is
that
in
aquarable?
Yes,
so
that
is
the
container
idea
in
an
eks
cluster.
C
Is
most
likely
going
to
be
related
to
the
to
the
eks
detector?
I
suppose
does
anybody
know?
Is
that
does
that
reside
in
the
in
the
query
book
I
think
as
far
as
I
recall,
that
is
in
concrete
right.
D
C
All
right
so
looks
like
they're
they're,
actually
using
both
and
yeah.
So
that's
the
container
detector
as
faster
recorders
are
so
part
of
country,
bread.
D
C
Yeah
so
I
guess
I
would
transfer
that
issue
over.
D
D
On
a
maybe
semi-related
node
I'm,
not
sure
how
all
these
detectors
are
going
to
be
updated
to
work
with
the
new
changes
on
on
Main
around
detectors.
Well,
probably,
some
work
to
be
done
here.
D
C
Yeah,
that's
also
how
it
works
as
far
as
I
record
right.
It
won't
make
all
of
the
attributes
resolved
as
asynchronously
by
default.
You
have
to
make
a
manual
effort
to
change
your
detectors
over
right.
B
B
It's
in
contrib,
it's
a
PR
one.
Three
one,
eight
hundred.
B
Yeah,
so
this
PR
introduces
a
text
in
the
readme
that
describes
that
the
order
of
a
detector
is
important
and
the
last
one
that
is
defined
at
the
wins
like
if
there
are
multiple
detectors
that
set
the
same
attribute
value.
So
it
it's
doing
it
only
for
AWS,
but
maybe
it
should
be
border,
like
maybe
some
other
detectors
that
Collide
and
detect
the
same
attribute
name.
B
So
just
wanted
to
bring
this
to
your
attention.
C
Yeah
I
will
definitely
give
this
one
a
look
and
see
it
I
guess
it
would
make.
It
would
definitely
make
sense
to
have
that
or
something
like
that
for
in
a
broader
sense,
for
all
the
detectors,
not
just
the
AWS
ones,.
B
Yeah
actually
I
tried
to
search
the
spec
for
like
a
specification
on
how
the
detectors
should
be
merged
into
a
single
resource
and
I
couldn't
find
it
and
I
think
that
the
collector
is
preferring
like
a
bio
resources
and
we
prefer
like
later
resources.
So
there's
some
inconsistency.
B
C
Yeah
I
also
have
not
not
really
checked
that
for
for
the
others,
I
think
it
will
have
most
likely
to
one
one
of
either
approaches
like
keeping
the
first
one
or
keeping
the
second
one
I
think
this
async
detection,
with
that,
the
approach
that
we
have
that
that
just
doesn't
exist
for
most
of
the
other
sdks
right.
B
No,
it's
not
related
to
I!
Think
like
if
you
have
multiple
resources
and
you
won't
urge
them
together
right
so.
D
I
think
I
think
it's
there.
It's
talking
about
merge
and
it
says
the
resulting
resource
must
have
all
attributes
that
are
on
any
two
of
the
input
resources.
If
a
key
exists
on
both
the
Old
and
updating
resource,
the
value
of
the
updating
resource
must
be
picked
even
with
the
updated
value
is
empty.
D
Doesn't
say
if
the
first
one
is
yeah,
if
you
merge,
if
you
start
with
the
first
and
then
merge
these
subsequent
ones
or
if
you
start
with
the
last
and
kind
of
work.
B
A
C
All
right,
then,
moving
on
to
the
next
one
seems
to
be
some
problem
with
it,
trying
to
read
and
and
environment
wherever
or
something
that
is
not
defined
in
the
browser.
C
And
that
seems
to
be
a
flannel
already
answered.
There
seems
to
be
some
stuff
going
on
with
bundling.
C
C
C
But
all
in
all
that
looks
like
either
a
P1
or
P2
to
me.
What
would
you
say
for
me
this?
This
looks
like
it.
It
throws
some
unhandled
error
and
some
promise,
which
we
probably
won't
don't
want
to
have
so.
C
All
right,
I'll
remove
the
the
triage
is
already
removed
automatically
all
right,
then
we
can
move
on
to
the
next
one.
C
C
Or
if
it
is
the
export
or
something
in
between,
but
it
causes
somehow
causes
Telemetry
not
being
exported
so
I
would
yes,
that's
a
p.
C
The
Prometheus
exporter,
and
if
it's
related,
something
related
to
that,
the
number
I
will
give
that
a
look
and
otherwise
I
would
just
do
some
digging
and
see
if
the
the
HTTP
instrumentation
has
some
problems
there.
C
All
right,
and
then
here
is
another
one
related
tool:
gcp
Cloud
functions,
which
I'm
less
familiar
with
than
I
would
like
to
admit.
So
there
seems
to
be
some
grpc
core.
C
Okay,
so
this
this
seems
to
be
around
because
there's
some
something
still
running
where
there's
a
timeout
not
too
familiar
with
this,
but
like
it's,
it's
shutting
down
and.
C
I
will
cycle
back
to
that
one
after
the
meeting
and
and
see
how
to
properly
prioritize
this
one
might
also
be
a
duplicate
of
of
the
other
one
here,
all
right
and
the
next
one
is
created
by
then
that
I
think
is
for
the
Milestone
that
he
created
for
yeah
open
Telemetry
SDK
2.0.
So
we
don't
have
to
take
a
look
at
that
one.
Now
all
the
others
are
already
assigned
I
guess
we
can
then
move
on
to
the
country
box.
C
C
B
C
That's
a
different
one,
I'm,
not
sure
if
that
one
actually,
let's
see
not
instrumenting.
B
C
C
I
will
just
add
a
link
to
what
was
the
number
on
that
one
1313.
C
All
right
and
now
we
should
be
through
I,
think
so
is
it
if
anyone
has
any
more
topics,
now
would
be
the
time
otherwise,
I
will
give
you
35
minutes
of
your
time
back.