►
From YouTube: 2019-07-16 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
Okay,
the
recording
has
started-
and
this
is
the
July
16th
2019
rook
community
meeting.
As
usual.
We
will
go
ahead
and
get
started
with
a
milestone
checkup,
so
we
did
a
one
dot:
zero
dot,
four
patch
release
at
the
end
of
last
week
and
so
I.
Don't
anticipate
that
there
are
issues
here
that
we
are
getting
ready
to
do
another
patch
release
for
Travis.
Is
there
anything
to
update
us
on
with
other
potential
one
dot
dot,
one
dot
o
patches
now.
A
B
A
Yeah,
that
was
that
was
nice
to
get
the
1.04
out
to
address
some
of
those,
those
lingering
quality
or
reliability
issues.
You
know
and
drive
up
the
quality
of
what
not.
Oh,
that's
great.
B
A
A
Said
yes,
awesome,
yeah,
that's
good!
It's
definitely
good
to
have
that
velocity
and
you
know
the
being
able
to
be
effective
at
getting
fixes
and
new
new
code
out
to
with
the
community.
That's
great
stuff.
So
tribes
I
know
that
you
had
a
a
1.1
roadmap,
update
and
timeline
proposal.
Do
do
you
want
to
focus
on
any
of
the
1.1
stuff
with
that?
Or
do
you
think
it's
worth
looking
at
the
milestone
a
project
board
right
now?
It's
probably
not
well.
B
Today,
yeah
so
I
took
a
quick
pass
through
the
1.1
board
and
just
tried
to
make
sure
what
was
in
there
was
somewhat
accurate
or
that
the
tickets
that
are
in
progress,
I
tried
to
make
sure
they're
in
progress
were
in
the
review
column.
So
I
did
some
basic
things
with
that,
but
it
needs
more
review
and
really
to
see
if
what
we
expect
to
be
in
their
leases
is
matching.
It
I
think
we're
yeah
we're
really
four
to
six
weeks
away
from
wanting
to
get
to
a
feature.
B
B
A
And
also
like
incorporating
the
newer
operators
that
we
have
as
well
there's
you
know
a
number
of
those
operators
that
are
being
worked
on.
That
would
be
included
in
that
one
right
like
in
like
Numa
and
I'm.
B
B
B
A
Yeah,
hopefully,
every
day,
because
there
was
a
great
start
with
that,
and
it
was
basically
a
full
implementation
that
we
started
in
ER,
a
ting
on
with
feedback,
and
you
know
getting
getting
the
right
patterns
in
place
and
some
consistency
across
the
codebase.
And
you
know
all
that
sort
of
stuff.
So
hopefully
we'll
be
able
to
drive
that
in
41.1.
I
would
love
to
see
that
in
this
milestone,
especially
since
a
significant
effort
has
already
gone
into
it,
be
nice
to
finish
that
off
yeah.
A
B
It's
related
I
mean
I,
don't
think
they're
necessarily
closely
tied
okay,
but
as
far
as
the
timeline
proposal,
the
the
thought,
in
my
mind
at
least,
is
and
talked
to
some
people
in
the
like
the
Ceph
huddle
about
it
for
early
feedback.
So
if
we
have,
we
want
to
do
a
feature
freeze
like
a
couple
of
weeks
beforehand,
the
before
we
actually
released.
B
A
A
We
would
need
to
evangelize
and
socialize
that
well
and
make
sure
that
there's
good
buy-in
from
that
and
it's
communicated
clearly
as
well,
because
we
don't
want
to
have
any
surprises
about
oh
yeah,
we're
had
a
feature
freeze
now
and
you
had
you
know
of
storage
writer
a
you
had
a
big
feature
that
you
wanted
to
get
in.
So
sorry
yeah.
We
don't
want
to
necessarily
do
that.
A
So
if
have
a
good,
no
mutation
here,
making
sure
that
there's
buy-in
it
will
be
very
important,
but
the
general
idea
of
like
basically
the
gold
here
of
your
proposal,
is
to
have
convergence
and
drive
towards
stability
and
minimize
churn
for
the
release,
and
that's
all
pretty
sane.
So
I
agree
with
that.
Yeah.
B
And
part
of
what
I
want
to
do
is
updating
the
the
roadmap
is
hey,
let's
get
in
the
roadmap,
what
we
plan
and
here's
the
timeline,
because
so
we
have
something
to
shoot
for
that.
So
it's
clear
for
everyone
involved
in
in
that
roadmap,
so
yeah
definitely
I'll
reach
out
to
others
edge
of
s.
Cassandra
just
haven't
heard
yet
on
the
roadmap
from
them
and
others
so
have.
B
C
Am
I
was
eyeing
me,
but
now
you're,
good,
okay,
yeah
I,
think
that
all
sounds
good,
we'll
see
how
it
goes
and
if
we
need
to
adjust,
then
we
can
next
release.
It
does
seem
like
with
like,
with
the
addition
of
of
you
know,
new
new
backends
I
think
it
might
be
worth
revisiting
thoughts
about
being
able
to
release
the
backends
independently,
especially
like
coordinating
like
you
know.
C
As
you
know,
back
in
X
or
Y
or
Z
like
do
I
feel
like
I,
am
at
a
feature
stable
place
where
I
can
release
it's
gonna
get
harder,
as
we
have
more
and
more
back-end,
so
I'm
a
little
concerned
that,
like
we
might
have
some.
You
know
that
are
like
okay,
I'm
ready
to
release
and
there
is
still
two
weeks
away,
but
I'm
kind
of
stuck
in
this
kind
of
code
freeze
place
or
you
know,
I.
You
know
some
other
release
kind
of
pushing
back
the
release
date
from
others
and
I
I.
A
C
B
Yeah
and
that's
yes,
we'll
work
through
it
as
it
comes
and
the
way
we
approach
this
keeping
them
all.
In
sync,
I
mean
it's
simpler
from
a
release
perspective
with
the
realities
of
having
the
different
storage
providers
needing
or
requiring
different
release
schedules.
That
seems
like
it
could
be
anyway,
a
longer-term
thing
to
figure
out
yeah.
C
C
A
I,
probably
yeah,
we
probably
want
it
to
be
something
that
we
plan.
You
know
not
as
part
of
a
release
window
like,
like
you
know,
remember,
doing
a
1
dot,
X
release.
We
probably
don't
want
to
be
planning
it.
You
know
during
that
time
you
probably
want
to
have
a
plan
set
up
and
then,
from
the
beginning
of
a
release,
you
know
entire
window
of
release.
The
lifecycle
Pro
release
like
have
that
implements
it's
at
the
beginning,
ok
cool!
It's
a
good
discussion
here,
guys
appreciate
it.
A
A
A
The
next
topic
that
I
had
on
here
was
just
a
quick
note
that
you
know
last
Friday
so
like
four
days
ago,
or
so
was
the
call
for
proposals
for
coupon
San
Diego
cute
con
North
America
here
has
collect
closed,
and
so
hopefully
everyone
got
some
submissions
in
that
you
are
excited
about
and
hopefully
we'll
get
some
talks
accepted
and
that'll
be
that'll,
be
cool.
Did
anybody
submit
any
talks
that
they're
particularly
excited
about.
A
A
A
A
Said
I
wish
that
I
was
able
to
submit
more
than
two
that's
the
maximum
I.
So
I
submitted
two
talks
as
well,
but
I
think
the
rook
or
other
topics
they
both
incorporate
of
some
rook
action
in
there
as
well
for
sure
I.
Think
I
have
a
feeling,
like
the
program
selection
committee,
is
getting
tired
of
seeing
my
talks,
though
I
think.
A
For
that
are
not
open
yet,
yes,
exactly
at
Nancy
lancaster
confirmed
that
the
maintainer
track
sections
like
rook,
intro
and
rock
deep
dive
has
not
opened
yet
it
has
a
different,
independent
deadline
from
the
regular
q
con
CFB.
So
when
we
went
my
Nancy
sends
information
out
about
that.
I'll
be
sure
to
share
that,
and
we
will
we'll
hopefully
get
both
an
intro
and
a
deep
dive
in
Shanghai
this
year.
They
had
to
cut
it
down
to
just
a
single
one.
A
They
did
not
have
space
at
the
event
for
every
project
to
get
both
intro
and
deep
dive.
But
hopefully
here
at
the
San
Diego
Convention
Center
the
home
of
comic-con,
we
will
have
enough
room
there
and
there
will
be
intros
ante
dives
for
everybody
all
the
CNCs
projects,
and
you
know
we're
we're,
definitely
looking
to
the
community
to
for
any
ideas
about
talk.
So
you
want
to
give
there
or
maybe
potentially
new
speakers
that
wants
a
co-present
or
you
know
be
they
have
a
hands
and
presenting
or
speaking
at
the
at
those
events.
A
So
you
know,
let
us
know
if
you
have
any
interest
in
participating
in
the
intro.
In
the
deep
dive-
because
you
know
we've
all
gotten
a
lot
of
turns
on,
presents
you
get
those
and
having
fresh
faces
and
more
community
involvement,
and
those
can
can
get
interesting
and
it
might
need
to
be
a
Co
speaking
thing
because
I
think
part
of
the
requirements
is
that
you
know
as
a
then
it's
a
maintainer
track.
So
you
have
to
have
some
maintainer
involvement,
but
we'll
look
into
that.
But
definitely
more
involvement
is
this
from
the
community.
Is.
A
A
A
That
he
just
shared
that
I
haven't
gotten
a
chance
to
look
at
yet
this
morning
since
we're
just
getting
started
here
on
the
west
coast,
but
any
other
eyes
on
that
would
Alex
leader
had
previously
been
looking
into
it
and
I
started
looking
into
it's
just
yesterday.
But
it
looks
like
it's
a
reproducible
scenario.
There
were
a
ton
of
information
from
logs
and
output
from
series
and
stuff
there.
So
hopefully
we'll
have
enough
information
to
get
a
little
bit
of
insight
into
this,
and
this
is.
B
A
Yet
I,
don't
think
I'd
only
seen
it
in
slack
so
far,
so
I
don't
think
that
there
is
a
github
issue
open
yet,
but
I
think
that
we're
at
that
point
where
we've
got
some
quality,
you
know
troubleshooting
information
collected
now
that
we
should
be
able
to
describe
the
scenario
described
the
repro
set
up
and
add
the
troubleshooting
information
to
get
it
tracked
in
github
I.
Think
there
were
at
that
point:
yes,
yeah
that'd
be
great,
yeah,
awesome,
yeah,
and
then
we
yeah,
then
we
can
follow
it
up,
follow
up
on
it.
A
Yeah
this
is
a.
This
is
a
good
section
here
for
the
community
topics
here.
Mixology
did
you
want
to
speak
about
that
now,
or
did
you
just
want
to
kind
of,
like
figure
out
logistics
like
being
able
to
talk
some
more
about?
You
know
how
like
what
the
process
is
for
adding
a
new
Rock
operator
or
some
of
the
foundational
information,
that's
needed.
What
do
you?
What
do
you
want
to
what
audience
do
you
want
to
talk
about
that
with
one.
D
B
D
A
A
Pager
right
up
that
you
know
kind
of
explains
what
what
you
bite
is
like
as
a
storage
operator,
sort
of
some
of
the
background
information
there
and
then
a
little
bit
about
the
proposal
of
how
the
operator
integration
would
would
proceed
or
what
are
some
of
the
steps
or
maybe
phases
of
it
to
kind
of
get
an
understanding
as
a
community
and
then
be
best
prepared
to
be
able
to
support
that
effort
is
well.
So
if
you
want
to,
you,
know,
open
an
issue
and
you
just
get
a
kind
of
a
one.
A
D
So
yoga
mat
is
basically
a
database
which
replicates
you
know
across
regions
across
ASEAN
region
highly
available,
and
it
gives
you
like
a
post
this,
like
interface
as
well,
and
there
is
Cassandra
like
interface
as
well,
but
under
the
hood
it
is
thought
to
be
as
one
of
the
angels
and
then
there
is
a
very
scalable
and
tolerant.
You
know
database
system
on
top
of
rocks,
TV.
A
D
D
Don't
know
like
there's
more
expected,
so
what
you
want
to
understand
about
my
tour
is
like:
when
is
enough.
A
D
So
well,
the
other
features
is
like
superior,
but
also
charger.
At
the
same
time,
so
you
could
technically
have
users
spread
across
multiple
geographies,
while
they'll
open
it
right
like
a
local
Utica
and
then
I
was
grossly
forget
to
other
places
so
to
speak
or
the
the
kind
of
user
that
we
have
seen.
I
would
say
and
I
divided,
but
I
would
say
a
pretty
cool.
Scum
Pisan
would
be
something
like
Google
spanner
but
out
in
the
open
source
in
a
way.
A
A
D
A
D
A
Cool
and
yeah,
and
if
there
are,
if
there
are
any
other
questions
here
from
the
community,
you
know
we
can
ask
those
now
or
we
can
continue
the
conversation
on
the
github
issue
and
little
one
pager
write-up
about
you.
The
bite
and
Brooke
operator
integration
be
happy
to
happy
to
keep
on
that.
There
yeah.
B
But
I
just
put
a
link
to
the
design
Docs
in
the
chat
there.
Just
you
can
look
at
other
design,
Doc's
been
created
but,
for
example,
design
Docs
for
the
Nuba
operator,
we're
just
merged
recently.
So
that's
an
example:
there's
a
PR
open
right
now
for
the
Apache
ozone.
Just
if
you
want
to
refer
to
what
previous
work
is
related.
D
A
B
A
B
Wanted
to
make
sure
it's
clear
to
people
that
in
the
1.1
release
a
removing
support
for
luminous
and
the
main
motivation
here
for
the
13
2.4
minimum
version
specifically
is
that
means
all
blue
store.
Osds
going
forward
will
be
created
with
SEF
volume
and
no
more
of
our
own
partitioning
scheme
to
be
created
for
4,
OS
T's
and
yes,
that's
pretty
much
it
just
to
make
sure
people
are
aware.
So
before
you
update
to
1.1,
you
would
have
to
update
your
version
of
SEF
you
mimic
or
Nautilus
and
then
upgrade
to
one
that
ones.
That'll.
A
A
That
is
that
something
that,
because
that's
I'm
trying
to
figure
out
like
if
there's
an
automated
story
there
for
the
upgrade
like
if
you
upgraded
the
rook
operator,
the
rook
staff
operators
1.1,
and
it
sees
that
you
don't
meet
the
minimum
staff
version.
Would
it
be
able
to?
You
know,
help
it
like
to
enforce
that
of?
Oh,
you
need
to
update
this
before
I
continue
to
update
any
of
the
other
staff
components
that
the
monitors
or
managers
and
stuff
yeah.
B
So
with
this
with
the
PR,
if
your
version
of
cepheid's
in
the
CR
isn't
is
not
met,
if
a
minimum
version
is
not
met,
the
operator
will
not
continue.
It'll
fail
the
orchestration
and
you
would
have
to
update
the
CR
at
that
point,
to
get
it
to
continue
past
that,
so
you
could
still
change
the
SEF
version
after
the
fact
and
get
it
going
again,
but
it
will
be
blocked
if
you
don't
meet
the
minimum
version.
Yeah
got
it
that.
A
C
A
And
then
so
in
terms
of
the
themes
of
all
the
stuff
you
know
named
releases.
You
know,
octopus
obviously
makes
a
ton
of
sense
for
all
this.
Those
themes,
but
it's
also
almost
a
little
bit
disappointing
like
it's
too
plain:
okay,
octopuses,
it's
a
simplified
sure,
but
this
one
it's
not
the
bus,
that's
boring.
Here's.
A
B
Trying
to
purge
at
all
yeah
this
one
yeah
hoping
we
don't
need
this
in
the
future,
because
this
CRD
versioning
story
is
better.
Since
was
that
k's
1.11,
I
believe,
maybe
112,
but
we
should
we're
planning
to
stick
with
v1
for
a
while
anyway,
I
hope
we
don't
need
v2
you
for
a
long
time
at
least
yeah.
That's.