►
From YouTube: 2019-07-02 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
A
B
B
B
A
B
C
Did
you
open
the
issue
to
because,
in
the
issue,
it's
kind
of
getting
more?
If
you
already
read
the
title,
I
think
it's
more
her
and
that's
the
point
that
the
whole
rgw
instances
the
deployments
are
getting
deleted
after
some
time
it
really
meaning
it
hold
the
object.
So
it's
not
usable
anymore
to
like,
for
whatever
reason,
I'm,
not
sure
if,
like
resetting
the
operator
was
affixed,
but
that
it's
completely
gone
in
the
first
place
is
well
a
critical
issue.
C
B
C
B
B
A
A
A
Got
it
oh
gosh,
okay,
copy
and
pasting
does
not
do
what
I
wanted
it
to
I
started,
dragging
the
column
off
the
screen,
so
just
basically
get
rid
of
that.
There
we
go
okay,
so
automation
should
work
now,
hopefully
yeah.
We
shall
see
all
right
so
will
so.
That
sounds
like
a
good
status
update
for
all
those
issues
there
and
we'll
talk
about
there
of
30
to
45,
with
a
little
bit
later
on.
It
sounds
like
two
or
will
need
follow-up,
a
Sebastian
Hahn
as
well.
It
sounds
like
for.
A
B
A
A
A
A
Oh,
you
know,
common
common
library
support
the
Brooke
framework,
and
so
Giovanni
was
working
on
that
for
the
summer
here,
and
so
once
you
have
fastest
design
of
approved
and
figured
out,
then
I'll
start
working
on
the
coding
of
it
for
the
rest
of
the
summer,
so
definitely
very
excited
to
have
him
taking
that
on.
Hopefully,
we'll
get
him
at
some
of
the
community
meetings
here
as
well,
so
we
can
get
more
involved
in
the
community.
A
A
You
know
the
approvers
and
reviewers
policy.
Is
there
it's
fully
outlined
in
the
contributing
documentation,
but
I
believe
this
has
been
used
a
couple
of
times
already
so
far,
but
people
that
are
part
of
the
approvers
list
here
can
now
merge
independently
merge,
pull
requests
into
master
that
so
that
should
have
unblocked
the
fair
amount
of
folks
for
being
able
to
search
fighters
to
be
able
to
make
independent
and
autonomous
progress,
and
hopefully
it's
all
working
as
expected
in
there
aren't
significant
roadblocks
still
with
that
policy
in
place
now,
yep.
A
All
right,
the
next
item
we
have
on
the
agenda
here
is
that
this
is
the
one
that
Blaine
added
and
it
looks
like
he's
still
not
on
the
call
here,
so
we're
just
gonna
assume
he
won't
be
able
to
make
it
today
so
Travis.
You
said
you
wanted
to
speak
through
this
then.
Instead,
since
Plains,
not
here,
yeah.
B
She
must
be
clicked
on
the
issue
there,
so
the
and
I
should
have
read
through
this
before
the
meeting,
but
I
we
had
a
discussion
about
it,
Blaine
and
I,
but
I'll
see
if
I
can
summarize.
The
the
plan
is
that
so
SEF
itself,
with
file
store,
basically
being
replaced
by
blue
store
set,
is
really
only
supporting
devices
for
OS,
D's
and
so
rook's
support
for
directories,
which
use
essentially
use
file
store.
That
also,
similarly,
should
be
deprecated
really
and
guide
equal
to
go
and
just
use
devices
with
blue
star
as
the
way
production
clusters.
B
It's
I
mean
it's
not
completely
deprecated
or
removed
yet,
but
it
really
will
be
at
some
point.
I,
don't
know
the
timeline
for
that,
but
the
as
far
as
what's
recommended,
it's
already
recommended
to
use
devices
with
boost
or
in
production
and
yeah.
The
file
store
support
will
eventually
be
going
away.
A
That
sounds
the
one
I
have
two
concerns
about
that.
The
first
one
is
that
I
actually
do
know
of
a
company.
That's
the
replicated
is
the
name
of
them,
and
they
what
their
business
about
is
about
providing
sass
services
like
bundling
them
up
and
running
them
on
premises
for
people
and
they're,
actually
they're.
A
What
they're,
using
for
storage
right
now
is
rook,
but
it's
actually
it's
directory
based
files,
the
file
store
based
so
they're,
using
that
in
production
for
their
their
customers,
and
so
you
know,
there's
at
least
that
example
of
somebody
that
would
have
non-trivial
impact
by
that,
because
you
know
that's
what
their
actual
solution
is
for.
You
know
in
customers
that
they're
selling
to
and
then
the
other
thing.
The
other
potential
concern
is
around.
A
Just
like
initial
kick
the
tires
type
of
experience
you
know
like,
maybe
maybe
you
know
blue
stone
or
devices
or
what's
supported
more
so
for
production,
but
for,
like
you
know,
just
be
somebody
without
any
access
to
raw
devices
just
being
able
to
easily
do
you
know
one
command
of
you
know,
cube
control,
apply
cluster
and
just
being
able
to
use
a
directory.
That's
you
know
on
the
local
file
system
just
to
try
it
out.
A
B
So,
as
far
as
being
able
to
just
kick
the
tires
and
try
it
out,
I
mean
blue
Stuart
can
run
in
a
file,
which
is
something
that
we
may
want
to
support
into
our
devices,
but
nobody
should
ever
use
that
in
production
type
of
things,
so
I'm
cher
concerns
about
removing
this
whole.
All
the
directory
support
cuz
there
other
things
like
people
want
to
use
LVM
and
configure
it
how
they
need
it.
You
know
the
only
way
to
do
it
right
now
is.
D
And
also
just
to
kind
of
pile
on
here,
even
if
it's
not
a
mainline
suck
scenario,
I
think
the
a
cloud
native
approach
where
it
just
runs
as
a
kubernetes
application
without
needing
a
lot
of
host
level
changes
is
desirable
and
worse.
If
you
look
at
other
projects
like
are
at
the
same
level
of
zerk,
they
support
those
scenarios
like
as
an
it
just
runs
without
needing
special
access
to
devices
or
I'm.
Sorry,
I
was
hoping
that
that
would
be
a
case
where
stuff
could
invest
in
20
forward.
D
B
B
B
C
B
A
So
that's
it
sounds
like
we
should
also
I
took
a
follow
up
notes.
You
add
a
comment
to
this.
This
issue
here
is
there?
Is
there
a
separate?
It
looks
like
it's
captured
directly
in
this
issue
right,
there's,
not
a
separate
one.
Pager
first
setting
comments
right
right,
there's
not
right
now:
okay,
yeah,
so
I
think
that
if
anybody
wants
to
add
their
opinion
to
this
issue
here,
so
that
it's
collected
in
one
place,
then
that
would
be
highly
desirable
cool
thanks
for
thanks
for
bringing
this
up.
A
A
So
we
kind
of
mentioned
this
already
here
about
back
Porter,
not
back
for
it,
and
we
mentioned
the
criticality
of
this,
how
it
affects
what
seems
to
be
the
most
concerning
here
is
that
it
affects
rgw
users.
After
already
getting
set
up
and
starting
to
consume
data.
It's
not
an
initial
blocker
to
set
up,
and
it's
actually
once
you're
running
and
have
data
stored
that
it's
then
it
then
breaks
and
I,
don't
know
if
there's
potential
data
loss
or
not,
but
there's
definitely
downtime
or
unavailability.
That
does
seem
concerning
and
so
Travis.
B
D
B
A
A
All
right-
and
then
this
was
also
on
here
about
refactor
operator
Cod-
is
that
that's
been
merged?
Is
that
I
think
that
that
got
I
think
this
good
was
on
the
agenda
late
like
after
we
adjourned
the
meeting
last
week
or
last
time
it
got
added
on
like
just
like
10
minutes
too
late,
and
so
it
looks
like
in
the
last
two
weeks
that
it
actually
got
resolved
emerged.
It's
so
doesn't
I.
Don't
think
that
this
topic
has
any
more
communication
or
discussion.
It
needs
to
happen.
A
B
A
I
noticed
that,
within
the
context
of
reviewing
the
Nuba
operator,
that
you
know
these
new
flags
here
about
running
locally
and
being
able
to
not,
it
seems
like
this
is
one
approach
to
tightening
up
the
dev
iteration
cycle
as
well,
and
so
not
just
debug
ability
but
not
having
to
build
a
package
and
deploy
that
package
or
that
image
to
the
cluster
and
starting
up
a
pod
with
it
and
all
that
stuff.
It
seemed
like.
It
was
a
very
quick
iteration
cycle
to
be
able
to
run
the
binaries
locally
on
your
Debra.
B
B
A
What's
the
latest
for
the
Ceph
dev
team
in
terms
of
your
mainline
development
experience
and
flow
I
know
that
mini
cube
was
largely
abandoned
by
a
lot
of
people
and
then
what?
What
are
you
guys
using
most
often
now
well.
B
A
A
B
C
A
C
C
C
A
It
have
a
different
focus
than
the
multi.
This
is
I
think
this
is
super
interesting
topic
here,
because
development
environment
to
developments
interation,
you
know,
efficiency
I,
think
is
super
important
to
be
able
to
enable
the
community
to
effectively.
You
know,
develop
and
create
fixes
and
few
new
features
for
Brooke,
so
in
the
chat
is
flooding
in
with
other
projects
as
well.
So
I
want
to
spend
maybe
five
minutes
on
this,
because
this
is
actually
super
interesting,
so
Sebastian
Wagner
suggested
this
one
so
we'll
get
to
that
in
a
second,
so
yeah.
A
C
A
You
do
you
think
the
tour
should
go
into
this
upstream
projects
here
row
hint
like.
Should
you
do
you
open
up
or
request
a
Noah's
project
and
add
them
there?
Well.
A
A
Sebastian
he's
off
the
corner:
okay,
people
throw
it
out
great
ideas
and
running
away
this
yeah,
okay
cool,
so
there's
it
I
didn't
know
that
there
are
all
these
other
potential
options
here.
That's
that's
super
interesting
I
know.
Some
people
really
at
least
saw
on
Mac
really
enjoy
docker
for
Mac
is
well
that
has
a
built-in
kubernetes
cluster.
That's
pretty
darn,
smooth
I
know
that
it
had
some
issues
when
you're.
A
If
you
do
things
like
you
know,
creating
RBE
devices
that
didn't
work
for
me
a
number
of
months
ago
and
I
haven't
revisited
that
so
maybe
it
might
be
a
you
know,
a
kernel
issue
thing
and
in
that
environment,
but
cool.
So
this
is
super
interesting
for
thank
you
for
everyone,
for
sharing
these
I'm
gonna.
Probably
add
these
to
the
the
agenda
doc
as
well,
so
that
there
at
least
persisted.
C
C
A
Awesome,
okay,
cool,
so
the
I
added
those
the
agenda
Docs,
so
people
can
refer
to
those
again
that
looks
like
some
interesting
projects,
maybe
to
check
out
to
see
if
they
have,
you
know
if
they
work
for
other
people's
development
flow
and
also
being
able
to
share
that
knowledge
and
the
community
to
get
everyone
more
productive
is
always
useful.
Alright,
so
that's
all
the
items
that
were
on
the
agenda
for
today
was
there
anything
else
that
any,
but
he
wanted
to
bring
up
before
we
adjourn
for
this
week,
I
put.