►
From YouTube: Patrick McGarry -- Ceph Community Update
Description
A
We're
coming
to
SEF
day
I,
really
appreciate
it:
I'm
Patrick,
hey
Gary
I'm,
the
community
monkey
here
at
Steph,
I
tear,
take
care
of
care
and
feeding
for
the
people
that
actually
do
all
the
hard
work.
So
if
you
have
any
questions
I'm
more
than
happy
to
answer
them,
if
it
gets
any
more
technical
than
kind
of
scratching,
the
surface,
I
will
probably
just
plug
you
in
with
someone
who
can
actually
answer
real
questions.
A
Those
of
you
that
don't
know
my
history
I
spent
six
years
at
slashdot
as
an
editor
and
then
eventually
running
the
slight
the
site
there.
After
that
I
kind
of
went
to
alcatel-lucent
tried
to
bring
them
the
open-source
gospel
and
decided
I
was
much
more
happy
in
an
open-source
company,
so
it
to
ink
tank
right
at
the
beginning
there
and
came
to
Red
Hat
with
the
acquisition.
So
that's
who
I
am
and
where
we've
been
so
now
we're
going
to
talk
about.
A
What's
going
on
in
the
safe
community,
there
is
a
ton
of
really
cool
stuff
going
on
the
Sikh
community
right
now.
The
the
nice
part
about
SEF
is
that
sage.
From
the
very
very
beginning,
you
know
more
than
10
years
ago
now
really
set
stuff
up
as
a
great
community
first
on
a
great
open
source
project
before
we
ever
started,
even
thinking
about
anything
commercial,
so
it's
got
a
great
kind
of
legacy.
You
know
one
of
the
things
I
told
him
when
I
very
first
started
ink
tank
was
you
know
this
is
wonderful.
A
You
did
all
of
the
right
things
for
community
which
I
have
never
seen
before,
and
now
all
I
have
to
do
is
not
screw
it
up.
So
so
far
we
have
not
screwed
it
up
for
the
last
couple
of
years
and
it's
been
good,
but
here
we
are
at
SEF
days
we're
definitely
kind
of
turning
up
the
heat
on
set
days.
This
year
we
I
think
we
had
10
SEF
day
events
next
year,
we're
going
to
we're
going
to
shoot
for
just
as
many
plus
a
few
more
hackathons,
so
that'll
be
good.
A
We
are
beginning
our
2016
planning.
We
don't
have
anything
solidified
4
2016.
Yet
so,
if
you
know
of
a
venue,
you
want
us
to
come
back
here
back
to
Tokyo.
China
was
another
one.
This
year,
that
was
really
good,
so
I
think
we're
going
to
be
spending
more
more
time
on
this
side
of
the
planet
this
year
that
we
did
last
year
so
that'll
be
exciting
for
us
all.
A
The
stuff
hackathon
I
mentioned
hackathons,
we
had
our
very
first
SEF
hackathon
Steph
is
10
years
old,
as
I
mentioned,
and
this
year
was
our
very
first
hackathon,
so
that
was
kind
of
exciting
Intel
hosted
us
in
their
hillsboro
oregon
facility.
We
had
about
20
people
depending
on
the
day
some
came
in
late
and
some
left
early,
but
we
had
some
amazing
performance
results
that
were
really
cranked
out
of
there.
That
was
that
was
excellent.
There
was
one
change
that
was
made
during
the
hackathon.
A
That
I
think
resulted
in
like
6x
back-end
performance
or
something
so
it
was
a
big
deal.
We
had
some
of
the
core
engineers
there
with
some
of
our
really
heavy
community
participants.
That
was
exciting,
so
we're
going
to
do
that
again
this
year,
we're
looking
at
least
probably
two,
maybe
three
hackathons
this
year.
I
think
the
current
one
that
were
we're
kicking
around
is
probably
going
to
be
a
little
bit
before
OpenStack
Dallas.
A
So
if
you
plan
on
going
to
OpenStack,
maybe
think
about
showing
up
a
little
early
and
come
play
with
us
at
the
hackathon
metrics,
this
is
another
cool
piece
of
our
community.
It's
been
around
for
a
little
while
those
of
you
that
are
familiar
with
the
OpenStack
dashboard.
These
are
the
same
guys.
It's
a
company
called
vector
gia.
There
are
really
small
company
data
services,
type
company
that
built
OpenStack,
dashboard
or
open-source
dashboards
for
open-source
companies,
and
they
do
a
really
good
job.
A
A
There's
been
a
lot
of
work
in
the
community
to
talk
about
performance,
results,
performance,
tuning
and,
and
some
of
the
hardware
configurations
that
go
around
with
that,
and
that's
going
to
be
something
we're
going
to
be
adding
to
our
metrics
dashboard.
So
in
the
future,
those
people
that
are
interested
in
you
know
hey,
let's,
let's
figure
out
what
we
need
to
do
to
build
a
set
cluster
for
a
particular
use
case.
Hopefully,
you'll
be
able
to
go
to
metric
staff
check
out.
A
Some
of
the
proof
of
concept
builds
the
commercial
builds
and
things
like
that
and
see
what
people
are
doing.
What
kind
of
performance
they're
getting
out
of
it?
You
know
so
do
I
want
an
all-flash.
Do
I
want
a
mix
of
sass
and
flash?
How
do
I
want
to
set
it
up
to
get
performance
for
what
it
is?
I'm
doing
you
know,
am
I
doing
cold
storage?
Am
I
doing
Sand
Gnats
replacement?
Am
I
doing
OpenStack
storage?
A
Hopefully
we'll
have
some
good
resources
that
will
be
able
to
get
you
well
on
your
way,
the
user
committee,
those
of
you
that
have
been
around
for
a
while,
I've,
probably
seen
the
user
committee.
This
was
a
completely
grassroots
movement,
not
started
by
ink
tank,
not
started
by
Red
Hat.
This
was
our
users
self-organizing
saying
we
want
some
infrastructure
in
the
community
to
help
us
do
things
like
meetups
shipping,
swag
stuff,
like
that,
and
now
the
most
recent
person
that
is
elected.
A
As
the
chair
of
the
user
committee,
vedotin
Hollander
he's
been
a
longtime
community
member
he's
from
42
on.
He
has
actually
started
building
a
mirror
network
for
Seth.
The
first
two
mirrors,
the
first
one
appeared
as
EU
dots,
f
com.
That
was
actually
in
vetoes
data
center,
and
now
we
have
an
au.
That's
F
com,
a
mirror
that
where
you
can
get
all
of
the
packages
everything
it'll
be
local
and
you
don't
have
to
wait
for
transit
times
or
anything
like
that
across
the
ocean.
So
good
news
more
will
be
coming
soon.
A
We're
actually
going
to
build
out
a
real,
real
mirror
network.
Now
it's
just
kind
of
an
ad
hoc
rsync
thing
that
that
lives
as
some
crazy
zombie
in
the
internet
somewhere,
but
we're
building
a
real
network.
Now
that
will
be
able
to
allow
people
to
sink
down
to
their
own
Hardware
somewhere
their
own
data
center
and
then
volunteer
and
say
I
have
a
I
have
sefa
mirror,
so
that
will
be
exciting.
Look
for
that
in
the
near
future.
This
user
committee
does
a
few
other
things.
They
do.
A
Some
of
the
release
notes
work
to
make
sure
that
you
know
people
are
getting
credit.
We
look
at
who
the
top
committers,
who
are
the
top
companies,
doing
commits,
and
things
like
that.
So
if
you
have
any
interest,
there's
a
mailing
list
always
hop
on
there
and
shadow
folks
and
see
what's
going
on
google
Summer
of
Code.
This
was
our
second
year
google
Summer
of
Code
our
first
year
we
had
two
slots
which
was
exciting.
We
have
some
good
stuff.
A
We
had
a
wire
shark
plug-in
and
a
really
good
kind
of
exploration
of
the
reliability
model
that
came
out
of
that
this
year
is
Google
of
summer
of
code,
just
wrapped
up
I
think
we
were
given
six
slots,
yes,
six
slots
and
we
used
four
of
them.
So
definitely,
if
you
know,
students
that
want
to
get
paid
over
the
summer
for
hacking
on
Seth
send
them
our
way.
A
This
next
year,
we've
already
started
kind
of
ramping
up
for
google
Summer
of
Code
2016,
and
you
know
we
didn't
have
enough
bodies
that
had
I
mean
we
had
10
or
12
different
submissions,
but
it's
a
matter
of
quality.
At
that
point,
we
need
to
have
somebody
who's
actually
going
to
be
dedicated
and
make
it
through
to
the
end.
So
you
know
dedicated
students
that
want
to
get
paid
for
coding
for
the
summer
rather
than
going
and
flipping
burgers
or
something
less
exciting,
definitely
send
them
our
way.
A
We
also
participated
in
the
outreach
program
which
is
very
similar
to
google
Summer
of
Code.
It's
run
through
the
kinome
guys
and
we
only
had
one
student
there,
but
it
worked
out
really
well.
They
helped
us
migrate,
our
wiki,
those
of
you
that
new
our
wiki
before
knew
it
was
pretty
Emma
on
the
front
but
less
than
usable
on
the
back.
So
now
we
traded
it's
not
as
pretty,
but
it
actually
gets
the
job
done.
So
that's
a
much
better
for
us.
We've
never
been
about
pretty
and
polished.
It's
always
been
about
what
what
works?
A
A
So,
if
you've
been
following
the
centos
world,
you
know
there's
the
the
new
special
interest
groups
with
new
ish
you've
got
like
the
server
cig,
the
vert
cig,
the
few
other
things,
but
the
gloucester
guys
and
the
SEF
guys
we
got
together
and
we
started
the
storage
sync
now,
there's
another
four
or
five
people
that
are
playing
those
of
you
that
don't
know
it's
basically
a
collection
of
packages
that
you
put
into
a
centos
repo.
That
makes
it
easy
to
create
something,
whether
it's
a
virtualization
box
or,
in
this
case,
a
storage
node.
A
We
actually
just
had
one
of
the
the
guys
working
at
one
of
the
CERN
Atlas
sites
in
Chicago,
name's,
Lincoln
Bryant.
He
just
volunteered
to
be
the
package
container
for
the
centos
storage
sake,
so
that'll
make
the
sentox
world
a
lot
better
force.
F
I
think
in
the
near
future,
we'll
get
a
little
bit
more
polish
on
those
packages
and
make
it
easier
to
use,
but
we're
definitely
kb
and
I
have
been
talking
a
lot
about.
How
can
we
make
the
centos
experience
force,
F
users,
more
seamless
and
more
polished
anyway?
A
So
he's
he's
going
to
help
us
do
the
Polish,
since
we
obviously
are
not
polished
people,
so
yeah
I.
Look
for
that
in
the
near
future.
If
you
want
to
get
involved,
I'm
sure,
there's
always
places
for
many
hands
to
make
light
work
stop
by
the
mailing
list.
Centos
has
a
cig
mailing
list.
I
think
we
meet
the
storage
guys
meet
once
every
two
weeks.
Just
to
talk
about.
What's
going
on
governance,
it's
here,
I
had
to
have
a
whole
slide.
A
Just
for
my
confetti
cannon,
those
of
you
that
have
been
following
along
with
the
SEF
governance
stuff
know
that
I
have
been
fighting
for
open
project
governance
for
the
better
part
of
two
years.
When
we
were
at
ink
tank,
it
was
all
about.
Ok,
let's
get
this
governance
done,
let's
get
it
out
there,
then
it
was
whoa
hold
on.
We
need
to
go
out
for
some
funding
to
pay
the
bills
and
make
sure
people
can
keep
working
on
Seth.
So
we
slowed
down
there.
We
restarted
it
up
again
and
then
the
acquisition
happened.
A
So
then
it
got
put
on
hold.
We
try
it
again
in
red
sets,
Red
Hat
said
well,
you
need
to
think
about
this
and
redo
it
so
that
it
fits
in
the
Red,
Hat
kind
of
sphere
of
influence,
and
so
now
we've
done
it.
We
have
a
soft
launch
that
just
happened
at
OpenStack
Tokyo,
which
I'm
really
excited
about.
We
have
10
board
members
as
kind
of
charter
board.
Members,
myself
and
sage
obviously
staged
as
the
project
lead.
A
Then
we
have
to
kind
of
non-commercial
members
was
Vito
from
the
user
committee
and
dan
van
der
stir
from
CERN
who's
going
to
represent
academic
interests.
Those
of
you
that
don't
know
that's
history.
Seth
was
originally
started
as
a
part
of
a
trilabs
grant,
so
it
was
the
Department
of
Energy
BOE
guys.
A
So
it
was
the
Lawrence
Livermore
sandhya
and,
let's
forget
the
third
one,
but
it
was
basically
research
and
academia
said
to
sage
like
look.
Let's,
let's
make
a
better
luster,
so
the
funny
part
is
obviously
thats.
Ffs
has
been
the
only
piece
of
staff
that
hasn't
been
quote.
Unquote
production-ready,
although
sage
did
announce
that
sefa
vest,
should
be
landing
2016.
So
we're
really
excited
about
that.
A
Yeah,
it
would
be
nice
to
have
stuff
of
s.
We've
been
waiting
a
long
time
for
that,
but
the
academic
guys
have
been
using
sacrifice
regardless
and
they've
been
very
involved
in
kind
of
hacking,
their
own
uses
of
SEF.
So
we
feared
we
figured
that
it
would
be
a
good
idea
for
the
academic
guys
to
have
representation
on
the
board,
because
they
do
kind
of
different
things
with
it
and
then
we've
got.
A
We've
got
six
people
from
different
commercial
interests
that
are
helping
to
to
kick
the
board
off
and
make
sure
that
the
the
project's
head
in
the
right
direction-
everybody's
talking
to
each
other.
And
if
you,
if
you
know
any
of
these
folks
on
this
list,
you'll
know
that
everybody
here
on
that
sheet
right.
There
is
very
technical,
very
involved
with
SEF
except
me:
I'm,
just
the
guy
who's
going
to
grease
the
wheels.
Now
we've
got
sandisk
and
then
we've
got
canonical
and
Souza,
which
was
a
big
win
for
us.
A
A
Intel
actually
donated
a
huge
pile
of
hardware
that
will
be
coming
available
to
the
community
very
soon,
it's
already
available
in
our
sand
sandy
lab
for
those
of
you
that
want
to
play
with
it.
You
just
need
VPN
access
right
now,
once
we
get
the
community
lab
up
in
our
Raleigh
data
center,
anybody
from
the
community
can
play
with
it.
Although
the
upside
is,
if
you're
interested
in
building
things
with
SEF
and
then
testing
them,
our
totality
framework,
which
is
our
testing
framework,
now
works
with
any
OpenStack
endpoint.
A
So
when
you
submit
a
fix
with
a
test,
you
can
point
it
at
like
an
OVH
cloud
in
in
Europe,
or
you
know,
dream
compute
from
the
dreamhouse
guys
and
say
run
my
tests
over
there.
Then
you
don't
have
to
wait
in
queue
with
everyone
else.
So
a
lot
of
exciting
things
coming
from
the
governance
world
definitely
been
a
long
time
coming.
A
Sef
Tech
Talks,
those
of
you
that
have
been
around
for
a
while,
probably
have
seen
our
SEF
Tech
Talks.
It's
a
monthly
talk
that
we
we
do
some
deep
dive
exploration
of
some
technical
stuff
related
topic.
These
can
be
anything
from
core
things
like
raid
offs
I
think
the
radius
was
our
first
SEF
Tech,
Talks
am
just
who's.
Our
CTL
component,
technical
lead
for
ray
dos,
came
in
and
did
a
deep
dive,
and
this
is
how
it
works.
This
is
where
we're
headed
that
kind
of
thing.
A
Josh
Durgin,
did
an
rbd
he's
the
CTL
for
our
BD.
So
we've
had
a
lot
of
the
core
guys
give
their
give
their
spiel,
but
we've
also
had
some
people
from
the
community.
Florian
haas
came
in
and
did
it
talk
about
what
our
placement
groups
and
how
do
they
work?
It's
basically
an
hour
online.
You
come
in,
you
watch,
you
can
ask
questions
and
if
you
aren't
able
to
get
there,
we
record
it
and
put
it
up
on
YouTube.
So
it's
on
the
YouTube
channel
so
definitely
go
check
that
out.
A
If
you
have
it
yet
it's
a
great
deep
dive.
Examination
of
stuff.
That's
ephrata
and
if
you
have
something
that
you
want
to
share
with
the
community,
you
know
maybe
you've
been
doing
some
sort
of
integration
or
you've
been
building
a
new
new
interface
to
liberate
us
or
something
it's
always.
It's
always
nice
to
get
community
members
to
come
talk
about
that.
So
yeah
email
me
just
just
poke
me
in
the
eye
and
I'll
make
it
happen.
They'll
be
good,
Steph,
developer
summit.
It
used
to
be
a
quarterly
thing.
A
We
used
to
release
a
new,
stable,
build
of
SEF.
Every
quarter.
People
told
us
to
slow
down.
They
didn't
want
us
to
move
quite
as
fast.
It
was
hard
for
them
to
keep
upgrading
with
us,
so
I
think
we're
down
to
about
three
stable
releases
a
year.
Every
time
we
get
ready
to
do
a
stable
release,
though
we
have
a
SEF
developer
summit
again.
This
is
on
line
one
of
the
things
early
on
in
the
SEF
timetable
that
we
decided
is.
A
We
wanted
to
make
sure
that
there
were
no
water
cooler
discussions,
force
F
every
engineer
whether
he
was
working
at
ink
tank
and
now
red
hat
or
outside
of
that
we
want
to
make
sure
that
it
was
an
equal
footing.
So,
on
our
staff,
developer
summits,
you
don't
have
a
bunch
of
internal
guys
around
a
conference
table
with
one
camera
or
something
every
engineer
has
to
have
their
own
laptop
their
own
camera
and
participate
as
an
individual.
A
It's
something
that
we
felt
very
strongly
about
for
a
long
time
is
that
you
know
we
we
like
to
make
sure
that
that
stuff
is
very
easy
to
participate
in,
despite
the
fact
that
it
is
a
deeply
technical
topic.
People
that
actually
want
to
get
involved
should
be
able
to
do
so
relatively
easily.
So
every
time
we
get
ready
to
do
a
stable
release,
we
start
talking
about
what's
coming
next,
so
obviously
we,
the
most
recent
summit,
was
the
one
talking
about
what's
coming
in
jewel,
which
will
hit
probably
februari
March
of
next
year.
A
So
once
once
we
get
ready
to
release
jewel.
Obviously,
we
will
be
talking
about
the
K
release.
Any
SEF
developer
summit
has
all
of
its
all
of
its
content
recorded.
So
you've
got
the
blueprints,
those
of
you
that
are
not
familiar
with
how
SEF
works
every
time
we
get
ready
to
do
these
SEF
developer
summits,
anyone
who
has
a
feature
or
any
kind
of
major
performance
work
or
anything
that's
going
into
SEF.
We
ask
them
to
fill
out
just
a
really
short
blueprint
this.
A
Basically,
this
is
what
I'm
doing
this
is,
who
I
am,
and
this
is
what
I
need
every
one
of
those
blueprints
gets
about
a
half
hour
session
with
sage
and
the
rest
of
the
community
to
say
what
it
is.
They're
about
sage
can
tell
them
all
right.
You
got
to
think
about
these
other
things
that
are
coming.
They
talk
about
kind
of
implementation,
details
how
it's
going
to
work
out
and
then
obviously
they
have
a
plan
to
go
forward
from
there.
A
A
If
you've
been
watching
the
mailing
list
this
year,
you
noticed
that
we
had
a
intrusion
into
SEF,
calm,
unfortunately,
part
of
being
a
start-up
also
means
that
you
don't
necessarily
have
all
the
resources
to
do
secure
infrastructure.
You
have
idiots
like
me
running
SEF
calm,
who
are
not
security
minded,
although
thankfully
it
was
not
my
fault,
it
was
actually
a
problem
with
WordPress
somebody
got
in.
We
fixed
that
cleared
cleaned
up
what
they
did,
although
we
missed
a
backdoor
that
they
left
for
themselves,
so
they
actually
had
SEF
calm
access
for
a
long
time.
A
So,
rather
than
trying
to
fix
it
patch
it
band
ate
it,
we
burned
it
to
the
ground
and
made
a
new
SEF
calm.
The
good
news
is
with
the
new
SEF
com.
We
have
a
lot
new,
a
lot
of
new
tools
and
things
that
are
that
are
going
to
be
put
into
the
new
infrastructure
and
redhat
is
running
it
for
us
and
making
sure
that
it
is
much
more
secure
and
friendly,
so
definitely
keep
an
eye
out
for
the
SEF
com
redesign.
A
A
You
don't
have
to
go
through
in
their
interface
of
any
kind,
and
it
exposes
some
really
cool
benefits.
Underneath
obviously
gives
you
better
performance,
but
it
exposes
things
like
mutable
objects
and
a
really
cool
API
that
you
can
do
some
some
fun
stuff
with
it's
the
same
interface
that
any
of
our
external
the
object,
block
and
file
interfaces
that
are
external
to
SEF.
That's
we
use
liberate
us.
So
it's
what
we
do.
A
Sef
of
s,
as
I
mentioned
landing
next
year,
very
exciting.
We've
done
a
lot
of
hard
work
on
sephis.
We've
actually
been
dogfooding
it
using
it
for
all
of
our
QA
workloads
and
everything
internally.
So
it's
come
a
long
way
since
then.
We
also
had
one
very
enterprising
individual
from
Intel
China
this
past
year,
who
at
one
point
had
a
lot
of
fixes
that
he
wanted
to
make
and
I
think
over
the
course
of
three
months.
A
He
submitted
something
like
seven
or
eight
hundred
patches,
so
it
has
come
a
long
way
thanks
to
a
small
group
of
people
and
a
much
larger
group
of
folks
that
are
that
are
hammering
on
the
work
that
they've
done.
I
threw
a
few
stats
in
here
to
look
at,
but
but
if
you
want
to
know,
what's
really
going
on,
obviously
check
out
metrics,
that's
F
object
classes.
A
This
is
one
of
the
most
underutilized
pieces
of
staff
and
I'm,
not
sure
if
it's
just
that
people
haven't
figured
out
what
the
best
way
to
use
it
is
or
if
people
just
don't
know
about
it,
but
the
the
ceph
infrastructure
actually
has
a
way
for
you
to
build
in
object
classes
to
your
storage
cluster.
So
this
this
allows
you
to
execute
code
directly
on
the
storage
nodes
themselves.
So
there's
a
lot
of
really
cool
stuff
that
you
can
do
with
this.
The
the
easiest
example
that
I
give
is
think
of
yourself.
Maybe
you're.
A
A
A
I
am
excited
for
the
day
that
somebody
builds
the
better
hadoop
with
SEF
object
classes,
so
I'm
excited
to
see
what
what
people
can
do
with
this
I
think
there
have
been
a
couple
of
SEF
developer
summit
talks,
discussions,
people
that
are
trying
to
extend
these
object
classes
into
new
and
better
things
so
definitely
check
out
those
videos
if
you're
interested,
but
but
this
is
this-
is
ripe
for
exploration
and
some
very
cool
stuff.
So.
A
We've
seen
some
stuff
from
a
couple
of
different
companies
where
they
want
to
play
with
this
I
just
think
that
it's
a
longer
path
than
something
that
they
wanted
to
do
right
way.
People
are
playing
with
it
at
this
point,
but
it
hasn't
gotten
past.
That
is
basically
what
it
comes
to
deployment.
Orchestration
I
usually
get
a
lot
of
questions
about.
How
can
I
deploy
SEF?
How
can
I
install
it?
What
can
I
do?
A
We
actually
play
with
pretty
much
all
of
the
deployment
and
orchestration
frameworks,
those
of
you
that
pay
attention
to
the
business
wire
news
and
things
like
that.
You'll
see
that
red
hat
just
acquired
ansible
this
year.
So
obviously,
I
think
the
the
set
financeable
relationship
was
always
quite
good,
but
now,
hopefully
it
will
be
even
better,
but
we
also
play
with
all
the
other
major
players,
the
chef
and
puppet.
A
Speaking
of
virtualized.
We
also
do
a
lot
of
stuff
with
docker
I
know
there
are
half
a
dozen
folks
out
there
that
are
doing
either
SEF
under
docker
or
SEF
on
docker,
so
we
were
on
there
both
ways.
Obviously,
containers
are
a
big
point
of
excitement
for
a
lot
of
people
and
SEF
is
no
different.
We
definitely
excited
about
what
that
means,
for
especially
our
build
and
test
infrastructure.
We're
looking
at
ways
to
use
that
testing.
As
I
mentioned,
we
have
a
test
cluster
right
now,
that's
that's
publicly
available
ish.
A
It's
you
have
to
get
access
with
the
VPN,
but
if
you
do
any
amount
of
SEF
work
and
you
want
some
place
to
test,
it
will
usually
throw
the
keys
out
to
whoever
needs
it.
We
also
have
an
internal
lab,
so
you
won't
have
to
share
most
of
these
resources
with
red
hatters.
It's
mostly
just
community
guys,
that's
where
the
Intel
performance
cluster
lives.
A
A
A
If
you'd
like
to
know
more
about
that
check
out
that
URL
Loic
daiquiri
use
the
guy
who
built
it
he's
been
a
SEF
guy
for
a
long
time,
finally
got
hired
by
Red
Hat
and
he
wrote
the
OpenStack
integration
for
technology
and
has
been
sharing
that
he
also
did
a
talk
at
OpenStack
Tokyo.
So
definitely
check
that
out.
I
think
that's
pretty
much
all
for
me.