►
From YouTube: 2023-03-07 Rook Community 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).
A
All
right
recording
has
started,
and
this
is
the
March
7th
2023
Rook
community
meeting,
let's
jump
on
into
Milestone
checkup
here
with
1.10
looks
like
since
the
last
community
meeting
out,
which
is
21st.
We
did
have
a
patch
release,
1.10.12,
which
kind
of
coincided
with
the
the
bigger
1.11
release.
So
I
doubt
I
will
pulled
up
the
release.
Notes
real,
quick,
but
I
bet
that
it's
more
important
to
talk
about
what
was
in
1.11
instead.
B
C
A
Sounds
good
and
then
we
don't
have
specific
plans
for
1.10.13
another
Pax
release
there
we're
suggesting
doing
it
later
in
a
couple
weeks
from
now,
because
there's
not
you
know
a
ton
of
things
going
on
there.
So
maybe
in
like
a
normal
Cadence,
a
normal
slowed
down
Cadence
for
1.10
patching.
C
A
Right
so
yeah,
so
let's
move
on
then
to
1.11..
Let's
take
a
quick
look
at
the
release.
Notes
first
for
the
1.11.0
release,
as
that
was
the
big
thing.
That's
happened
since
the
last
community
meeting,
so
yeah
go
ahead,
Travis
and
get
us
updated
on
the
big
highlights
for
1.11.
C
C
So
a
lot
of
good
things
in
here
and
the
release
notes,
don't
capture
all
the
great
work
everybody's
done,
but
just
a
few
really
small
breaking
changes
just
as
far
as
up
the
minimum
version
of
kubernetes,
so
we're
supporting
the
last
six
that
are
out
there
from
1.21
and
newer
there's
a
new
CSI
driver.
So
we
up
the
main
version
of
that
removed
the
machine
disruption
budgets
which
nobody
was
using.
C
Goaling
versions
updated
so
really
housekeeping
for
all
those
things
and
then
feature
wise
yeah.
Just
I
don't
know
if
you
want
to
talk
through
all
those
but
happy
to
have
all
these
features
out
there
as
far
as
enabling
this
fxporter
doing
more
with
messenger
2
and
allowing
people
to
enable
that
by
default,
even
if
they
don't
have
encryption
on
The
Wire
enabled
marrying
data
across
clusters
with
overlapping
networks
is
an
important
one
for
people
mirroring
when
they're
not
using
host
networking.
C
The
last
point
there,
Seth
managers
standby
using
the
Readiness
probe,
talk
more
about
that
with
an
issue
you've
got
we
need
to
fix
in
the
patch
release,
but
probably
we'll
revert
that
one,
actually,
that's
the
plan
and
yeah,
but
overall
great
release.
Thank
everyone
for
contributions.
A
Like
async
follow-up
for
for
this.
C
C
Yeah,
so
we
did
have
a
few
issues
in
the
release
that
a
couple
have
been
have
been
fixed
already,
but
just
to
talk
through
them
and
what
happened
so
the
exporter,
so
the
new
stuff
exporter
demon
is
enabled.
Now,
if
you
have
the
Prometheus
monitoring,
enabled
then
we'll
start
this
exporter
demon
on
each
node
well,
the
node
is
currently
crashing.
If
it's
on
a
node,
where
there's
no
OSD,
that's
what
it
looks
like
for
the
Repro.
C
So
still,
this
is
the
an
open
one.
Still
don't
have
an
exact
or
no
PR
open
yet
for
the
fix
but
I'm
hoping
today
to
go
ahead
and
find
a
a
small
fix,
at
least
even
if
it's
disabling
the
thing.
So
we
can
get
the
1.1
release
to
be
a
stable,
build.
That's
my
goal:
I'd
rather
have
it
not
enabled
the
whole
export
or
disabled
then
have
a
crashing
exporter
demon
so.
A
And
Travis
in
the
in
the
comments
here
I
saw-
maybe
maybe,
as
you
mentioned,
something
about
large
crds.
Is
that
like
a
red
herring,
tangential
sort
of
thing
or
is
that
related
here.
C
D
Yeah
and
that's
too
big
so
yeah,
this
is
I.
Think
particularly
this
was
the
NFS
crd,
which
I
I
suspect
has
to
do
with
the
new.
D
We
allow
users
to
specify
volume
sources
for
some
of
the
config
files
so
that
they
can
pick
whether
they
want
like
a
host
Mount
volume
or
a
secret
or
a
config
map,
or
you
know
what,
wherever
else,
their
config
might
be
stored
and
then
have
that
applied
into
the
Container.
But
those
volume
sources
can
be
pretty
pretty
long,
and
so,
if
you
do
apply
the
like
last
applied
configuration
is
set
on
an
annotation
on
the
resource
and
that
annotation
has
a
limited
size.
You
know
the
resource
itself
does
it.
C
A
Right
yeah,
thanks
for
that
update
on
that
that
detail
as
well
Lane.
A
All
right,
okay,
so
then
there's
it
sounds
like
the
Travis
is
a
few
things
to
track
down
there
or
you
know,
potentially
make
the
decision
to
disable
the
whole
exporter.
But
it
looks
like
there's
a
good
lead
there
on
why
it's
crashing
in
terms
like
if
it
lands
on
a
node
that
doesn't
have
an
OSD
presence.
So
that
sounds
like
something
to
follow
up
on
to
see
if
we
can
get
that
fixed
in
quickly
enough.
If
that
is
the
root
cause.
Okay,.
C
Yeah
then,
the
next
issue
there
is
the
is
what
I
was
referring
to
in
the
release,
notes
that
we
have
an
issue
with
so
in
in
the
release
1.11,
we
changed
the
manager
so
that,
with
the
active
standby
configuration
we
had
been
managing
that
with
a
sidecar,
that's
running
with
the
manager,
and
then
the
sidecar
would
update
the
services
as
when
the
active
manager
changes.
C
But
we
changed
that
to
use
a
Readiness
probe,
which
was
working
really
smoothly
from
Luke's
perspective,
because
the
Readiness
probe
pretty
quickly
detects
a
stuff
manager,
changes
from
active
to
standby
or
vice
versa,
so
we've
merged
that
I
got
into
release,
but
then
it
turns
out
there's
a
couple
of
external
components
which,
like
a
Prometheus
alert
that
was
failing
and
some
other
deployment
tools
which
didn't
like
the
fact
that
any
pod
is,
is
in
a
not
ready
State,
because
basically
it
means
the
standby
manager
is
always
not
ready.
It
always
fails.
C
C
That
where
he
opened
a
PR
to
revert
all
the
related
fixes
for
that
feature,
we'll
just
revert
back
to
the
sidecar
and
then
we've
been
discussing
another
approach
to
improve
the
sidecar
implementation
and
so
and
he's
got
a
PR
open
for
that.
But
we
want
to
do
some
more
testing
and
verification
of
the
approach
before
we
merge
it.
So
yeah
we'll
just
revert
it
and
then
get
the
dot
one
release
stable
again.
From
that
perspective,.
A
Yeah,
that
sounds
that
sounds
good.
Travis
I
think
it's
good
yeah
that
we've
identified
these.
As
you
know,
like
they're,
in
the
blocking
release
column,
they
are
the
things
that
we
want
to
address
with
the
first
patch
we're
on
top
of
both
of
them.
It
sounds
like,
and
it
need
to
play
in
here
is
to
do
something
you
know
to.
A
Does
this
still
feel
accurate?
That
Travis
you're,
saying
that's
that
today,
for
a
patch
release
is,
is
reasonable,
still.
C
You
know
the
the
manager
standby
that's
ready
for
marriage.
I
just
want
to
do
some
validation
on
that
revert
and
then
the
other
exporter
issue
that
may
be
a
stretch
to
get
in
the
release.
I
know
I
want
to
think
about.
Well,
do
we
still
release
about
it
or
do
we
or
do
we
delay
till
tomorrow,
Maybe
to
get
that
fixed
too?
So
I'll
see
how
that
fix
goes
today
and
keep
the
chat,
updated.
A
C
C
Well,
a
couple
of
people
are
using
an
interim
release,
build
with
that
fix
in
it,
because
host
Network
osds
were
using
are
declaring
part
ports
that
we
didn't
intend
to
declare,
and
so
the
ports
are
conflicting
but
yeah,
so
that
issue
it's
fixed,
so
we
just
want
to
get
that
fix
out
there
and
then
also
there
was
a
couple
Prometheus
alerts,
firing
that
we
just
disabled
since
they
don't
seem
to
work
well
in
the
costumes
we
pulled
the
self
monitoring
the
latest
alerts
from
the
zeph
repo
and
yeah
a
year
ago
we
disabled
them
after
they
didn't
work
well,
and
so
we
just
needed
to
redisable
them
after
we
pulled
again
yep
so.
C
A
C
A
A
And
perhaps
next
community
meeting
sometime
in
the
near
future,
we'll
be
start
start
talking
about
1.12,
but
you
know
supporting
1.11
after
if
it's
just
released
here
is
the
primary
focus
right
now,
exactly
all
right
and
speaking
of
111
release.
Let's
go
ahead
and
move
to
the
community
topic
section
and
do
a
little
retrospective
here
looks
like
Travis.
Here
got
some
ideas
already
to
talk
about.
C
Yeah,
so
the
we
got
everything
out
and
I,
don't
think
it
was
a
ref
release.
Just
wanted
to.
You
know
talk
about
a
couple
of
things
that
came
up
so
the
blog
image
I
always
forget
to
ask
Matt
for
it
soon
enough,
and
then
it's
always
a
fire
drill
and
it's
terrible
that
it
ends
up
being
that
every
time,
because
I
don't
remember
to
do
things
in
advance
that
I
only
do
every
few
months
so
any
anyway
I
when
I
had
since
I
didn't
have
the
chance
to
get
an
image.
C
We
just
went
ahead
with
the
blog
without
an
image,
and
so
it's
just
a
boring
blog
without
an
image.
Yesterday
he
emailed
and
said:
oh,
do
we
still
need
this
and
I
said
yes,
we'd
still
love
to
have
it
I'd
still
love
to
add
an
image
to
the
blog,
but
then
I,
yes,
I,
don't
know
if
he's
planning
on
it
or
or
what
his
schedule's
like,
but
we.
A
I
think
that's
what
prompted
him
to
go
ahead
and
follow
up
yesterday,
so
yeah
I
think
that
he
should.
He
should
I'll
make
sure
that
we
give
a
little
nudge
right
now
to
remind
him
that
it's
still
a
good
thing
to
do.
C
Yeah
cool
and
then
you
know
for
a
future.
The
alpha
I
can
follow
up
with
him
too
to
say
yeah.
What's
the
best
way
to
work
with
this,
should
we
just
ask?
Is
it
more
efficient
for
him
to
say,
hey
I'll,
just
create
three
for
the
next
three
releases
and
we'll
just
save
them
right.
So
I
don't
have
to
ask
them
every
few
months
or
just
give
him
or
just
finally
giving
them
a
week
or
two
notice,
and
then
that's
fine
or
I
I
just
want
to
know
what
works
best.
A
Yeah,
maybe
maybe
one
thing
that
could
be
useful,
is
having
a
tracking
issue
for
it
and
have
it
on
the
board,
so
that,
like
as
the
release,
is
winding
down
and
we're
kind
of
moving
things,
you
know
to
the
right
side
of
the
board.
You
know
that
tracking
issue
will
still
be
there
and
that'll
kind
of
give
us
a
bit
of
a
heads
up
before
the
release
goes
out.
You
know
before
it's.
C
A
C
Okay,
yeah
so
we'll,
hopefully
yeah.
We
need
to
put
it
in
the
notes
of
things
to
do,
for
the
release
sooner
yeah,
all
right
and
then
I
was
we
were
about
to
release
on
the
Thursday
and
then
yeah
most
of
us
had
Friday
off
for
for
our
team,
and
we
said:
oh,
let's
wait
till
Monday
just
so
we're
not
out
the
day
after
the
release.
In
case
things
come
up,
so
we
delayed
till
Monday
then-
and
that
was
all
good
and
such
wanted
to
comment
on
that.
A
Beats
and
I
think
that's
even
reinforced
further
the
value
of
doing
that
when
you
know
we
have
like
the
couple
of
of
you,
know,
issues
or
potential
regressions,
whatever
that
we're
aiming
to
get
into
1.11.1
patch
release,
it's
good
that
we
have
Folks
at
hands
on
deck
and
you
know
be
able
to
field
those
and
investigate
into
them
and
come
up
with
Solutions
Etc
right
after
the
release,
so
that
was
that
was
proven
to
be
the
right
call
here.
I
think
that.
C
Yeah
yeah
I
think
it
was
good,
yeah
and
and
overall
there
I
think
I
would
just
put
in
a
plug
also
for
asking
all
maintainers
to
kind
of
be
on
deck
the
for
a
few
days
after
the
release
in
case
things
come
up.
I
think
this
time,
even
there
were
a
couple
people
out
of
PTO
or
sick
or
something
it
just
felt
like.
Where
is
everybody?
We
need
more
more
Hands-On
deck
type
of
thing,
just
in
general,
so
yeah
I
know
we're
all
busy
and
have
things
so
I'll
just
leave
it
at
that.
C
And
then
the
last
point
I
had
there
is
that
this
issue
we
had
that
blocked
clusters
with
host
networking
enabled
we
haven't
ever
had
a
host
networking
enabled
CI
test,
and
so
I
opened
an
issue
there
to
say
hey.
We
really
need
this
test
so
that
we
don't
have
that
common
configuration
blocked.
D
C
A
We'll
see
it
all
right,
any
other
thoughts
that
folks
wanted
to
share
in
terms
of
their
their
experience,
or
you
know,
retrospective,
look
back
at
the
1.11
release.
D
I
I
guess
I
have
well.
The
thought
that
I
have
is
the
reminder
recently
that
there
are
the
like
large
or
you
know,
big
GitHub
Runners-
that
we
I
haven't
really
had
a
lot
of
time
to
look
into
that
support,
nested
virtualization.
D
D
It
would
be
pretty
easily
easy
to
do
host
networking
on,
and
we
could
potentially
also
give
it
like
three
four
discs
that
I
think
would
exercise
the
failure
that
we
were,
or
at
least
one
of
the
failures
we
were
seeing.
D
And
then
I
Alexander
made
the
note
that
we
could
have
a
pre-cooked
blog
images,
I
wonder
if
that
might
not
actually
be
a
really
good
idea,
foreign
and
we
could
yeah
I
mean
I
I'm,
not
a
designer,
but
if
it's
something
where
we
have
a
pre-cooked
image
and
we
can
just
put
some
text
over
it
in
a
way
that
is
like
kind
of
guided
or
obvious
for
us,
then
that
might
be
kind
of
helpful.
B
Is
that
what
you're
mean
I
to
be
honest,
I
kind
of
just
brought
it
up
as
like
a
joke
prepared
images,
but
I
I,
don't
know
like
having
at
least
like
one
motive
or
something
where
we
in
the
worst
case
can
just
yeah
I,
don't
understand
some
text
over,
it
seems
tempting
if
we
should
forget
it
again
or
sort
of
can
just
okay
emergency
backup
would
take
it,
put
put
the
text
or
update
the
text
in
some
way,
but
I.
B
D
I
also
had
the
thought
that
the
whatever
font
file
that's
used,
might
also
be
a
good
thing
to
have
I.
D
B
But
at
least
we
have
it
written
down.
Thank
you.
We
had
the
same
thing
for
occur,
going
on
also
where
it's
like,
so
what
fun
did
we
use
again
wait?
What
color
was
the
prep,
oh
and,
and
we
just
wrote
everything
down
then,
and
it's
I
think
if
maybe
you
can
just
yeah
ask
him,
then,
for
the
information
like
that
and
like
edit
to
the
readme
of
the
repository,
the
is
it
the
media
or
the
artwork
one
I
think
it's
called
artwork
for
Rook.
That
makes
sense.
Yeah.
C
A
Yeah
definitely
with
enough
time
Matt
is
Matt's.
Skills
do
come
into
are
very
useful.
Let's
say:
okay
cool!
These
are
all
great
ideas,
guys.
So
then
Travis
do
you
want
to
go
ahead
and
move
on
to
kindergarten
Amsterdam
or
is
that
actively
soliciting
ideas
for
the
the
talk
still.
C
Yeah
well
actually,
just
finally,
yesterday
I
I
posted
in
slack
to
say,
hey
any
ideas.
What
do
you
want
to
hear
about
in
the
talk?
I
haven't
seen
any
responses
yet
so
we
yeah.
If
anybody
has
ideas,
definitely
looking
for
I
mean
we'll.
Do
the
the
standard
talk
with
background
or
overview
and
then
talk
about
the
latest
release,
features
and
things
like
that
talk
about
the
road
maps.
So,
if
there's
anything
else,
people
want
to
see
curious
to
hear
ideas
so
yeah
otherwise
I
mean
I.
C
B
Yeah
I
should
be
there,
I,
don't
know
if
you
saw
the
message,
but
we're
still
figuring
things
out
at
the
moment
for
Deepika
right,
yeah
hoping
to
have
a
yes
or
all
good,
listen.
C
A
Wait
sounds
good,
definitely
looking
forward
to
Amsterdam.
For
with
everybody,
there
should
be
great
all
right.
So
that's
everything
that
was
on
the
agenda
then
so
we'll
do
an
open
call
here
for
any
other
agenda
items
before
we
adjourn.
A
All
right
sounds
good,
all
right,
yeah,
so
once
again
thanks
everybody
for
the
efforts
on
getting
in
the
1.11
release
out
and
following
up
on
getting
the
first
patch
out
as
well
and
continuing
to
maintain
and
support
that
release
so
yeah
great
to
see
everybody
today.