►
From YouTube: 2020-02-04 :: Ceph Crimson 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).
B
B
And
we
have
a
welcome
here
last
week,
I
was
I
was
reviewing
unless
you
had
to
bring
the
support.
I
took.
Let
me
talk
it
to
to
crimson
and
and
and
did
appear
with,
some
I'm
to
the
cover
it
with
the
context
and
some
minor
changes
and
also
reviewed
the
fixed
core
story.
Socket
change,
vibes,
banking
and
I
think
it
close
and
I
said
from
MIT.
I
was
also
reading.
James
timber
working
progress
here
to
bring
to
implement
sister
stood
still
not
need
and
making
no
progress.
C
I
started
appear
just
a
shred
up
here
today
and
I
saw
it
already
marked
in
so
to
fix
the
PG
collection.
So
it's
it's
weird
for
the
memory
style
casting
I
fell
can
continue
if
our
bench
can
continue
and
you
get
it
without
it
and
the
fall
alien
Brewster.
It
is
for
C
equal
to
1
it
that
our
benchmark
can
continue,
but
of
our
OS
equal
to
2.
C
B
C
It's
not
the
better
beginning
H,
when
I
prayed,
the
poor,
OS,
leepu
and
I
said
the
pity
number
equal
to
126
and
then
act
out
some
the
code
continued
on,
sometimes
better
than
in
some
cases.
The
collection
is
empathy,
it's
non
pointer,
so
I
still
trace
it.
I
don't
think
it
is
the
matter
data
because
it
is
very
beginning.
I
I
mean.
C
B
C
Yeah,
okay,
so
SQL
has
the
problem,
so
sometimes
that
collection
is
empty
and
sure
if
you
sent
with
me
and
I,
can
verify
it
if
it
can
fix
the
issue
and
I
verify
the
the
C
star
you
committed
it.
It
didn't
test
out
the
CBO
says
easier.
So
I
I
tried
the
memory
store.
All
the
members
dog
cases
is
running
on
the
CPU
0,
so
I
will,
when
I,
fixed,
18,
plus
tau
issue
and
and
then
I
will
fix
that
part.
C
B
C
D
B
B
E
Hi
Thank
You
of
this
work
I've
been
a
goner.
They
tell
little
bit
that's
because
I'm
a
finals
period
now
I
just
have
some
exams
in
university,
but
before
I
started
doing
that,
I
was
finalizing
some
stuff
with
CBT
I'm
gonna
need
I
need
another
round
of
review
on
current
for
requests
with
the
CPU
cycles,
feature
and
I'm.
E
We
had
to
implement
more
functions
that
support.
Currently,
it's
only
support
in
our
VD
thing,
so
I'm
gonna
start
implementing
some
more
of
the
functions
there
for
other
plugins
and
that
was
also
discussed
with
with
Radek
I.
Think
that
I'll
start
working
on
that
on
probably
on
Thursday
I'm,
just
still
kind
of
though
some
have
some
more
files
to
finish
with,
but
I'm
just
fighting,
sometimes
lots
to
work
on
work
on
the
project
and
Dunham's
that
we
could
be
back.
A
E
B
A
Hello,
I'm
I'm
working
the
poor
request
ringing
what
mental
support
accordingly
to
the
reviews
from
commented,
a
comments
from
review.
However,
my
mind
focus
at
the
moment
is
a
memory,
corruption
issue
that
is
a
bite
our
are
hopefully
similar
to
openly
related
to
our
recovery
mechanism,
which
attempt
map,
for
instance,
recited
that
just
a
couple
of
small
tears
with
its
batteries.
That's
me.
B
B
D
D
B
D
I
will
show
the
way
I'm
investigate
in
that's
one
part
in
another
part
of
what
I'm
doing
is
starting
to
read
probing
as
Josh.
My
boss
asked
me
to
take
a
look
at
that.
The
part
of
it
is
the
part
of
scrubbing
that
turn
do
not
involve
recovery,
which
we
don't
have
yet
well
more
or
less.
That's
the
what
I
was
doing
this
week.
B
A
A
B
A
I
think
yes
about
the
first
work
of
the
well.
We
have
to
implement
to
the
PG
log
machinery
in
the
right
in
the
crimson
right
path.
First,
because
everything
is
walking
on
that
I
think,
okay
and
then
we
can
make
I
think
we
can
new
recovery
and
backfill
in
parallel.
Make.
G
Yes,
I
saw
the
C
star,
is
up
streamed
and
updated,
and
I
was
addressing
the
the
put
pull
request
and
mostly
I
just
but
just
still
song
tamesha.
The
first
is,
it
seems,
there's
a
performance,
a
question:
I,
don't
know
why?
Because
I
didn't
change,
I
think
I
didn't
change
the
the
critical
I'll
pass
and
the
second
is
this
gives
us
some
new
errors
in
that
with
the
latest
code
may
be
related
to
buffer
I'm
still
looking
at
it,
and
that's
all.
B
G
B
B
G
A
D
D
F
Think
for
a
c-store
it
will.
Such
a
thing
will
make
a
great
deal
of
sense
in
about
six
weeks.
Maybe
eight,
maybe
eight
weeks,
because
at
that
point,
it'll
be
it'll,
be
easy
for
other
people
to
contribute
all
right.
Imagine
the
recovery
work
will
have
similar
milestones
along
the
way
so
yeah.
We
should
think
about
scheduling
something
like
that.
The
merge.
F
B
Can
have
a
good
video
or
t9
didn't
review.
B
G
F
March
it'll
still
be
appropriate
I'm,
not
as
thrilled
with
the
concept
of
the
code
walkthrough,
so
much
by
that
I
hope
it'll
be
in
a
base
where
people
can
actually
review
it,
but
I
am
happy
to
talk
about
the
design
and,
more
importantly,
we
should
talk
about
how
we
want
to
coordinate
work
on
it
going
forward
since
we're
gonna
want
to
rewrite
the
major
parts
of
it,
a
lot
of
its
going
to
be
very
simple:
to
get
a
prototype
out
the
door
at
that
point,
we'll
have
to
start
discussing
where
the
performance
bottlenecks
are
and
where
we
want
to
invest
effort.