►
From YouTube: Fluent Community Meeting Feb 10th, 2022
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
Hey
everyone
welcome
to
the
february
10th
fluent
community
meeting.
Today
we
have
a
pretty
cool
agenda.
I
see
a
lot
of
topics
here,
I'm
going
to
share
my
screen
and
we
can
get
started.
We
we
just
a
couple
of
like
logistic
updates.
We've
now
updated
this
meeting
to
be
45
minutes
long.
We
found
that
the
30
minutes
was
a
little
too
short.
A
A
lot
of
folks
had
questions
topics,
especially
last
minute
ones,
which
is
great.
We
love
community
participation
and
then
the
second
thing
is
we
are
going
to
now
do
these
meetings
every
two
weeks
instead
of
every
month,
so
folks
wanted
to
chat
more
frequently
about
stuff
that
we
were
doing
and
and
contribute.
So
that
was
always
always
welcome.
A
I
think
that
the
first
topic
just
to
jump
straight
into
it,
is
kind
of
centralizing
all
of
the
places
that
you
can
get
help
in
in
the
fluent
community.
We
have
quite
a
few
places.
We
have
what
are
the
google
groups?
These
are
like
thousand
folks,
150
folks
forums
on
on
google.
A
A
What
what
we're
finding
happening
is
that
things
are
getting
really
really
disjoint
and
especially
with
the
google
groups.
There's
a
ton
of
spam
there's
every
every
few
days
I
have
to
go
in
and
there's
like
giant
messages.
I
don't
know
what
it,
what
language,
what
the
thing
is,
but
there's
there's
something
going
on
there
and
because
we,
you
know,
we
want
to
allow
anonymous
posting
in
google
groups.
You
still
need
an
email.
The
good
github
discussions
might
just
be
the
the
better
course.
A
I
think
we've
had
actually
really
good
discussions
in
the
github
discussions
that
go
beyond
just
asking
for
help
and
the
slack
channel
is
still
still
really
good.
My
take
is,
we
should
just
deprecate
and
archive
the
google
groups.
This
is
what
the
kubernetes
folks
did,
and
they
also
moved
away
from
discourse
and
cncf
as
a
total
is
moving
towards
more
of
the
github
discussions.
A
D
A
Yeah
yeah
and
discourse
as
well
so
discourse.
Let
me
just
show
here
flinky
discussions
this
one,
which
is
it
has
a
lot
of
good
info.
But
honestly,
it's
it's
quite
a
bit
of
of
management
here
too,
and
not
finding.
A
E
Yeah,
taking
a
look
at
the
google
group,
a
we
have
240
members,
so
I
think
that
more
than
interaction
that
has
been
used
for
announcements
or
releases,
so
I
don't
know
I'm
thinking
that
I
don't
think
that
these
people
will
sign
up
when
they
go.
We
have
discussion
because
this,
like
us
announcements
list
also
maybe
we
can
use
it
for
that
purpose,
only
and
encourage
a
everything
that
is
about
support
or
technical
problems,
go
to
the
discussion
instead
of
just
archiving.
A
Right
totally
we
like
in
what
we
could
do
archive
is
just
make
it
read
only
and
then
only
moderators
can
post.
So
we
can
still
give
announcements
of
like
hey,
announcing
new
version,
but
then
no
one
can
ask
questions.
We'll
just
have
a
banner
there
hey.
If
you
want
to
ask
a
question,
please
don't
google
a
good
github
discussion
and
then
just
so
just
moderators
of
the
google
group
can
say
announcements
that
is
possible
yeah
and
then
we
can
also.
We
should
use
the
newsletter
list.
I
think
there's
we.
E
A
The
newsletter
has,
I
think,
a
thousand
folks
on
it
too:
okay,
cool
yeah,
and
I
mean
I
think
at
announcements
side.
We
should
probably
have
announcements
in
slack
and
stuff
too,
because
there's
like
6
hundred
folks
in
there
so
probably
good
ways
to
to
do
that
all
up.
Okay,
but
I
think
we
have
a
good
good
at
least
initial
action
plan.
D
Yeah,
so
you
guys
can
probably
cover
a
little
bit
of
it
as
well,
but
I
just
wanted
to
to
bring
up
a
couple
of
things
one
of
the
it's
not
really
1.9,
but
as
we
move
into
bi-weekly
meetings,
would
it
I'm
just
wondering
if
it
would
make
sense
to
have
them
in
alternate
time
zones
or
something
like
that
to
try
and
encourage
maybe
more
people
to
attend
and
and
stuff
like
that,
it's
a
bit
of
an
awkward
time
for
me.
D
A
D
It's
probably
maybe
ask
people,
what's
the
people
who
can't
attend
this
one?
What
would
their
preference
on
on
the
other
one
and
then
see
there
yeah
just
just
to
get
get
a
bit
more.
You
know,
then,
it's
a
bit
more
a
bit
more
global
as
well
and
stuff
yeah
on
the
1.9
stuff.
I
just
wanted
to
highlight
a
few
things.
D
We
do
have
a
one-line
install
script
now
it
actually
works
for
1.8,
it's
fine,
but
the
the
link's
only
in
the
1.9
documentation
but
yeah.
If
people
want
to
give
it
a
go,
feel
free.
So
that's
yeah.
It's
just
a
simple
script.
That's
in
the
repo!
Now
that
does
the
installation
and
also
the
other
thing
that
I'm
not
sure
how
visible
it
is
from
1.9
onwards,
we're
deprecating
the
td
agent
bit
package,
so
we're
going
to
build
both
of
them.
B
A
D
Yeah
so
so
we'll
default
to
the
new
one,
but
for
well
for
my
testing
of
the
script.
The
new
one
doesn't
exist
in
the
release
repos.
So
so
just
pulling
back
the
td
agent
bit.
D
But
the
there
is
one
in
the
repo
install
or
something
I
can't
remember
what
it's
called
the
other
questions
so
actually
about.
Well
this
afternoon,
I've
been
looking
at
nightly,
builds
for
for
testing
it
to
to
build
all
the
different
targets
and
then
run
the
tests.
D
There's
a
few
there's
a
couple
of
issues
at
the
moment
where
we'll
we'll
resolve
those
and
also
some
windows.
Ci
I've
been
working
on
as
well.
Currently
the
windows
stuff's
all
in
app
there,
but
we're
moving
it
into
github
actions
and
then
adding
some
smoke
tests
and
stuff
like
that,
because
there's
been
a
few
issues
recently
with
the
windows
side
which
could
be,
you
know,
the
default
config
wasn't
quite
right
and
stuff
like
that,
which
we
really
should
catch
much
earlier.
D
So
we're
adding
that
there's
the
question
on
centos
eight.
What
what
do
we
do
with
those
targets?
Do
we
still
support
them?
What
do
we
want
to
move
to
rocky
or
alma
linux?
What
are
people's
feelings
on
that
which
architectures
those
kind
of
questions
and
same
with
macos?
D
So
I
just
wanted
to
try
and
generate
some
discussion
or
get
some
feedback
on
what
people
might
want
on
those
targets.
I
don't
know
how
to
drive
that
here.
It's
anyone.
Anyone
got
any
preferences
or
ideas
what
they
want
to
do
on
those
on
those
ones.
Centos
is
probably
probably
the
main
one.
Do
we
want
to
transition
to
like
rocky
or
alma
linux,
but
I
think
rocky
only
does
one
architecture
at
the
moment.
D
So
do
we,
you
know,
is
that
okay.
Does
anyone
care
those
kind
of
questions.
D
So
the
problem
we've
got
so
we
support
it
now,
so
I'm
building
it
now
for
1.9,
but
obviously
it's
end
of
life,
so
it
won't
be
getting
any
updates
and
which
is
not
going
to
be
good
from
a
security
perspective.
D
So
it's
that
question,
and
so
I
can
build
it
now,
but
I
have
to
use
all
the
archive
repos
to
do
that
and
then
obviously,
as
you
move
forward,
those
aren't
going
to
be
updated.
So
what
do
we
do?
Should
we
migrate
everything
to
rocky
thenx
linux.
A
I
think
we
should
but
probably
need
to
wait
to
see
what
folks
ask
for
I'd
say:
let's
wait
on
on
this
end,
like
what
folks
asked
for
I
have
so
I
guess
the
question
is:
what
do
I.
D
A
Oh,
oh
well,
okay,
we're
building
amazon
linux
separately,
but
for
centos
eight,
it's
also
compatible
with
red
hat.
Eight
well
I'll,
be
compatible
with
that
version.
That.
A
A
A
D
Yeah,
so
I
think
we've
got
instructions
now
to
build
it
from
source.
I
could
might
put
those
into
ci.
I
guess
the
question
is
we
probably
only
want
to
target
the
native
architecture
of
them?
The
m1s,
I
don't
know
yeah
how
how
much,
how
much
effort
we
want
to
do
there.
E
I
think
this
is
mostly
a
calipia
interest
to
ship
a
macos
versions
like
signet
packages,
yeah,
so
right
now,
our
ci,
it's
already
built
in
on
macos.
I
think
that
is
not
building
on
m1
architecture,
but
we
just
push
the
introduction
for
m1
on
the
documentation.
So
if
anybody
wants
to
build
it,
it's
pretty
straightforward.
D
C
B
A
D
A
D
A
A
B
Oh,
can
you
share?
Can
you
see
the
browser
yeah?
Okay?
A
couple
months
ago
we
decided
to
add
the
blindly
spot
to
current
flight
bit
operator.
So
we
have
some
progress
now.
Basically,
the
front
desireds
are
added.
Now
we're
still
reflecting
back
fixing,
but
it's
working
now.
This
part
is
we
have.
We
are
planning
to
rename
the
flint
flame
bit
group
to
fluent.io,
so
this
part
has
been
done
and
we
also
planning
to
change
the
name
specially
to
class
ones
for
flint
bay,
and
this
is
still
ongoing.
B
We
are
still.
There
is
pr
we
are
still
working
on
it
about
the
fluenty
therapy.
We
have
some
example.
We
can
see
the
example
how
it
works.
B
First,
you
you
defend
our
front
bait
output
to
flindy
yeah
and
we
have
example
a
cluster
fluency
cluster
output.
It's
a
there.
The
fluency
scrd
is
you
to.
You:
can
define
the
input
where,
where
the
logs
comes
from
it's
basically
a
forward
or
atp
like
you
can
defend
the
replica
and
then
you
defend
a
fully
config
selector.
B
We
have
a
cluster
wide
selector.
You
can
depend
like
this,
you
can
you
can.
You
can
depend
which
filter
and
which
output
to
use
and
logs
from
which
namespace
should
be
filtered
for
this
config.
If
you,
if
you
didn't
define
this
one,
all
the
logs
from
all
namespaces
will
be
sent
to
the
filter
and
output.
B
B
B
So
if
you,
if
you
have
a
namespace,
you
you
don't,
you
can
you,
you
can
define
a
cr
flambe,
config
and
output
in
your
namespace,
and
only
the
logs
from
yearlings
will
be
sent
to
this
output.
So
this
is
the
this
is
a
kind
of
tenant
control
mechanism,
and
you
can
also
you
can
also
use
them
together.
I
think
that
there
is
an
example
here
you
can
also
come.
B
You
use
them
together
at
the
cluster
30
and
defend
the
name
speciality
and
the
clustering
will
only
select
the
global
output,
so
it
will.
It
will
send
all
the
clusters
logs
to
this
cluster
output
and
the
name
specialty
will
only
send
logs
from
this
namespace
to
this
users
and
code.
So.
B
This
this
one
also
have
the
user's
label,
so
you
can
also
send
the
you
this
users
log
to
the
the
same
class
output,
so
this
kind
of
hybrid,
a
combination
of
cluster
crd
and
means
srd
usage.
B
Actually,
we
we
have
the
we
have
the
flint.
We
have
the
flint
bit
operator
as
here
in
our
future,
and
it
it
will.
It
can
create
a
little
bit
demonstrate
and
friendly
cylindrical
side,
and
previously
we
we
implement
bleeding
as
a
deployment,
but
it's
not
okay,
because
department
can
only
have
one
replica
to
use
the
previously.
So
we
have,
we
have
changed
to
use
a
steel
site
and
you
can
define
multi
replica
of
friendly
and
in
and
we
have
implement
this
feature
you
can
in
a
single
friendly
port.
B
You
can
create
multiple
workers,
multiple
flindy
workers
on
in
the
same
different
friendly
part.
So
this
is
basically
how
it
works.
So
we
have,
we
still
need
some
bug
fix
reflector,
but
it's
working
it's
working
out
and
I
think
after
all
this
is
done.
We
can
rename
the
product
to
a
flint
operator.
A
B
B
B
This
is
already
config
generated
and
we
use
a
plugin
called
label.
Router
is
from
banzai
cloud
login
operator,
so
you
define
a
label,
and
this
is
in
this
label.
Is
you
can
filter
your
logs?
So
all
the
crd
configuration
will
be
generated
into
one
from
the
configuration
file
like
this
yeah.
It's
basically
how
it
works.
B
So
after
we
have
refactoring
the
flame
base
crd
and
we
can
rename
the
product
to
flinthop
retargeting
yeah,
that's
great.
A
B
Yeah
yeah
we
are
actually
this
is
you
know
you
can
define
multiple
crds,
so
they
will
not
interfere
with
each
other
with
this
liberal
router.
So
this
is
unique,
id
highly
unique
id,
so
nice
yeah.
B
A
A
Okay,
so
that's
the
flint
operator
update
yeah.
Let's
say:
let's
talk
a
bit
about
the
new
website
for
the
fluent
community.
A
Yeah,
I
guess
I
I
can-
I
can
leave
it
because
I've
talked
with
women
before
about
this,
so
we're
building.
We
built
a
new
fluid
bit
website,
we're
just
in
the
midst
of
finishing
the
last
pieces
of
it
to
get
it
launched.
I,
like
all
all
of
these
projects
that
it's
all
open
source
on
and
you
can
modify
pull
from
it
if
you
like
anything
of
it
but
yeah,
it's
a
brand
new
website.
A
A
Also,
just
like
you
know
the
community
stuff
that
we
have
here
and
yeah
the
slack
channel.
That's
probably
another
one
of
the
main
things
that
we
want
to
want
folks
to
join,
so
this
will
be
coming
coming
soon.
It's
hosted
on
netlify
and
then
everything
is
open
source
and
when
we
chatted
about
it
last
time,
the
big
thing
was:
how
do
we
also
do
pieces
around
the
fluentd
website?
Now
the
fluency
website
is
a
much
larger
I'd
say
task.
It
has.
A
It
is
a
bit
older
as
well,
so
we'll
probably
want
to
do
some
updates
to
this
that
are
represented
more
of
like
affluent
community
and
so
point
more
users
to
say:
hey
flynn,
community,
here's
what
flintbit
is
and
here's
how
these
things
work
together.
Here's
the
fluid
operator,
here's
the
fluid
release,
nodes
everything
everything
kind
of
around
that,
but
we
have
we
have
yet
to
build
out
a
kind
of
template
there.
A
So
it'd
be
great
to
hear
if
folks
are
interested
in
working
on
that
and
we're
happy
to
help
drive
a
lot
of
that
as
well
from
from
ecliptic
side.
A
Yeah,
I
feel
like
still
be
awesome.
If
yeah
we,
we
could
start
with
some
initial
designs
and
thoughts
around
it
and
and
basically
do
a
separate
meeting
around
like
a
fluent
website.
I
think
that'll
be
great.
C
A
Cool
cool
yeah-
and
I
guess
you
know
one
one
thing
is
we:
we
generally
build
these
in
hugo
to
follow
the
kubernetes
website,
and
some
of
the
other
projects
is
that
is
that,
like
a
framework
that
you're
not
they're,
familiar
with
hugo.
B
A
Okay,
yeah,
otherwise
I
think
we
could
also
as
long
as
open
source.
The
main
thing
is:
it
just
has
to
be
open
source
the
website,
so
people
can
contribute,
modify.
Take
so
yeah
docusaurus,
I
know,
is
another
really
popular
one
but
yeah,
I
think
that'll
be
good.
We
could
we
could
kick
off
another
one
and,
and
we
could
help
with
some
of
the
design
stuff
too
and
yeah.
So
I
think
that
will
be
that'll
be
great.
A
A
Okay,
cool
cool;
let's:
let's
create
a
new
one
and
I'll
I'll
ping
ping.
Everyone
to
join
that
and
I'll
add
that
here
in
the
notes.
So.
B
A
A
Maybe
we
make
maybe
we
add
it
as
part
of
the
fluent
bit
docs
and
we
have
a
kubernetes
section.
So
then
it
can
be
like
kubernetes,
section,
helm,
charts,
stamen
sets
and
influence
yeah
yeah.
I
think
that
makes
that
makes
the
most
sense.
A
Yeah
we
have
stream
processing,
but
we
could
put
like
kubernetes
yeah
section.
E
Actually
they
yeah
that's
perfect
to
have
it
there
and
one
of
them.
The
goal
is
like.
Maybe
then
we
can
revamp
or
make
some
changes
in
the
website
to
get
a
theory
access
to
here,
because
actually
users
get
to
the
documentation.
But
it's
it's
not
like
a
clear
step.
E
B
B
I
have
one
question
we
have.
Thailand
has
some
problem
when
building
the
m64
image
of
lundy.
Is
that
possible
to
build
arm
image
for
from
d.
A
B
Is
there,
is
there
a
container
image
from
arm
64.
B
E
A
E
B
A
A
Then
yeah
we
can
end
it.
We
can
end
it
there.
So
thanks
everyone
for
joining
we'll
see
everyone.
In
two
weeks.