►
From YouTube: OpenStack Austin 3/20/13
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
B
B
So
what
we're
going
to
talk
about
with
season
studio,
so
susan
studio
is
a
tool
that
we
built
it's
based
off
of
an
open
source
project
called
kiwi.
I
don't
know
if
you've
heard
about
it.
I
had
not
heard
about
it
until
I
actually
started
working
that
had
to
be
paid.
I
was
paid
to
know
about
it.
B
The
kiwi
is
a
cool
tool.
It
actually
surfaced
from
our
point
of
sale,
environment
and
our
thin
client
environment,
where,
in
those
cases
you
have
to
be
able
to
very
you
have
to
have
a
very
tightly
controlled
granular
process
for
building
an
image
that
can
then
be
mass
deployed
over
a
large
number
of
systems
sound
familiar.
So
we
we
worked
on
that
code
for
our
our
pos
systems,
and
so
a
couple
years
ago
we
actually
had.
Basically,
we
have
like
a
hack
week
that
happens
at
our
labs.
B
Utah
and
some
of
the
guys,
I
believe
it
was
a
number
sat
down
and
wrote
an
initial
thing
as
soon
as
the
studio
essentially
put
a
web
front
end
on
top
of
the
qe
project.
So
what
this?
What
this
does
then
is
this
gives
you
a
nice
pretty
web
interface.
You
can
go
into
and
you
can
very
quickly
and
easily
build
your
own
customized
linux
workload
for
going
for
deploying
out
to
a
large
variety
of
different
platforms.
B
Some
of
this
actually
by
coming
in
and
showing
you
what
looks
like
so
there's
two
versions
of
susan
studio,
there's
a
paid
version
that
can
be
installed
in
your
data
center.
It's
called
susan
studio
on-site,
I'm
going
to
focus
on
the
free
version,
since
this
is
a
community
event,
I
don't
want
to
be
a
marketing
drone.
B
I'm
going
to
talk
about
susastudio.com,
they're
they're,
essentially
the
same.
The
the
free
version
is
more
kind
of
our
sandbox
and
development
area.
So
there's
gonna
be
new
features
introduced
there,
we'll
test
them
out
before
they
get
merged
into
the
commercial
product.
So
there's
lots
of
new
things
that
you
can
do
inside
of
here.
B
So
first
thing
we're
going
to
do
is
after
I
log
in.
I
can
see
a
whole
bunch
of
different
things
that
I
have
built
in
the
past.
You
can
see
that
I
built
a
image
from
my
kids
to
do
it.
It
has
an
educational,
three
called
free
that
so
it
launched
this
image
and
it
will
say
a
live
cd
and
you
just
boots
up
off
of
it
loads
up
all
the
stuff
for
the
kids
and
they
get
to
play
with
that's
their
heart's
content.
B
I've
got
a
long
time
ago.
I
was
playing
with
the
essex
release.
I
haven't
touched
this
one
since
april
of
last
year,
so
it's
been
a
little
while
and
a
bunch
of,
and
some
other
things
that
I
built
this
one
I
just
built
a
couple
of
months
ago
for
the
dell
tech
center
campus.
B
B
B
A
Yeah,
I'm
just
I
I
having
worked
with
that,
what
that
installs
is
much
less
than
the
minimal
install
that
you
can
get
out
of
a
cd
install,
so
the
package
there's
differences
in
the
packages
and
the
dependencies.
B
Yeah,
I
want
to
say
it's
like
540
megs
worth
of
stuff,
that
it's
a
very
small
and
thin.
B
B
You
can
download,
but
if
you
actually,
if
you
I'll,
show
you
right
here
well
in
the
next
one,
so
besides
just
the
juice,
we
do
have
just
a
server
type
template
a
desktop
template
for
kde
or
gnome,
and
then
we
also
have
you
can
base
it
off
of
the
enterprise
version.
If
you
want
to
do
the
commercial
stuff
and
have
the
the
support
and
all
that
kind
of
fun
stuff,
let's
go
ahead
and
build
one
here
off
of
open
season,
12,
twelve
three.
B
If
you're
gonna
have
something
that
the
end
user
is
gonna,
be
spending
time
on
in
some
sort
of
graphical
fashion,
say
you
can
do
like
a
vpn
endpoint
for
being
able
to
provide
a
virtual
desktop
that
somebody
can
log
into
for
the
most
part
if
you're,
especially
if
it's
like
a
web
type
application
somebody's
accessing
through
an
api
or
for
a
web
page
or
something
like
that,
you
wouldn't
need
much
more
than
that
juice
template,
especially
if
it's
something
where
it's
not
going
to
take
a
lot
of
care
and
maintenance.
C
B
This
is
one:
that's
that's
kind
of
surfaced
in
cloud
computing.
So
if
you
have
a
pet,
you
give
it
a
name,
you
nurse
it
back
to
health.
When
it
gets
sick,
you
love
it.
The
member
of
the
family
is
wonderful
right.
If
you
have
cattle
one
of
it
gets
sick.
You
take
it
behind
the
woods,
then
what
should
you
shoot
it
right?
So
when
we
made
it
back
to
what
and.
B
Back
exactly
so
so
the
answer
to
that
would
be.
If
I
have
a
web
type
service,
that
is
a
cattle
type
server
that
is
just
gonna,
be
if
it
goes
down,
then
I'm
just
gonna,
kill
it
and
send
up
a
new
one.
Then
I
would
just
leave
it
in
juice
type
environment.
If
you're
gonna.
C
B
To
do
day-to-day
administration
task
server
and
you're
gonna
want
to
you're
gonna
wanna.
Have
things
like
you
know,
check
the
fsck
right
you
wanna
be
able
to
do
a
file
system
check
on
it
or
you
know.
A
Just
bare
minimum
system
tools.
B
A
Particular
particular
appliance,
in
other
words,
if
you're
using
it
essentially
as
a
virtual
private
server.
You
would
set
it
up
with
something
more
exactly
what
you're
doing.
If
you're
treating
a
cloud
appliance
as
a
cloud
appliance
right,
then
you
would
probably
start
with
just
juice.
Yes,
that
is
correct.
Gee.
Did
you
have
that
discussion
last
night?
B
Over
there,
it's
almost
done
all
right,
so,
let's
create
all
right.
So
first
off
we
can
see
some
well.
Some
of
us
here
in
the
front
can
maybe
see
this.
I
don't
know
about
you
guys
in
the
back.
I'm
gonna
have
to
put
on
right
here.
Okay,
so
you
can
see
between
the
speckles
right
here
and
it
says
180
100
megabytes
of
download.
For
my
this
is
this
is
just
a
juice.
We
haven't
done
anything
with
it.
B
So
180
megabytes
of
download
580
megabytes
uncompressed,
it's
a
little
bit
smaller
with
the
enterprise
version,
but
the
580
megabytes
uncompressed,
that's
what
the
actual
bare
minimum
this
size
this
can
go
on
to,
and
this
is
actually
really
neat.
So
we
talked
about
earlier.
C
B
B
B
I've
done
you
know,
typically
for
for
with
these,
if
I'm
using
the
qcal
2
image
that
has
compression
built
in
stuff,
like
that,
I'm
looking
at
two
to
three
hundred
megs
that
gets
passed
around
the
pipe,
which
is
fairly
quick
and
easy
you're
talking
about
seconds
to
transfer
that
from
glance
over
into
your
server
and
then
it's
just
the
I
o
availability
to
be
able
to
write
that
onto
the
vm
and
spin
it
up
so
you're.
B
Looking
at
having
a
fully
functional
vm
within
30
seconds
to
a
minute,
maybe
a
little
bit
more.
If
you're,
I
o
bound,
if
you're,
just
assuming
that
q,
which
is
not
already
over
there,
because.
B
Again,
with
just
the
767
configuration
options
for
openstack
or
probably
more
now,
okay,
so
that's
just
this
size.
So
let
me
try
and
show
what
you
said:
yeah
so
putting
it
before
use
case.
That
means
that
I
want
to
have
seven
different
server
flavors
with
different
whatever
informal
disk
sizes.
I
don't
actually
have
to
create
seven
different
images,
because
the
image
is
smart
enough,
that's
correct!
B
All
right
so
software,
so
there's
a
couple
different
ways.
We
can
do
this,
so
we've
got
all
the
all
the
all
the
packages
from
the
distribution
here
so
say
we
wanted
to
do
a
port
scanning
appliance.
I
can
do
a
search
for
nmap.
B
If
I,
for
whatever
reason,
didn't
know
anything
about
it,
I
could
click
on
here
and
get
some
details
about
the
package
we
have.
You
know
the
typical
groups.
We
also
have
what
are
called
patterns.
A
pattern
is
a
group
of
one
or
more
packages
that
provides
a
specific
service.
So
you
can
just
say
you
know,
set
me
up
as
a
lamp
server
or
set
me
up
as
an
oracle
server,
and
it
will
do
the
different
pieces
that
needs
to
get
you
ready
to
install.
B
There's,
there's
a
lot
of
intellectuals,
new,
surgical,
crowbar,
I
don't
think
crowbar
is
in
here
as
part
of
the.
B
We
actually
from
so
our
sousa
cloud,
our
commercial
product,
it's
openstack
base.
We
include
crowbar
in
there
and
we
have
a
tool
called
the
open,
build
service
that
can
talk
about
a
little
bit.
It's
actually
fairly
interesting
for
developers
that
we
have
builds.
That
is
what
we're
using
for
building
the
crowbar
pieces
and
stuff
like
that
program
for
usage,
and
so
you
can
go
to
the
open,
build
service
and
find
stuff.
B
So
so
you
can
pull
those
packages
that
are
part
of
the
distribution
you
can
also
see
up
here.
You
have
the
ability
to
add
repositories.
So
if
you
have
your
own
custom
repositories
or
individual
rpms,
you
can
add
those,
and
then
you
can
make
them
be
part
of
your
appliance
and
your
life
is
wonderful
and
your
boss
gives
you
a
raise,
and
everybody
acknowledges
that
you're
awesome
all
right.
I
thought
that
was
pearl.
B
B
Okay,
so
we're
gonna
start
customizing
it.
So
we
can
set
our
locale
time
zones,
the
firewall
network
settings
add
in
users
and
groups.
If
you're
gonna
do
a
local
authentication.
You
can
do
that
right
here,
scooch
downs
that
people
can
see
over
the
top.
For
some
reason,
people
get
a
big
kick
out
of
this.
You
have
the
ability
to
upload
your
own
custom
logos
and
backgrounds.
B
Yes,
and
no
so
it
will,
it
will
configure
the
first
it'll
configure
the
first
one
by
default,
just
as
a
standard
template,
it
will
still
recognize
the
rest
of
the
of
the
interfaces,
and
so
you
can
actually
you
can
do
scripting,
which
we'll
get
to
in
just
a
second.
You
can
do
scripting
where
you
can
detect
if
there's
multiple
mixtures
and
so
set
up
the
fill
over
the
first
time.
A
B
That
so
straight
out
of
the
box,
so,
for
example,
I
you
know
I
I
do
a
lot
of
the
cloud
stuff
for
our
team
and
so
crowbar
is
a
wonderful
tool
but
they're
quite
often
instances
where
you
have
to
go
through
and
rebuild
it
for
whatever
reason.
So
I
actually
built
a
image
in
studio
that
had
crowbar
installed
on
there
where
and
it
was
on
a
preload
isos.
But
I
could
basically
preload
isil
had
said:
is
a
disk
image
wrapped
up
around
in
a
bootable
iso?
B
So
you
feed
off
the
iso
and
we'll
say:
are
you
really
sure
you
say
yes
and
then
it
just
drops
the
image
onto
your
whatever
system
that
you're
you're
on,
and
so
I
set
it
up
so
that
I
could
boot
off
that
that
crowbar
preload
iso
and
then
fire
off
a
script
when
it
gets
done.
That
would
then
go
create
all
the
directory
structures.
I
need
rsync
copy
and
the
ssh
takes
all
the
files
into
place
and
then
get
everything
set
up.
B
So
I
could
in
theory,
then
just
run
the
install
install
script
for
actually
setting
up
chef
and
doing
all
the
back
end
stuff
and
crowbar.
So
you
can,
you
can
get
very
granular
in
there
and
so
for
my
by
default
it
only
it
will
set
up
on
the
first
on
the
first
ethernet
card
in
h0,
but
there
are
I
I
haven't
set
so
that
will
actually
go
in
and
add
in
stuff
configurations
for
the.
B
B
You
can
do
it,
it's
not
as
straightforward
as
just
assigning
an
ip
to
to
whatever
and
telling
it
hey
we're
going
to
multipath.
So.
B
B
Click
the
button
to
add
it
all
right
server.
So
if
you're
doing
postgres
or
mysql,
you
can
check
the
box
right
here
and
you
can
actually
tell
it
to
upload
data
files.
B
So
if
you
have
a
mysql
dump
file
or
mysql
users,
you
can
just
add
them
in
right
here
and
those
will
be
just
sucked
into
your
blinds.
Pretty
cool
stuff.
B
Let
me
take
that
off
that
nag
warning
over
there
on
the
side,
so
at
desktop
you
have
the
ability
say
you
want
to
set
up
a
kiosk,
build
a
kiosk
image.
You
can
tell
it's
automatically
log
in
launch
firefox
in
kiosk
mode
and
do
whatever
launch
whatever
application.
It
is
that
you
have
you're
running
through
there.
B
All
right
getting
into
some
of
the
appliance
stuff.
If
it's
going
to
be
a
virtual
machine
in
the
case
of
zen
and
vmware,
it
will
actually
create
a
config
file,
the
vmx
for
for
vmware
and
the
xml
file
for
zen.
So
you
can
tell
it
how
much
memory
this
size?
If
it's
going
to
be
a
disk
image,
how
much
of
a
swap
partition
do
you
want
to
do?
You
can
set.
A
B
On
there,
you
can
do
lots
of
different
things.
If
I
was
in
the
enterprise
version
of
this,
there
would
also
be
tie-ins
where
I
could
actually
tell
to
integrate
with
cloud
and
integrate
with
a
tool
we
have
called
susan
manager
lets.
You
manage
manage
souza
as
well
as
other
rpm
based
distributions
that
shall
not
be
named
all
right
and
then,
once
you
sign
up
for
all
that
kind
of
stuff,
you
can
do
run
custom
scripts.
You
can
run
a
script
at
the
end
of
the
build
process.
B
You
can
tell
it
to
run
whenever
the
appliance
boots
and
that
can
be
either
you
can
see.
We've
got
a
little
if
statement
right
here.
Thank
you.
If
you're
inside
of
the,
if
statement,
then
we'll
do
it
the
first
time
it
boots?
If
you're
outside
of
that,
if
statement
then
we'll,
do
it
every
single
time
it
boots.
So,
just
depending
on
what
you,
what
you
want
to
do
with
your
life,
you
can.
B
You
can
go
through
and
customize
this
quite
a
bit,
so
you
can
see,
there's
quite
a
few
options
for
being
able
to
customize
this,
according
to
your
needs.
The
next
step
up
for
for
that
customization
is
overlay
files.
So
say
you
have
you
know
your
security
settings
for
pam
and
for
okay,
you
know
you're
using
sc
linux,
because
you
hate
yourself
or
whatever
you
can
drop
in
those
config
files.
B
Here,
if
you
have
tarball
based
applications,
you
can
upload
the
tarballs
or
zip
zip
files
tell
it
where
to
extract
them
to
so,
for
example,
on
my
sousa
studio
on
site
that
I
I
run
off
my
laptop,
I
actually
have
an
oracle
appliance
where
I
uploaded
the
the
multi-gig
oracle,
install
files
and
everything
like
that
had
it
all
set
up
so
that
the
oracle
application
is
there
and
ready
to
go,
and
you
can
do
things
like
moving
stuff
around
you
can
set
permissions.
You
can
set
ownership,
you
can
do
all
sorts
of
fun
stuff
questions.
B
There's
the
version
control
number,
so
I
was
talking
earlier
in
the
difficulties
you
have
like
your
multiple
gold
master
images,
and
then
you
have
your
multiple
applications
and
you
have
to
you
have
to
maintain
that
whole
compatibility
matrix.
B
What
studio
allows
you
to
do
is
take
those
applications
and
merge
them
directly
into
the
gold
master
image
and
then
version
control
that
as
an
integrated
whole,
so
your
developers
can
can
produce
version
001
right
then
virtual
one,
you
just
simply
hand
the
disk
image
over
to
qa
qa
can
fire
up,
however
many
instances
they
need
to
it.
They
test
it.
They
go
back
and
forth
version,
003
gets
blessed,
and
then
you
just
have
a
disk
image
for
oo3
that
you
run
out
to
production.
It
greatly
simplifies
that
whole
process
of
your
dev
test
qa.
B
It
also
simplifies
the
process
of
getting
rid
of
that
entire
compatibility
matrix
right,
because
it's
baked
right
in
right
and
you
just
you-
just
maintain
your
versions
of
this
specific
appliance.
So
is
that
that
a
failed
string?
Excuse
me
it's
a
plain
old
string
or
is
that
a
real
version
or
the
semantics.
B
No,
it's
just
basically
a
number.
It
will
it'll
bump
the
last
number
automatically
every
time
you
build
so
there's
no
other
lock
there
just
so.
There
are
patterns
that
you
can
establish.
No,
we
thought
about
that.
But
basically,
while
there
are
generally
accepted
practices,
there
are
about
10
million
different
deviations
from
those
those
and
that's
probably
being
generous,
and
so
we
just
it's
just
essentially
a
straight
string
field.
B
So,
but
you
can
see
in
here
we
can
target,
builds
for
a
large
variety
of
platforms.
So
we've
got
our
hard
disk
image.
You
can
build
a
live
cd.
You
can
build
a
zen
guest,
a
preload
usb.
So
I
talked
about
that
preload
iso.
You
can
do
the
same
thing
on
a
buildable.
Usb
drive
you
boot
off
the
usb
drive
it
says.
Are
you
sure
you
really
want
to
do
this?
Yes
and
we'll
wipe
out
your
first
disc
and
system
drop
image
on
there
and
launch
it
up?
B
Yeah
you
can
build
it
for
hyper-v,
not
necessarily
gonna,
be
popular
in
this
room,
but
for
some
people,
that's
very
important.
You
can
build
that
pre-loaded
iso.
B
You
can
build
a
vmdk
which
is
consumable
by
vmware
virtual
box
and
of
course,
kvm
can
consume
that
as
well,
and
then
we
have
a
cubica
2
image
that
we
is
for
we're
labeling
that
soon
you
know
that
most
of
the
components
are
on
kvm,
and
so
that
gives
the
ability
to
do
that
there
and
then
you
can
also
build
an
image
for
azure
and
then
because
they're
not
on,
because
I'm
not
on
the
enterprise
version
on
the
enterprise
version.
You
can
actually
do
there's
an
amazon
ec2
as
well.
B
So
you
could
build
all
these
different
cloud
platforms
openstack
as
well
as
ec2,
as
well
as
azure
all
from
a
single
config
right
here,
and
then
you
just
push
it
up
to
the
appropriate
place
in
the
case
of
azure
and
ec2.
You
actually
can
provide
your
credentials
inside
of
here,
and
so
once
once
it
gets
built,
you
just
click
a
button
and
it
will
come
over
and
say:
okay,
where
do
you
want
to
launch
it
up
dude?
So
you
pick
your
your
location.
B
B
Way
we're
doing
integration
with
with
susan
clouden
with
openstack
in
general,
so
we
introduced
recently
and
hopefully
they
haven't
changed
things.
B
All
right,
we
introduced
the
concept
of
a
web
hook,
so
a
web
hook
is
essentially
whenever
you
complete
a
build,
it
will
send
a
notification
to
a
an
arbitrary
point
that
you
specified
here
and
say:
okay,
I
built
this
appliance
with.
It's
got
this
name
and
this
version
number
and
it's
got
this
download
url,
and
this
is
the
build
target
that
we
we
chose
and
you
can
have
a
listener
sitting
at
that
location
that
listens
for
that
message
and
then
turns
around
and
performs
an
action
on
it.
B
So,
for
example,
we've
got
a
ruby
script
that
can
be
that
sits
there
and,
for
example,
you're
on
the
devil
team.
Well
I
built
this
over
in
your
dsc.
We
we
did
so
you
build
the
images
in
studio
and
automatically
post
a
notification
and
pulls
that
down
sticks
in
the
advanced
server
and
it's
automatically
available
for
anybody
in
the
cloud
to
consume
so
you're
telling
us
we
should
memorize
this
api
key.
B
There
we
go:
oh
look,
a
new
key
to
memorize
I'll
randomize
it
after
we
get
done,
but
so
so
that
gives
you
the
basic
idea
of
what
you
can
do.
Oh
you
know
what.
B
Produced
image
so
sorry
I'm
not
a
suicide
guy.
So
sorry,
for
example,
with
centos
for
the.
A
B
Sense,
you
would
have
to
set
that
up
ahead
of
time
so
in
the
image
so
well
on
the
live
cd.
It
will
it'll
give
you
the
ability
to
drop
it
in,
but
it
will
drop
in
the
image
of
your
system,
just
not
a
not
flight,
cd
version.
This
is
doing
it.
Does
everything
this
everything
this
is
generating
is
not
going
to
be
using.
What
we
call
auto,
yes,
which
is
our
our
quick
start
kickstarter,
but
everything
these
are
doing
is
all
image
related.
A
If
you
included
the
appropriate
packages
in
your
build
and
your
configuration
in
your
configuration
management
system,
whether
that
was
shaft
or
puppet
or
whatever
you
could
put
onto
the
image
yeah,
the
ability
to
at
first
boot
go
run
chef,
client
or
puppet
agent
or
whatever,
yes
and
finish
configuring.
The
machine.
B
Yes,
you
could,
although
that
kind
of
breaks
the
model,
so
so
there's
there's
two
different
models
right.
There's
the
model
of
I
build
everything
up
ahead
of
time
so
that
essentially
I
have
a
usable
machine
in
the
image.
So
as
soon
as
I
launch
up
that
image,
it's
ready
to
go.
It's
it's
got.
You
know,
say
it's
oracle
server!
It's
got
the
oracle
server
installed.
It's
got
all
the
user
accounts
set
up.
It's
got
all
that
kind
of
stuff.
B
It's
got
a
script
that
goes
on
first
through
that
joins
it
up
to
the
ldap
directory,
and
you
know
all
that
kind
of
all
that
kind
of
stuff,
so
that,
ideally
it's
a
fully
self-contained
image
that
you
launch
it
up
and
once
it
finishes
booting
it
is
a
fully
functional
piece
of
equipment.
B
Then
there's
the
there's,
the
other
side
of
it,
which
says
just
launch
up
an
operating
system
and
then
we'll
do
come
in
through
chef
or
puppet
or
whatever
way
to
come
in
and
and
or
see
it
and
manipulate
it
to
get
it
to
a
point
of
of
where
I
wanted
to
be.
Yes.
Well,
ideally,
it's
and
what
you
hear
from
the
chef
guys
is
the
balance
of
the
two.
Yes
really
it's
that
it's
two
panda
sandwich
and
you
can.
You
can
do
a
hybrid
approach.
A
B
Just
saying
that,
ideally,
you
would
get
as
much
done
in
the
in
the
image
as
possible
for
the
simple,
the
simple
point
that
you
want
as
much
as
possible
to
remove
complexity
and
to
have
the
fewest
number
of
moving
parts.
That's
the
biggest
sell
to
doing
an
image
base
versus
just
you
know
letting
chef
or
pepper
do
it
all
together
is
that
you
know
you've
still
got
processes
going
and
relying
on.
A
External
resources-
and
that
makes
sense
to
a
point,
but
if
you're
doing
some
sort
of
a
configuration
change
that
involves
changing
a
configuration
file
on
the
server
you
don't
want
to
have
to
go,
generate
a
new
image
to.
B
Then
sorry
I
misunderstood,
but
yes,
absolutely
you
can
once
it
is
launched
up.
It
is
a
full-blown
linux
machine
that
can
do
all
the
stuff
that
linux
can
do.
So
if
you
want
to
manage
it
through
puppet,
if
you
want
to
do
all
that
kind
of
stuff,
you
don't
necessarily
have
to
do.
You
know
you
could
have
you
know
periodic
releases
through
the
through
studio
and
then
in
between
maintain
it
with
you
know,
whatever
your
tool,
yeah,
do
you
have
a
plan
for
vapor
vagrant
yeah,
it's
getting
a
little
popular
as
you
say.
B
Well,
if
you
have
virtual
box,
you
can
use
microsoft
and
there's
I'm
not
familiar
with
vagrant
to
be
able
to
answer
that
question.
Okay,
so
for
the
ignorant
amongst
us,
can
you
enlighten
us?
Vagrant
is
now
like
premiere
tool
for
doing
single
box
development,
virtual
machine
automation.
B
You
can
use
popular
shopper
scripts
and
you
set
up
what's
called
a
big
ring
box
which
is
a
which
is
a
juice,
basically
very
medium-sized
juice,
and
it's
initially
targeted
at
just
virtual
box
and
just
developer
workstations
yeah
for
within
your
land.
For
your
wear
and
you've
got
your
your
golden
image
box,
which
is
a
limited
box,
and
then
it'll
run
chef,
solo
or
puppet
on
its
own.
B
Interesting,
I
can
see
how
that
would
not
be
terribly
difficult
to
to
incorporate.
I
have
no
idea
but
effectively
to
support
a
vagrant.
You
need
to
create
a
box
that
has
a
base
box
that
has
pocket
and
ship
in
it
for.
B
I
would
say
that
any
integration
was
probably
not
gonna
be
part
of
the
studio.
It
would
be
somebody
writing
a
project
for
adapting
vagrant
to
be
able
to
consume
those
images
for
being
able
to
deploy
up
onto
know
your
building,
which
brings
up
the
questions.
Yeah
one
is:
is
there
a
community
collaboration
site
beyond
sharing
the
image
around,
for
example,
the
web
books
script?
You
describe
to
upload
images
into
plans
or
this
type
of
stuff.
For
the
most
part,
the
answer
is
no
I'd
be
thrilled
to
death.
B
If
somebody
wanted
to
start
one
we
do
have
I
mean
there
are
there.
Are
there?
Is
community
support
around
open
souza
for
like
the
kiwi
project?
That's
the
base
underpinnings
for
this.
So
there's
there's
lots
of
work
that
goes
on
there
and
lots
of
community
involvement
that
goes
on
in
there
in
terms
of
the
web
hook
receivers.
B
We
actually
haven't
officially
released
that
yet
it's
only
visible
in
here,
because
I'm
an
employee
who
has
early
preview
permissions,
the
gold
master
for
our
actual
studio
onsite,
will
make
that
be
a
support
thing.
It's
only
supposed
to
come
out
last
week.
It
might
be
coming
out
this
week,
it's
it
hasn't
been
released
yet.
But
I
guess
the
the
question
was
a
little
more
general
because
there's
a
set
of
yeah,
for
example,
best
practices
around.
A
Community
sites,
where
they
share
cookbooks
and
recipes.
B
And
manifests
et
cetera,
and
I
would
say
that
the
the
space
is
new
enough,
that
there
has
not
been
a
whole
lot
built
up,
and
the
second
question
was
the
one
that
you
promised
me
an
answer.
Okay,
repeat
the
question,
because
I
don't
remember
what
cloudiness
type
integration?
Okay.
B
So
if
we
had
done
this
on
the
using
the
enterprise
version,
there's
a
little
check
box
that
will
install
the
api
tools,
allow
you
to
do
like
the
ssh
key
injection
and
and
that
will
enable
you
to
integrate
with
with
an
openstack
cloud
at
that
point
beyond
that,
we
haven't
done
a
whole
lot
in
terms
of
integration
with
and
any
of
the
other
stuff.
B
We
prefer
to
externalize
that
to
the
web
hook,
receiver
and
just
you
know,
have
an
image
in
there
and
let
the
web
receiver
do
whatever
it
needs
to
do.
I'm
not
yeah,
it's
not
a
very
good
answer.
The
short
answer
is
that
we
do
have
a
limited.
You
can
roll
your
own
with
the
scripts
yeah
you
can.
You
can
run
around
with
the
script
section.
That's
the
fallback,
because
if
it's
not
an
official
feature,
you
write
it
in
script
and
it'll.
B
So,
okay,
so
that's
a
fantastic
question,
so
say:
you've
built
a
your
own
custom,
rpm
or
a
package
back
here
at
the
very
beginning.
B
Under
software
you
can
click
on
upload
rpms
and
you
can
actually
upload
either
from
a
web
url
or
or
from
an
if
you
have
a
firewire
desktop,
but
you
can
also,
if
you
have
repositories
of
those-
and
you
can,
you
can
add
a
repository
as
well.
B
So
whatever
you
add
will
be
automatically
installed
by
the
image
on
the
image
on
the
image,
and
so
one
key
thing
to
keep
in
mind
here
is
that
when
studio
builds,
it
builds
the
latest
versions
of
all
the
packages
that
have
available
to
it
at
that
point
in
time.
So
if
there
are
new
packages
released
into
your
repositories
or
into
the
souza
repositories,
the
next
time
you
do
a
build,
it
will
incorporate
those
packages
in
there
is
there
a
way
of
taking
it
to
a
known
set
of
versions
on
the
community
side?
B
No,
I
do
that
on
my
I
do
that
on
my
on
site,
but
we
have
another
tool
you
can
you
can
source
it
from
an
alternate
tool
and
then,
within
that
tool
you
can
create
staging
versions
and
stuff
like
that.
So
just
somewhere
there
was
somewhere
on
the
start
page
there
was
upload
and
appliance
definition.
Yes,
can
I
download
an
appliance
you
sure
can
okay
was
that
that
would
be
a
way
of
freezing.
B
B
C
B
Okay,
so
all
right
here
we
have
a
configuration
link
right
here.
If
you
click
on
that,
that
will
give
you
the
the
actual
config
file
for
what's
going
on
in
there,
and
you
can
export
this.
B
Export
your
appliances,
kiwi
configuration
and
you
can
download
the
source
and
it'll,
give
you
the
zip
file
that
has
all
that
information
on
there
and
then
you
could.
You
could
re-import
that
into
a
new
version.
If
you
want
from
scratch
for
the
new
version,
so
can
you
face
move
down?
I
think,
can
you
base
the
downloaded
version
when
you
import
it
on
a
different
version
of
yes,
the
operating
system
yeah
as.
B
B
So
once
we
have
our
disk
image
built,
let's
get
into
the
customization
stuff
right
here
we
have
what's
called
a
test,
drive
and
accidentally
already
clicked
it
test
drive.
What
that
will
do
is
it'll
actually
launch
it
up
here
in
the
network
on
the
net
on
the
studio
server,
and
you
can
sit
here
and
we
can
watch
it
boot
when
it
gets
finished
booting
we
can
actually
log
in
at
the
console
we
can
come
up
here.
We
can
turn
on
networking
and
then
be
able
to
ss
sshn.
B
C
B
You're
doing
all
of
that,
it
will
come
over
here
and
it
will
produce
for
you
a
list
all
the
files
that
are
modified
in
your
session,
and
then
you
can
just
select
the
check
box
for
all
the
ones
you
want
to
pull
back
in,
and
you
can
make
that
be
a
part
of
your
of
your
overlay
files
and
your
appliance.
B
B
That
it's
the
fastest
model
as
in
using
the
not
blanking
out
it's
not
that
convenient.
No!
No!
No!
It's.
B
B
B
We're
not
necessarily
talking
about
data
integrity
on
the
disk.
At
this
point,
all
we're
talking
about
is
monitoring
what
files
config
files
that
I
change
or
what
file
to
introduce
the
system,
so
I
can
pull
those
back
in,
as
I
would
not
think
I
think
I
know
the
monitoring
of
that
level
is
overkill.
B
B
B
Can
we
sure
we
can
we
can
do
it?
We
can
do
it
right
now.
Qv
supports
building
for
lots
of
different
things.
Our
open,
build
support,
building
for
all
the
major
distributions.
It's
just
a
question
of.
Is
that
something
we
want
to
do
to
enable
our
competition
we
feel
like
this
is
something
that
is
very
unique
to
souza
and
it's
a
great
differentiator
for
us,
and
we
don't
necessarily
want
to
open
that
up
to
everybody.
It's
it's.
B
A
balancing
act
right,
obs
is,
is
unique
to
us
at
the
build
service
and
when
we
do
open
that
up,
this
is
one
that
we've
chosen
to
to
hold
back.
Does
it
do
a
cool
animation
once
you've
reached
your
one
hour
limit?
I
don't
know
I've
never
gotten
to
the
one
for
them.
B
B
You
got
the
test
drive.
You've
got
the
download
if
it
was
the
the
amazon
ec2
or
the
azure
image
instead
of
download,
it
would
be
upload
to
whatever
you've
got.
If
you
were
on
the.
If
we
were
on
the
commercial
side
again,
there
would
be
actually
a
supportability
report
where
it
would
generate
and
say:
okay,
this
is
the
source
for
all
different
packages,
so
these
ones
over
here
that
are
part
of
the
souza
distribution.
We
cover
these
ones
over
here.
You'll
need
to
have
a
third-party
support
contract.
B
B
B
Generally
speaking,
the
answer
is:
it's
not
part
of
our
just
it's
not
our
code,
so
we're
not
supporting
it,
whether
you're
gonna
sell
support
or
whether
you're
gonna
pay
somebody
else
to
support
it.
You'd
have
to
go
that
route
now.
Obviously,
if
you
put
a
big
enough
stack
of
money
on
the
table,
there's
lots
of
things
that
can
be
done
right.
B
It
just
depends
on
how
onerous
it
is
and
the
biggest
step
and
how
big
the
saturn
is
right
right.
The
the
semantics
there
were
not
exactly.
B
That's
covered
by
your
support
contract
with
us,
and
this
is
the
stuff
that
you
would
have
to
find
support
for
in
some
other
form
or
fashion,
whether
it's
self
support
or
you
know,
going
to
oracle
and
getting
a
mysql
support,
contract
or
I'd
rather
cut
out
my
hands.
Probably
yeah.
You
have
a
similar
thing
for
licenses,
so
what's
gpl
gpl
apache,
you
can
actually
see
when
so
when
you
go
into
selected
packages,
although
I
meant
that
there
was
a
summary
report,
it
was
dependencies.
B
No,
I
think
it's
a
very
valid,
very
valid
question,
because,
if
you're
going
to
redistribute
that
you
pldt.
A
B
Mtp
project
has
recently
run
into
the
problem
of
oh
crap.
Now
we've
got
to
go
rip
out
all
these
gpl
things
that
have
been
contributed.
A
That
still
had
the
original
sgi
copyright
blocks
and.
B
A
B
The
if
you
export
the
kiwi
configuration
you
can
do
a
diff
between
the
keyword,
config
files
and
get
that
there's
also
easier.
B
Wait
for
these
to
finish
building
and
then
we'll
go,
create
we'll
bump
the
version
and
rebuild
it,
and
I
believe
there
is
a
change
log
that
will
tell
you
what
the
difference
is.
This
may
be
kind
of
a
feature
asked,
but
do
you
generate.
A
B
The
image
we
do
generate
md5
hashes
with
this
there's
nothing
that
I
know
of
in
terms
of
certification.
B
Get
you
you
know,
you'll
know
somebody's
tampered
with
it
in
the
form
of
fashion
right.
Well,
yeah,
there's
always
exceptions
to
every
rule.
Right,
I'm
curious
what
this
hill
will
be?
Well,
what
you
could
mean
what
we.
A
Do
is
you
stick
a
list
of
all
the
files
that
were
installed
their
hashes
and
then
you
cryptographically.
B
B
B
On
the
install
media,
so
if
you
have
that
iso
or
the
raw
image
or
whatever
like
that,
you
would
be
doing
the
verification
on
that
on
that
image
before
you
actually
launch
it
up
in
your
cloud
once
you
launch
it
up,
all
bets
are
off
because
you're
going
to
be
modifying
it,
just
just
a
simple
act
of
booting:
it
creates
files.
B
A
B
B
Has
the
has
the
cryptographic
signature
in
it
and
then
each
of
the
each
of
the
rpm
files
themselves
have
have
their
have
their?
I
believe,
if
I
remember
correctly
md5
is,
is,
is
incorporated
as
part
of
the
rpm,
so
it
can
now
you
can
check
itself.
Yes,.
C
B
B
Are
valid
so
that
potentially
I
don't
know
if
that
matches
your
use
case
better,
because
I
know
because
remember
you
you
take
the
image
and
then
it
does.
It
runs
a
post
processing
pass
on
it
and
anytime.
You
do
that.
You
have
to
have
it
yeah.
You
need
it
before
that.
You
need
to
be
able
to
look
at
the
image
and
say
yes,
this
is
the
image.
This
is
the
correct
image.
Then
you
deploy
it
and
then
once
you
deploy
it,
you'll
have
to
have
some
sort
of
other.
B
B
Actually,
the
studio
on-site,
which
you
can
tell,
looks
significantly
like,
like
the
dot-com
version
right
there,
you
can
see
I've
built
all
these
different
things
and
you
can
see.
We've
got
change
logs
that
are
available
for
all
the
different
versions.
So,
for
example,
I've
got
I've
got
it
for
between.
B
B
I'm
sure
you
could
the
most,
if
not
all
this
functionality
is
available
through
an
api,
a
rest
api
too,
so
you
could,
you
could
pull
it
out
out
of
the
api
and
then
tell
it
to
you
know:
do
whatever
it
is.
You
want
to
do
to
transform
it
and
get
into
the
state.
Are
we
really
sufficiently?
B
B
To
go,
get
your
pizza
you've
been
sitting
here.
Listening
to
me,
ramble
on
for
ages.
On
end
I
mean
this
is
fantastic.
I
had
lots
of
backup
filler
materials
to
go
back
to
if
you
guys
are
just
kind
of
standing
on
it
sitting
on
your
on
your
hands.
It's
been
great,
I'm
really
happy
so
that,
in
a
nutshell,
is
susan
studio.
B
Oh
one
thing
I
almost
forgot
to
show
you
guys,
so
one
of
the
cool
things
you
can
do
is
you
can
actually
come
up
and
once
you
have
something
built
say,
for
example,
I've
got
this
base
server.
I
can
clone
that
configuration
over
to
a
to
a
new
system.
So
say
I
want
to
have
ic
come
in
here
and
build
up.
You
know,
put
in
all
the
different
policies
for
security
and
for
networking
all
that
kind
of
fun,
stuff
and
and
fulfill
again
layers.
B
The
politics
level
of
the
osi
model
get
all
that
stuff
built
into
here.
Instead
of
having
to
recreate
that
on
every
single
image
you
simply
clone
it
over
add
in
your
application
or
in
the
case
of
you
know,
with
your
cloud
integration.
If
you
have,
you
set
up
all
your
cloud
integration
stuff
in
your
base
and
then
you
simply
clone
it
over
and
you're
good
to
go.
B
B
A
B
However,
if
there
is,
I
mean,
if
it's
an
application,
they
can
run
on
students
linux.
Then
there
is
no
reason
why
it
cannot
be
incorporated
when
you
say
notified
in
a
very
clean
way.
B
It
tends
to
be
the
case
that
some
applications
have
been
integrated
with
with
things
like
flex
lm
in
one
of
its
modes,
and
if
they
don't
do
it
in
a
nice
mode,
where
you
could
take
the
image
duplicate
it
and
the
key
that
that
someone
decided
that
they
were
going
to
put
in
the
image,
isn't
happy.
It
will
lock
out
all
of
the
images
right.
B
So
if
you
have
to
have
individual
keys,
if
you
have
individual
keys
like
that,
you
put
into
the
first
boot
script
to
catch
the
appropriate
key
from
whatever
repository
and
inject
it
into
the
game,
you
wouldn't
be
able
to
do
that
at
the
yeah,
somewhat
nice
and
decide.
Oh,
you
have
a
hundred
user
key
as
long
as
100
aren't
running
simultaneously
you're
good.
A
B
So
I
know
that
rob
wanted
to
talk
a
little
bit
about
the
openstack
summit
and
some
of
the
upcoming
events,
so
I'm
gonna
just
wrap
up
real
quick.
You
guys
have
been
wonderful.
I
really
appreciate
the
feedback.
It's
been
great,
feel
free
to
go
out,
run
out
to
studentstudio.com
start
building
up
to
your
heart's
content.
If
you
want
to
do
it
in
your
data
center
and
you
want
me
to
come
help,
you
install
it
I'm
more
than
happy
to
come.
Do
that
too?
B
If
I
want
to
contribute
the
ability
to
make
a
vagrant
box,
as
you
contributed
to
kiwi
and
caesar
studio,
pick
it
up?
B
Yes,
in
a
way,
it
will
not
necessarily
be
exposed
to
the
web
interface,
but
it
would
be
part
of
the
kiwi
underpinnings,
there's
actually
functionality
for
kiwi,
and
it's
present
state
right
now
that
is
not
exposed
in
studio.
You
can
get
so
a
lot
of
like
the
functionality
you
were
talking
about,
like
the
multi-pathing
and
some
of
these
other
things
like
that.
B
You
can
actually
do
a
lot
of
that
stuff
inside
of
kiwi
if
you're,
if
you're,
just
using
the
command
line
tool
but
studio
itself,
the
web
ui
does
not
have
all
those
pieces
exposed.
B
Osatx,
if
you
want
links
some
of
the
things
they
talked
about
like
supervisor
sandwich
stuff
like
that,
so
it's
traditional
for
us
to
talk
a
little
bit
about
the
summit
things
coming
up
in
the
summit,
so
how
many
people
are
planning
to
go
to
the
openstack
summit?
C
A
A
B
B
But
so
it's
in
the
fall
so
make
make
if
you're
gonna
attend
the
next
next
in
the
ball.
Playing
playing
big
travel
should
be
a
lot
of
fun.
How
many
people
have
been
will
how
many
people
have
been
to
a
summit?
What
the
second
wow,
so
they
actually
have
a
good
number
of
newbies,
which
is
really
cool.
So
openstack
summit
is
two
different
things.
B
It
is
a
openstack
conference
and
there's
a
complete
track
of
four
days
of
openstack
content
using
it
deploying
it
commercial
things
going
on
with
it
user
sessions,
there's
a
lot
of
things,
matt
bray
and
I
run
the
ops
tracks
and
there's
actually
a
three
day
effectively
three
days
of
ops,
josh
volunteered
who
was
roped
into
hosting
one
of
our
panels,
which
I
appreciate,
and
so
there's
there's
a
lot
going
on,
but
it
is
also
a
design
side.
B
So
one
of
the
things
that
people
find
confusing
about
openstack
is
that
it's
actually
the
havana,
which
is
the
next
release
design
summit.
So
the
purpose
of
bringing
everybody
together
is
there's
actually
a
parallel
conference
that
goes
on
where
we
talk
about
the
design
and
features
that
are
coming
up
for
the
next
release.
B
So,
while
a
lot
of
people
are
there
doing
business
around
the
current
grizzly,
the
latest
there's
a
lot
of
us
who
also
focus
on
what's
coming
in
havana,
which
is
the
upcoming
release
and
the
discussions.
If
you
look
at
the
list,
you'll
see
we're
electing
ptl's
for
the
next
release
cycle
and
doing
a
whole
bunch
of
things
like
that.
So
what
are
ptls
sorry
project
technical
leads
so
that
each
openstack
is
is
evolving
into
a
lot
of
different
projects.
B
There's
some
really
interesting
board
activities
around
which
are
core
and
which
ones
are
integrated
projects
and
all
sorts
of
fun
politics.
We
have
a
whole
meeting
just
about
that,
but
those
every
one
of
those
projects
has
a
technical
lead
and
then
there's
actually
a
board
that
is
made
up
of
those
leads
and
some
other
people
who
are
on
the
board
elected,
oh
committee
committee,
so
there's
a
board
which
oversees
the
foundation
and
then
there's
a
technical
committee
which
oversees
the
code,
the
projects
there's
also
a
user
committee
which
focuses
on
user
things.
B
At
some
point,
I
suspect,
we'll
have
an
operations
committee,
because
we
have
enough
people
who
have
a
specialization
hops
alone
around
that
stack.
B
So
we
don't
have,
I
don't
want
to
spend
too
much
time.
Do
people
have
questions
about
the
summit,
sometimes
we'll
like
survey
and
we'll
see
what
people
what
things
people
are
interested
in
topics
and
what
what
we?
What
agendas
we
want
to
push
in
for
the
openstack
summit,
but
we
don't
have
quite
enough
time
to
do
that.
So
we'll
just
do
a
summary
meeting
afterwards.
B
The
people
have
questions
about
the
summit,
but
they
think
maybe
we'll
highlight
the
continuous
deployment
track
session.
That's
part
of
the
it'll
be
easy
to
speak
without
the
seminar
that
continuous
deployment
at
the
design
side.
B
B
We
typically
are
very
ops
focused,
and
so
we
usually
go
to
the
openstack
summit
with
an
ops
agenda,
which
is
why
matt
is
also
austin
based,
and
I
typically
like
to
chair
the
ops
track.
So
we
can
focus
on
that.
So
continuous
op
continuous
deployment
has
been
the
real
theme
for
this,
which
that
and
what
that
means
is
that
the
people
deploying
openstack
are
consistently
moving
to
deploy
the
trunk
code
for
openstack
right.
So
the
code
that
we
work
on
the
crowbar
project
actually
deploys
for
trunk,
which
is
an
important
thing.
B
There's
a
lot
of
major
deployments
out
there
that
have
really
made
an
effort
and
a
lot
of
made
a
lot
of
flood
publicity
where
they
deployed
the
trunk
components
right
so
you'll
see,
there's
hosts
who
are
deploying
openstack.
B
B
Thing
it's
not
simple,
but
doing
a
continuous
deployment
infrastructure
where
you're
committed
to
running
the
code.
The
trunk
of
openstack
code
drives.
All
these
behaviors
into
the
community
has
been
a
really
significant
thing.
So
at
this
session
we
have
a
lot
more
ops
focus
tracks,
you'll,
see
a
lot
more
deployment,
information,
you'll,
see
a
lot
more
user
stories
and
user
stories
and
you'll
actually
see
two
days
of
the
conference
in
the
design
summit
side
which
are
dedicated
to
nothing
but
process
and
infrastructure,
which
is
a
new
thing.
B
So
we're
not
just
talking
about
compute
and
network
and
storage
and
all
the
code
we're
actually
dedicating
engineers
who
do
nothing
but
talk
about
process
and
infrastructure.
It's
a
big
deal,
yeah
yeah!
That's
that
you
know
for
every.
Hopefully
everybody
here
how
many
people
are
actually
deploying
openstack
internally.
B
So
the
fact
that
we're
doing
that
we
have
that
type
of
focus
on
deployment
and
infrastructure
is
major.
So
one
of
them
sorry,
the
continuous
deployment
for
upstream
openstack
is
the
name
section
I
was
thinking
about
and
one
of
the
things
that
did
right
on
what
you
just
said.
So
the
code
is
stable
but
can
deploy,
which
I
guess
your
comment
just
summarized
it,
because
there
is
no
prescriptive
way
of
saying.
B
All
kinds
of
random
folks
that
try
to
make
their
name
by
writing
books
about
like
this
one,
but
the
thing
is
that
one
of
the
things
that
I
think
rob
you've
been
trying
to
put
with
push
with
fits,
but
then
a
lot
of
community
a
lot
of
folks
that
we
cuddle
with
try
to
push
it.
She
can't
deploy
it
trump
is
broke.
Stop
the
horses
fix
it
move
on
right.
B
So
the
idea
of
that
part
of
the
continuous
integration
and
testing
of
saying
is
the
code
saying
it
needs
to
include
that's
effectively
on
this
session.
I
don't
know
the
guy,
but
I
like
what
it
started.
So
that's
what
this
session
is
trying
to
say
is:
let's
modify
the
process
of
openstack
the
development
process
to
stop
the
wagon.
If
you
can't
deploy
it
so
the
code
trunk
has
broken.
B
You
can't
submit
new
code
into
trunk
if
that
submission
breaks,
whatever
canonical
deployment
mechanism
is
used
to
verify
that
you
can
still
deploy
something
and
that
that's
a
big
move
in
maturity
in
the
community
to
say
it's
not
just
about
the
code,
it's
actually
about
using
being
able
to
use
them.
Yeah
a
year
ago,
I.
B
Proposal
is
not
in
the
room,
the
sessions
will
be
streamed
and
one
of
the
things
that
we
will
do
is
I
need
to
look
at
where
we
are
with
agendas
and
topics,
but
you
know
we
do
typically
spend
a
session
talking
about
what
was
interesting.
B
B
B
B
B
B
Say
just
do
pip,
it
doesn't
doesn't
matter.
No
one
needs
to
see
someone
walking
back
and
forth.
Like
some
4k,
we
can
hear
them.
We
see
the
slides
and
we
can
add
those
stitches
together.
One
of
the
things
I
would
suggest
is
watch
the
ether
pads
one
of
the
ways
that
I
find
it
engaged.
So
if
people
aren't
used
to
this,
it's
really
it's
really
important
to
understand
a
lot
of
the
design
discussions
for
openstack
summits
and
actually
a
lot
of
the
activities.
A
B
If
you're
there,
I
highly
encourage
you,
pull
up
an
ether
pad
help,
take
notes
because
it
really
builds.
It
has
a
very
dynamic
effect
in
meetings
to
have
people
on
these
ager
pads
and,
if
you're,
remote,
being
part
of
the
ether
pad,
can
really
help
you
follow
along
with,
what's
going
on,
because
people
will
want
to
like
notice
the
ethercat
attached
to
the
well.
B
If
you
go
to
summit.openstack.org,
that's
the
design
card,
you
should
go
to
openstack.org
summit.
That's
the
conference
part!
You
might
want
to
complain
about
that
as
well.
I
will,
but,
but
as
some
of
the
interface
are
not
yet
live,
another
quantum
team
was
just
still
going
over
the
mailing
list
before
they
put
it
in,
but
that
is
that
is
one
of
the
best
ways
if
you're
there,
even
if
you're
in
a
session,
you
can't
monitor
these
really
people,
don't
realize
how
it's
one
of
those
things
where
it's
like
a
twitter
back
channel.
B
If
you're
on
it,
you
see
how
active
it
is.
The
same
thing
there's
a
lot
of
activity
in
openstack,
but
if
you're
not
aware
that
people
are
doing
it,
you
can't
tell
it's:
can
folks
jump
on
irc
during
that?
I
would
expect
they
will
they
do
yeah,
because
irc
is
also
really
active
for
that
it's
going
to
get
very
chatty
the
trick,
and
so
a
lot
of
because
it's
such
an
international
project,
a
lot
of
openstack
activity
happens
on
irc,
including
meetings
and
things
like
that
and
even
board
meetings.
B
We're
now
making
the
board
has
a
lot
of
lawyers.
We
have
a
couple
of
developers
and
then
we
have
some
lawyers.
We
have
some
corporate
types
and
it's
a
healthy
mix,
but
the
developers
are
driving
the
etherpad
and
irc
channels
so
that
you
can
actually
start
participating
remotely
and
seeing
some
of
the
stuff
going
on
board
meetings
too,
because
we're
really
working
towards
transparency
and
there's
actually
a
transparency
subcommittee
for
the
board.
B
So
it's
there
there
are,
I
think,
they're
going
for
is
it
incubated
status
is
that
this
is
so
okay,
so
we're
yeah
we're
there's
a
lot
of
people
who
want
to
get
incubated
and
what
I
can
tell
you
is
that
there's
we're
starting
to
add,
I
wouldn't
say,
barriers
to
incubated,
but
we're
starting
to
be
much
more
specific
around
what
creates
an
incubator
project.
B
It's
a
lot
of
people
want
to
be
incubated
because
they
want
to
ride
under
the
open
stack
barrier,
and
so
we're
actually
being
very
trying
to
be
very
specific
to
make
sure
that
projects
that
are
core
are
actually
required.
Projects
that
are
integrated
are
components
that
must
be.
You
know
that
are
part
of
the
the
coordinated
release
cycle
and
things
that
consume
the
api
and
want
to
be
part
of
the
project
are
part
of
the
ecosystem,
and
so
we're
helping
distinguish
those
things.
B
I
can
tell
you
that
there's
a
lot
of
a
lot
of
projects
that
want
to
be
part
of
core
that
really
aren't.
We
have
things
that
are
in
core.
That,
probably
are
not
so
it
was
this
whole
integrated
versus
core
versus,
and
that's
there's
a
technical
there's,
a
there's,
a
committee
about
that
specifically
and
we're
working
through
what
those
things
are,
because
it
comes
back
to
what's
required
to
call
something,
an
openstack
deployment
which
matters
a
lot
if
you
are
looking
for
an
interoperable
host.
B
So
if
you
deploy
these
things,
you
should
be
able
to
interrupt
with
interrupt
with
another
site.
So
it's
very
important
to
say
this
is
core
or
non-core,
and
then
it
gets
even
funnier
because
you
get
into
somebody
like
seth
is
a
really
good
example.
Sef
wants
to
have
a
different
implementation
of
object,
storage
and
swift,
but
use
the
swift
api,
and
so
there's
you
get
into
a
really
funny
question.
B
B
Okay
and
it's,
it
sounds
sort
of
mundane
and
legalistic,
but
it
ends
up
being
very
important
to
drive
from
the
day-to-day
technologies
even
down
to
does
openstack
have
to
be
python
or
not,
and
it's
a
technical
community
decision,
but
I
think,
don't
think,
and
the
feds
and
api
versus
implementation,
if
you
end
up
with
lots
of
different
ports
effectively
for
so
non-interoperable
openstack
public
clouds,
you
just
shot
yourself
in
the
cloud
right.
So
the
importance
of
having
this.
B
This
is
what
the
logo
means
is
important
to
every
anybody
who
cares
about
openstack,
because
otherwise,
if
the
bubble
means
nothing,
you
just
must
and
part
of
another
oil
in
the
fly
in
the
ocean.
Sorry
is
swift
because
it
says
we
want
to
be
able
deploy
just
another
storage
platform.
We
don't
want
keystone
horizon
well,
keystone,
maybe
but
not
horizon.
No,
but
don't
care
about
all
of
this.
Will
you
not?
Let
us
use
the
logo
of
openstack,
sorry,
but
you're
about
to
say
it's
about
capabilities.
I
mean
I
can
deploy
quantum
and
nationality.
B
Those
are
we're
actually
all
those
sales,
so
there's,
there's
concepts.
There's
concepts
of
reference
architectures,
and
one
of
the
things
that
monty
taylor-
and
I
are
trying
to
talk
about-
is
using
heat
to
actually
specify
reference.
Architectures,
which
brings
up
flavors,
which
brings
up
these
are
layers
and
layers
of
conversations
which
makes
them
sort
of
fun
to
talk
about,
but
it
does
at
the
end
of
the
day,
if
the
implementation
matters
configuration
of
different
options,
shouldn't
necessarily.
B
B
B
But
I
I
can
tell
you
for
right
now.
The
implementation
will
still
continue
to
drive
the
definition
about
this
deck
so
right.
This
is
I
this
actually
is
not
my
preferred
model.
I
would
rather
see,
as
fits
testing
and
apis,
very
that
I
think
over
time.
That's
that
will
ultimately
be
where
we
go,
but
in
the
short
term,
openstack
is
defined
by
the
implementation
of
the
code.
B
There's
behaviors
that
that
drives
that
are
positive
for
us
and
one
of
the
things
that
openstack's
been
good
about
doing
is
creating
a
plug-in
model
so
that
you
can
have
different
providers
underneath
the
scheduler
or
a
networking
system
or
sender,
or
things
like
that,
and
so
having
an
implementation
with
a
plug-in
model,
drives
a
lot
of
behaviors
for
people
to
collaborate
on
the
base
code
and
not
create
forks
and
then
argue
about
their.
How
they've
implemented
the
api.
B
We
still
have
those
arguments,
but
we'll
see
we'll
see
how
we'll
see
over
time,
I'm
sure,
as
the
apis
become
more
standard.
That
will
have
that
discussion
and
I'm
personally,
my
part
in
this
is
trying
to
make
sure
that
we
create
the
definition
of
ecosystem
so
that
you
can
have
api
compliance
or
api
use
and
be
considered
openstack.
B
B
Actually,
I
should
stress
if
you
are
really
motivated
by
these
things,
these
discussions
and
debates,
all
the
openstack
meetings
are
open,
they're,
not
all
participatory,
meaning
that
you
can't
show
up
at
the
technical
incubator
meeting
and
speak
speak
your
mind
as
part
of
the
meeting,
because
that
would
be
chaos,
but
you
can
participate.
You
can
be
part
of
the
back
channel.