►
From YouTube: #3 OpenCrowbar v2.1 ReadyState
Description
Adding Machines (VMs) to OpenCrowbar
A
Hello,
this
is
Robert
all
welcome
back
to
the
curb
our
install
video
series.
We
are
on
our
third
video.
We
are
now
at
the
booting
note
nodes
phase.
Hopefully,
you've
been
able
to
walk
through,
if
not
the
bootstrapping,
to
sort
of
see
a
UI
tour.
That's
where
I
toured
all
the
all
the
menu
options
and
then
probably
hold
off.
If
you
haven't
hope
you
can
hold
off
on
doing
the
installs
until
they
ready
to
walk
through
the
installs.
So
at
this
point
we've
already
got
the
system
running
I've
gone
through
an
initial
annealer
past.
A
See
so
that's
three
go
crazy
and
you
do
for
slaves
and
then,
from
this
point
you'll
see,
is
that
we're
basically
bringing
up
a
whole
bunch
of
KBM
instances.
The
KPM
slave
here
is
really
handy
for
testing.
If
you
checked
out
the
code,
I'd
automatically
build
a
kvm
and
attach
it
so
I'm
fan
of
being
able
to
do
it
like
that.
It's
it
saves
a
lot
of
time
in
repetition,
then
fundamentally
we're
just
going
through
this
process
crowbar
by
default.
A
When
a
new
system
comes
up,
it
owns
the
dhcp
server
that
DCP
server
is
going
to
direct
it
to
do
a
pixie
boot
pixie
boot
then
brings
in
our
our
sledgehammer.
Our
in
memory
image
does
discovery
and
management
and
inventory
from
there.
We
then
have
full
control.
When
we
can
do
ran
bios
configurations,
we
can
set
the
out-of-band
management.
If
there
is
such
it
is
pretty
much
a
generic
pattern.
A
We're
currently
working
on
bringing
other
patterns
in-
and
one
of
the
things
to
note
here-
is
that
we
don't
rely
on
dhcp
to
tell
us
the
system's
there.
We
actually
allow
system
to
boot.
All
the
way
have
the
image
itself
tell
us
it's
a
little
bit.
It
doesn't
tell
us
as
quickly
that
there's
a
system
there,
but
we
found
that
it
was
much
more
generic.
A
Let
us
not
be
as
tied
into
the
DHCP
and
trying
to
sort
of
backdoor
the
dhcp
server
in
this
case,
servers
come
in
through
the
front
door,
the
API
of
crowbar,
and
so
it's
possible
to
pre-populate
information
and
determine
names
and
IP
addresses
and
all
sorts
of
cool
things
by
pre
populating
the
database
or
allowing
you
to
bring
it
out
of
band
discovery
into
the
system.
We've
done
that
before
improve
concept,
it's
not
part
of
the
open
source
release,
but
it's
only.
A
We
talked
about
all
the
time,
so
an
example
would
be
the
out-of-band
management
ports
come
online.
They
beacon
in
one
way
or
another,
to
service
that
service.
Then
registers
in
crowbar
using
the
API
kovar
could
then
control
the
system's
out-of-band
before
they've
ever
been
booted
before,
in
fact,
for
their
powers,
you
can
enter
it
on
so
so
really
interesting
ways
that
you
could
scale
up.
Hardware,
discovery,
config
and
things
would
be
excited
to
talk
to
people
who
want
to
see
that
work
done
for
the
community.
A
A
So
these
are
the
nodes
that
have
booted
and
it's
important
to
realize
what
we've
done
here
is
we
actually
have
discovered
the
systems
and
before
we
touch
them,
we
go
in
and
we
make
corresponding
DHCP
entries
and
dns
entries.
This
is
exactly
the
type
of
system
orchestration.
That's
really
important.
Some
of
the
changes
that
we're
making
in
the
next
release
cycle
would
allow
us
to
notify
multiple
DHCP
servers
for
next
boot
information,
dns
errors
and
really
start
to
be
very
flexible
as
a
platform,
because
every
type
of
system
is
brought
up
or
changes.
A
These
entries
are
critical
control
points.
Every
time
a
nick
is
added
or
configured
in
a
system.
Dns
records
have
to
be
created,
both
ipv4
and
ipv6.
We
support
ipv6,
natively
and
then
use
IP
for
as
an
overlay.
We're
really
proud
of
that
choice.
It's
proven
to
be
a
good
strategy
for
us.
What's
going
on
right
now,
is
we
basically
pulling
all
the
pieces
together
and.
A
A
You
can't
hear
it
or
see
it,
but
my
hard
disk
is
spinning,
like
mad
from
the
virtual
machine
settling
in
my
video
recording.
So
you
can
see.
Cromer
gives
you
very
clear
guidance
when
it's
doing
things
to
different
nodes.
In
this
case,
you
can
see
exactly
what
it's
doing
where
it
is
in
progress
when
something
fails,
you'll
get
a
very
clear
indication
exactly
what
failed.
It's
always
fun
to
look
at
the
logs
for
these
these
steps.
So
you
can
see
within
this
individual
node.
You
can
see
the
attributes
that
are
passed
back
and
forth.
A
You
can
see
the
work.
That's
been
done,
the
work,
that's
in
progress,
so
very
clear
idea
of
what
the
workflow
is
for
the
nodes
and
then
some
very
detailed
logging,
so
that
you
can
actually
see
exactly
what
was
done.
Troubleshoot
do
diagnostics
and
if
something
hangs
up,
you
can
retry
and
go
back.
There's
another
view
for
this.
This
more
linear
shows
you.
The
things
are
actually
being
done.
Cromer
has
ten
simultaneous
threads,
that's
configurable,
so
a
larger
system.
You
can
boost
the
thread
counts.
A
It
shows
you
what's
pending,
so
you
can
only
do
one
sequence
of
work
in
each
node
in
multiple
nodes.
You
can
parallel
eyes,
and
so
this
work
is
waiting
to
be
done
after
some
other.
The
steps
can't
start
until
the
the
steps
can't
start
until
the
the
steps
that
are
ahead
of
them
in
line
go,
and
you
can
actually
look
at
individual
nodes
and
make
the
same
assessment
on
a
note
by
note
basis.
So
it's
really
the
core
of
the
curb
our
system.
A
Is
this
orchestration
and
annealing
that
goes
in
behind
the
behind
the
scenes
and
helps
figure
out
what
has
to
be
done?
What
order
in
a
physical
infrastructure
system?
This
can
include
out-of-band
actions
that
can
include
things
in
the
sledgehammer
in
solves
and
then,
of
course,
in
helping
coordinate
things
in
the
operating
system
itself
and
then,
when
you're
done
with
all
that
handed
off
to
another
infrastructure,
chef,
puppet
salt
ansible,
once
you're
above
the
readystate
layer.
A
In
this
case,
you
can
see
I've
completed
these
steps.
I
have
these
steps
remaining
I'm
in
an
error
state
and
then
I
have
a
log
souviens
DNS
client
failed,
which
is
what
I've
looked
at.
It
wasn't
able
to
do
it
and
is
telling
you
that
we're
looking
at
some
404
object
not
found.
There's
two
precondition
problems.
A
A
I'm
going
to
pause,
let
it
get
it
flush
through,
it's
actually
going
to
do
it
faster
than
I
can
pause
it.
So
we're
completing
the
last
couple
steps
the
discard,
node
discovered,
is
actually
no
op
roll,
so
it
doesn't
take
any
effort
to
do
it
several
lease
or
no
op
what
we
call
no
ops
or
milestone
rolls
or
targets
in
the
workflows.
Then
things
you
build
things
before
after
they
career
they
sort
of
create
these
stable
points
that
allow
you
to
be
very
flexible
at
how
the
orchestration
works.
A
A
Wizard
really
combines
multiple
steps
that
are
available
in
the
UI,
but
we
we
found,
we
did
them
all
the
time
in
a
certain
sequence,
and
so
we
built
a
wizard
that
allowed
you
to
do
that
and
fundamentally
it
lets
us
create
a
new
deployment.
That
deployments
are
scoping
boundaries,
so
I
can
have
multiple
boundaries
where
I
say
these
components
like
a
sequel,
server,
OpenStack
infrastructure
or
a
SEF
infrastructure
are
in
a
deployment.
A
They
have
a
cross
dependency
on
each
other
and
you
can
have
pair
of
parallel
deployments
that
don't
that
don't
sort
of
bump
into
each
other,
they
could
have
set
some
of
the
similar
networks
and
things
like
that.
Each
node
belongs
to
a
deployment.
So
this
will
put
these
nodes
in
this
deployment.
We
have
a
network
mapping
1g
one
is
the
firt
is
the
second
one
big
Network
Ten
g0
would
be
the
first
10
gig
network
and
there's
some
very
sophisticated
logic
and
ways
to
build
teamed
and
bonded
pairs
and
vlans,
and
things
like
that.
A
All
dynamically
I
named
that
range
and
then
I
give
it
a
IP
range
behind
the
covers,
there's
also
a
ipv6
range,
and
if
you
wanted
to
build
these
using
the
API
or
then
more
detailed
networks,
you
I
you'd
actually
have
a
lot
more
control.
We
found
most
people
want
just
these
basic
ones
and
then
from
here
I
have
these
discovered
nodes.
I
can
pick
different
operating
systems
that,
if
I
want
to
install
different
operating
systems
on
it,
it's
sort
of
fun
to
pick
make
different
choices
here,
creates
more
colorful
display.
A
As
the
install
goes
so
I
can
pick
those
different
choices.
I
can
run
my
wizard
running.
The
wizard
doesn't
actually
start
the
installs.
What
it's
doing
is
it's
creating
the
deployment,
the
network
setting
up
the
nodes,
putting
them
in
this
in
the
state,
and
then
we
have
a
review
step.
So
in
this
review,
step,
I
can
come
in
and
see
I'm
going
to
build
a
pilot
network
I'm
going
to
get
the
hardware
configured
we're
going
to
install
the
operating
system,
and
then
this
is
a
milestone,
roll
and
hardware
configured
to
milestone.
A
If
this
was
physical
gear,
I
would
have
a
lot
of
raid
and
bios
options
and
things
like
that
in
front
of
hardware
configured,
because
those
things
have
to
be
done
before
I
install
an
OS
and
I'll
show.
You
would
come
back
for
the
packs
back,
deploy
you'll
see
us
actually
add
things
into
the
ready
state.
In
this
case,
when
I'm
ready
for
this,
I
can
just
hit
commit.
The
commit
process
tells
the
annealer
that
it
can
start
working
on
these
nodes.
So
they
now
can't
be
edited.
A
A
Okay,
now
so
we've
been
running
ourselves
to
the
place
where
we
can
run
the
readystate
wizard
I
have
chosen
the
three
different
operating
systems:
I'm
going
to
run
the
wizard
this
case,
I've
got
my
different
bits
and
pieces
installed
if
I
wanted
to
add
additional
components.
I
could,
for
example,
if
I
picked,
OpenStack
I
was
doing
the
OpenStack
deployment.
A
A
We've
got
different
operating
systems
ready
to
go
when
I
commit
that
turns
it
over
to
the
annealer
and
the
annealer
is
going
to
start
doing
its
thing,
which
means
that
it's
going
to
reboot
the
different
notes
that
I
have
set
up
and
then
it's
going
to
start
the
install
on
each
one
of
these
nodes.
So
one
thing
to
note
that
crowbar
does
is
first
before
it'll
do
anything
else:
it
recompete
the
environment.
So
it's
rebuilt
my
pile
in
network
and
then
it's
going
to
go
back
through
and
do
the
reboot.
A
So
at
this
point
is
doing
the
hardware
configured
it's
installing
the
OS.
All
this
work
really
sends
us
back
into
the
system
configuration
it's
going
to
take
a
while
to
get
back
into
pilot,
so
the
real
action
here
is
in
system
is
very
important.
To
understand.
Crowbar
recognizes
that
it
has
to
apply
reapply
a
whole
bunch
of
logic
and
activity
against
the
machines
that
have
built,
and
so
as
it's
going
through.
A
What
you'll
see
is
I'm
basically
going
to
go
through
and
do
three
and
OS
installs
here
and
what
what
is
happened
is
crowbar
recognizes
all
right,
I'm
still
I
new
OS
I
have
to
go
through
and
reapply
all
of
this
work
to
reattach
this
new
operating
system
into
my
environment.
That's
a
significant
step!
So
I
went
to
sledgehammer
made
sledgehammer
part
of
my
environment,
fully
integrated
logging,
dns
dhcp.
A
All
of
the
things
I
need
to
actually
treat
that
the
machine
is
a
first
class
citizen
and
then
I
went
back
and
when
it
was
time
to
change
the
operating
system,
which
you
would
do
not
just
for
a
sledgehammer
discovery
but
for
any
OS
return.
When
I
install
that
OS
again,
I
have
to
reapply
all
of
the
system.
Integration
components
that
I
had
before
that
could
include
burnin.
It
could
include
hooking
into
your
monitoring
system
or
your
inventory
system.
A
I
could
install
include
installing
other
infrastructure
components
and
so
you'll
see
here
we're
actually
installing
three
different
operating
system
since
and
since
seven
this
is
65,
and
this
is
a
bun
to
all
coming
in
and
doing
doing
their
appropriate
operating
system.
Installs
it's
going
to
take
a
little
while
it's
literally
going
to
go
through
once
those
installs
are
complete,
crowbars,
going
to
reapply
all
this
work
and
then
jump
back
into
the
pilot
deployment
and
finish
it
all
the
way
through
down
into
this
and
what
I?
A
A
So
we're
back,
we've
completed
the
system
deployment
here
pilot
deployment,
so
everything
ran
through
I
click
on
one
of
these
machines,
you'll
see
they
have
different.
I
can
look
up.
Their
IP
addresses
this
one's
82.
Here's
the
US
v
6,
here's
the
pilot
network
IP
addresses
via
to
nicks,
so
we
got
found
on
both
can
also
look
at
that
using
my
network.
Node
map
and
I
can
find
the
different
IP
addresses
that
are
bound
you'll
see
the
admin
is
on
this
network,
not
on
these
networks.
A
Let's,
let's
tour
that
a
little
bit
by
the
way,
here's
my
med,
different
machines,
the
different
VMs
in
this
case.
What
I
really
want
to
do
is
I
want
to
log
into
them.
From
my
admin
note,
so
here's
my
admin
node
perfectly
normal
to
geta,
not
converge,
which
just
means
it
timed
out
waiting
for
the
production
script
to
run
alright.
So
now,
let's,
let's
play
on
a
little
bit.
I,
should
be
able
to
ping
192
168
1
2481
good.
A
A
81
I'll
do
a
couple
things.
First,
I
want
to
come
back
and
connect
my
bridge,
so
I
want
to
do
my
bridge.
Bridge
control.
I
want
to
add
an
interface
to
dr
0
in
this
case
I'm,
going
to
add
my
hardwired
port
80
you're,
going
to
want
to
have
a
port
that
you're
not
using
or
anything
else
or
if
you're
using
it
for
anything
else.
A
Just
make
sure
you
don't
have
IP
addresses
and
then
I
want
to
make
sure
I
add,
do
it
as
admin
IP
address,
add
I
want
to
add
192,
168,
1
24
dot
to
dev
docs,
dr
0,
and
what
that'll
do
is
that
will
allow
me
to
connect
in
to
that
network.
So
I
should
be
able
to
ping
it
now,
because
I've
installed
my
p
address
on
it.
A
A
And
now
I
am
connecting
to
that
machine.
Melissa
didn't
ask
me
to
login.
This
is
because,
when
we
set
up
crowbar
using
the
I'm
using
the
doctor
admin
right
now
when
I
do
that,
it
includes
my
keys.
If
you
include
your
keys
into
your
crowbar
container
and
that
list,
it
will
propagate
them
to
the
nodes
and
you'll
be
able
to
log
in
from
anywhere
you've
quit
your
public
keys.
We
do
that
as
a
convenience,
because,
when
you're
dealing
with
lot
of
machines,
you
need
to
propagate
these
keys.
A
So
that's
that's
pretty
handy
so
from
my
machine.
I
can
now
get
to
all
those
machines.
That
would
be
true
if
not
justified
VMs.
If
I
set
up
physical
notes
from
here,
my
my
system
would
now
be
able
to
access
those
house.
It's
really
important
when
you
start
playing
with
additional
deployment
stages
and
steps
like
with
ansible
or
chef
or
salt
or
above
it,
okay.
A
So
I'm
already
logged
in
keys
have
already
been
exchanged,
everything's
good
and
now
I
can
ping
1010
1010,
and
I
have
access
to
that
system.
So
now
that
I'm
in
that
this
is
a
little
bit
of
a
tortured
path,
but
now
that
I've
ssh
into
the
slave
I
have
access
to
both
networks.
They've
both
been
configured
correctly.
For
me,
a
really
important
thing
right,
crowbar
will
set
up
the
network,
so
you
ask
it
to
set
up
on
a
dynamic
basis.