►
From YouTube: Fluent Community Meeting August 19th 2021
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
Perfect,
okay,
so
we
are
here
with
our
second
meeting
around
the
fluent
operator,
which
has
now
been
moved
to
the
fluent
organization
so
thumbs
up
to
that,
I
believe
everyone
who
is
a
maintainer
of
that
project
has
also
been
added
to
the
fluent
organization
as
either
a
member
or
a
maintainer,
and
specifically
part
of
the
new
fluent
operator
group
or
team
that
I've
created,
and
secondly,
now
we
have
it
in
the
fluent
organization.
What's
next,
how
do
we
go
from
here?
A
We
definitely
want
to
create
some
buzz
around
the
the
project
moving
to
the
phone
organization.
What
does
that
look
like?
How
do
we
also
prioritize
a
roadmap
on
what
what
it's
going
forward
and
then
maybe
just
even
establish
some
logistics
with
how
community
outreach
goes
or
how
community
would
interact
or
ask
for
features
and
etc?
On
the
fluent
operator,
that's
what
I
thought
for
the
agenda.
Does
anyone
have
any
other
pieces.
B
B
B
Can
I
share
my
screen
absolutely
I'll
pass
it
to
you.
Okay,.
B
Actually,
we
have
make
some
significant
change
to
the
initial
proposal.
The
first
one,
I
think,
is
we
need
to
reflect
the
original
front
base
rd
because
they
are
originally
name-spaced
crds,
but
we
use
the
frame
base
area
as
we
use
one
bit
as
a
class-wide
agent.
So
it's
collect
system
logs
and
continue
also
it's
better
to
be
clustered
by
crds.
So
I
think
we
can
create
this
series
under
the
new
group
under
the
new
flint
organization,
crd
group.
B
B
It's
more
clear,
so
the
flintbait
can
send
logs
directly
to
these
destinations
and
it
can
forward
to
friendly
and
the
following
dcrds
we
we
have
initially,
then
the
fluently
series
and
the
phoneticists
can
forward
logs
to
these
destinations
as
well.
So
and
we
have
for
the
friendly
crds.
We
have
make
some
changes.
The
there
are
config
to
config
one
class,
provide
one
namespace
and
two
output
series
and
one
namespace
and
one
cluster
and.
B
The
initial
proposal-
I
I
don't-
I
will
not
go
through
the
details-
I
just
make
some
changes
this
input.
We
we
have
removed
the
tl,
the
tier
input
because
it
doesn't
mix
us
to
add
the
tier
input
for
our
friendly
deployment,
so
it
just
accept
a
network
inputs
like
forward
or
hdp
or
syslogs,
etc.
B
A
B
So
it's
a
namespace,
the
crds
you
can
define,
we
have
review
or
restrict
log
sent
to
the
output.
Only
only
the
only
the
logs
in
this
namespace
will
be
sent
to
the
final
output.
We
will
add,
some
restricts
there
yeah
and
this
is
the
nymph
specialities
and
we
have
cluster
crds.
I
think
we
have
cluster
crds.
You
can
send
everything
you
you
you
received
to
the
final
destinations.
B
There
are
no
name
space
there.
There
is
one
to
you
can
cite
here.
You
can
control
you
can
select,
which
namespace,
which
blockers
are
in
your,
which
name
please
to
assign
to
the
final
destination.
So
yeah
sure
is
some
adjustment
and
I
think
either
buffer.
B
You
can
possess
the
to
the
previously
oh
nice
yeah,
that's
great
yeah.
So
we
add
a
cluster
directive
crd,
so
we
can
define
some
common
user
directives
here.
This
is
this
is
new.
B
It
can
be
reused
in
other
crds.
I
think
yeah,
and
this
is
some
examples
of
sync.
This
is
an
example
to
use
the
different
crds
common
directive
is
different
here.
B
So
from
da
actually
is
a
more
complex,
so
it
takes
time
to
design
crds
yeah.
A
One
note
about
the
the
crd
names
that
I
would
worry
about
it's
like
and-
and
I
just
want
to
make
sure
my
understanding
is
correct-
is
when
a
user
wants
to
look
at
config
for
flue.
Indeed,
they
would
look
at
the
config
objects,
they
would
say
like
cube
ctl,
and
if
I
saw
the
kind
name
it
was
config,
so
they
would
say
get
config
or
yeah.
B
Yeah,
could
we
control
could
be
custom
edit
config
and
there
will
be
a
group
name:
dot,
flan,
dot,
dot
io,
maybe.
A
A
A
B
Exist
all
right:
this
is
the
entire
fluenties
rd.
They
can
defend
the
staff's
for
4d
deployment
yeah.
B
So
the
biggest
difference
between
fluent
of
reader
and
the
login
operator
is
the
login
embedder
required
the
4d
to
send
logs,
but
I
think
we
for
fallen
operator
is
it
doesn't?
It
doesn't
require
things
for
fluently,
so
you
can
use
base
only
you
can
forward
different
d.
So
that's
that's
our
principle
and
this
theory
define
the
frontier
department.
I
think
yeah.
B
So
I
think
this
is
great.
There
is
a
service
for
the
flindy
and
this
cluster
and
and
and
limb
space
config
information
previously
input,
filter
output
and
you
can
add,
label
select
labels,
select
name
space
to
collect
and
fall.
Config
is
all
restrict
to
this.
This
namespace
yeah.
B
So
I
think
it's
make
more
sense
than
the
last
version.
Yeah.
A
B
Okay,
you
can
differ,
you
you,
you
need
to
define,
create
a
multi
of
nd
department.
Yeah,
you
mean
that
well,
it's.
A
Like
a
at
the
process
level,
what
we
did
with
fluentd
was,
we
said,
we'll
take
this
socket
and
use
it
and
use
the
kernel
and
then
we'll
share
it
between
two
ruby
processes.
So
you
can
define
like
in
fluentd,
config,
okay,
there's
five
workers
for
this
config
four
workers
for
this
config
multi-process.
B
I
understand
that
so
we
will
add
that
I
think
we
we
didn't
consider
that
I
I
understand
that
there
are.
There
are
worker
concepts,
influential.
We
didn't
notice
that.
A
I
I
don't
know
if
we
need
to
so
this
is
the
the
interesting
piece,
because
the
idea
might
be
that
this
is
a
one
worker
operator
and
if
you
need
more,
you
just
increase
the
replicas.
The
replicas
themselves.
Do
the
yeah
the
one
yeah
that's
another
way
that
might
be
simpler
and
we
don't
have
to
worry
about
both
doing
a
process
in
a
pod
and
multi
replicas.
B
Yeah,
I
think
it's
more
suitable
for
the
cognitive
world,
exactly
many
replicas,
it's
okay,
yeah,
but
I
think
we
can
add
that
parameter.
It's
our
default
one,
maybe.
B
Think
otherwise
looks
input
section
just
forward
atp
and
we
call
it
transformer
and
graph.
B
So
I
and
so
complex
directives
and
or
graph
it's
difficult
to
define
series
four
from
the
password
as.
B
A
A
I
I
think
one
really
important
thing
is
so
this
this
proposal
looks
great
nothing
nothing
about
this
is
we
want
to
start
showcasing
the
kubernetes
operator
within
our
normal
con,
our
normal
documentation,
so
our
fluent
documentation
that
that's
pretty
popular
and
that's
where
folks
will
discover
how
to
run
within
helm
how
to
run
with
the
daemon
set,
and
I
think
we
need
to
start
adding
some
docs
about
kubernetes,
and
I
know
it's
not
going
to
be
like
an
immediate.
A
Let's
just
take
all
the
docs
from
here
and
and
add
it
here.
So
maybe
we
can
add
some
pointers,
but
I
was
hoping
someone
one
of
the
maintainers
might
be
able
to
take
that
on
or
like
propose
a
way
to
to
add
to
the
documentation.
There.
A
And
that
that
should
hopefully
also
increase
awareness
so
both
of
the
documentation
sites
combined,
get
36,
000
unique
users
per
month.
There's
a
yeah.
It's
a
good
disagreement.
B
Yeah,
I
think
for
sure
we
need
documentation,
but
I
think
it's
it
can
be
started
after
the
development.
It's
almost
finished
sure
that
okay,
so
yeah.
A
Yeah,
it
could
even
be
like
we've
moved
the
project
over
and
we
just
point
to
the
documentation
that
exists
already.
Maybe
yeah
like
from
the
fluent
bit
operator.
Sorry,
not
the
not
the
full
fluent.
A
B
Oh
okay,
you
mean
either
the
documentation
for
the
film
beta
operator.
Okay,
that's
that's!
Okay!
I
I
think
we
we
just
think
you
mentioned
to
add
some
prefix
to
to
the
config
yeah
right:
okay,
okay,.
B
B
So
does
anybody
have
any
other
questions.
B
A
Not
from
my
side,
I
think
only
the
name
piece:
we
I
think
we
can
come
to
a
decision
multi-worker.
We
just
leave
that
out
and
insist
on
replicas
and
and
then
I
think,
if,
if
we
could
do
like
some
sort
of
benchmark
or
something
that
could
help
incentivize
users
to
to
try
this
we're
starting
to
do
that
with
just
the
projects
themselves,
benchmark
them
just
more
frequently,
we
have
integration
tests
and
long-running
tests.
The
only
downside
is
it's
just
expensive.
B
Okay,
we
can
do
some
benchmark
tests
now,
ci,
okay,
so.
B
Okay,
it's
much
complex
than
the
original
flambeat
operator,
yeah
yeah.
A
B
A
Year
old
project
yeah
just
stop
the.
A
I
think
yeah,
so
that's
that's.
The
primary
things
I
wanted
to
discuss
is
like
getting
some
stuff
in
the
documentation,
the
just
the
status
of
what
we
we
have.
We
we're
we
are
going
to
do
a
quick
update
at
fluentcon
just
to
the
general
audience
about
all
the
projects
like
how
projects
are
progressing,
and
I
know,
there's
a
dedicated
session
to
the
the
fluent
operator.
A
So
we,
what
we
could
do
is
maybe
a
one
minute
pointer
and
say:
hey.
You
can
come
to
the
session
to
to
learn
a
lot
more
about
it,
but
this
is
just
a
quick
summary
of
what
we've
done
so,
okay,
yeah,
I
was
thinking
I
could
just
put
together
something
really
quickly.
I
believe
eduardo
will
be
presenting
that
from
the
fluent
community
side
hold
yeah.
A
Yeah
yeah,
so
it
said
eduardo
will
be
doing
the.
I
believe
the
keynote
and
updates
for
about
10
to
15
minutes,
and
it
will
include
roadmap
for
fluent
bit
and
then
just
updates
from
the
general
project
perspective.
This
would
be,
of
course,
very
good
to
have
this
as
well,
and
I
do
hope
that
folks
will
be
able
to
attend
fluentcon
as
well,
so
at
least
virtually
or
or
or
in
person
will
be
great.
B
That's
that's
exciting.
For
us,
the
product
we
create
on
the
phone
call,
it's
so
exciting.
A
Yeah
and
then
the
I
I
know
there's
that
session
too.
I
believe
it
was.
You
know
it
was
accepted,
so
well,
we'll
find
out
here
in
in
the
next
two
days,
so
yeah,
perfect.
Okay,
that's
all
I
had
from
an
action
site.
Any
anyone
else
have
anything
what
they
want
to
discuss,
chat
about.
B
B
I
think,
as
you
remember,
that
you
have
mentioned
a
fluence
community
website
proposal
and
we
have
completed
the
the
initial
proposal
in
the
google
doc,
so
I'm
not
sure
how
to
move
forward
this
this
proposal
in
our
community.
I
think
we
can
have
some
web
builders
on
board
and
to
join
this
new
website
building.
So
I
think
we
can
announce
this
proposal.
A
Yeah
absolutely
and
then
yeah,
so
so,
there's
there's
a
couple
of
things
that
we
were
doing
behind
the
scenes
and
I
should
have
updated.
But
let
me
I
can
update
now
is
one
the
cncf
provides
a
bit
of
a
budget
so
that
we
can
go
hire
folks
and
and
then
pay
for
the
work
of
doing
some
of
this
website
design.
A
The
timelines
can
range
because
it
is
something
the
cncf
commissions.
So
we
were
waiting
a
little
more
for
for
information
on
how
to
proceed
with
that.
But
in
the
meantime
we
found
a
designer
that
could
just
do
a
quick
like
a
lot
of
mock-ups,
so
the
next
community
meeting
I'll
I'll
be
actually
very
keen
to
share
with
everyone
hey.
These
are
all
the
mock-ups
that
this
new
designer
has
done
and
and
see
if
folks
would
be
interested
in
in
building
yeah.
A
So
absolutely
with
the
proposal,
I
think
you
can
share
it
in
the
the
slack
channel
and
the
community
meeting
like
next
week.
I
can
share
what
the
designers
kind
of
come
up
with
as
like
a
proposal
of
like
how
we
showcase
information.
How
do
we
get
more
users?
And
I
really
want
everyone,
here's
feedback.
It
will
be
really
good
to
make
sure
that
website
is,
is
really
nice
and
then
on.
On
top
of
that,
we're
going
to
try
to
build
the
websites
together
so
fluent
bit
and
fluency
right
now
are
too
disjoint.
A
So
it'll
kind
of
look
like
we'll
redo
flip
it
we'll
redo
fluent
d
and
then
at
some
point
we're
going
to
try
to
merge
it.
So
the
design
language
we
want
is
consistent,
but
then
the
overall
website
will
be
integrated,
so
I
I
think
we
want
to
keep
pushing
more
and
more
users
to
understand
the
fluent
ecosystem
versus
just
okay.
I
know
what
fluent
d
is
and
I'll
have
no
idea
what
fluid
bit
is,
or
I
only
know
what
flip
fit
is.
I
didn't
know
what
fluent
d
was
so
yeah?
A
That's,
so
I
I
think
there'll
be
some
good
stuff
coming.
A
Yeah
exactly
so,
the
designer
will
do
the
mock-ups
or
has
done
some
mock-ups,
but
then
it
will
be
the
yeah.
Then
we
got
to
figure
out
how
who's
going
to
put
it
into.
Actually
like
a
hugo
format,
all
the
website
has
to
be
open
source.
So
that's
our
requirement.
It's
like
our
websites
are
all
open
source
data
deployed
on
like
a
provider
that
basically
just
deploys
the
source
code
and
runs
so
right
now
our
website
is
in
jekyll
and
then
I
believe
the
fluent
d1
it's
in
some
other.
A
A
Okay,
then
I
we
do
have
the
we
do
have
it
next
week,
so
that
will
be
I'll,
be
keen
to
show
all
the
designs
and
everything
with
everyone.
Any
any
questions.
Thoughts.
B
A
Everyone
yeah,
perfect,
okay.
Well,
then,
next
week,
looking
forward
to
it
bye
for
now
see
you.