►
From YouTube: 2019-04-29:: Ceph Orchestration 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
Welcome
to
the
April
29th
orchestrators
team
meeting:
we've
got
the
agenda
there
in
the
chat.
Today
it
started
I
guess
how
about
we,
the
quick
rook
update.
We
are
pretty
much
ready
for
a
rook
one
dot.
Oh
we've
got
all
the
code
changes
in
no
blocking
issues
on
the
code
side.
The
one
thing
we're
waiting
for
now
is
we're
gonna.
A
There's
a
new
website
being
worked
on
by
the
Uptown
guys
and
according
to
our
discussions
on
Friday
the
you
know
it's
about
ready
and
supposed
to
be
ready
by
tomorrow,
I'm
trying
to
review
that
today
get
the
latest
fight
to
look
at
and
review,
but
that's
the
last
blocker.
So
we
can
just
what's
the
one
dot
oh
release.
It's
like
this
big
looks
like
a
major
version
number
change.
We
want
the
new
website
at
the
same
time
about
ready
anyway,
it'll
be
exciting
thing.
Then
your
look
and
feel
do.
A
The
one
one
more
question,
though,
we're
gonna
find
the
link
for
the
manager.
There's
an
issue
open
or
a
PR
open
right
now
to
fix
the
path
or
setting
the
server
editor
I
want
to
know
Sebastian.
Hopefully
you
know
in
the
four
Nautilus
it
does.
The
manager
name
need
to
be
in
the
setting
path.
If
you
just
click
on
the
PR,
that's
a
pretty
simple
example
to
look
at.
C
A
B
B
B
B
D
A
Some
of
the
output
is,
in
the
conversation
conversation
on
that
PR
wasn't
clear
what
would
be
broken.
Actually,
though,
it
should
be
safer
to
set
that
then
yeah.
B
B
A
C
A
A
D
A
B
Yeah,
as
I
said,
some
tests
in
the
totality
test
looked
very
much
like
they
would
also
run
in
a
Cuban
Etta's
environment,
but
they
are
so
tightly
integrated
into
a
tautology
that
it
is
basically
a
rewrite
to
try
to
pull
out
these
tests
and
I
would
focus
on
some
tests
that
really
focus
on
the
integration
between
terminators
and
to
Eternity
asari
cuban.
It
is
rook
and
ancef
I.
D
B
D
Right
just
so,
you
have
some
context
for
why
I'm
asking
these
annoying
questions.
I
work
in
the
perfect
scale
team
and
we
recently
have
been
trying
to
deploy
rook
and
SEF.
You
know
on
a
reasonably
reasonably
sized
cluster
with
you
know
real
hardware
and
giving
an
example
of
the
kind
of
thing
that
we
ran
into
a
year,
but
in
the
chat
window.
D
So
there
was
this
problem
with
out
of
memory
kills
happening
quite
a
bit
because
the
cephalo
SD
memory
target
was
set
to
the
same
thing
as
a
secret
limit,
so
it
would
just
basically
commit
suicide
when
you
put
a
large
load
on
it,
though
I'm
not
saying
that
you
know,
we
have
to
do
all
our
tests.
That
way
you
know,
but
it
would
be
great
if
there
was
some
small
fraction
of
them
that
operated
it.
D
D
We
have
some
workloads
yeah,
though
including
you
know,
things
like
FIO
and
we'd.
Be
willing
to.
You
know,
help
supply
those,
but
but
we
don't
have
the
automation
ci
that
that
you
were
talking
about
you.
B
D
B
D
D
B
D
D
A
A
D
A
C
B
D
D
D
B
D
D
All
right
take
a
look
at
this.
Thank
you.