►
From YouTube: 2020-02-11 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,
the
recording
has
started
and
I
don't
know
if
I'm
showing
my
screen
yet.
But
this
is
the
February
11th
2020
Rick
community
meeting
and
let
me
get
this
screen
share
started
here
there.
It
is
there
we
go
okay,
so
we
can
get
started
here
so
super.
Since
the
last
community
meeting
we
had
a
one
point,
two
point
four
patch
release
and
we
also
did
a
one
dot:
two
dot.
Sorry,
when
I
turn
a
three
and
a
one,
dot
2.4
as
well,
and
then
so
we're
targeting
a
potentially
a
1.2.5.
A
B
A
B
A
A
B
Well,
for
the
schedule
and
thinking
you
know,
we
want
to
do
it
by
mid
to
late
march
in
time
for
coop
con
really
yeah.
That
would
give
us
a
quarterly
cadence
anyway
yeah
and
as
far
as
the
Ceph
storage
provider
that
aligns
with
when
octopus
is
supposed
to
be
coming
out,
and
we
want
1.3
to
support
that.
That's
one
of
the
major
things
so
yeah
mid
kind
of
mid
March
is
the
current
goal.
:.
A
A
Yeah
I,
don't
think
Jarrod
thinking
released
named
as
exciting
was
probably
not
part
of
the
criteria.
Yeah,
that's,
okay,
all
right
cool!
That
sounds
that
sounds
good
for
all
the
milestones.
I
think
that's
all
all
sounds
reasonable.
Were
there
any
major
features
merged
that
we
wanted
to
briefly
bring
up
here?
A
A
B
A
Okay
sounds
good,
alright,
so
in
the
community
section
you
know
quick
updates
on
Cuba
and
sterday
Amsterdam,
which
will
be
less
than
two
months
now
starts:
March
30th,
like
the
last
day
of
March,
and
we
do
have
confirmed
the
Thank
You
Travis
for
the
grammar
fix
there.
We
do
have
confirmed
the
maintainer
track
sessions
for
work,
so
we
have
a
separate
intro
in
deep
dive
session
this
time
in
Amsterdam
and
then
the
CNCs
project
pavilion
sign
up.
A
B
C
B
A
D
D
A
Hit
the
road
doing
the
circuit-
oh
thanks,
all
right
cool.
So
then
the
next
item
we
have
here
is
an
update
on
the
graduation
process.
So
we
have
continued
to
be
tracking
all
the
to-do
items
in
the
this
document
here
that
just
links
from
the
agenda.
Let's
see
some
some
updates
since
last
time
is
that
we
confirmed
with
the
CNCs
that
we
only
need
one
sponsor
and
I
have
a
meeting
a
call
tomorrow
afternoon,
Asad
Ali,
the
you
know,
chair
of
these
students
of
the
kubernetes
sink
storage.
A
He
is
a
new
member
of
the
technical
Oversight
Committee
and
he
is
a
storage
expert
in
Kerber
Nettie's
and
has
known
about
the
rug
projects
for
a
while.
So
he
would
be
a
fantastic
sponsor
for
us
and
he
seems
receptive
to
it
so
I'm
meeting
with
him
tomorrow
afternoon
to
talk
through
a
sponsorship
for
the
technical
Oversight
Committee,
and
we
also
confirmed
that,
even
though
the
process
is
not
publicized
or
documented.
A
Yet
if
we
do
have
to
go
through
a
special
interest
group
first
before
presenting
to
the
technical
oversight
committee
or
before
a
vote
could
be
called
by
the
COC
at
least
so.
The
schedule
for
those
is
the
second
and
fourth
Wednesday
of
every
month.
So
tomorrow
morning
is
the
second
Wednesday
of
February
and
so
a
10:00
a.m.
Pacific.
We
will
be
presenting
to
the
CNCs
storage
sig
tomorrow
morning
and
the
expectations
have
been
set.
I
believe
that
you
know
we're
not
ready
for
a
full-on
presentation.
A
We
weren't
planning
on
doing
that
until
like
March
for
the
TOC.
But
it's
what
we're
gonna
start
the
conversation
we're
going
to
go
through
some
of
the
progress
and
this
that's
around
the
project
and
some
of
the
accomplishments
and
get
the
conversation
going
about
where
the
gaps
in
the
due
diligence
that
the
sig
needs
to
see
before
before.
Moving
on
to
the
technical
oversight
committee,
so
that.
C
A
Which
would
be
really
cool
to
not
have
to
do
two
separate
presentations
once
the
sig
and
do
it
again
for
the
technical
oversight
committee
so
right
that
would
be
cool
to
be
a
little
bit
more
efficient.
That
way,
we
will
see
how
things
go
with
the
sig
and
what
the
recommendation
is.
The
people,
people
on
the
sig
or
a
lot
of
people
are,
we
act.
We
are
friendly
with
and
have
interaction,
good
interactions
in
the
ecosystem
and
community
with
Aaron
Boyd
is
one
of
the
chairs,
I.
A
Think
of
the
steak
and
Quinten
cool
and
Alex
Kirk
da
li
Shang
yang
Lewis
Lewis
Obama
bond
from
Port
works
is
on
there
too.
So
a
lot
of
people
we
all
know
and
have
good
interactions
with.
So
you
know
it
should
be
a
productive
conversation
and
I'm
excited
to
have
that
tomorrow.
A
Let's
see
ok,
so
the
we
have
about
40
responses
to
the
user
survey,
which
is
which
is
good
but
I
I
would
love
to
you
know.
If
there
are
you
know,
high
profile
or
bigger
companies
that
we
know
are
using
broken
production.
It
may
not
have
seen
the
survey
to
specifically
go
reach
out
to
them
and
get
their
their
responses
as
well.
So
I
don't
know
if
we
have,
you
know
done
that
effort
yet
or
if
you
just
kind
of
been
putting
it
out
in
slack
into
Twitter.
B
A
They've
been
pretty
good
spread.
Actually,
one
thing
that
kind
of
its
kind
of
tough
is
that
you
know
a
lot
of
people
are
not
okay
with
sharing
that
don't
respond
to
the
server
but
they're,
not
okay,
with
their
for
their
company
name
and
stuff
being
shared
publicly.
So
you
know
40
responses,
I,
think
like
it's
about
60%
or
so.
Maybe
that
will
not
we're
not
okay,
with
having
their
name
shared
publicly,
so
that
kinda.
B
A
A
Okay,
so
that's
everything
I
know
about
graduation.
So
far,
Travis
and
I
are
gonna
sync
up
after
this
meeting
to
just
talk
through
some
more
some
more
stuff
there.
You
know
like
getting
all
of
our
stats
together
in
our
plans.
So
that's
a
not
a
closed
session
or
anything
it's
just
not
near
my
bore
some
people
on
the
general
community
call.
So
if
anybody
wants
to
hang
around
after
the
call
is
done
to
talk
graduation
stuff,
that
that's
perfectly
fine
all
right
Seb,
do
you
want
to
tell
us
about
your
OSD
changes
again.
D
D
So
the
latest
ones
are
for
so
or
where,
when,
when
running
away
these
on
PVCs,
we
also
want
to
be
able
to
have
a
dedicated
device
to
store
blue
stores
metadata.
So
now,
when
you're,
creating
creating
a
cluster,
not
only
you
can
obviously
create
PVCs
for
the
main
block
device,
but
you
can
also
add
other
PVCs
and
those
PVCs
will
be
used
to
store
the
metadata
device
for
each
individual
oasis.
So
that's
what
the
patch
is
about.
B
B
B
D
A
D
Yeah,
yes,
so
the
goal
is
as
of
1:3,
we
won't
be
supporting
leak
anymore,
so
the
supported
version
will
start
for
anonymous
and
more
precisely
for
14.2
dot
5,
because
known
address
has
been
around
for
more
than
a
year.
I
think
well
actually
a
year,
and
it
has
numerous
numerous
pinpoint
releases
already.
So
it's
already
14
to
2.7
and
yeah
I
guess
the
idea
is.
D
D
A
D
D
Guess
this
brings
an
interesting
question
is
that
we
have
to
make
sure
that
octopus
or
1/3
is
available
before
octopus
comes
out
because
I'm
sure
people
will
rush
it
once
it's
available
and
if
they
run
1
2
in
1.
3
is
not
there
there
and
then
we'll
be
like
oh
wait:
I
wanna,
I
wanna
grade
two
octopus.
D
D
B
B
D
B
I
did
add
that
yeah
just
a
note
that
I
think
the
next
highest
common
issue
in
our
integration
test.
Right
now
for
random
failures
is
the
Flex
volume
mounting
and
unmounting,
sometimes
as
unpredictable.
So
in
any
case,
let
me
see
a
side
driver
is
preferred
and
we're
moving
away
from
flex,
so
I'm
changing
the
integration
tests
so
that
we
have
most
of
them
running
on
CSI
and
only
a
few
of
them
on
flex,
so
kind
of
shift
that
the
test
matrix
to
do
more.
A
B
B
A
Well,
I
was
happy
with
where
we
landed
with
me.
Let
me
open
the
PR
in
this
edition
that
Josh
suggested
I.
You
know,
isn't
what
we
had
agreed
upon
before
when
we
had
met
and
and
I
I'm
not
sold
on
on
this
edition,
yet
myself,
so
so,
yeah
I
need
to
I,
don't
know
how
to
I
want
to
be
able
to
vocalize
or
explain
it
well,
instead
of
just
I,
don't
like
it
so
trying
to
try
to
figure
that
out
and
then
bossam
stake
on
it.
A
B
A
Engagement
from
them
to
really
drive
this
I
think
that
you
know
they
had
some
interest
and
we
didn't
really
it
didn't.
I
mean
it's
a
good
seems
like
a
nice
idea,
but
it
seemed
like
the
work
to
actually
not
just
incorporate
it
but
be
able
to
intelligently
utilize.
The
data
at
surfacing
like
that's
the
hard
part
and
I,
don't
see
anyone
who's
gonna,
pick
that
that
part
up
to
may
actually
make
this
useful
any
anytime
soon.
So.
A
A
A
B
B
C
I
can
give
a
brief
a
kind
of
a
brief
so
in
in
SEF
thirsting
concept
of
the
the
orchestrator
interface
and
wanting
to
add
add
things
to
work
from
from,
like
that.
Stuff
was
like
SF.
Dashboard
is
kind
of
the
end
goal,
and
Steph
also
has
this
concept
of
Drive
groups,
which
is
a
declarative
intent
for
what
OSD
is
to
create,
which
is.
It
doesn't
like
slot
nicely
over
top
of
rooks
current
thing,
like
current
methods
of
adding
adding
drives.
C
So
the
proposal
is
to
add
in
the
ability
to
specify
Drive
groups
in
the
CR
D
additionally
to
to
what
currently
exists
and
there's.
The
proposal
is
also
some
more
about
kind
of
the
interactions
between
SEF
and
Sykes
F
volume,
and
things
like
that
to
make
everything
work
nicely
from
a
like
UX
perspective,
but
I
think
I
think
having
the
drive
groups
added
in
is
going
to
be
I.
C
Think
it's
going
to
be
a
win
overall
because
is
declarative
and
it
is
like,
maybe
a
little
over
engineered,
but
it
is
like
it
does
allow
for
a
lot
of
specification
about
like
what
you
want
for
your
cluster
in
a
in
a
very
general
sense,
even
down
to
like
I
want
to
use.
You
know,
nvme
drives
that
are
larger
than
ten
terabytes
for
this
and
I
want
to
that
I
mean
that
seems
ridiculous.
B
B
A
Yeah
I
mean
the
more
automation
they
put
in
there
because,
like
the
you
know,
local
volume
provisioning
still
has
a
fair
amount
of
manual
steps.
As
far
as
I
understand
you
know
so
right
any
automation,
you
can
add
to
streamline
process
processes.
Even
if
it's
not
like
this
explicit,
you
know,
upstream
kubernetes
local
volume
way
is
I,
think
a
win
for
user
experience
in
my
book
right,
so
cool,
okay,
all
right!