►
From YouTube: 2019-02-20:: 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
I've
shared
my
development
environment
in
the
other
pet.
It's
like
we
pasted
inside
yeah.
It's
based
on
the
centralized
version
done
by
Noah,
but
was
one
two
Bionic
haven't
yet
looked
into
getting
that
and.
B
A
A
A
Resume
the
completion
of
the
command
handler
when,
when
calling
white
that
will
then
in
the
end,
all
the
weight
function
of
the
orchestrator,
then
I'm
not
sure
about
the
error
handling,
because
if
you
are
having
multiple
Orchestrator,
completions
and
almost
and
finish
successfully,
some
of
them
finish
with
an
Internet
era,
and
some
of
them
thinner
for
example,
and
returning
a
multiple
multi
status
result
is
not
easy
and
not
that
straightforward.
C
Yes,
I
understand
what
what
you
say
last
time
when
I
was
taking
a
look
today
to
the
review
in
this
moment.
Okay,
is
it
very
well
I,
like
it?
Okay,
but
well,
I,
think
that
we,
we
should
I
think
that
in
the
case
of
the
state
or
white
method,
well,
we
should
be
only
focuses
in
make
progress
in
the
operations.
Okay
and
the
operation
has
Feeny
said
we
can
finish
the
operation
in
two
estates:
the
the
ratio
happiness
it
successfully.
C
C
What
you
are
going
to
have
is
several
different
multiplayer
pursuit
and
I'm
different,
a
star
to
suburbs.
So
I
think
that,
for
me,
the
weight
method
is
only
for
make
operations
to
progress
and
then
in
the
in
decompression
object.
Then
I
think
that
we
should
have
the
result
of
the
operation
and
if
the
client
of
the
the
consumer
of
the
completed
object,
the
responsible
of
making
any
decision
with
a
with
a
result
of
the
observation.
A
B
Question
on
the
safe
orchestra,
a
white
thing
that
you've
got
there,
that
that's
that's
a
change
to
how
the
CLI
works
right!
No.
B
A
A
Let
me
have
a
look
I
think
it's
creator
was
D
or
create
service,
I,
don't
know
it
first
gathers
all
hosts,
and
then
it
mitten
makes
another
core
to
adds
a
specific
service
and.
A
At
first
gathers
or
host,
it
called
self
that
get
host
wait
for
the
completion
and
then
afterward.
It
creates
the
OSD
and
those
during
the
completion,
and
that
does
not
really
solves
a
problem
of
properly
returning
immediately.
If
you
sue
chronically,
where
for
for
the
host.
B
C
C
A
C
But
in
my
case,
not
not
too
much
progress,
okay,
in
fact,
what
I
have
playing
with
a
with
a
duration
test?
Okay,
but
well
it
either.
Yesterday
it
was
when
the
question
of
the
identification
in
the
answer
board.
Runner
said
these
arrows
are
again
okay.
What
I
have
probably
we
are
going
to
start
with?
C
Modify
Asians
in
the
unseeable
run
a
service
in
order
to
provide
a
TLS
mutual
authentication
and
when
I
will
do
the
these
change
and
they
and
the
deletion
testing
in
parallel,
but
yesterday
I
was
with
a
potential
donor
service
most
of
the
time.
So
no,
no,
not
too
much
progress
with
with
my
path.
Okay,
perfect.
B
This
is
what
I
mentioned
separately.
Do
you
Sebastian
on
that
call
the
other
day
where,
when
deep
sea
deploys
their
fit,
deploys
the
Mons
and
then
it
deploys
the
managers?
And
then
I
was
immediately
at
that
point
telling
ya
I
was
having
deep
sea
invoke:
SEF
manager,
module
enable
Orchestrator,
CLI
and
certain.
A
B
A
B
B
For
one
second,
on
every
loop
iteration
effectively,
just
palling
SEF
manager
dumped
to
wait
until
it
becomes
available
and
then
I
can
proceed
with
enabling
and
configuring
the
module.
So
I
have
this.
This
actually
works.
Now
it
made
me
wonder
if
we,
if
there
would
be
any
sense
in
having
a
safe
manager
command
that
you
could
run.
You
know,
SEF
manager
is
available
or
something
like
that
to
know
if
it
was
completely
started
up.
B
B
Mean
the
I've
just
if
you're
using
the
dashboard,
all
of
Ceph
managers
modules
are
already
loaded.
So
the
dashboard
can
talk
to
any
anything
at
once.
I'm
with
deep
sea
I've,
only
just
the
safe
manager
team
and
has
only
just
started
and
hasn't
loaded
any
of
its
modules.
Yet
and
in
then
one
or
two
seconds
before
SEF
manager
has
completely
finished
starting
like
when,
when
you
run
distant
control,
start
staff
manager
that
returns
immediately
with
success,
even
though
SEF
manager
itself
has
not
yet
so
the
safe
manager
process
is
running,
but
it
hasn't
actually
loaded.
B
It
hasn't
finished
iterating
through
what
modules
are
on
disk
to
even
find
out
what
modules
are
available
yet
so
it's
possible
at
that
point
in
time.
I've
I've,
just
deployed
safe
manager
and
now
I
want
to
enable
some
modules
and
I
can't.
Yet
because
I
have
to
wait
a
couple
of
seconds
for
safe
manager
to
actually
have
loaded
itself
properly.
It.
A
C
B
B
B
B
B
B
Anyway,
okay,
so
to
summarize,
I
have
something
now
that
that
does
enable
the
deep-sea
Orchestrator
module
when
Dipsy
deploys.
If
it
may
not
be
my
little
loop
in
there
to
make
it
work
is
possibly
may
or
may
not
be
the
cleanest
thing,
and
so
I'll
try
a
couple
of
other
things
with
with
force
and
config
set
to
see.
If
I
can
make
that
a
bit
neater.
But.