►
From YouTube: 2020-05-27 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
A
B
B
B
B
B
B
B
B
C
C
C
B
B
B
So
for
those
who
don't
know
me,
I'm
Morgan
MacLean
at
Google,
and
we
have
a
number
of
other
prolific,
open,
telemetry
contributors
who
will
be
presenting
different
topics
at
each
of
these
meetings,
including
today,
so
the
first
one
and
by
the
way,
there's
a
shared
agenda
on
the
calendar.
So
the
first
topic
we
have
is
that
open
Salama
tree
had
the
second
highest
number
of
contributions
for
any
CN
CF
project
over
the
past
90
days
on
just
behind.
Well,
the
only
one
head
of
us
being
kubernetes,
which
of
course
is
a
very
large
project.
B
So
this
is
definitely
worth
celebrating.
I
think
it
was
URI
or
Liz
who
had
brought
this
up
on
at
one
of
the
governance
meetings
and
although
I
know
that
the
community
had
been
growing
rapidly,
I
was
really
impressed
just
to
see
the
number
of
committers
and
the
number
of
commits
that
we
had
had
vaulted
us
over
some
other,
like
really
substantial
CNCs
projects.
So
it's
definitely
worth
celebrating.
There's
a
link
in
the
meeting
notes
to
the
CN
CF
dashboard.
B
C
So
we
started
a
meeting
weekly
meeting
every
Monday
for
maintainer.
Of
course
it's
an
open
meeting,
so
everyone
can
come,
but
this
was
to
address
the
fact
that
we
have
this
monthly
community
meeting
and
we
have
a
spec
meeting.
But
we
didn't
have
a
meeting
where
the
cig
maintainer
could
get
together
and
talk
about
sort
of
non
spec
related
cross
across
language
issues
like
testing
and
CI
and
releasing
and
theta
stuff.
D
B
B
B
B
Obviously,
logging
is
a
new
date
of
touch
for
open
salami
tree.
It's
and
to
be
clear,
like
we've,
the
focus
for
the
project
has
historically
been
traces
and
metrics
and
those
are
what
we're
still
going
to
GA
with
and
we're
not
going
to
slow
down
that
plan.
So
logging
will
be
on
a
different
cadence
in
a
different
schedule,
but
you
know
some
points
at
some
point.
This
year
logging
will
be
added
and
who
go
through
alpha
beta
and
then
GA
status
according
to
its
own
schedule.
B
There's
a
whole
work
stream
around
it,
and
a
number
of
new
contributors
have
joined,
including
people
from
Amazon
and
elastic
who
companies
that
have
not
usually
at
least
historically
being
major
participants
in
the
community.
So
it's
really
exciting
to
have
something
and
I
don't
know
if
Tigran
I
think
you're
on
the
call
or
I
don't
know
if
Wesley's
on
the
call,
if
either
he
wanted
to
just
say
anything
about
logging.
B
E
That
I
thought
was
kind
of
cool,
seeing
lots
more
contributions
now
from
the
community
as
a
whole
excited
to
announce
that
the
light
Steph
and
honeycomb
exporters
were
merged
into
contributing
I,
also
see
a
New
Relic
PR
up
Alibaba
cloud
PR
up
and
spun
clog
PR
up
so
generally
seeing
lots
more
people
get
involved
here,
which
is
pretty
cool.
Also,
the
host
based
metrics
has
been
merged
into
core,
so
there's
no
basic
support
for
CPU
memory.
E
Disk
network
also
super
exciting
to
see
that
getting
rich
support
just
like
the
tracing
side
has,
and
there
are
plans
now
to
kind
of
get
the
missing
processors,
so
the
processors
today
primarily
only
work
for
tracing
the
goal,
is
to
get
that
added
for
all
the
metrics
I'd
hear
over
the
next
month
or
two
so
overall
really
appreciate
everyone's
contributions
to
date
and
just
apologies
to
folks.
If
your
PR
took
a
little
bit
longer
to
get
reviewed,
but
we
are
trying
to
get
through
that
backlog
pretty
quickly.
Yeah.
B
And
I'll
add
on
to
that
there's
a
lot
of
work
going
on
for
capturing
like
sort
of
pre
packaged
or
third-party
application,
metrics
on
Linux
and
Windows
and
I
know
like
I.
Think
Splunk
has
made
it
clear
over
the
past
several
months
that
they
plan
to
use
this
to
eventually
replace
the
signal
FX
metrics
agents.
The
same
is
true
for
us
at
Google
will
be
using
this.
The
OT
collector,
starting
with
Windows,
to
capture
both
the
host
metrics
as
well
as
metrics
from
prepackaged
applications
for
all
of
the
MGP,
are.
B
A
E
Guess:
question
on
the
packaging
things
for
folks
in
the
calling
in
case
there's
any
input
right
today,
there's
just
there's
a
kubernetes
config,
there's
a
docker
container
and
then
you
have
the
binaries,
we're
looking
to
add
more
packaging
options,
so
helm
open
shift,
rpm
Deb.
If
people
have
preferences
here
or
strong
opinions
on
what
the
missing
things
are
they'd
like
to
see
all
that
would
be
great
feedback,
as
we
can
try
to
prioritize
and,
of
course,
be
are
as
welcome.
So
would
love
people
to
help
help
out
with
that
as
well.
C
E
Have
both
you
can
get
just
the
core
or
you
can
get
the
core
and
contributed
ooh
kind
of
one-off
so
selectively?
We
don't
have
that
yet
you
could
do
it
yourself,
but
typically,
it's
kind
of
an
all-or-nothing
four
core
contributor,
a
shin,
so
by
default
kind
of
everything
is
turned
off.
You
can
then
select
which
things
you
want
to
enable
there's.
B
B
F
E
Another
different
topic
so,
on
the
instrumentation
side,
I
know
Trask
is
on
and
Ted
Pryor.
Some
input
here.
I'm
here
is
status
of
auto
instrumentation
across
the
board.
I
know,
Java
is
pretty
advanced,
it
looks
like
Python
is
pretty
advanced
from
what
I
can
tell
the
other
ones.
I
don't
see
my
own,
maybe
node
has
some
I
haven't
played
with
it
yet?
Can
someone
provide
a
quick
update
on
that
on
thoughts
as
well
as
gaps?
I
know,
that's
definitely
a
bit
behind
on
the
auto
instrumentation,
but
there's
kind
of
a
plan.
C
Yeah
I
I
think
you're,
correct,
so
isn't.
Java
has
a
Java
agent,
that's
pretty
great.
The
latest
version
of
it
had
something
like
a
4x
speed
boost.
We
noticed
over
here.
So
the
point
3
release
is
going
strong
in
Python
and
nodejs.
The
way
the
agents
work.
I
believe
you
still
currently
have
to
put
a
bit.
You
still
have
to
paste
a
bit
of
code
into
your
application
to
start
I
believe
in
Python
they're,
adding
a
wrappers
that
you
won't
have
to
do
that
in
the
future.
C
They
will
only
run
modules
that
you
have
downloaded,
and
so
we
are
looking
at
a
sort
of
distro
model
for
those
languages
kind
of
like
can
trim
where
you
have
like
an
install
all
command
or
a
package
that
just
has
a
reference
to
everything
that
seems
reasonable,
but
there's
this
sort
of
question
of
what
what
does
all
mean
that
kind
of
gets
back
to
this
issue
of
like?
Where
does
all
this
instrumentation
live
who's
on
the
hook
for
maintaining
it
long
term,
so
that
that's
just
an
open
question
in
these
different
communities?
C
We're
mostly
moving
to
the
model
that
you
have
the
core
repo
and
then
you
have
a
contribute,
oh,
where
all
this
extra
stuff
lives
and
because
it's
a
distro
model,
someone
could
make
other
distros
if
they
wanted
to
include
things
that
we
don't
maintain
as
well.
So
that's
how
it's
going
for
python
in
nodejs
and
ruby,
it's
not
fully
implemented
in
ruby,
but
there's
a
spec
there.
That's
similar
to
Java
JavaScript
in
and
Python
for
the
other
languages.
I'm,
not
sure
I,
know,
there's
been
a
big
switch
in
direction
on
dotnet.
E
Yeah
I
know
there's
some
back
and
forth
on
that
we've
been
plugged
in
a
little
bit
and
the
Python
one
I
think
they
finally
removed
the
dependency.
So
no
code
changes
now,
it's
just
a
pip
install
and
it
works.
I
tried
out
at
least
a
few
other
packages.
It
appears
to
work
out
of
the
box,
which
is
pretty
sweet.
Great.