►
From YouTube: OpenCrowbar Provisioning & Annealer ROUGH CUT
Description
Tour of OpenCrowbar provisioning VMs and Docker Containers using the UI.
A
Well,
this
is
reversed,
felt
and
I
wanted
to
pull
together
a
short
demo
of
up
and
crowbar
to
wrap
things
together
and
show
you
the
type,
the
capabilities
and
technology
that
we've
been
blowing
together.
This
is
the
system
overview
screen,
I'm
right
at
the
end
of
the
earlier
demo,
about
setting
up
the
dev
system
that
we're
going
to
jump
into
a
run,
the
crowbar
in
docker
admin
node.
This
actually
did
a
full
probar
install
and
then
to
automatically
launch
the
api's
to
install
the
admin
server.
A
This
is
the
roles
view
of
the
annealer,
where
I'm
watching
this
admin
node
setting
up
all
each
roles
that
are
necessary
to
run
the
ops
infrastructure
for
data
center
DNS
NDP,
DHCP
provisioner,
the
pixie
boots
server,
all
those
components
getting
set
up
now
on
this
docker
admin
note
over
here.
I
can
actually
show
you.
So
this
is
the
docker
node
which
is
running
centos
right
now
and
we're
letting
it
converge.
This
will
automatically
exit
the
script
when
we
complete,
although
I
can
stop
it
right
now.
A
So
the
US
and
the
api's
are
already
up
by
docker.
Show
the
processes
you'll
see
that
this
is
the
admin
note
right
here.
That's
come
up
up
and
running
it
for
about
eight
minutes
it
as
it
goes
through
a
little
complete
process
and
there's
a
lot
to
show
and
tell
about
open
crowbar.
The
number
one
thing
that
I
really
want
to
focus
on
with
this
is
sort
of
you
I
how
it
works.
A
We'll
show
some
of
the
CLI
also,
but
it's
really
the
annealer,
because
the
annealer
is
the
magic
that
makes
crowbar
orchestration
work
in
a
predictable,
reliable
way.
It
is
really
the
heart
of
what
we
think
it's
going
to
help
provide
upgradable
physical
infrastructure
based
on
some
of
the
core
tenants
carve,
are
likely
finding
where
we
don't
have.
We
don't
make
decisions
about
the
infrastructure
until
you've
actually
discovered
the
infrastructure
so
braid
and
BIOS,
and
networking
configuration
is
done
really
after
you've
discovered
what
the
infrastructure
is.
A
You've
made
decisions
about
how
you
want
your
system
to
look,
then
you
can
make
decisions
about
what
roles
the
system
takes
on.
The
view
in
front
of
us
is
the
annealer
backlog
and,
when
it's
showing
is
I,
have
one
role
that's
being
set
up.
In
this
case,
the
provisioner
by
our
chefs
hola
jig,
when
that
completes,
it'll
continue
moving
forward
and
all
these
these
items
are
blocked.
So
one
of
the
really
important
things
we
did
in
open
crowbar
was.
A
A
This
infrastructure,
where
each
action
is
very
atomic
so
from
any
one
of
these
the
column
no
trolls
is
the
intersection
of
a
roll
and
a
node
I
can
get
very,
very
specific
information.
I
can
see
what
had
to
be
done.
First,
I
can
see
what
is
coming
next.
I
can
see
its
state
if
there
was
something
broken
up.
That
would
flag
up
as
an
error
very
quickly
and
I
can
see
exactly
what
the
system
did
in
processing
that
action.
A
So
each
one
of
these
is
very
small
father
done
by
what
we
call
jig,
which
is
the
running
action
and
jigs
are
really
extensible
cmdb
agnostic.
So
we
do
a
lot
of
Chef
solo.
We
have
a
puppet
jig.
We
have
a
script
jig
which
actually
just
itches
in
and
runs
bash.
We
could
run
API
calls.
This
is
really
it
could
be
done
as
know
how
to
ban
process,
so
it
really
is
open.
For
extension,
as
mayor
of
fact,
we
have
a
list
of
the
jigs.
A
You
see,
there's
quite
a
quite
a
few
of
them
in
what
we're
doing
then
same
same
thing
with
ropes,
so
each
roll
is
tied
to
jig
and
you
can
see
that
we've
built
quite
a
number.
We
have
test.
Jakes,
no
objects
are
effectively
milestone,
points
that
we
use
to
create
checkpoints
within
our
deployment,
so
I'm
going
to
notice,
discovered
or
if
a
node
is
managed.
Your
operating
system
is
installed.
We
also
have
a
docker
roll
which
allows
us
to
have
docker
come
in,
and
the
doctor
note
actually
can
provide
other
roles.
A
So
as
we
as
we
watch
the
systems
literally
going
through
and
applying
each
one
of
these
roles.
Building
things
up
in
some
cases
we're
actually
building
building
pixie
boot
images
from
Isis
in
our
system
to
cash
them.
But
you
can
see
I've
installed
a
month
to
incent
us
and
the
actual
information
that
we've
collected
from
those
reposes
is
a
lot.
We
hide
some
of
that
under
our
envy.
You
I,
you
know
our
function.
Our
UI
function
is
very
focused
on
exposing
capabilities
for
crowbar.
A
We
haven't
reached
a
generation
point
yet
or
we're
really
trying
to
create
use
case.
Focused
UI,
/,
app
and
you'll
watch
me
walk
through
the
UI
you'll
see
they
exposed
a
lot
of
capability,
and
but
it's
one
of
those
places
where
we
obviously
know,
there's
a
lot
of
improvement
to
come.
So
let
me
walk
you
through
what
we've
done.
I
think
we've
just
about
completed
our
deployment
here.
We
have
systems
reached
a
stable
point
and
at
this
point,
what
I?
A
A
So
my
doctor
admin
is
actually
connected
to
the
same
virtual
bridge
as
my
management
interface
on
this
node
and
I'm
going
to
do
the
second
one,
and
these
are
literally
being
pixi
booted
from
the
crowbar
pixie
infrastructure,
so
crowbar
set
up
a
DHCP
server
that
dhcp
server
is
directed
to
pixie
boot,
these
nodes
by
default
and
then
as
it
as
it
pixie
boots
them
it'll
bring
up
our
discovery
image.
Our
discovery
image
will
talk
to
the
crowbar
api's
and
set
up
register
the
node
and
put
it
into
adult
default
configuration.
A
But
what's
really
important
for
us,
especially
with
capabilities
like
docker,
is
that
you
don't
have
to
use
our
discovery
image
to
add
nodes
into
crowbar.
The
discovery
image
does
that
by
calling
the
api's,
but
you
could
pre
load
images
into
crowbar
using
the
api's
when
we
build
containers,
we
actually
use
the
api's
to
create
the
container
as
a
doctor
note
and
then
spawn
allow
the
crowbar
itself
to
do
the
docker
run
to
build
containers.
A
A
Sorry
I
just
created
one
called
the
fault.
Let
me
create
another
deployment
called
demo
and
the
idea
with
deployments
is
we
found
that
this
is
very
important.
Concept,
for
people
is
that
we
can
have
multiple
deployments.
These
are
children
of
system
system,
as
our
discovery
deployment
all
knows
come
into
that
initially.
A
But
the
idea
here
is
that
we
could
actually
have
a
test
and
a
dev
and
a
production
deployment
that
had
similar
things,
duplicate,
start
databases
or
message
buses
or
things
like
that.
Our
networks
that
we're
just
used
for
that
one
deployment
scope-
and
this
is
a
really
important
thing
using
something
like
chefs
church
like
we
used
to
do,
made
it
very
hard
for
us
to
create
multiple
configurations,
multiple
deployments
within
the
same
system
architecture,
and
we
really
made
a
point
in
open
crowbar
to
not
repeat
that
mistake.
A
Sorry,
here's
my
networks,
so
I
have
one.
My
admin
network
is
my
default
and
I
have
certain
ranges.
I'm
going
to
configure
a
new
network
for
purpose
of
the
demo
I'll
call
it
steph,
because
I
actually
have
some
cest
severals.
I
can
put
in
I'm
going
to
go
ahead
and
put
it
in
the
demo
deployment.
That
means
it's
scoped
to
that
capability,
just
on
vm,
so
I
don't
have
a
VLAN.
A
I
could
specify
that
about
crowbar,
open,
crowbar
uses
ipv6
as
its
default
and
ipv4
effectively
as
an
overlay,
and
so
that's
been
really
important
for
us
to
be
able
to
enable
a
lot
of
a
lot
of
interesting
capabilities.
A
lot
of
network
resilience
by
using
ipv6
is
our
primary
Network
and
then,
of
course,
ipv4
allows
allows
us
to
work
a
lot
of
legacy
capabilities.
I
can
set
up,
bring
some
teams
and
once
I
add
this
network.
A
A
So
what
I've
done
here
is
I've
set
up
this
network,
so
I
now
have
a
public
network
on
this
range
and
that
will
let
me
actually
provision
this
extra
network.
Even
if
I
don't
plan
to
use
it
I'll
show
this
demo
how
we
provision
it
well,
the
note
notes
have
come
up,
so
I
can
go
back
to
my
system
deployment
and
what
you'll
see
is.
These
are
the
two
nodes
that
I
just
brought
up,
and
this
is
the
roles
that
I'm
adding
into
the
system
and
crowbarred
has
brought
in
new
roles.
A
So
we
have
a
logging
server
on
the
admin,
node
and
logging
clients
on
the
other
nodes
and
TP
clients,
basically,
as
we
bring
in
new
things
and
new
capacities,
new
roles
as
we're
working
towards
this
node
discovered,
which
is
one
of
our
milestone
roles,
but
it
brings
in
all
these
other
roles
and
capabilities.
So,
let's
let
this
thing
run
so
I.
Let
the
system
run
all
the
way
through
to
getting
all
the
components
provisioned.
A
That
means
my
dealer
now
has
nothing
going
and
now
we
can
jump
right
into
actually
doing
some
interesting
work
and
letting
this
discontinued
forward.
So
I'm
going
to
go
into
my
demo
deployment,
then
I'm
going
to
do
two
things.
First
is
I'm
going
to
put
in
the
crowbar
the
OS
install
role
in
this
case.
This
is
one
of
our
milestone.
Rolls
the
only
thing
we
show
in
this
list
or
milestone
rolls
or
no
opera,
as
we
call
them
internally
and
then
I'm
going
to
need
to
add
some
role.
A
A
Great
so
now,
when
I
come
back
to
this
deployment,
I
now
have
nodes.
I
have
a
role
and
if
I
add
that
role
to
this
to
this
note,
so
I'm
literally
assigning
I
want
this
node
to
reach
this
milestone.
Crowbar
says:
hey,
wait!
A
second!
If
you
want
this
milestone,
I
also
need
the
installed.
Os
I'll,
give
you
an
example
of
another
one:
the
Ceph
config.
A
If
I
add
that
role
here,
I
can
choose,
I
can
say:
well
I
just
want
that
config
on
this
one
notice,
it
doesn't
add
it
for
my
others,
but
it
brings
in
the
network.
So
my
SEF
config
says:
hey.
You
need
to
have
a
network.
This
is
the
network
that
I
configured
earlier
and
it's
going
to
go
in
and
set
it
up
on
the
system.
So
the
wrench
indicates
that
I'm
in
a
proposed
state.
So
I
can
go
and
change
things.
A
So
this
node,
let's
go
in
and
he's
on
centos,
let's
go
ahead
and
let
him
or
her
I
don't
know
which
sexist
noticed
or
if
it
matters
to
the
node,
it
be
in
a
bun
to
node
and
this
one
we're
going
to
leave
as
the
default
for
sent
us
and
you'll
notice.
I
can
set
it
by
the
node.
I
also
have
a
deployment
level
value
and
I
could
also
change
that
if
I
wanted
to
so
I
can
set
things
either
at
the
deployment
system
or
individual
node
roles.
A
So
now
I've
got
a
deployment
that
looks
like
it's
in
pretty
good
shape.
It
says
I'm
going
to
install
the
OS
for
this
one
and
for
this
node
it's
going
to
go
a
little
bit
further
and
start
running
running
to
our
set
config.
The
SEF
config
right
now
is
only
going
to
set
up
the
network
on
the
second
nic.
That's
remember
when
I
set
that
network
up
so
now
hasn't
hasn't
done
anything
yet.
The
annealer
just
has
some
proposed
new
state
I
go
ahead
and
commit
that
that
takes
this.
A
These
no
trolls
proposal
and
sends
it
to
do
sends
it
to
be
configured
by
the
annealer,
and
at
this
point
the
system
says
all
right:
I
need
to
do
some
provisioning.
Most
of
the
action
is
going
to
happen
over
here
in
the
on
the
system
deployment,
because
this
is
where
it's
actually
going
to
do.
The
provisioning
and
the
first
thing
it's
figured
out
that
has
to
do-
is
change
the
dhcp
boot
records,
and
that
makes
perfect
sense.
A
I'm
about
to
tell
these
these
nodes
over
here
to
reboot
now,
which
is
exactly
what
they're
doing
right
now
doing
a
shutdown
and
reboot,
and
so
in
this
case
it's
saying.
Aha,
instead
of
running
our
discovery,
image,
which
is
called
sledgehammer
I've,
now
set
that
to
be
a
centos
65.
If
I
jump
back
to
the
other
node.
This
one
you'll
notice
that
this
one
has
my
abundant.
So
each
one,
depending
on
the
configuration,
has
been
given
a
different
set
of
boot
media.
The
system
configures
those
notes
off
until
they
report
back
in.
A
So,
even
though
they're
busy
doing
an
OS
install
they're
not
available
for
crowbar
and
they're
considered
off,
and
so
if
I
look
now
at
my
annealer
I
can
look
at
my
background
work
and
see
all
right
I'm
waiting
until
the
nodes
come
back
on,
so
those
are
in
a
to-do
state,
basically
on
hold
until
they
complete
the
the
OS
install.
You
can
see
in
the
background
over
here,
I
have
two
OS
installs
running
very
happily
pulling
down
a
whole
bunch
of
resources.
A
If
you
build
something
like
this,
a
gig
of
ram
is
a
minimum
for
an
OS
install.
We
recommend
for
gigs
for
the
crowbar
dev
admin
server.
If
you're
gonna
actually
run
containers
which
I'll
show
you
how
to
do
more.
Ram
is
better
because
containers
do
take
some
incremental
ram.
As
you
said,
you
spin
them
up.
You
do
like
some
of
our
devs
like
to
do,
which
is
sped
up
30
or
more
containers
to
test
your
system
out
more
RAM
better.
We
have
some
32
gig
ram
systems.
A
Of
course,
that's
a
lot
lighter
than
if
you
were
to
build
31
game
VMs
on
your
system,
so
basically
right
now
we're
back
in
in
a
wait
state
as
we
build
up
the
capacity
time
to
show
you
a
couple
things.
So
I
showed
you
networks.
We
actually
have
a
network
node
map
that
shows
you
IP
addresses
in
am
assigned
two
nodes.
We
haven't
gotten
all
the
way
there.
Yes,
so
you
can.
We
don't
have
that
much
in
here.
Thus,
in
our
faces
is
really
important.
A
When
we
set
up
these
one
g,
11
g,
0
nix,
the
enumeration
for
each
system
is
controlled
from
here,
because
what
we
found
is
that,
as
you
add,
Nix
into
systems,
the
enumeration
can
vary,
and
so
we
had
to
provide
a
way
for
you
to
create
custom,
enumerations
and
there's
actually
a
regex
pattern
that
we
used
to
match
system
systems
generally
or
specifically,
I
should
your
system
overview.
I've
showed
you
roles
bar
clamps.
If
you're
used
to
crowbar.
A
While
we
still
have
the
concept
of
bar
clamps,
we
it
helps
sort
of
prove
to
take
up
these
big
concepts
and
break
them
into
smaller.
So
we
still
have
sort
of
bar
clamps
as
concepts
about
roles,
but
you
don't
highlight
them
to
the
same
extent
that
we
used
to.
We
really
cut
the
crowbar
repose
you
go
crowbar.
A
We
really
cut
the
number
of
repose
way
way
down.
So
if
you
look
at
the
open,
crowbar
repo,
as
opposed
to
the
old
crobar
repo
you'll,
see
it's
pretty
logical,
we've
got
a
SEF
OpenStack
core
hardware
Hadoop
a
much
smaller
number
of
targets
for
people
to
develop
against
it
and
core
contains
all
of
the
core
pieces
for
crowbar
and
we've
been
integrating
documentation
into
that
repository.
So
you
looked
at
crobar,
performed
and
confused
about
all
the
wiki's
and
things
like
that.
A
We
really
have
worked
hard
to
train,
put
all
the
documentation
in
the
code
base
in
github.
It's
easy
enough
to
edit
through
the
github
you
eyes
and
he's
ended
much
faster,
a
highlight
our
Quick
Start
documentation
with
docker
that
lets
you
sort
of
bring
up
the
system
really
good
things.
So
we've
got
some
some
systems
that
are
busy
booting,
I'm
going
to
go,
I'm
going
to
pause,
not
have
you
watch.
A
That's
the
system
goes
through
his
deployment
because
till
the
nodes
actually
discovered
fully
booted
through
this,
not
much
to
watch
except
OS
is
being
laid
down.
Sorry
different
and
docker
doctors,
a
little
bit
quicker,
show
you
that
come
back
from
the
break,
so
I'm
checking
back
in
on
on
our
install
after
a
couple
minutes,
and
what
you'll
see
is
that
this
this
system
here
has
completed
the
install
and
the
behind-the-scenes
script
is
checked
into
crowbar.
It
said:
hey
the
powers
on.
A
So,
if
I
look
at
the
system,
you'll
see
it's
on
and
now
you
can
actually
see
the
work
that
needs
to
get
done.
One
things
I
really
enjoy
watching
with
the
way
the
annealer
works
and
how
atomic
we
are
in
our
actions
is
it.
I
can
see
exactly
the
work
that
needs
to
get
done
in
the
order
it's
going
to
be
performed
by
crowbar.
So
I
can
see
it's
building
the
network,
the
admin
network,
right
now.
It's
installing
the
dns
logging
and
TP
clients.
A
It's
setting
up
the
repos,
so
it
can
do
in
this
case
app
get
by
setting
up
bar
the
deployer
client
and
then
the
cheeves,
the
crowbar
management
noted
state,
and
then
it's
going
to
go
ahead
and
complete
the
provision
or
OS
install
a
bar
installed
it
at
that
point.
It's
going
to
say:
hey,
I
am
a
complete
note
and
since
we
don't
have
any
agent
at
that
point,
we
could
actually
detach
crowbar
and
just
use
this
system.
A
A
This
is
these
are
the
ranges
that
I
said
so
I
have
my
admin
nodes
get
pulled
out
of
this
range
dhcp,
so
we
have
a
69
range
for
dhcp
addresses
but
actually
is
used
by
the
dhcp
server
when
it
sets
sets
up
th
the
dhcp,
and
so,
if
you
change
this,
the
dhcp
server
will
be
changed,
and
this
is
the
house
network.
So
these
these
are
the
statically
allocated.
Node
ranges
for
the
host
and
you'll
see
we
actually
have.
This
is
a
22
range.
A
So,
with
the
doctor
notes
coming
in
and
setting
up
a
hundred
nodes,
we
actually
needed
to
blow
this
out
a
little
bit
more
and
then
in
the
SEF
network.
I
did.
I
did
a
different
range,
so
this
actually
gives
me
if
you're
used
to
curve
are
one
we
used
to
do
this.
This
was
called
in
the
network.
Jason
file
in
this
case
we're
much
more
adjustable,
meaning
that
you
can
make
these
changes
and
update
it.
You
don't
have
to
do
it
all
posts
before
you
even
do
the
install
so
a
major
major
improvement.
A
How
r
operates
in
essence
so
about
two
still
installing
over
here
I'm
losing
patience
in
that
I'm,
going
to
go
ahead
and
demo
or
dr.
Ned,
bring
out
she's
going
to
put
some
more
stress
on
the
machine,
but
I
think
should
be
able
to
handle
that
so
you'll
see
I'm
still
in
my
daugher
infrastructure.
I've
only
put
you
through
22
minutes,
but
54
minutes
of
there
for
you
something
minutes
at
last.
Time
has
gone
on
and
what
I'm
going
to
do
is
we
have
a
tool
for
dr.
Slade
notes
with
this?
A
What
this
script
will
do
is
it
actually
creates
nodes
using
the
car,
/
AP
I
and
then
creates
a
screen
session
for
each
one
of
those
actually
team
up
session
she's
like
screen
if
it
doesn't
have
a
40
forty
screen
as
forty
session
limit
like
screaming
I'm,
going
to
just
go
ahead
and
create
six
slaves
here
and
kick
that
off
and
you'll
see
it's
actually
bringing
up.
These
are
different
screens.
It's
creating
is
hitting
the
API
to
create
that
soccer
a
container
and
then
that
tells
docker
to
tells
crowbar
to
go,
run
docker
containers.
A
So
it's
going.
I'm
going
to
let
it
kick
through
the
six
that
I
asked
for
work:
five,
your
six
returns
controlled
to
me.
So
now
I'm
back
to
the
system,
and
if
I
do
a
dr.,
PS
you'll
see
it's
been
it's
crowbars
actually
now
spinning
up
these
docker
containers
and
not
you'll
see
them
in
the
in
the
UI.
So
one
thing:
that's
really
nice
about
docker
is
unlike
the
process
that
I
followed
to
bring
up
a
new
admin.
Note.
A
The
docker
containers
are
very
quickly
and
have
available
OSS
really
quickly,
so
where
it's
taken
me
40
minutes
to
provision
up
some
VMS
through
crowbar
that
I
can
actually
use
for
deployments.
My
doctor
does
we'll
be
ready
in
just
a
couple
of
minutes,
so
actually
less
than
a
minute
as
it
works
through
the
annealing
process,
crowbar
can
operate.
We
have
ten
workers
in
the
kneeler,
so
it
can
work.
You
ten
simultaneous
tasks.
/
note,
so
we
single
thread
through
a
node
and
we
paralyze
as
much
as
possible.
A
If
I
look
at
the
dealer
here,
my
laptop's
day
will
warm
as
I'm
putting
some
stress
on
it.
You'll
see
that
we're
bringing
we're
working
a
whole
bunch
of
actions
in
parallel
in
the
infrastructure
and
then
we're
tracking
a
whole
bunch
of
work
to
be
done
in
the
background
that
can
go
into
an
individual
node.
A
Ok,
I
got
an
error,
so
I
can
show
you
what
that
looks
like
to
correct
and
I
can
see
my
attributes
are
set
and
docker,
but
we
don't
can't
do
the
same
type
of
inventory.
I
can
actually
see
what's
what's
in
process,
so
I
can
see
I'm.
I
aven't
network
failed
to
get
set
up
and
then
these
items
are
blocked
and
then
what's
happening
right
now
is
my
docker
container,
sir.
That
fail
simply
because
my
system
is
too
busy,
and
so
that's
really
what
I
really
like
to
see
here
is
it
I.
A
Have
a
failure
in
my
atomic
operations
and
I
actually
can
see
exactly
what
the
note
operations
were.
I
can
get
very
detailed
diagnostics
and
way
the
annealer
works.
That
one
thing
fails
and
it
stops
downstream
work.
So
it
gives
me
a
chance
to
fix
and
correct
other
things
will
flow
around
it
if
this
error
isn't
walking.
A
So
if
this,
this
node
role
isn't
required
by
other
things
and
what
I
can
do
the
I'm
not
going
to
retry
just
yet
I'm
gonna,
let
things
settle
down
a
little
bit
magma
systems
just
a
little
bit
bogged
down
so
retrying.
This
error
is
not
gonna
not
going
to
help
me
recover
any
faster,
actually
I'll.
Do
it
just
to
see
what
happens.
I
can
take
my
error
correct
to
the
issue,
in
this
case
pixie
dust
on
my
machine
to
give
them
your
cpu
and
ram
hit
retry.
A
A
And
it's
going
to
go
back
and
try
again
so
here's
dr.
to
its
back
in
this
role.
You
can
see
all
my
other
containers
have
advanced
beyond
that
and
they're
moving
through
the
actual
deployment
and
they'll
get
to
a
usable
state
very
quickly.
If
I
go
into
the
level
editor
I
can
start
moving
those
into
my
other.
My
deployment,
so
I
can
put
them
in
different
deployments,
depending
on
what
I
wanted
to
do.
A
The
reason
why
this
is
really
important
capability
for
us
is
that
it
lets
us
do
multi-node
scale
deployments
without
needing
a
lot
of
multi
a
lot
of
lot
of
VMs
or
a
lot
of
physical
infrastructure,
so
I
can
build
real
op
scripting
that
behaves
as
if
it's
a
scaled
out
system
do
hae,
build-outs
things
like
that
and
from
a
developer's
perspective,
even
a
QA
perspective.
I
can
go
through
those
samples,
much
much
faster,
a
lot
less
resources.
This
has
always
been
a
hurdle
for
us
in
any
of
our
testing.
A
A
A
A
Then
I
can
take
action
on
khobar
using
its
our
CLI
from
the
container
or
I
can
ssh
as
route
info.
So
I
just
want
to
check
back
in
I've,
got
all
my
systems
up
and
going,
and
so
basically
I've
now
got
a
system
that
I
can
continue
doing
to
Plymouth's
and
playing
with
and
otherwise
enjoy,
as
a
ops
environment
for
deploying
things
like
Seth
do
for
OpenStack,
that's
worth
noting.
Now
we're
still
working
on
bringing
those
workloads
and
other
workloads
into
the
curve,
our
environment.
A
The
point
of
these
demos
is
to
show
the
environment
that
deploys
the
base
core
for
thing
for
us
to
work.
On
top
of
that's
really
our
key
milestone
for
right
now,
so
hope
you've
enjoyed
this
demo
and
gotten
something
useful.
Please
join
us
on
the
curb
our
mailing
list
bit
ly.
Crowbar
dash
list
or
also
join
us
on
freenode
crowbar
is
our
IRC.
Thank
you.