►
From YouTube: 2023-01-18 meeting
Description
cncf-opentelemetry@cncf.io's Personal Meeting Room
B
C
C
C
All
right,
then
I
guess
we
can
get
get
started.
C
So
let
me
share
my
screen.
Real
quick.
C
So,
first
topic
of
the
day,
we
have
a
new
release
for
the
API
SDK
and
experimental
packages
out
the
performance
clock
drift
PR
for
both
traces
and
metrics
made
it
in
so
yeah.
Hopefully
that
is
now
fixed
once
and
four
and
yeah
the
various
improvements,
metrics
instrumentations
and
more
also
out
then
moving
on
to
the
me
asking
you
for
for
reviews,
the
exponential
histogram
PR
is
still
out
there
and
open
Matt
put
a
lot
of
work
into
this.
We
currently
have
two
reviews
in
on
it.
C
One
more
would
be
would
be
highly
appreciated.
So
if
anybody
has
some
time
to
read
into
the
exponential
histograms
yeah
one
one
more
review
would
be
highly
appreciated.
We
otherwise
I
think
Dan
has
mentioned
that
we
were.
We
were
move
on
with
just
two
reviews
and
merge
it
so
that
we
can
continue
the
review
process.
This
is
split
into
three
parts
and
yeah.
We
have
we've
talked
about
this
a
few
times
already.
C
C
Then.
The
next
topic
is
one
that
I
put
on
the
agenda
is
basically
I
have
noticed
that,
when
using
the
instrumentation
packages
and
also
having
a
matrix
export
in
there,
we
tend
to
export
a
lot
of
empty
Scopes.
So
basically
we
export
one
scope
for
each
instrumentation
package.
So
if
you
load
the
auto
instrumentation
package,
you
you
get
a
bunch
of
yeah
empty
Scopes
that
are
being
exported
without
any
metrics
in
them
and
I'm,
not
sure.
C
If
we
talked
about
this
before
or
not,
but
I
realized,
we
could
delay
the
creation
of
the
meter
until
the
first
time
the
the
get
ice
card
for
the
instrumentation,
so
I'm,
hoping
if
somebody's
here
who
is
familiar
with
the
instrumentation
packages,
could
chime
in
and
basically
tear
their
thoughts
on
that.
If
you
have
any
otherwise
I
would
just
go
forward
and
open
the
pr
for
it,
as
that
would
greatly
reduce
the
the
amount
of
stuff
that
we
export.
C
All
right,
so
if
it,
if
there's
no
opinions,
then
I
guess
I
would
just
go
ahead
on
that
one
write
any
other
topics
for
today.
C
All
right,
thank
you.
So
this
is
the
the
draft
PR
that
you,
oh
no,
that
is
a
different
one
reviews
on
the
eager
exporting
for
batch
spam
processor.
Do
you
want
to
talk
about
this
very
quick.
D
C
All
right,
yeah,
so
I
guess
this
is
call
for
reviews
on
that.
A
C
I
will
put
that
on
my
list
and
and
see
if
I
can
review
that
today
or
tomorrow.
E
Yeah
my
last
comment
on
this
is:
it
looks
mostly
good,
except
for
the
last
bit
where
the
thing
the
time
is
running
all
the
time
we
shouldn't
have
a
timer
running
unless
there's
something
in
in
the
queue
descent.
So
if
that's
fixed,
then
I'd
be
happy
to
start
that
off.
C
All
right,
thank
you,
and
the
next
one
is
from
Martin.
C
This
is
the
pr
that
you
opened
recently.
I
already
had
a
quick
peek
at
that
one.
Do
you
want
to
talk
about
it.
A
Yeah,
just
it's
it's
in
draft.
Currently,
it's
basically
separating
the
events
API
from
the
logs
API
in
into
a
separate
package.
This
is
following
a
recent
discussion
in
the
logsig
and
also
there
is
a
similar
implementation
in
Java.
That's
also
in
being
reviewed
right
now,
but
that's
the
direction
that
I
think
we're
heading
currently
having
a
separate
events,
API
and
events
SDK,
which
which
would
make
a
clean
separation
between
events
and
logs.
A
In
this,
please
take
a
look
at
it.
Take
a
look
at
it.
This
PR
specifically
I
mean
nothing
really
complicated.
It's
just
separating
the
existing
API
into
a
separate
package.
C
So
the
the
the
general
approach
will
be
that
there
is
an
API
for
logs
and
events
and
they
will
use
the
same
SDK
as
files.
I
recall
right.
A
So
there
will
be,
there
will
actually
be
separate
events,
SDK
and
separate
logs
SDK.
All.
A
C
All
right,
any
other
comments
or
topics
that
you
want
to
bring
up.
C
If
there
are
none,
then
I
guess
we
are
headed
towards
a
short
meeting
today,
I
yeah.
If
there
are
none,
then
I
will
move
on
to
the
15
minute
back
triage
session.
C
Causing
build
failures,
she
seems
to
be
coming
from
the
country
repo,
so
I
will
assign
this
P2,
or
is
that
a
P1,
even
I,
guess
I
guess
we
can
bump
that
to
P1
it's.
C
B
C
Is
fixed
all
right,
so
that
is
definitely
a
P1.
Then
so
is
anyone
here
who
is
familiar
with
the
grpc
instrumentation?
Who
would
like
to
pick
that
up?
Otherwise,
I
could
also
have
a
look
at
it,
but
I
am
I
have
to
admit
I'm
not
too
familiar
with
this
specific
package.
Here.
C
All
right,
then,
I
will
just
sign
myself
for
now
yeah.
If
anyone
else
has
some
free
cycles
and
wants
to
take
a
look
at
it
feel
free
to
do
so
all
right.
Moving
on
to
the
next
one.
C
C
There's
a
PR
open
for
that.
But
there
is
a
second
one,
also
open
here
and
I,
guess
that
one
is
going
a
bit
further
in
fixing
it
yeah.
So
there
there
seems
to
be
some.
C
More
more
issues
other
than
just
using
a
window
in
this
event
listener
that's
causing
troubles,
so
this
BR
I
think
doing
some
more
I
will
put
that
as
a
P2
as
it
is
causing
Telemetry
not
to
be
exported,
but
I
don't
think
it's
causing
any
application
crashes
which
would
warrant
a
P1.
So
we'll
put
it
at
P2.
C
All
right
moving
on.
C
Where
start
server
is
not
not
working,
as
expected,
I
think
we
had
a
similar,
similar
issue
to
this
one.
In
the
past.
C
C
I
can't
seem
to
find
it.
I
will
leave
it
at
that
and
then
follow
up
after
the
meeting
to
see
if
this
is
a
duplicate.
But
as
far
as
I
recall,
there
was
some
problem
that,
if
the
part
is
already
in
use
the
end
user
application
would
crash
and
I'm
not
sure
if
that
has
been
addressed
yet
or
not
so
we'll
follow
up
on
that
issue.
After
the
meeting
right
and
the
next
one
seems
to
be
related
to
AWS
Lambda.
C
Yeah
anybody
on
the
car
who
knows
what
this
is
about
or
has
some
some
ideas
about
this
I
have
to
admit
I'm
not
too
familiar
with
that.
So
can't
really
can't
really
say
anything
from
the
top
of
my
head.
C
And
I
will
mark
this
up
for
grabs
in
case
someone
is
interested
in
it,
picking
it
up.
C
All
right
and
moving
on
to
this
right
here,
I
see
we
have
quite
a
few
old
ones,
but
nothing
new
in
the
country,
people,
so
that
is
it
for
the
triage
session.
I
guess
yeah
any
more
topics
to
talk
about
from
your
side
or,
if
not,
then
I
guess
we
can
call
it
a
meeting
and
I
will
give
everybody
45
minutes
of
the
time
back.
C
All
right,
if
there's
no
more
topics,
then
wish
you
all
a
great
day
and
see
you
next
week.