►
From YouTube: 2022-02-23 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
C
C
B
B
A
D
D
F
E
E
Okay,
anybody
else
from
the
community
like
whether
tom
or
essan,
any
of
you
received
the
proposal
earlier
because.
E
D
D
E
F
D
D
I
mean
where
exactly
if
you
really
want
to
contribute,
which
which
is
a
place
to
do
it
and
what's
the
process
we
follow,
so
I
mean,
if
you
want
to
do
it,
we
can
do
it
right
now,
probably
before
we
go
through
our
routine
agenda,
I
mean
I
don't
want
you
guys
to
waste
your
time
on
those
things
you
may
not
be
interested.
Probably
you
can
just
tell
us.
E
Briefly
about
what
exactly
it
is
sure
perfectly
fine,
so
I
do
have
the
sabrin,
who
is
the
product
manager
and
also
very
active
contributor
on
the
oss
community
side.
So
several
probably
would
you
like
to
give
the
brief
and
then,
after
that,
lux
quickly
can
go
through
the
document
which
we
have
already
shared.
G
Yeah
sure,
can
you
hear
me:
can
you
see
me
awesome
just
just
to
to
to
to
give
you
some
background
on
that?
So
so
we
have
a
set
created
and
an
instrumentation
for
for
the
apache
web
server
with
open
telemetry,
and
we're
fully
aware
that
there's
an
existing
solution
for
that
in
in
the
c
plus
plus
contrib
repository.
G
We
also
looked
into
the
possibility,
let's
say
to
to
provide
a
merge
of
both
solutions,
but
but
that's
unfortunately
not
possible,
and
what
we
wanted
to
discuss
with
you
today
is,
let's
say:
give
you
some
insights
on
what
we
have
created
and
what
we
want
to
donate
back
to
the
community
and
and
we
we
are
curious
to
to
to
get
your
feedback.
G
If
this
is
a
contribution
you
you
would
be
appreciating
what
would
let's
say
you
you
would
expect
for
from
from
us
to
to
change
on
that,
so
so
that
we
can
yeah
decide
to
give
this
back
to
the
community
as
a
donation,
like
other
things,
have
been
donated
to
the
community
like
the
sql
community,
commenter
or
the
flowmill.
G
So
so
we
would
like
to
to
follow
that
process,
but
first
of
all,
we
just
wanted
to
yeah
get
get
in
touch
with
with
the
c
plus
plus
community
since
since
you're
owning
this,
this
module
right
now-
and
this
is
the
c
plus
plus
commun
implementation-
and
get
your
feedback
on
that.
G
Yeah,
so
so
so
did
the
existing
one
lives
in
the
open,
telemetry
c,
plus
plus
contrib
repository
what
the
differences
are.
I
think
my
my
colleagues
can,
let's
say
point
this
out
better,
so
so
we
have,
I
think
we
have
a
few
slides
if
you
have
time
for
that,
we
we,
we
can
let's
say
give
us
some
insights
into
that.
D
H
Sure
sure
I'm
audible.
H
H
Yeah
is
my
slide
visible
to
you
guys,
yeah,
yes,
sure,
okay,
so
I'll
just
you
know
cover
it
quickly
because
we
are
like
we
have
a
short
time
and
if
time
permits
like,
I
provide
a
brief
demo
also
about
what
we
are
doing.
H
So
let
us
first
of
all.
H
Like
about
our
what
our
contribution
open,
telemetry
existing
module-
so
you
know
what
we
are
doing
is
like
our
apache
module.
It
captures.
You
know
the
request
processing
time
through
each
and
every
module
like.
As
we
all
know,
the
http
request
passes
through
different
modules
in
its
lifecycle,
and
you
know
we
capture
the
response
name
for
each
of
the
module
by
injecting
our
you
know,
instrumentation
into
the
apache
server
at
runtime.
So
this
is
our.
You
know,
module
which
we
are
planning
to
contribute
to
the
space.
H
So
this
is
just
an
importance
like
why
it
is
so
important
to
monitor
web
servers
these
days,
because
you
know,
as
we
all
know,
the
capacity
and
the
number
of
requests
are
increasing
day
by
day
and
it
becomes
a
you
know,
sheer
importance
to
monitor
the
requests
so
that
in
case
of
any
breakage
or
something
we
are
able
to,
you
know
rca
that
so
this
is
a
architecture
diagram
of
a
module.
H
Through
the
architecture
diagram,
so
this
is
the
client
client
makes
a
you
know,
a
request
to
the
runtime
and
with
all
other
http,
con
modules
like
which
are
the
default
modules.
Our
module
is
loaded
with
them.
So
when
a
request
comes,
it
will
go
through
each
of
the
module
and
apache
modules.
Are
you
know
our
app
module
sits
at
the
first
one?
So
whenever
a
request
comes,
it
will
first
go
through
our
module
and
then
it
will
go
through
the
subsequent
modules
like,
for
example,
we
have
other
modules
like
authentication,
phase
or
php.
H
You
know
to
market
the
php
request
and
other
such
modules.
So
each
way,
each
by
you
know,
step
by
step.
It
will
go
through
each
of
the
modules
and
then
you
know
as
I'll
talk
about
the
logic.
Later
as
we
are,
you
know
capturing
the
response
name
for
each
of
the
modules
we
will
have
a
you
know,
a
different
spans
created
for
each
of
the
particular
module
like,
for
example,
for
a
php
request.
If
you're
catering,
php
requests
we'll
see
a
span
for
mod
php.
H
And
we,
how
much
that
you
know
the
time
to
what
was
its
contribution
in
the
whole,
you
know
processing
the
request,
and
then
we
are
using
c
plus,
plus,
you
know
hotel
sdk
to
set
this.
You
know
spans
and
traces
to
the
collector.
This
is
the
architecture
diagram.
H
So
these
are
just
you
know,
simple
steps
like
how
do
our
module
gets
loaded.
So
one
important
point
that
needs
to
be
pointed
out
here
is
that
our
appd
module
the
web
server
module
needs
to
be
loaded
as
the
last
in
the
http
dot
configuration
file,
which
you
specify.
H
The
reason
for
the
same
is
because
you
know
all
the
modules
like
if
you
are
aware
of,
like
we
specify
in
that
configuration
file
all
the
modules
which
are
loaded
before
our
app
dynamics.
Module
will
be
instrumented
correctly,
any
module
which
will
be
loaded
after
that
will
not
be
instrumented
because
of
the
property
by
apache
that
you
know
it
loads.
The
module
in
the
reverse
order
in
which
they
are
you
know,
specified
in
the
configuration
file
and
third
point
here
is
that
you
know
they.
H
There
are
certain
common
required
for
the
internal
starting
or
functioning
of
the
apache
server
we
can
see
and
we
have
excluded
them
from
our
list.
Just
because
you
know
it
doesn't
make
me,
you
know
it
is
not
a
much
importance
to
monitor
each
of
them
because
they
don't
directly
take
part
in
the
request
processing.
H
Their
only
role
is,
to
you
know,
start
and
spawn
the
apache
server.
So
here
is
a
list,
and
it
is
there
in
the
you
know,
proposal
doc
as
well.
H
So
this
is,
you
know
the
basic
comparison
which
we
were
discussing
earlier,
like
what
it
is
different
from
the
existing
module.
So
if
the
main
point
here
is,
you
know
the
existing
model,
only
like
monitors
exit
call
through
mod
proxy,
like
you
know,
when
the
part
is
functioning
as
a
reverse
proxy.
H
Whereas
if
I
talk
about
our
module
on
the
right
hand,
side
we'll,
you
know
we'll
capture
the
response
name
for
each
of
the.
You
know
the
module
through
which
the
request.
H
Goes
through
and
we
have
the
module
like
the
importance
of
this
particular.
You
know
this
particular
function
is
like
I
have
a
use
case
over
here
yeah.
So,
for
example,
if
you
know
a
customer,
if
you
know
a
user
is
using
authentication
module
for
you
know
its
web
application
and
you
know
for
if
we
go
through
the,
if
we
use
our
existing
hotel
module
we'll
miss
some.
H
You
know
a
monitoring
because
like
because
it
doesn't
provide
any
capability
to
for
the
module
wise
monitoring,
like
you
know,
if
a
user
faces
any
you
know,
blockage
or
you
know
the
you
know,
the
the
authentication
module
is
taking
much
longer
than
expected.
H
Then
that
will
be
you
know
correctly,
captured
by
our
you
know
our
abdi
module
and
we'll
able
to
show
them
show
that
you
know
a
particular
response
time
in
the
form
of
spans,
and
I
have
a
you
know
small
demo,
to
show
this
like
to
explain
this
bed.
D
So
the
current
existing
hotel
apache
module
that
module,
which
we
have
right
so
that
that's
not
that's,
only
going
to
intercept
the
request
going
to
mod
proxy,
not
not
to
all
the
other
modules
which
are
configured.
H
Correct
correct,
correct
right,
so
this
demo
explains
it
better.
So
here
I'm
running
two
simple
scripts,
as
you
can
see
over
here.
So
first
one
is
a
simple
html
file
which
takes
you
know,
milli
or
micro
seconds
to
serve,
and
the
second
one
is
a
php
file
which
I've
added.
You
know
a
sleep
of
one.
Second,
this
I've
done
it
intentionally
so
that
you
know
we'll
be
able
to
capture
it
and
provide
it.
You
know,
compare
it
differentially
from
this
html
script.
H
If
you
see
the
duration
first
of
all,
it
is
clearly
visible
that
you
know
the
php
file
is
taking
one
plus
second,
and
whereas
the
html
file
is
taking
only
micro
milli
seconds,
and
if
I
open,
you
know
any
of
them
like
this
is
a
php
one.
So
you,
as
I
was
explaining
later
so
these
are,
you
know,
module
wise
contribution
for
into
the
total
response
time
now,
as
you
can
see
this
mod
php,
this
is
taking.
H
You
know
a
99
percent
of
time,
even
more
than
99
percent
like
one
second
is
clearly
coming
from
this
mod
php
we
can
see
which
is
contributing
to
this.
This
total
response
time,
others
contribution
as
well,
is
very
negligible.
H
38
microseconds,
so
this
because
you
know
you
will
be
able
to
capture
the
the
right
and
the
correct
area.
Further.
Your
rca.
D
D
Okay,
how
how
it
will
work
coming
in
the
kind
of
I
haven't,
really
tried
the
missing
hotel
module.
That
will
just
give
me
the
request
it
has
flown
through
mod
proxy,
even
though
the
request
has
gone
through
other
modules,
which
is
not
going
to
show
me
that.
H
This
is
a
the
proposal
dock
which
we
prepared
like
we
can
share
it
with
you
guys,
which
was
talking.
D
D
H
E
D
H
D
Okay
got
it
now:
I've
got
it.
This
is
this
module
exists,
along
with
all
the
other
modules.
I
thought
this
is
exactly
correct.
G
D
E
So
let
me
understand
this
is
the
first
time
you
are
going
through
all
these
details,
but
only
thing
we
were
thinking.
Probably
you
guys
have
the
context
already,
because
that
was
the
information,
but
definitely
you
can
ask
any
kind
of
question
or
we
will
be
able
to
provide
you
all
the
details.
Wherever
there
is
a
confusion.
H
Yeah
so
allow
it,
as
you
can
see,
so
I
have
opened
the
configuration
file
for
http.
So
now
what
we
are
doing
is
we
are
loading
our
own
configuration
file
in
it.
You
can
see
app
dynamics
agent.conf.
H
H
D
D
And
this
is
not
going
to
intercept
any
of
the
default
apache
modules
like
like
ssl
module.
It
is
not
going
to
intercept
any
traffic
which
is
going
through
that
it
is
going
to.
I
mean
just
correct
me
like
if,
if
there
is
https
request,
will
that
also
get
intercepted?
And
it
will
show
us
how
much
time
the
actor
has
to
be
here
and
s?
Handshake
is
taking.
H
So
sorry,
my
net
is
breaking.
Can
you
repeat
that
so.
H
So
I
don't
remember
correctly
like
if
mod
ssl
is
included
or
not,
but
if
it
is,
you
know
part
of
that
included
modules
which
we
specify.
Then
it
should
show
that.
H
Yes,
what
I
wanted
to
say
is
like
currently,
there
are
some
of
the
modules
which
are
already
been
handled
and
we
are
maintaining
a
list
of
modules
which
which
are
not
instrumented,
like
those
lists.
Are
there
in
the
are
mentioned
in
the
document,
so
you
were
asking
about
mod
ssl.
H
So
if
I
quickly
search
there
ssl
and
that
is
in
the
excluded
list,
so
we
are
not
monitoring
that
mod
ssl
as
of
now,
but
if,
if,
if,
if
the
community
requires
it
and
wants
to
instrument
that
we
can
anyway
remove
it
from
the
excluded
list
and
that
that
will
be
automatically
monitored.
D
E
The
whole
idea
is
now
we
can
start
at
the
cncf
reviewer
level,
like
with
you
guys
and
also
the
owner
of
that
module,
so
that
we
can
discuss
the
overall
proposal
and
find
out
what
is
the
right
approach
so
overall,
based
on
our
understanding
and
that
insight
which
we
have
about
the
existing
module.
We
definitely
believe
the
module
which
we
have
developed,
which
include
all
the
functionality
of
the
existing
one
plus
additional
thing.
So
direct
merging
is
definitely
doesn't
look
very
feasible
or
the
right
approach.
D
E
D
That's
what
I
was
thinking
I
mean
normally
the
the
current
maintainer,
so
we
have
components
so
we
I
mean
how
it
is
organized
is
that
we
have
the
complete,
open,
telemetry,
cpp
level
maintainers.
So,
like
me
and
tom,
we
both
are
the
maintainers
of
the
open
telemetry
and
then
we
have
component
level
maintenance
for
the
contributors.
D
D
So
normally
those
maintainers
do
join
the
meeting
now
and
then,
if
they
have
something
to
really
say
and
some
questions
are
there
to
come,
so
I
think
let
me
see
if
they
can
join
next
week
meeting.
I
think
next
week
would
be
joining
from
india
time
right.
It
will
not
be
feasible
for
you
to
join
the
next
meeting.
That's
it.
I
think.
I
know
that,
because
middle.
D
To
next
week,
in
that,
if
we
can,
we
can
ask
them
if
they
can
join
the
meeting,
and
probably
you
can
discuss
it
over
there.
I
can
give
you
the
contact.
I
mean
if
you're
already
in
the
slack
channel
I
can.
I
can
give
you
the
contact
of
those
those
maintainers.
I
think
you
can
start
discussing
with
them.
D
D
Yeah,
I
think
I'll
give
you
the
contact
of
them
feel
free
to
discuss
directly
with
them
here.
I
think
definitely
it
will
make
sense
and
let's
we
can
invite
them
to
join
in
next
to
next
week
meetings
and
then
it's
comfortable
for
you
to
join
in
discuss
over
there
and
I
think,
let's
communicate
over
slack,
I
mean
any
discussion
we
want
to
have
over
this
and
sure
that
looks
good.
To
me
I
mean
I
was
not
aware
that
that
the
existing
apache
module
is
only
for
mod
proxy,
not
for
others.
D
E
Absolutely
yeah
that
will
make
sense,
so
I
will
make
those
honest
details
from
your
site.
We
will
have
an
interaction
and,
if
required,
maybe
a
quick
meeting
with
them
and
then
week
after
next
we
can
again
have
a
follow-up
meeting
and
if
you
suggest
probably,
we
can
share
this
proposal
document
in
cpp
slack
channel
also
so
that
not
only
you
even
rest
of
the
people
can
also
go
through
it.
Do
you
suggest
that
you
have
here
yeah.
D
D
D
I
D
E
Yeah
again,
definitely
I
leave
it
to
rest
of
my
teamwork,
but
probably
I
will
continue
because
this
will
be
a
good
start
for
us
to
start
understanding
rest
of
the
things
which
is
happening
on
cpp
and
he
would
be
definitely
regular
on
the
some
of
the
signatures.
So
not
only
we're
looking
for
this
apache
module,
which
we're
discussing.
Definitely
our
commitment
and
focus
is
towards
contributing
on
many
other
thing.
On
the
c
c,
plus
plus
community.
D
F
C
If
they
need
to
be
updated,.
C
D
D
C
D
And
I
wanted
to
talk
about
the
existing
pr,
which
owen
is
probably
that's
something
we
can
start
discussing
right
now,
because
I
think
there
are
lots
of
review
comments
coming
on
that.
That's
basically
for
a
concurrent.
D
D
I
mean
I
do
see.
This
is
definitely
going
to
improve
yeah
the
exporter
performance
for
sure
I
mean
at
least
throw
in
the
multi-core
systems,
where
we
can
have
multiple
threads.
If
they
can,
you
can
mentally
manage
multiple
sessions
here.
Definitely
that
is
going
to
help.
The
only
concern
I
have
already
put
that
review
comment
here
is,
which
I
see
is
I
mean
I
have
a
couple
of
concerns
here.
Actually,
one
one
is
that
this
is
not
in
accordance
with
the
specs,
because
the
specs
says
that.
D
That
the
exporter
should
result,
it
should
return
success
only
if
the
transfer
over
the
wire
is
successful,
which
is
something
we
are
not
doing
it
now
in
this.
So
this
basically
returns
successful.
Even
if
we
are
able
to
add
the
session
in
the
list
of
pending
sessions,
the
exporter
will
return
success,
which
is
not.
D
Yes,
yes,
so
you're
saying
we
should
make
exporter
asynchronous
right.
D
D
If
we
really
want
to
have
these
changes,
I
mean
it's:
basically,
the
the
exporter
should
return
a
callback
to
the
processor
and
let
processor
handle
success
or
the
or
failure
inside
that
product,
so
that
that's
how
that's?
How
java
and
both
both
java
and
javascript
are
also
doing
this
so
javascript
already
uses
that
asynchronous
pattern
which
they
have
and
even
java,
also
uses
java
already
have
something
called
as
collectible
interface
completable
features,
so
it's
using
the
future
returns
the
future
to
processors
and
that
processes
handle
the
future.
D
Other
thing
I
was
thinking
is
that
this
is
very
tightly
coupled
to
otp
to
otlp
http.
If
we
can
somehow
make
it
a
separate
component
which
can
be
used
in
both
zipkin.
Somehow,
if
you
can
do
that,
I'm
not
sure
how
we
can
do
it,
but
I
think
if
we
want
to
implement
the
same
thing
in
chicken,
it's
again
going
to
be
the
repeatable
food.
All
these
things
we
need
to
add
again
in
the
zip
pin,
also,
which
does
not
make
sense.
C
D
D
Yes,
what
should
we
do
with
this?
Should
we
have
can't
we
mock?
Can't
we
mock
the
concrete
test
scenario.
C
A
A
D
D
D
This
is
it's
not
a
bit
complicated,
but
straightforward
change.
I've
just
come,
get
the
list
of
matching
views
from
the
view
registry
and
create
the
evolution
for
that,
and
those
aggregation
stores
would
be
used
to
do
the
actual,
recording
or
actual
actual
consumption
of
the
matrix
and
the
collection
so
for
the
consumption
part,
it's
already
there
for
collection,
which
I
said
that
I
have
to
raise
another
pr
which
is
still
pending
this,
because
I
am
I'm
right
now
working
on
similar
pr
for
asynchronous
instruments,
which
is
bit
complicated
right
now.
D
We
have
to
create
the
aggregation
store
for
asynchronous
instruments
also
and
that
that's
getting
a
bit
complicated,
I'm
just
trying
to
see
how
how
I
want
to
do
it.
I
think
once
once
that
is
done
I'll
raise
this,
the
pr
for
the
consumption
and,
let
me
see-
but
I
think
I
mean
it
will
have
some
time.
Probably
I
can.
I
can
discuss
offline
with
you
if
you
want
to
take
some
of
the
work
and
yeah
yeah.
D
D
D
C
For
benchmarks,
I
collected
the
failing
ones:
okay,
total.
We
had
41
failures.
A
C
C
All
right,
it's
the
name
of
the
test,
I'm
looking.
C
Not
we
we,
we
can
disable
the
error,
failure
for
sdk
test.
So
we
don't
see
this.
D
D
B
I
haven't
seen
our
changelog
there's
no
new
new
new
vr
added
to
our
released
changelog,
but
I
think
we
have
many
fixes
we
still
have.
I
think
I
think
your
patch
release,
I
think,
makes
sense
for
for
this
week.
How
do
you
think.
D
B
C
Small
one
which
one,
what
was
it
static
function.
D
F
D
D
B
D
B
D
D
In
fact,
it's
good
good
to
have,
I
mean
more
time
within
releases.
That
shows
that
our
our
the
I
mean
our
sdk
is
more
stable
or
if
they
are
lower.
D
B
I
think
this
works
for
me
when
we
can
release
and
do
a
release
next
next
month.
Yeah,
it's
fine.
D
B
B
Yeah,
I
think
then
most
parties
align
their
metrics
release,
matrix
prs
to
their
to
the
date.
E
E
D
A
D
D
D
F
D
D
D
D
D
C
D
D
D
D
Okay,
I
think
this.
This
is
a
problem.
No
shady
span.
This
is
backward
reported
from
rib
std,
which
is
gnu
gpa,
license
nothing.
So
we
have
no
std
variant,
which
is
back
quoted
from
epsilon,
which
is
the
apache
license,
but
we
have
no
studies,
fan
which
was
backloaded
from
deep
stdc
plus
plus,
which
is
gnu
version.
3,
and
probably
I
have
to
see
if
this
version
is
as
imminent
as
apache
or
if
it
is
causing
any
constraints
on
not
able
to
use
open
sdk
as
a
file.
D
This
is
gpa,
license
version
2
used
to
be
most
restricted,
but
I
think
this
is
gpu
version
3,
which
I
had
a
quick
look.
That
look
looks
to
be
similar
to
mit
and
apache.
I
may
be
wrong,
but
I
have
to
see
once
again,
but
the
existing
gpa
license
version.
1
version
2
used
to
be
a
bit
strict.
We
cannot.
If
we
are
using
that
in
the
source
we
have
to
release
the
source.
D
A
D
D
D
D
If
it
is
using
some
of
the
code
from
gnu,
nice
and
serious,
we
should
not
be
so
that
something
we
have
to
change
it.