►
From YouTube: CNCF SIG Observability 2020-05-26
Description
CNCF SIG Observability 2020-05-26
A
A
B
A
C
F
C
E
D
C
E
E
E
Good,
so
the
other
thing
is
just
updated
from
from
CN
CF.
We
had
a
scene
CFTC
and
SEC
chair
called
last
week
and
basically
the
TOC
is
fine
with
us
suggesting
changes
to
pretty
much
whatever.
We
should
just
suggest
it,
and
they
it
didn't
sound
as
if
they
were
too
much
set
on
any
particular
parts
of
the
question
error
of
the
of
the
template
it
gut
feeling.
Is
there
probably
we
understand
whatever
we
whatever,
because
always
a
me
how?
How
did
you
perceive
this
I.
E
Okay
and
the
other
thing
sorry
I'm,
jumping
outside
of
the
agenda,
but
I
literally
have
back-to-back
meetings.
Today
there
was
a
user
sorry,
bye-bye
cortex
in
preparation,
also
for
the
due
diligence
and
such
Gotham
shared
the
results
with
me.
It's
like
one
of
those
questions
are,
are
rather
text
heavy
and
open
for
interpretation,
and
not
so
not
so
metric
heavy
editor,
so
I
was
thinking
to
maybe
even
around
a
second
one,
based
on
the
Prometheus
survey,
which
was
sent
out
last
Friday
I.
E
Think
because
that
also
has
like
scaling
numbers
about
how
many,
how
many,
how
many
metrics
we
actually
run,
what
is
scraping
interval
how
many,
how
many
data
points?
Do
you
get
per
X
amount
of
time
and
questions
like
that?
Oh
yeah
other
hand
running
two
surveys
directly
behind
each
other.
It's
kind
of
weird,
but
like
learning
from
this,
and
also
feeding
this
back
adds
yet
another
template
to
see
in
CF
to
to
basically
have
something
to
work
with
for
future.
Due
diligence
probably
makes
sense.
Question
is
what
could
you
ever
think
about
this.
E
E
My
thinking
is
that
in
part
we
are
there
spearheading
at
least
part
of
the
processes
within
CN
CF,
and
we
already
had
several
walls
there.
That
process
can
be
improved,
let's
say,
and
if
we
are
able
to
give
a
template
for
something
which
worked
also
in
regards
to
questionnaires
and
user
surveys,
I
think
that
actually
makes
sense
and
creates
more
benefit
within
CCF.
So
that's
why
I'm
considering
to
do
it
twice.
I
I
H
I
think
I
agree
like
let's
get
those
learnings
for
the
future
surveys,
but
let's
not
spam
users,
essentially
I
think
those
numbers
are
on
Sirius
I
think
there
are
larger
than
prom
cues
from
what
I
have
seen
anyway.
So
I
think
there
are
good
enough
for
incubation.
That's
what
we
would
like
to
get
from
this
survey
for
our
purposes.
So
I
don't
think
we
need
to
repeat
that.
Yeah,
that's
my
opinion,
but
we
should
have
those
questions
for
next
ones
for
sure.
E
Also,
if
I
mean
just
one
point
to
note,
I
fully
expect
ELC
to
turn
challenge.
Why
that
new
template
hasn't
been
used
for
cortex,
which
has
an
obvious
answer,
but
just
to
would
you
anticipate
the
future
I'm
I'm
willing
to
bet
that
this
challenge
will
happen?
And
it's
like
that's
the
logical
thing
anyway.
B
B
B
E
Yes,
I
go
back,
there
are
the
differences,
but
if,
if
cortex
has
consensus
not
running,
another
survey
is
the
way
to
go,
then
we
already
have
consensus.
No
one
else
is
speaking
up
so
from
the
from
the
working
groups
point
of
view.
There
is
no
reason
to
to
charge
the
decision
by
cortex
I
just
do
it
just
to,
of
course,
obviously
I
had
my
head
of
suggesting
this,
but
none
with
the
chair
and
on
I'm
completely
fine
with
this
I'm
fine
with
this
I
don't
have
a
strong
opinion
on
this
anyway,
and
if.
F
E
Fine,
either
way,
okay,
so
I
hope
everyone
has
read
the
document
which
we
are
about
to
dive
into
the
due
diligence
document
getting
much
more
time
since
last
time
we
are
on
page
11
unless
there's
any
substantial
change
above
which
we
should
be
taking
into
account.
There
were
some
to
do
those
being
addressed,
call
them
I,
honestly,
don't
know
yeah.
H
E
E
I
E
E
F
E
D
This
is
here
for
half
a
definition:
it's
just
not
in
there,
so
there
would
be
like
specific
criteria
like
it
has
a
declarative
API,
so
it
can
be
deployed
on
a
coordinated
platform.
It
it
is
available
wire
containers,
so
that
would
be
like
actually
a
number
of
criteria
that
the
CNC
F
itself
defines
for
culinary
certifications,
they're,
just
not
part
of
this
application.
So
technically
they
all
say:
yes,
it
can
be
shipped
as
a
container.
It
can
be
deployed
either
virus,
C
or
D
and
operator.
C
E
Make
the
core
consensus
for
this
one
right
now,
then,
should
we
wait
for
got
them
to
drop
the
document
in
the
background,
or
should
we
just
say
that
in
the
scope
of
that
question,
you're
happy
I'm,
basically
happy
with
all
of
them?
I
just
hope
that
we
are
able
to
actually
either
say.
Yes,
we
happy
with
the
complete
document
or
unhappy
with
the
complete
document
at
the
end
of
this
I.
D
E
Okay,
yeah,
but
then
still
this
question
needs
to
be
tightened
up
and
actually
refer
to
that
blah
blah
blah
also
becomes.
It
becomes
a
checklist
because
currently
it's
at
its
sentiment
on
the
plus
side.
We
all
agree
with
the
sentiment.
So
whatever
okay
document,
that
the
project
has
an
affinity
affinity
for
our
scene,
see
if
operates
and
understands
expectation,
has
been
CVF
project.
Yeah.
H
C
E
E
I
E
E
E
E
E
E
Okay,
do
we
believe
that's
a
growing
and
thriving
project
yet
tonight
with
the
scenes?
If
values,
then
it
could
mean
creation
criteria,
but
it
should
started
sandbox
I.
Think
incubation.
Well,
that's
kind
of
the
question
in
this
context
that
it
has
a
sound
documented
process
for
source
control,
issue
tracking
at
least
my
trend.
But
it's
it's
documented
process
for
ending
committers
that
there
is
a
governance.
There
are
committees
for
multiple
organizations
that
there
is
a
code
of
conduct
that
does
have
a
license.
That's
easy!
Oh,
can
you
add
which
license?
E
Very
good,
could
you
put
the
license?
Yes,
so
users
who
uses
two
projects
we
have
as
a
reference,
so
obviously
their
end
users,
also
by
the
way,
as
we
had
that
copied
last
time
end.
Users
are
clearly
defined
within
CDF
as
people
who
don't
provide
services
fathers,
so
service
providers
are
not
defined
as
end
users.
I
happened
about
two
tweeters
sometime
few
days
ago
and
I
just
remember
this
employee.
E
I
E
And
taking
off
microphone
on
my
chair
head-
and
he
put
him
on
microphone
of
self
for
a
second
I-
know
that
there's
actually
actual
time
by
a
tech
writer
being
committed
to
watch
this
as
well.
So
I
know
it's
true
just
for
the
benefit
for
the
others,
on
the
call
so
firmly
taking
off
the
poor.
Father
like
that
again
and
putting
back
on
the
chair
head
so
as
of
today,
sick
observability
is
happy
with
the
answers
in
the
About
section
all
on
read.
E
D
And
maybe
the
only
comment
that
I
could
maybe
make
of
this
is
at
this
point
here.
It
might
be
interesting
and
that's
like
an
end
user
point
of
view,
others,
if
I'm
an
end
user
and
I,
want
to
choose
between
tannaz
and
cortex.
What
would
be
the
right
in
your
IV
form
project
over
letter
so
right
now
the
feedback
is
very
technical
and
related
openly
to
some
of
those
bits
and
pieces.
D
D
Would
I
mean
about
just
a
proposal?
I
might
work
with
you
to
the
document.
If
you
have
somebody
just
reading
it,
how
would
you
like
that
question?
Like
should
I
pick
a
OB
and
is
there
a
volume
setting
those
and
back
to
it?
If
that,
like
to
project
I,
do
pretty
much
the
same
thing
from
an
end
users
perspective.
Look
sorry,
why
would
I
be
able
was.
E
This
be
okay,
I
mean
we
can.
We
can
request
that
that
this
is
put
into
the
section
as
workers
I
can
see
the
value.
The
question
is:
are
we
still
happy
with
this
section
even
without
this,
or
should
we
call
off
the
consensus
call
until
until
that
is
added,
or
are
we
happy
as
long
as
partic
and
work
on
through
edit.
I
H
Think
is
harder
every
day
because,
like
we
are
collaborating
so
much
that
we
are
literally
starting
to
use
the
same
API
and
having
the
same
features
like
very
detailed
features
like
we
care
the
same
things.
We
we
use
exactly
the
same
format
in
nowadays,
and
you
know
saying
that
hey
you
should
use
this
versus.
That
is
just
just
some
another
argument
for
separation,
whereas
we're
trying
to
go,
though
kind
of
together
and
hear
the
differences
are
very,
very
small
and-
and
they
are
getting
even
smaller.
H
G
H
H
I
E
Also
I
think
for
honestly,
this
is
outside
the
scope
of
the
discussion.
We're
having
about
the
due
diligence
of
cortex,
I,
think
I,
think
our
eyes
should
I,
never
point
make
sense
and
I
just
added
some
herbage,
which
which
gives
gives
some
some
encoding
beyond
that
I
think
it
shouldn't
be
any
technical
document
for
the
technical
diligence
on
a
member
project.
I
think
it
should
be
in
kind
of
an
overview
or
maybe
best,
current
practices
document
or
something
which
goes
more
into
depth,
about
the
trade-offs
and
about
the
design
goals.
E
H
G
H
E
E
Yeah,
well,
that's
that's
actually
a
good
question.
It's
again
unless
it's
something
which
blocks
from
moving
into
incubation,
it's
not
something
which
should
be
discussed
within
within
the
due
diligence
call.
So
unless
Stefan,
unless
you
disagree,
I
would
actually
tend
towards
jumping
back
into
the
document
finalizing
the
document
and
then
jumping
to
a
good
question.
It's
a
valid
question.
I
think
there
is
a
good
answer
to
it,
but
I
would
just
like
to
close
the
other
thing.
First,
objections:
yeah.
D
E
D
Thank
you.
The
party
says
that
there
should
be
at
some
point,
a
distinction
between
those
projects,
or
you
might
consider
at
some
point
version
that
what
happened
to
all
sensors
and
open
tell
them
to
tracing
at
some
point.
If
this
is
something
or
if
there
is
a
clear
distinction,
might
so
persist.
Well,.
B
J
So
if
you
allow
me
just
this
is
I
found
just
one
comment:
I
I
am
new
to
the
process
here
and
by
no
means
do
I
suggest
that
you
know
it
is
something
that's
a
problem
or
not
all.
What
I'm
suggesting
is
that
that
justification
or
that
acknowledgement
that
the
project
of
the
lab
and
why
they
they
overlapping?
It's
not
an
issue
should
be
stated
in
the
document.
That's
all
I.
E
E
J
H
E
D
E
E
E
Okay,
so
the
first
cultural
consensus
on
this
level,
SiC
observability,
is
happy
with
the
user
survey
and
does
with
the
cortex
user
survey
and
does
not
request
another.
No
I
think
we
already
have
agreement,
but
I
just
want
to
make
it
explicit
all
agreed
good
and
now
the
first
actually
non
janitorial
thing,
which
we
as
a
signature,
which
I
think
we
are
doing
in
the
second
or
third
officially
in
spent
instantiated
call,
which
is
quite
nice,
sick,
observability,
accepted
the
cortex
to
purchase
document
or
accept
sorry
accepts
the
diligence
document.
E
H
E
E
E
F
Happy
to
so
I,
by
design
actually
have
not
shared
this
out.
Broadly,
I
will
ask
people
to
review
it
for
the
next
cig
meeting,
so
we
can
discuss
it
in
more
depth.
I
just
wanted
to
talk
about
it
at
a
high
level
to
kind
of
ensure
the
same
agreement
understanding
so
there's
a
doc.
That's
attached,
please
feel
free
to
comment.
Everyone
should
have
rights
on
and
hopefully
make
edits
modifications,
so
we
can
discuss
it
more
in
depth.
F
It's
very
high
level,
it's
more
like
a
charter
than
it
is
like
a
deep
dive
on
the
specifics
by
design
to
ensure
that
it's
aligned
with
the
direction
that
the
sake
believes
that
we
should
go
in.
So
basically,
the
proposal
that
was
made
at
a
very
high
level
was
around
kind
of
promoting
and
providing
more
awareness
of
sig
of
observability
projects
in
CN
CF,
and
what
that
might
look
like
the
goal
is
not
to
duplicate
any
efforts
that
CN
CF
already
has
or
or
like,
promote
a
single
solution
or
like
best
practices
or
what-have-you.
F
So
there
are
some
like
non
goals
listed
at
the
bottom
as
well,
but
the
idea
is
just
to
kind
of
show
off
specifically
the
observability
products
and
how
they
tie
together
and
the
types
of
problems
that
they
can
solve.
So
I
listed
a
few
goals
here
at
a
very
high
level
like
how
would
content
be
shared?
How
would
we
track
that
content?
What
would
be
the
guidelines
for
contributing
contents
because
we
don't
want
it
to
kind
of
be
a
free-for-all?
For
example,
it's
not
up
like
vendor
pitches.
F
That
should
absolutely
not
be
the
goal
and
needs
to
be
explicitly
called
out
the
different
types
of
content
that
we
might
want
to
share,
whether
it's
kind
of
basic,
getting
started
stuff
or
a
much
more
deep
dive
of
like
solving
real
world
problems
that
exists
as
a
result
of
using
these
technologies
and,
of
course,
kind
of
opening
this
up
to
a
broader
audience,
so
that
anyone
can
really
contribute
and
get
involved.
It
shouldn't
be
just
about
maintainer,
x'
or
just
about
like
no-name
people
that
typically
do
blog
posts
or
what
have
you.
F
The
goal
here
is
really
to
to
share
and
collaborate
and
to
provide
more
visibility
as
to
what
is
possible
and
what
people
are
doing
with
these
different
technology
stacks.
On
the
second
page,
again,
very
high
level,
I
listed
a
few
open
questions,
I'm
sure
there
are
more
so
please
kind
of
post
them
in
here,
so
we
can
get
them
at
it
as
well.
Some
of
the
big
ones
that
come
to
mind
would
be
just
any
overlap
with
CN
CF.
So,
for
example,
I
already
crossed
out
one
of
the
potential
categories
as
it
turns
out
today.
F
Cn
CF
already
has
a
way
to
promote
new
projects,
new
releases
and
graduation
type
stuff,
I,
don't
think
duplicating
any
of
that
effort
makes
sense,
and
then
the
second
one
if
people
are
not
aware,
there's
a
proposal
out
to
change
how
sandbox
is
done
in
CN
CF
that
could
have
repercussions
as
to
what
is
possible
from
a
social
media
type
type
stuff.
For
example,
sandbox
projects
may
not
be
eligible
to
participate
in
this
as
a
result
of
some
of
the
upcoming
proposals
that
are
being
made
at
the
CN
CF
level.
F
So
I
guess
at
a
high
level.
Does
anyone
have
any
immediate
comments
or
suggestions
or
do
people
think
it's
a
good
idea,
a
bad
idea
heading
down
the
right
path,
the
wrong
path,
I'm,
just
trying
to
get
some
initial
feedback
as
to
whether
folks
think
this
makes
sense
and
whether
it's
worth
pursuing
more
depth.
H
F
F
F
And
if
people
are
aware,
like
hey
my
projects
already
doing
XYZ,
it's
been
good,
it's
been
bad,
like
that's
helpful
to,
or
people
were
aware,
like
I've
worked
with
CN
CF
and
they
suggested
the
following.
I've
talked
a
little
bit
to
them
in
regards
to
what
they
have
in
terms
of
blogs
and
webinars,
but
I,
don't
know
all
the
specifics,
so
I'm
sure
there's
a
lot
of
knowledgeable
people
here.
Please,
please
add
to
this.
F
Cool
so
again,
I
want
to
keep
this
very
high-level
just
to
kind
of
socialize.
The
idea
I
will
share
this
on
the
list,
so
people
can
review
it
for
the
next
cig
meeting
and
hopefully
we
can
spend
some
time
to
get
to
some
real
specifics
as
to
who
would
be
interested
in
kind
of
working
on
this
subgroup.
You're
gonna
define
the
criteria
and
anyone
that
might
be
interested
in
sharing
content
that
could
be
used
initially.
F
H
We
didn't
use
this,
even
though
you
really
created
those
documents
and
I
think
the
problem
is
that
yeah
we
kind
of
I
don't
either
the
idea
is
either
we
don't
know
how
to
use
it,
how
to
properly
get
advantage
of
it
or
it's
actually
an
use
for
for
last
month.
So
I
would
propose
actually
to
stick
to
github
issues
where
you
can
label
them.
I.
Think
it's
kind
of
you
know
you
can
have
some
discussions,
so
it's
not
that
bad,
so
I
think
that
we
actually
even
easier
to
spot
the
areas.
H
E
Yeah
that
the
one
thing
which
which
I
don't
want
to
fall
into
a
trap
of
basically
hiding
to
use
and
stuff
in
in
the
document,
because
it
will
just
as
soon
as
it's
all
of
you,
people
will
forget
about
it-
will
stop
being
tracked
and
updated.
So
I
I
can
see
the
idea
working,
but
then
I
think
we
need
someone
to
volunteer
to
actually
follow
up
and
go
through
this
living
document
and
make
sure
that
all
turtles
are
either
handled
within
a
certain
amount
of
time.
E
D
You
know
somebody
guessing
for
actual
action
items.
Different
issues
work
better.
It
don't
need
somebody
to
courage
this,
because
I
think
all
of
us
have
too
many
different
projects
that
they're
already
getting
issue
from.
So
it
would
be
like
part
of
the
video
side
before
every
meeting
that
you
go
over.
Those
issues
that
would
have
to
assign
them
I
also
agree
that
the
dock
is
good
for
collaboration.
D
If
you
want
to
discuss
them
because
he
made
quite
a
white
paper
and
issue
would
be
kind
of
available,
you
can
handle
it,
but
if
there's
like
this
week,
I
guess
that
you
want
to
handle
and
that
they
want
to
work
on
I
think
that
issues
are
the
more
natural
approach
to
that.
The
key
is
great
in
current
needs
who's
going
to
push
it
to
people,
because
that's
what
I
start
to
see
as
as
people
these
days
are
part
of
so
many
different
projects.
D
The
amount
of
issues
is
just
overwhelming,
so
somebody
still
need
some
time
to
get
stuff
done.
Unfortunately,
assigning
it
gives
up
issue
these
days
in
some
extent,
is
not
just
going
to
cut
it,
but
so
good
at
here
things
that
somebody
needs
you
rated
and
for
I.
Think
that
discipline
for
me
is
an
actual
working
document
and
it's
fine
for
certain
tasks.
But
that's
the
or
put
already
fact
just
following
up
on
the
tasks
and
the
issues
better
I.
H
E
I
think
we
actually
overrun
by
two
minutes
we
actually
overrun
by
by
12
minutes
course
this
meaning
and
it's
ten
minutes
early.
E
D
D
After,
like
the
practice
of
having
like
10
minutes
in
between
two
meetings
that
are
kind
of
like
those
meetings
run
over
you,
fine,
but
otherwise
you
run
over.
People
will
just
randomly
start
to
disappear
because
they
have
to
be
never
meetings
again.
That's
why
I
like
this
idea
of
like
every
this
10
minute
buffer
in
there
and
all
that
I
could
be
planning
for
it,
which
would
get
rid
of
like
a
serum,
a
buffer.