►
From YouTube: Ceph Crimson/Seastore Meeting
Description
Join us weekly for the Ceph Crimson/Seastore meeting: https://ceph.io/en/community/meetups
Ceph website: https://ceph.io
Ceph blog: https://ceph.io/en/news/blog/
Contribute to Ceph: https://ceph.io/en/developers/contribute/
What is Ceph: https://ceph.io/en/discover/
A
All
right,
let's
get
started
this
week,
I
just
merged
a
PR
rearranging
the
Crimson
Suite
A
little
bit
mostly,
it
moves
the
c-store
stuff
into
a
Crimson
rados
experimental
Suite,
since
it
doesn't
quite
pass
the
remaining
three
tasks
that
aren't
passing
yet
are
the
RBD
tests,
but
erratic
is
working
on
them.
It's
mainly
a
matter
of
the
additional
snapshot.
Support
I
also
pushed
a
pull
request,
for
that
adds
a
bunch
of
safety
features
for
crimson
and
non-corpson
users.
A
It
adds
an
OSD
map
flag
which
will
prevent
the
monitors
from
booting
any
Crimson
osts
if
it's
not
set
and
a
pool
flag
that
will
prevent
Crimson
osds
from
creating
pgs
from
that
pool
without
that
flag.
This
allows
the
monitor
to
also
gate
certain
features
from
being
used
on
Crimson
pools,
specifically,
the
features
that
Crimson
does
not
yet
support,
mostly
changing
the
number
of
pgs,
but
also
others
I'm
working
on
adding
a
config
option.
A
So
we
can
change
the
dash
dash
s
p
option
without
having
to
pass
in
a
command
line
argument,
since
some
of
our
tooling
doesn't
like
doing
that
so
much
and
then
I'm
going
to
move
on
to
working
on
scrub,
King
John
how's
it
going.
B
Last
week,
I
I'm
I
reviewed
the
random
block
manager
and
I'm
also
looking
at
the
messenger
V2
protocol
implementation,
and
my
current
understanding
is
that
to
enable
mod
core
we
we
will
need
to
split
the
protocol
handshake
pass
and
the
protocol
ready
pass
so
that
the
handshake
can
remain
in
the
main
core
to
look
up
and
replace
connections.
Atomically
like
before
and
after
connected
and
to
then
to
distribute
the.
B
I
o
pass
to
a
multiple
calls,
including
the
message
in
code
code
and
the
event
dispatching
and
I
think
in
order
to
do
that,
probably
need
further
factors
and
cleanups.
B
That
seems
reasonable.
For
me,.
A
All
right,
junior.
C
D
D
Are
working
in,
say,
5020
I
found
this
is
the
duplicated
or
Humanity
after
I
after
the
legislative
compiler
to
the
latest.
Writing
if
I
met
the
same
message
for
your
profile
before
during
the
mechanic
FS
process
about
the
openness
Excel.
D
A
Got
it,
can
you
add
the
GCC
version
you
were
using
I
haven't
had
time
to
look
into
it.
It
might
be
a
good
thing
for
you
to
look
into,
though.
D
Okay,
I
typed
it
oh.
A
E
Last
week,
I'm
working
on
the
articles
memory
stuff,
so
the
pr
is
provided
and
memory
store.
Can
you
use
the
article
on
this
his
dog
still
the
that
they
use
phone
call
I,
try
to
use
the
partition
or
or
multiple
device
to
the
to
give
it
to
system,
but
the
there
is
I
think
there
are
little
bugs
and
I
need
to
debug
on
it.
A
A
So
your
change
makes
it
so
that
the
OSD
refers
to
a
c-star
sharded
instance.
When
talking
to
the
store
right.
E
E
Oh
so
no
I
mean
the
80s
store
and
the
Sea
straw
still
use
one.
One
call
currently.
A
A
Yeah
again,
I,
don't
think
that's
actually
true
for
alien
store.
The
alien
story.
Implementation
is
actually
perfectly
happy
to
dispatch
dios
from
any
distort
C
star
core,
but
again
I
don't
want
so
I
I
don't
want
the
OSD
to
have
to
go
through
a
c-star
started
instance
to
get
to
the
backing
store
because
for
c-store,
for
example,
we
will
need
to
do
something
much
more
complicated
than
than
this.
A
A
A
Oh
no,
oh
that's
what
I
did.
Okay,
that
makes
a
great
deal
of
sense.
I'll
copy
it
back
over,
wait
which
PR
was
it.
B
A
A
Oh
thanks
I
appreciate
it
that
explains
that
mystery.
Okay,
did
you
have
anything
else.
A
All
right
run
and
old.
A
F
This
week,
I've
finished
debugging
the
intra
LBA
tree
pointers.
I'm
now
trying
to
add
pointers
between
LBA
leaves
nodes
and
The
Logical
extents.
That's
all
for
me.
A
Okay,
so
we're
so
the
I
expect
Reef
code
freeze
to
be
something
like
the
end
of
December,
not
exactly
sure
when,
but
it'll
probably
be
the
end
of
December
or
the
beginning
of
January.
A
My
goals
for
Reef
for
Crimson
from
this
point
until
then
so
for
the
next
two
months
are
to
finish
the
snapshot:
implementation
to
stabilize
the
Crimson
rados
test,
Suite
to
make
sure
that
stuff
ADM
install
a
documentation,
that's
all
squared
away
and
to
get
as
much
as
scrub
done
as
possible.
A
So
to
the
extent
that
you
guys
can
contribute
to
those
goals.
That'll
be
good,
but
the
other
stuff
is
important
as
well.
In
January
the
focus
will
shift
back
to
Performance
and
c-store
type
stuff,
but
I
think
it's
important
for
Reef
to
try
to
get
a
stable,
functional
version
of
crimson
that
people
can
do
performance
testing
on.