►
From YouTube: 2023-02-20 meeting
Description
Open Telemetry Meeting 1's Personal Meeting Room
B
Yeah
we'll
see.
C
A
Yeah
I
think
there
was
a
ongoing
discussion
with
something
that
should
I
see
mentioned.
A
A
Yes
and
then
I
think.
A
And
yeah
I
I,
I
I,
also
agree
with
you.
I
think
we
should
have
like
maybe
even
open
source
I
mean
in
the
same
way
that
we
have
like
vendors,
Forks
yeah
of
Open
Source
box,
where
the
community
added
stuff.
You
know
yeah.
B
We
could
call
them
Community,
develop
the
back
end
force
or
something
like
that.
You
know
and
just
lists
them
all
in
a
table
or
or
list
or
whatever
it
is
in
GitHub,
and
let
those
people
maintain
their
own
Forks.
B
True
true,
but
yeah,
that's
you
know,
and
we
can
certainly
provide
additional
support
up
front,
at
least
but
I
think
most
notably,
if
you
just
link
to
them
so
they're
they're
discoverable,
that
solves
a
lot
and
if
you
know-
and
maybe
one
day
this
all
makes
it
to
the
docs
and
then
it's
just
it's
in
the
docks.
Instead
of
a
readme
I,
don't
know.
B
That
that's
my
opinion
on
the
various
backends
and
I've
resisted
people
like
Hey,
we're
gonna,
add
Tempo
and
I'm
like
we
have
Jager.
What
are
we
doing.
B
So,
okay.
C
B
All
right,
we
should
I
should
slide
in
nobody
actually
updated.
This
thing
to
add
a
new
section
here
hold
on.
A
C
A
I
I
see
a
like
every
day,
a
new,
a
new
book
or
a
new
request
or
a
new,
whatever
issue
being
created
and
I'm
like
oh,
come
on
Apple.
Why.
B
I
put
in
a
request
for
an
M2
with
my
company,
I
I,
initially
asked
for
64
gigs
of
RAM.
That
machine
will
be
several
weeks
away
from
availability,
so
I
settled
on
just
give
me
a
32
gigabrand
box
instead
and
then
we'll
go
from
there
and
they
put
the
order
through.
But
then,
when
I
looked
at
the
order
form,
it
came
back
as
an
M1.
I
was
like
stop.
No
I
asked
for
them
too.
So
I'm
not
sure
what
the
status
of
it
is.
B
It'll
have
at
least
32
gigs
of
RAM
it'll
be
decked
out
and
I
will
spend
much
more
effort
in
building
armed
stuff
cleaning
up
that
hole
like
I'll,
run
kind
on
my
local
machine
and
just
figure
it
out.
What
are
our
challenges
and
get
over
them?.
A
Yeah
I
know
that
I
know
that
Mikko
has
aimed
one
and
he
runs
the
demo
locally,
like
I,
think
in
every
PR.
He
he
tries
to
build
and
everything
and
runs
everything,
yeah
I'm,
just
not
sure
about
kubernetes
deployments,
because
with
Docker
you
can
simply
Docker
compose,
build
no
cash
and
that
rebuilds
the
image
for
you
in
your
architecture,
but
with.
B
Kubernetes,
that's
not
the
case
yeah,
so
we
will
need
to
adjust,
extract
and
I.
Think
it's
that
we
said
when
the
dependents
were
going
to
do
for
the
next
release.
It
was.
C
A
A
I'm
not
sure
because
but
for
me,
I
just
friend
the
the
docker
one
and
to
use
on
kubernetes
I
I
just
deployed
on
a
ETS
cluster.
So.
B
B
Should
work
with
16.
yeah
I
personally
use
it
like
twice
my
entire
life,
where
you
just
like,
run
some
binary
and
everything
works,
and
then
you
get
a
cube,
cuddle
interface
to
it
as
far
as
I'm
concerned.
As
long
as
I
got
two
cuddle
I,
don't
care
I'll
come
down
to
this
as
long
as
I
can
run
K9
I.
B
So
yeah
we.
C
A
I
mean
if
we,
if
we
give
people
a
way
of
building
the
the
containers
locally
and
just
running
it
in
a
cabernetes
cluster
like
a
local
kubernetes
cluster,
then
that
would
be
a
nice
well
easy
work
around
at
the
moment.
I
only
see
an
easy
way
on
that
using
mini
Cube,
because
then
you
can
point
to
the
to
the
local
Docker
registry.
A
B
Yeah
but
then
you
have
people
who
run
eks
on
arm
stuff
like
that
or
they
have
I,
don't
know,
they've
got
an
arm
machine
somewhere
that
they
just
wanna
and
it's
not
so
much
kind
anymore.
It's
just
I've
got
my
coupons
cluster
running
an
arm.
I
want
to
just
Helm
install
this
thing
and
it
works,
and
that
really
should
be
the
experience
we
should
be
getting
right.
You.
C
B
To
do
anything
else
right
now,
because
we
don't
publish
arm
images,
it
doesn't
work.
If
we
publish
arm
images
with
the
proper
manifest
for
them,
then
you
could
just
do
a
poll
like
film
installed.
It
should
just
work.
B
So
this
is
actually
part
of
our
release
process.
We
need
to
fix
it.
C
B
D
Know
it's
a
series
of
pipes,
yeah
yeah,
I
haven't
I
had
an
unrelated
question.
D
A
Do
we
have
a
charge
to
help
charts
concrete
foreign.
D
B
By
the
way,
if
the
example
value
files
are
non-bender
specific,
we
need
way
more
CI
for
the
helm
chart.
So
if.
C
B
D
B
Specific,
you
know
you
have
a
folder
with
your
various
values.
Examples
right
there
yeah
as
for
maintaining
a
fork
of
it,
I
I,
don't
think,
there's
a
clean
way
to
do
it
unless
you
grab
the
whole
repo
and
then
just
delete
the
other
charts
yeah.
B
B
It's
weird
yeah,
so
it's
like
a
process
that
runs
and
it
creates
an
index
file
under
GH
Pages
attaches
and
manifests
as
a
GitHub
artifact
and
it's
cool
how
it
works.
But
it's
just
there's
two
branches
for
every
GitHub,
repo
of
a
a
hub
chart,
yep
or
a
Helm
repo
I
should
say
yeah.
Sorry,
I,
don't
have
a
great
answer
for
you
there
to
be
honest,.
D
No,
that
was
helpful.
Thank
you.
I
think
I'll
for
now,
I'll
put
the
example
values
and
then,
as
far
as
adding
Services
it'll,
probably
just
be
a
little
bit
of
a
lift.
Every
time
the
the
home
chart
gets
updated.
B
Yeah,
technically,
you
could
add
additional
services.
B
B
No,
no,
no,
the
schema
does
not
specified
component
names,
it.
C
B
B
D
B
D
B
B
So
we
have
components
and
then
components
contains
component
right
and
the
way
this
schema
says
is
here's
your
properties
and
here's
a
name
and
it's
a
type
component.
I
wonder
if
I
could
just
say
any
property
underneath
here
moment:
I
wonder:
if
there's
a
way
to
do
that,
and
if
there
is
then
we
would
be
able
to
remove
the
strong
type
nature
of
the
component
names
right.
B
Yeah,
at
the
same
time,
though,
it
checks
people
who
try
to
do
like,
maybe
they
screwed
up
cakes
right
like
right,
Accounting
Service
all
lower
case,
the
the
schema
will
pick
it
up.
B
Screwed,
this
up
I
think
you
meant,
you
know
something
different.
B
That
might
be
better
and
then
it's
an
array
of
component
that
fits
underneath.
It.
C
B
B
Yeah
make
an
issue
for
Ability
for
vendors
to
add
components.
Yeah
to
this,
somebody
else
asked
about
adding
sidecar
containers
as
well
and.
B
B
Okay,
anything
else
that
is
bothersome.
B
B
B
Have
we
even
changed
the
Proto
namespace
that
we
could
just
say
since
everything
under
open
Telemetry
demo
and
then
have
the
linter
check
for
licenses
I?
Think
it's
more
of
just
a
chore
and
then
somebody's
got
to
write
a
letter
or
I'm
sure
linters
exists,
you
just
gotta,
add
it
to
to
our
PR
checks,
yeah.
B
B
This
is
Austin's
and
he's
not
here.
He
wants
to
chat
about
this,
so
I
will
leave
it
alone,
and
the
last
one
was
resource
metadata
here.
These
were
like
things
to
look
into
container
information
for
dental
apps.
B
B
Should
we
I
think?
Yes,
we
should
because
it
really
helps
complete
the
picture
yep,
but
we
also
have
two
different
deployment
models
and
collecting
metrics
for
each
of
those
models
is
vastly
different
one
to
the
other,
and
your
dashboards
are
going
to
be
different
because
your
metric
names
are
different.
B
Maybe
we
could
do
one
where
you
know
the
docker
one
deploys
a
dashboard
that
has
only
the
docker
metrics
and
the
kubernetes
one
of
the
plays.
The
dashboard
only
has
it
for
me
this
map
or
the
the
kubernetes
metrics,
but
that
does
you
know
how
do
we
go
about
doing
that
and
then
yeah
doing
it
right
and
doing
it
in
a
way
so
that
we
use
open
Telemetry
components
as
well
like
can
a
collector
collect,
Docker
stats
about
itself
and
anybody
else
in
this
compose.
D
C
D
To
me
to
me
this,
like,
maybe
this
is
just
my
personal
philosophy
and
how
I
use
the
demo,
but
I
treat
the
docker
as
like
for
use
When
developing
an
individual
application
and
if
I'm
going
to
run
the
thing
in
a
production
right
away,
it's
on
kubernetes,
so
I
kind
of
conceptualized
it
as
like.
There's
the
batteries
included
version
with
kubernetes
and
there's
like
the
you,
can
work
on
the
apps
and
it
will
run
inversion
with
docker.
D
We're
not
going
to
be
like
is
that
gonna
have.
Is
that
gonna
be
different
depending
on
how
people
like
people
have
very
different
Docker
setups,
because
that's
still
going
to
work.
C
D
B
I
don't
know
I'm
sorry,
okay,
I
mean
I.
Guess
we
just
have
to
try
to
find
out
yeah,
but
I
I
do
see
one
here
and
it
seems
like
you
just
point
to
a
Docker
socket.
Okay,
that's
easy
enough
and
it
goes
and
Scoops
up
all
your
stuff.
There's
a
way
to
exclude
images,
but
not
too
include
images
which
I
think
is
what
we'd
want
to
do,
because
I
don't
want
to
collect
these
guys
with
all
your
Docker
stuff.
Just
about
the
thing
that
we
care
about.
B
And
then
for
kubernetes
there
is
in
there's
a
cubeless
stats.
Receiver
I
think
it's
called,
but
that
I
don't
think
it
could
restrict
it
to
namespace,
which
would
stuck
if
we
can't.
B
B
It's
I'm
looking
at
it
doesn't
look
like
a
public
holiday
to
me.
It's
not
to
me.
B
A
We
moved
the
docs
to
to
the
open
telemetry.io
page.
Is
there
a
way
we
can
kind
of
automatically
create
an
issue
whenever
something
is
changed
on
the
demo
or
kind
of
I?
Don't
know
force
people
to
update
the
the
dogs?
A
Well,
first
people,
it's
it's
a
strong
word,
but
advise
people
to
update
the
docs
as
soon
as
they
update
the
demo,
because
before
we
had
that
in
one
PR
and
now,
depending
on
how
many
PRS
we
have
that
may
get
out
of
control
and
we
may
have
an
updated
docs
in
the
open
television
I.
C
B
If
we
add
a
label
to
a
thing,
need
doc,
updated.
For
example,
part
of
the
merge
process
that
we
have
in
GitHub
actions
creates
an
issue
on
the
docs
repo
linking
back
to
our
PR
and
titles.
It
some
way
that
works
with
the
docs
team
and
then
the
science
to
the
demo
team
as
well.
So
we
know
to
jump
on
it.
B
I
feel
like
I've,
seen
this
happen
elsewhere
before
I,
can't
remember
where
I
seen
stuff
like
this
before,
but
I
yeah,
it's
I
think
that
would
be
a
good
approach
we
could
take
now.
We
would
need
permissions
from
the
demo
team
to
do
all
this
stuff
and
I'm
sure,
there's
API
keys
and
weirdness
involved
and
all
that
and
who
owns
like
it'll,
be
some
kind
of
weird
bot.
B
That's
making
the
the
Issue
request,
I
guess
but
I
I
think
we
need
to
do
something
like
that
yeah
for
sure
otherwise
than
that
we're
gonna
get
ourselves
into
a
pickle
I'd
like
to
have
the
same
thing
by
the
way
for
Help
Church
anytime.
We
make
an
update
that
says
hum
whatever
required.
We
posted
an
issue
in
the
helm,
repo
as
well.
A
A
B
B
Pr
because
I
remember
seeing
one
fly
across
like
oh
crap:
that's
not
set
for
help.
Maybe
I
went
back
through
and
I
added
to
help
label
to
it
afterwards,
but
it
would
be
good
if
we
were
just
a
little
bit
more
proactive
that
things
that
change
anything
with
the
deployment
we
have
to
make
sure
a
home
gets
caught.
C
B
If
it's
any
change
to
Docker
file
or
to
a
resource
which
is
included
at
runtime
like
like
a
dashboard,
Jason
or
collector
config,
we
have
to
update
the
helm
chart
so
changes
to
environment,
Docker,
compose
and
any
resource.
That's
loaded
at
runtime.
B
C
B
Thank
you,
okay,
I'm,
going
to
finish
this
Issue
request
here
for
this.
But
yes
definitely
that
sounds
like
we
should
tackle.
C
A
Yeah
I
have
a
Meetup
tomorrow
and
the
co-organizer
from
the
cncf
Meetup
here
in
Maine's
with
me.
He
has
GitHub
actions.
Experience
oh
I'll
ask
him.
Maybe
he
knows
something
and
can
help
us
out.
B
Foreign,
if
he
does
know
something
of
maybe
another
repo
that
does
this
that'd
be
great
I,
think
it's
going
to
the
governance
committee,
I'm,
sure
and
and
working
at
that
level
can
make
it
work.
But
this
makes
a
lot
of
sense
given
how
dependent
we
are
on
other
things,.
B
A
A
Think
the
problem
with
the
penderbot
is
that
he
it
doesn't
have
the
it
doesn't,
has
the
what
is
the
requirement
that
we
need
like
the
agreement,
whatever
oh.
B
B
I
think
we
have
two
issues
here:
one
dependent
because
we
we
had
dependencies
that
are
external,
the
other
one
is
just
SDK
dependencies
like
when
an
SDK
releases
I
would
love
for
an
issue
to
show
up
inside
of
our
GitHub.
That
said,
hey
no
jssdk
released
update
whatever,
and
then
we
go
pick
that
up
and
do
it
it'll
make
our
issue.
It's
probably
really
noisy,
but
I,
don't
know,
it'd
be
let.
A
B
A
It
would
also
make
the
the
the
demo
more
up
to
date,
yeah
I,
like
the
approach,
even
though
it's
noisy
I,
like
the
approach,
because,
like
at
the
moment,
we
just
know
whenever
we
use
it,
and
sometimes
whoever
is
just
playing
around
with
the
demo.
They
do
not
pay
attention
to
to
like
versions
and
if
it
works,
people
just
enjoy
it.
B
Okay,
something
we
didn't
think
about
new.
We
could
chat
about
it.
B
In
two
weeks,
or
let's
find
out
what
you
can
learn
and
we
could
probably
even
read
the
slack
but
I
think
this
makes
sense
so
I
know
create
an
issue
as
well.
For
this.
C
B
Right
guys
enjoy
your
week
and.
B
D
So
we'll
see
I
mean
I,
don't
know
if
if
one
of
my
talks
gets
accepted
for
observability
day,
then
well.
B
B
B
Okay,
I,
don't
want
a
dog
in
any
vendor,
I,
don't
care
what
your
science
is
on.
Whatever
you
want
to
do.
Okay,
that's
fine
all
right,
but
it
will
be
about
open,
Telemetry.
I
know
we
are
in
talks
of
figuring
out
how
to
perhaps
do
a
sponsorship
ourselves
at
honeycomb,
maybe
even
joining
forces
with
other
vendors
to
to
pull
our
cash
together
and
share
a
table
even
I.
B
Don't
know
you
know,
belt
tightening
blah
blah
blah
those
kinds
of
things,
but
we're
definitely
on
board
and
there's
a
slack
Channel
dedicated
in
our
slack
about
what
we're
gonna
do
for
open
observability
today,
awesome
so
nice
yeah,
so
well
we're
going
to
have
a
presence
for
that
at
least
I.
Think
I'm
going
like
I
said:
I
just
don't
know
how
money
honeycombers
are
going
to
be
there
besides,
myself,
so
I
guess:
cool.