►
From YouTube: 2022-02-28 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).
B
B
C
The
last
week
we
had
this
kind
of
demo
and
presentation
from
cisco
apache
http
plugin,
so
I
mean
I
remember,
I
think
really
has
told
me
that
la
I
mean
really
told
me.
C
I
think
three
four
weeks
back,
that
cisco
is
want
to
come
up
with
some
some
contribution
and
they
want
to
probably
contribute
to
open
telemetry,
but
I
was
not
aware
of
what
exactly
it
is,
but
that
was
a
surprise
that
for
something
from
apache,
which
we
already
have
it
and
I'm
not
very
sure
I
mean
the
difference
between
what
they
want
to
contribute
and
what
is
already
there
is
as
far
as
something
something
which
is
already
the
current
apache
module,
which
is
there
in
the
content.
Repo
is
only
for
mod
proxy.
C
C
So
I
think
if
that
is
the
case,
then
probably
this
the
the
contrib
the
model
from
cisco
is
something
which
can
be
used,
irrespective
of
whether
we
have
any
proxy
in
between
the
server
and
the
client
so,
but
that
that's
my
understanding
and
probably
told
them
to
discuss
more
with
the
current
owner
of
the
model
which
we
have
and
somehow
he
is
not
reachable.
B
And
and
their
change
is,
gonna
have
performance
impact
right.
C
So
that's
that's
something
that
was
the
reason
why
it
was
better
if
they
can
discuss
whether,
whether
so,
whether
both
should
exist
simultaneously
or
or
instead
of
having
both
the
different
modules.
If
we
can
modify
the
existing
model,
which
we
have
to
support,
I
mean
that's
something:
probably
they
have
to
discuss
it,
because
it's
not
it's
not
that
they
want
to
propose
some
case.
They
already
have
done
their
own
implementation,
different
from
what
is
already
existing
and
they
want-
and
I
think
it's
something
we
have
to
see
what
should
be
the
next.
C
It
should
be
a
replacement
or
which
can
exist.
You
know,
probably
they
have
to
discuss
and
come
up
with
a
strategy.
What
to
do
in
that
case,
and
I
mean,
as
the
maintainers
and
approvers
of
the
main
repo
I
think
I
am
our
take-
should
be.
We
can
facilitate
the
environment,
but
we
cannot
take
a
decision
here.
I
mean
which
model
should
exist
or
it
should
be
existing.
Probably
they
have
to
discuss
and
come
up
with
a
certain
thing
so
that
that
was
my
thinking
but
probably
feel
free
to.
C
C
I'll
talk
with
rally
also
really
probably
better
to
talk
with
him.
What
is
his
take
on
this?
He
will
also
probably
I
think
he
will
also
say
that
probably
it
should
be
best
discussed
among
both
the
teams
and
let
them
come
up
with
a
strategy.
We
would
probably
good
to
have
a
discussion
with
you
and
yeah.
That's
for
this
this,
depending
what
dependent
battery
so
I
mean,
I
think
I'll,
probably
keep
it
to
you.
I
think
you
want
to
see
how
you
want
to
do
for
work
for
this
and
yeah.
C
That's
apart
from
that
concurrent
is
total
http
exporter
yeah,
that's
a
pr
which
owen
has
raised
and
if
we
can
make
it
reusable-
and
I
think
so
in
the
current
form,
I
think
it's
something
which
we
cannot
take
it.
So
probably
it
has
to
be
modified
and
that's
that's
something
probably
we
have
to
do.
How
can
how
can
we
make
it
work?
Probably
if
you
want,
is
going
to
work,
do
the
changes
on
this
pr,
it's
well
and
good.
C
C
C
C
We
mentioned
that
which
would
be
a
tricky
how
to
how?
How
do
we?
How
do
we
do
that
without
affecting
any
existing
exporters,
whether
part
of
open,
telemetry,
main
repo
or
if
somebody
has
developed
any
other
exporter
externally,
it
should
not
break
with
any
change
which
we
are
doing
in
processor.
So,
if
we,
if
we
add
a
new
interfacing
processor,
which
makes
asynchronous
call
to
exporter
export
method,
how
how
should
we
support
it?
C
And
yeah,
that's,
let's,
and
how
to
implement
also,
it
would
be
something
we
probably
have
to
think
would
be,
whether
it's
a
call
back
or
whether
we
should
use
async
await
pattern.
The
problem
with
the
async
await
pattern
is,
as
I've
been
seeing,
that
it's
not
very
mature
still.
C
Yeah,
that's
that's
a
a
sync
await
pattern,
which
c
plus
plus
has
so
normally
dotnet
and
javascript.
Most
of
the
languages
have
this
async
of
it,
and
this
one
so
javascript
has
c
sharp
has
so
which
so
this
was
something
which
was
missing.
So
we
don't.
C
I
mean
if
we
have
to
have
multiple,
multiple
tasks
where
we
want
to
run
in
a
given
application
either
we
can
spoil
multiple
sets
or
we
can
use
a
asic
await
pattern,
so
this
is
added
in
c
plus
11
using
features,
but
it's
as
I
have
been
reading
from
the
implementation
provided
to
multiple
compilers
or
platform.
It
looks
like
it's
not
still
very
stable.
C
We
do
have
one
more
pr
which
was
released
long
back,
support,
async
networking.
I
think
this
is
something
if
we
have
this
completed.
Probably
this
is
something
which
you
can
use
it.
I'm
not
haven't
seen
this
pr
completely
with
this.
This
is
this.
Is
this,
this
is
in
another
implementation
of
the
async
of
it
using
the
lib
event,
I
think
it
uses
the
I
o
polling
device.
C
I
do
see
it
definitely
usable
for
meters,
we'll
be
needing
it
for
sure
in
traces,
although
we
can
use
it
as
a
communication
pattern
between
exporter
and
processor,
but
that's
something
probably
we
can
see
if
either,
if
you
can
use
this
but
yeah,
just
just
some
thoughts
which
and
as
of
now,
I'm
getting.
C
C
C
Yeah
and
that's
we
can
we
can
discuss
on
that
I'll,
open
and
open
up.
If
you
want,
I
can
open
one
task
or
somebody
can
open
that
task
to
explore
more
on
changing
the
interface
between
exporter
and
processor.
If
there's
something
we
can
do
it
and
what
patterns
we
can
use
for
that,
I
think
that
should
be
a
good
start
if
we
do
want
to
make
something
asynchronous,
which
we
can,
after
that,
we
can
use
that
in
metrics.
Also.
C
Unfortunately,
I'm
sorry
sorry,
I
think
I
think
I
was
supposed
to
have
a
discussion
with
you
somehow
somehow
I
was
involved
last
couple
of
days
of
last
week
I
mean
it
was
on
some
other
tasks:
non-open
telemetry
work,
but
let's
let's
discuss,
I
think,
let's
have
it
talk.
C
Maybe
tomorrow
around
I
mean
whenever
I
mean
I
don't
want
to
give
a
date
whenever
you
feel
I
think
tomorrow
or
any
other
time
when
you
think
and
let's
have
a
discussion,
I
have
some
idea
now,
probably
because
we
can
have
we
can
have
you
can
figure
it
out.
I
mean
how
to
divide
the
work.
C
B
C
Data
point
yeah,
I
think
I
can.
I
can
probably
talk
about
that
so
delta
metrics,
I
mean,
is
not
four.
Therefore,
four
different
scenarios
we
can
have,
we
can
have
a
synchronous
aggregation
configured
and
we
have
a
delta
matrix
collection
for
synchronous
matrix.
So,
basically,
when
we
create
a
reader,
we
specify
whether
we
want
to
somewhere.
I
mean
I
think
I
can
show
you
quickly.
I
won't
take
much
time
here.
We
can
discuss
afterwards,
but
whenever
you
create.
C
Yeah,
so
we
specify
whenever
we,
whenever
we
create
a
reader,
a
collector
collectively
specify
that
what
kind
of
I
mean,
what
what
kind
of
data
matrix
we
want
to
read
whether
we
want
to
read
only
the
data.
The
data
means
that
only
the
measurements
which
have
been
created
after
the
after
the
last
collection,
which
we
did
but
cumulative,
would
be.
We
want
a
more
of
aggregation.
Cumulative
aggregation
of
all
the
metrics
which
have
been
created
to
know
so
cumulative
is
bit
complicated.
C
The
current
implementation,
which
I
did
for
synchronous
and
also
I
raised
a
pr
for
asynchronous
that
is
somewhere
handling
the
data
it
just
that
creates
the
matrix
and
it
will
clear
up
aft
after
doing
a
fetch,
so
that
we'll
get
a
fresh
matrix
after
every
collection.
C
C
So
I
just
created
these
four
tasks
for
that,
and
then
there
would
be
one
task
which
we
already
have
it:
probably
this
matrix
reader,
an
exporter
that
would
be
more
of
a
collection.
So
after
this,
I
think
this
will
create
the
complete
pipeline
for
the
collecting
the
matrix.
C
C
C
C
C
Yeah,
okay,
so
I
think
metric
implementation.
I
I'll
I
raised
one
pr
for
asynchronous
storage.
I
think
I
saw
I
think
sony
had
given
some
comments.
I
didn't
I
didn't
go
through
that,
but
let
me
go
through
that
and
probably
I'll
dissolve
that
and
yeah,
and
apart
from
that
tom,
I
think
I
don't
see
any.
I
mean
this.
Probably
you
can
take
make
a
call
on
this.
If
you
feel
whether
you.
C
C
B
One
more
thing:
that's
a
bit
unrelated
I'm
using
this.
I
don't
know
how
to
call
it.
I'm
using
git
part
for
development.
B
B
C
B
And
and
you
can
compile
open
telemetry
with
it
and
you
do
your
development
and
the
machines
are
really
fast
like.
C
B
C
B
C
C
C
C
C
B
C
C
Anyway,
good
to
have
any
one
of
the
rpm
environment,
if
you
can
have
some
somewhere
in
our
test,
I'm
not
sure
how
much
changes
it
could
be
required
anyway.
I'll
create
one
task
for
that,
and
let
somebody
want
to
take
it
need
not
be
as
if
anybody
want
to
contribute
any
anyone
who
is
using
our
rpm
based
environment
for
open
telemetry
or
their
application.
They
can
do
that.
B
C
C
C
C
C
B
C
C
C
The
problem
is
both
with
noise,
led
span
and
sd
variant.
So
some
part,
all
this
part
is
coming
from
from,
I
think
std
implementation
from
gnu,
which
has
a
license
for
gp
and
version
3,
with
a
specific
condition
that
it's
it's
okay
to
build
it
that
it
can
be
used
in
the
binary
form.
Also,
so
not
that
we
are
not.
We
are
not
using
the
correct
license
file.
We
are
using
the
correct
license
file,
but
how
should
we
put
it
here?
C
C
C
C
An
issue
that
it
was
always
getting
static,
initialization
of
satellite,
which
was
not
consistent.
It
was
always
creating
a
new
technical
storage.
Every
time
we
are
calling
that
method,
but
okay,
this
is
this
is
the
build
which
experience
it's
crashing
into,
not
not
the
oh,
it's
not
even
fairy.
It's
crashing.
Sorry.
The
build
is
successful.
C
C
Tom,
can
you
just
look
into
that?
If
it's,
I
think
the
srs
resolve
the
comment
which
you
had.
Can
you
just
go
through
that
if
you
feel
it's
I'll
also
go
through
this
today,.
C
C
That's
all
anything
else,
you
want
to
add.
C
A
bit
of
context
here,
so
really
you
really
one
of
the
maintainer
here
in
the
for
this
web
for
this
repository
open
telemetry.
If
you
see
here,
billy
tom
and
alex
he's
overseeing
the
complete
open,
telemetry
work,
which
is
happening
from
microsoft-
that
basically
both
dot
network
and
the
secret
space
work
he's
going
to
move
to
technical
committee
for
open
telemetry,
so
he
his
focus
is
going
anyway.
He
was
from
past
few
months.
He
was
not
very
much
active.
C
He
used
to
do
the
stuff,
but
it
is
not
very
active
and
he
is
going
to
move
to
technical
committee,
and
I
think
his
focus
is
going
to
change
quite
a
lot
after
that,
and
I
think
he
so
he
told
us
offline
that
I
think
he
wanted
you.
If
we
can
move
you
as
a
maintainer
when
he
will
move
out
by
mid
mid
of
this
month.
C
So
he
will
he.
He
won't
be
maintained
anymore,
and
probably
you
move.
I
think
he
proposed.
I
think,
he's
probably
seeing
your
work.
He
was
very,
very
happy
to
see
that
and
he
I
think
he
wanted
that,
let's
probably
remove
you
as
a
maintainer.
I
wanted
to
check
before
before
doing
that.
What
are
your
thoughts?
C
C
Won't
have
any
other
approvers,
so
did
I
I
I
think
tom
I
discussed
offline
with
owen
on
this
okay
and
I
think
he
he
is
happy
to
move
as
an
approver.
He
had
shown
interest
earlier.
Also,
probably
we
were
not
doing
that,
but
I
think
he
has
shown
the
interests
again.
C
That,
because
of
the
time
zone,
he
cannot
join
the
meetings,
but
that
is
fine
if
we
can
go
through
the
our
our
google
doc
for
all
the
agenda,
notes
which
we
are
taking
and
you
can
go
through
the
video
recording
for
this
meeting,
which
is
fine.
C
C
C
C
So
we
have
to
add
here
moment
in
as
part
of
us
move
out
as
I
am
from
approver
to
antennas.
C
A
C
A
If
move
out
of
remain
which
group,
which
role.
C
A
C
Somebody
he
will
be
still
like
when
he's
the
only
person
who
should
have
the
admin
responsibility.
He
is
the
person
who
create
all
the
somebody
who
creates
all
these
sick,
the
reports
and
he
handles
all
those
changes
related
to
that.
So
probably,
I
think
I
am
thinking
that
he
would
should
be
the
only
one
as
I
had
been
yeah.
A
C
So
that's
what
I
did
put
that
thing
that
we
do
have
to
do.
I
mean
on
and
off
ad
hoc
changes.
C
If
we
want
to
enable
any
of
the
github
actions
temporarily
or
something
we
have,
we
can
do
it
from
here,
so
we
do
need
someone,
someone
who
needs
to
have
it.
I
think
it
should
be
a
collective
decision.
If
it
has
to
be
done,
it
could
be
across
all
the
six,
not
specifically
because
he
was
blessed.
So
I
think
okay.
A
C
I
think
like
let's
be,
let's,
let's
be
as
of
now:
let's,
let's
be
there,
I
think
if,
once
once
a
decision
is
taken,
probably
we'll
see-
I
mean
I
don't
I
mean
I
know
until
now.
I
think
only
three
four
times
we
have
to
use
that
admin
role,
so
I
don't
think
it
would
be
required,
even
if
it
is
removed,
but
yeah
branches.
Also,
I
think,
moving
from
master
to
main
was
one
of
the
things
where
it
was
required.
C
Github
action
is
something
which
we
can
see
if
we
have
to
enable
and
disable
something
for
temporary
purpose.
That's
that
that's
the
reason
I
gave
that
we
may
require
it
for
that,
but
we
haven't
used
it
so
yeah.
Let's
see,
I
think
it's
not
something
we
have
to
move
it.
I
think
we
all
can
move
otherwise
we'll
be
there.
B
A
C
So
but
let's
target
I'm
targeting
to
complete
some
by
mid
at
least
a
prototype
by
mid
of
march
with
prototype,
I
mean
that
end
to
end.
We
can
have
one
into
one
scenario
for
collection
and
we
can
say
primitius
or
something
we
can
use,
I'm
not
not
primitive.
I
think
we
want
an
exporter,
but
I
don't
have
to
console
end-to-end
console
or
in-memory
exporter.
If
you
can
have
it.
I
think
that
would
be
good
to
showcase
by
in
the
payment
of
this
march.
C
B
B
I
mean
if,
if
you
decided
to
use
it,
I
can
prepare
the
environment
and
put
it
to
the
repo.
Add
it
to
the
repo
and
it's
really.
A
A
B
And
then
for
intellisense,
I
can
show
you
we
have
to
like
configure
with
cmake
and
that's
also
a
one-time
thing.
I
have.