►
Description
The recording of the network programmability stream which occurred on 2018/02/17
On this stream I explored NRE labs – the platform where you can learn network automation by doing labs directly in the browser https://labs.networkreliability.engineering/
The main creator of the project Matt Oswalt (https://twitter.com/Mierdin) also joined us in the chat and was answering questions!
I was doing Robot framework and StackStorm labs.
Timecode:
0:00 – Introduction
1:23 – Top of mind this week
3:25 – NRE Labs
8:00 – NRE Labs architecture
13:30 – NRE Labs – Robot Framework: Introduction
33:20 – NRE Labs – Robot Framework: Testing cases for Junos
43:10 – NRE Labs – Robot Framework: Best Practices
1:03:59 – NRE Labs – StackStorm: CLI and Basic Concepts
1:29:00 – NRE Labs – StackStorm: Actions
1:52:20 – NRE Labs – StackStorm: Workflows
2:15:05 – NRE Labs – StackStorm: Automating the Troubleshooting Chain
2:24:35 – NRE Labs – StackStorm: Sensors and Triggers
2:41:40 – NRE Labs – StackStorm: Events
2:48:15 – Impressions about StackStorm and Robot Framework
2:52:20 – Conclusions, summary and wrap-up
Resources:
Interview with David Bombal (about my background, network automation, Python, Nornir, Ansible, machine learning and certs): https://www.youtube.com/watch?v=U9hw0gustB8
NRE Labs: https://labs.networkreliability.engineering/
Please note that this content is stream-first and it is slow-paced by design. I recommend increasing playback speed in the player settings.
Don't miss my upcoming streams at https://twitch.tv/dmfigol
The code is on my GitHub: https://github.com/dmfigol/network-programmability-stream
Twitter: https://twitter.com/dmfigol
Blog: https://dmfigol.me
Background music (royalty-free):
https://www.pretzel.rocks/
A
Hello
and
welcome
everyone.
This
is
Dimitri
Amir,
hammock,
Nasser,
Network
programme,
unity'
stream,
and
today
we
are
going
to
cover
an
interesting
project
which
is
called
any
relapse
network,
reliability,
engineering
collapse
where
I
think
the
main
person
behind
them
is
Matt,
but
probably
it's
more
than
one
person
I
guess.
So
the
idea
behind
those
laps'
is
to
give
you
some
hands-on
knowledge
directly
in
the
browser.
A
How
would
you
approach
Network
automation
or
net
DevOps,
so
today
we
are
going
to
take
a
couple
of
laps
or
maybe
more
than
a
couple
and
go
through
them
together
and
I
will
provide
you
some
of
my
feedback
as
we
will
go
through
them
so
before
before
we
will
actually
do
that.
As
always,
I
will
share
a
couple
of
things
that
are
on
top
of
my
mind
this
week,
so
there
there
aren't
many
so
I
guess
it
will
take
just
five
minutes
by
the
way.
A
If
you
are
watching
a
recording
on
YouTube
in
in
the
description,
you
can
always
find
the
time
code.
So
if
you
are
not
interested
in
any
in
some
section
of
the
stream,
then
you
can
always
jump
into
the
what's
most
relevant
for
you.
So
a
couple
of
things:
the
I
had
an
interview
with
David
one
ball
during
Cisco
life.
Europe
David
is
just
an
amazing
person.
I
had
so
much
fun
talking
to
him
and
we
had
an
interview,
and
it
was
just
posted
last
week.
A
So
if
your-
and
we
were
talking
about
a
little
bit
about
my
background-
about
certifications
about
Network
automation,
Python
ansible,
a
little
bit
about
machine
learning
and
if
you're
interested
in
that
go
ahead
and
check
it
out
it's
available
on
YouTube
on
his
channel
now.
Another
thing
that
I
would
like
to
share
is
this
a
series
of
articles
I
read
just
yesterday,
which
is
the
state
of
Python
package.
A
It
was
just
recently
published
it's
serious
of
three
articles
and
it
goes
through
like
history
of
like
how
packaging
is
done
and
and
the
newest
paps
and
newest
additions
to
peep.
So
I'm.
If
you
are
a
library,
developer,
I
really
really
encourage
you
to
go
through
this
series.
It
will
be
really
helpful
for
you
to
understand
all
of
the
different
moving
pieces
inside
like
for
Python
packaging.
It
Sam
us
today,
shion's
and
how
it
can
change
your
water
flow
a
little
bit
with
a
new
way,
newest
PAP,
5.7
and
55.8.
A
So
I
do
really
encourage
you
to
go
through
this
article.
If
you
are
a
library
developer,
if
you
are
not,
you
will
still
benefit
from
understanding
like
how
piping
packaging
is
done,
but
it
will
be
just
a
little
bit
less
relevant
for
you.
I
really
enjoyed
this
read
so
encourage
you
to
check
it
out.
Okay,
that's
pretty
much
it
that
I!
That
was
on
top
of
my
mind
this
week.
So
let's
actually
jump
into
today's
topic.
A
So
as
I
told
you
before,
if
I
understood
correctly,
the
inner
relapse
is
hands-on
labs
that
you
can
do
inside
of
the
browser
where
you
can
get
familiar
with
some
of
the
network,
automation
and
some
of
the
net
DevOps
concepts
and
yeah.
The
majority
of
the
stuff
was
developed
by
Matt
Oswalt.
Who
is
right
now
in
the
chat,
so
you
can
say
hi
to
him
see
that's
the
benefit
of
you!
Being
on
the
live
stream,
you
can
actually
talk
to
some
of
the
creators
of
the
stuff
that
we
are
covering,
which
I
find
amazing.
A
A
If
you
go
to
this
portal
well,
if
you
look
in
the
Google
for
any
real
apps,
you
will
end
up
on
this
portal.
Let
me
also
unblock
all
of
the
JavaScript
here
to
make
sure
that
nothing
is
broken
and
we
will
go
to
find
lesson
content
and
we
will.
We
will
try
to
select
some
of
the
laps,
so
I
will
give
you
just
a
minute
in
the
chat.
A
If
you
have
I
do
have
some
of
my
personal
preferences,
but
if
you
would
like
to
some
of
these
laps
to
be
covered
today
on
the
stream,
please
do.
Let
me
know,
though,
please
also
keep
in
mind
that
some
of
these
topics
I
already
covered
in
one
way
or
another,
on
the
stream,
for
example,
like
Yama,
for
example,
we
had
ginger,
we
we
don't
really
have
kids,
we
definitely
had
wrested
behive
with
netbooks,
for
example,
so
I
may
not
necessarily
cover
the
topics
that
you
selected,
but
I
will
do.
A
I
will
try
doing
my
best
in
in
selecting
slaps
one
of
my
that
one
of
the
things
that
I
haven't
covered
at
all
was
and
I
really
want
to
was
robot
framework.
I.
Do
think
this
is
something
that
we'll
try
doing
today,
but
if
you
have
some
other
preferences,
let
me
know
probably
we
will
cover
several
of
them
today.
There's
questions
chat.
Can
I
ask
you
a
question
of
topic
because
I
need
to
go
to
sleep.
Have
you
shared
the
specs
of
the
server?
A
A
Doesn't
doesn't
really
matter
really
depends
like
how
much
how
many
devices
you
you
want
to
run.
Ok,
please
folks,
go
to
this
catalog
and
the
share
share
with
me
your
preferences
on
the
labs.
So
probably
we
will
not
go
through
the
fundamentals
on
the
left,
because
almost
all
of
this
we
have
covered
on
the
stream
in
one
forum
or
another.
Now
here
from
this
list,
I
don't
think
I
cover
G&G,
jeaious,
Jason,
IP
I,
don't
know
how
it's
pronounced
correctly.
I'm,
sorry
I
think
it's
it's
part
of
the.
A
A
We
did
have
something
about
salt,
but
we
recovered
it.
Very
briefly,
though,
we
haven't
covered
robot
and
this
stuff
on
the
right,
I'm,
not
yet
sure
what
exactly
it
includes
so
matching
the
troubleshooting
chain.
I'm,
not
gonna,
harp,
isn't
sure
what
kind
of
tools
you
are
going
to
cover
there.
This
sounds
interesting
to
me.
A
Stick
stick
compliance
well!
No!
So
I'm
not
sure.
What
is
that,
so
my
personal
preference
right
now
is
the
robot
framework,
so
we
will
start
with
that.
But
if
you
have
some
other
preferences,
what
we
should
do
after
that
lab,
please
let
me
know
in
the
chat.
Okay,
so
let's
go
and
I
think
it
would
actually
take
some
time
to
load.
Hopefully
yeah.
A
A
A
There
is
common
chat
in
an
interview
on
the
podcast
in
the
relapse
mentioned
that
there
are
options
on
any
real
apps
to
kind
of
build
your
network
topology,
but
I
didn't
see
such
options.
I
may
not
remember
it
well,
but
I
believe
you
said
in
your
labs
support
multiple
and
there's
hopeful.
You
will
cover
today.
Well,
I,
don't
know
I,
don't
know
because
I
well,
yeah.
B
A
Mean
I'm
not
really
part
of
this
project,
so
I'm,
just
the
user
here,
if
I
will
find
something
like
that.
I
can
definitely
try
doing
it.
I'm
sure
huge
architecture
based
in
the
docs.
Ok.
So
while
the
lab
is
loading
in,
it
seems
that
it
actually
has
already
loaded
I
think
it
would
be
good
to
go
through
the
swing
that
matt
has
shared
and,
let's,
let's
try
to
understand
a
little
bit
about
the
project
architecture.
A
Ok,
so
the
Google
compute
engine
is
being
used
on
top
of
it.
We
have
Burnett
is
now
I'm,
not
a
sure
what
syringe
and
antidote
web
is
being
used,
how
it's
described
the
right
here:
cloud
infrastructure,
terraform
instance,
config,
ansible,
now,
ok,
so
I
guess
the
ansible
is
being
used
just
to
provision
saying
is
ok,
so
I
guess
this
is
something
like
docker
and
cube:
CTL
installation
and
cube
admin
and,
like
that
I
think
so,
container
scheduling,
kubernetes
lesson,
provisioning,
sewage.
A
Too
costly
its
ingests,
a
very
simple
of
definition,
Chris
necessary
resources
in
the
cluster.
It
also
provides
an
API
for
upfront
and
to
know
how
to
provision
and
connector
up
resources:
okay,
also,
okay,
so
this
is
kind
of
like
how
you
do
on
the
back
end,
the
creation
of
the
proper
lap
back
end,
I,
guess
and
well
fronted
with
our
custom
web
application
based
on
Tomcat
and
guacamole
salsa,
fronting
guacamole
does
mean
see.
Jupiter
in
notebooks
helps
to
pick
through
the
lab
exercises.
A
A
Okay,
so
we
have
curriculum.
We
have
course
which
is
category
we
have
lessons
which
is
just
step
a
set
of
exercises,
and
we
have
labs
a
set
of
related
steps
forming
a
workflow
in
the
lesson
courses
who
is
the
list
of
course,
is
currently
under
consideration.
Development
troubleshooting
can
fix
elementary
security
testing
validation,
interest
code,
even
driven
infrastructure.
C
A
A
Yeah
syringe
an
antidote
web
are
the
custom
bits
we
develops,
which
talks
to
kubernetes
to
provision
resources
and
provide
an
API
and
to
that
web
is
what
you've
seen
the
browser
consumes
vary
in
shape.
Vi,
okay,
makes
sense
perfect.
Thank
you
so
much
Matt,
okay.
So
let's
actually
go
through
the
through
this.
Let
me
see
how
many,
how
many
laps
are
here,
so
we
have
entered
to
robot
framework
writing
test
cases
for
joonas
and
robot
framework,
best
practices,
and
it's
being
contributed
by
these
two
folks.
A
Chapter
1,
introduction
to
robot
fan
works.
Robot
framework
is
a
Python
framework
which
is
used
for
performing
to
perform
acceptance.
Testing
provides
a
simple
interface
to
write
test
cases.
Robot
is
keyword-driven,
testing
framework
will
or
all
actions
functions
of
the
destitute
are
performed
using
simple
human,
understandable
keywords.
A
Architecture,
shrining
diagram
and
so
get
the
user
guide,
so
it's
very
I
have
never
used
the
robot
framework
before,
but
I
definitely
heard
about
it
and
as
far
as
I
remember,
the
main
use
case
is
to
develop
DM,
this
kind
of
like
testing
interface.
If
you
will,
where
you
could
actually
write
all
English
sentences
which
are
tuned
or
converted
to
test
case,
so
you
could
write
something
like
check:
5,
bgp,
neighbors
on
device
acts
or
something
like
this
directly
in
like
English
in
English
right
and
then
this
the
fragment
itself.
A
B
B
A
A
A
A
A
A
A
Have
to
right-click
copy
my
environment
variables
used
to
store
string
environment
variables
and
our
reference
using
identifiers?
Okay.
So
this
is
directly
from
your
from
your
machine.
You
can.
You
can
specify
the
environment
variable
with
the
percent
sign.
Okay,
let
me
just
check.
Is
there
any
comments
in
the
chat?
Oh,
there
are.
There
were
many
comments
and
I
missed
all
that
hello
there.
Everyone
thanks
me
tree
for
streaming
first
time
here,
looking
forward
to
learn
paisa
Network
automation.
A
Thank
you
for
joining
I
hope.
You
will
learn
something
useful
today
on
and
on
from
my
previous
recordings.
Great
question
so
Matt
is
is
answering
great
question.
In
this
case,
all
network
devices
are
provisioned
on
demand,
as
virtual
machines
running
qmu
in
docker
can
in
docker
a
while.
So
you
actually
the
actual
built
a
container
risk
you
and
inside
of
it
you
run
the
the
virtual
network
device.
That's
pretty
awesome
since
we're
running
kubernetes
is
just
way
to
get
those
devices
running
without
calling
out
to
a
separate
hypervisor
makes
perfect
sense.
A
B
A
Can
be
Python
jaw,
okay,
to
use
the
keywords
defined
by
the
standard
external
private
libraries
first
need
to
be
important
in
settings
table
of
the
robot
file.
Built-In
keywords
can
be
used
out
of
the
box
without
the
need
to
import
any
library
in
the
settings
section.
Also,
it
should
be
noted
that
keywords
behave
like
functions.
They
can
take
in
arguments
and
can
return
that
was
back.
Okay,
let's
see
examples,
so
built-in
should
be
equal
change
on
this
built-in
keyword
should
be
cool
check.
A
A
I'm,
though
interested
you
say
that
it's
keyboard
is
a
subsidy
is
a
substring
automatically,
so
oh
I,
guess
it
just
what
removed
the
underscores
and
capitalizes.
This
final
is
desk.
Use
case
section
contain
the
test
cases.
You
would
like
to
execute
using
keywords:
variables.
Libraries
test
cases
are
executed
in
the
same
order
during
the
test
file.
A
A
You
could
kind
of
like
group
them
and
assign
the
names
for
these
tests
and
these
keywords
and
those
will
go
to
like
test
cases
so
I
guess,
for
example,
you
could
put
like
you
have
a
class
in
PI
test
which
is
named
with
something,
and
then
you
have
a
number
of
an
actual
tests
right.
So
something
like
class
name
in
this
case
would
be
a
name
of
this
use
case
name
of
this
test
case,
and
all
of
these
keywords
would
be
kind
of
like
your
functions
inside
of
that
class,
for
the
PI
test.
I
guess.
A
Okay,
few
important
syntax
points
to
remember:
there
should
be
at
least
two
spaces
between
the
following
key
words
and
key
words:
key
words
and
arguments,
arguments
and
arrogance.
Okay,
so
here
I
should
have
at
least
two
spaces
rhinock
you've
a
first
robot
file.
In
our
example,
we
have
a
file
called
chapter,
1,
eg,
1
robot
and
the
Python
file.
Substring
do
I;
okay,
let's
actually
examine
this
and
I
guess.
If
I
run
run
this
snippet,
it
will
execute
it
for
me
in
this
shell,
no.
B
B
B
A
A
So
I
mean
that's
perfect,
though
I
personally,
don't
like
it
since
I
like
to
invest
some
time
into
like
muscle
memory,
so
I
I
do
like
to
retie,
because
then
I
I
remember
more
what
exactly
I'm
doing
in
this
case
it
doesn't
really
matter
because
it's
just
a
cat,
so
I
think
it's
alright,
but
usually
when
I'm
learning,
new
stuff,
I
really
try
to
to
retype
stuff.
If
it's
not
super
big,
ok,
so
let's
take
a
look
on
this
function.
So
is
a
sub
string
str1.
A
If
those
two
strings
are
not
the
same,
raise
exception
is
not
a
sub
string.
Oh,
if
star
1
non
is
not
in
the
serger
okay,
so
this
makes
sense.
So
we
are
just
checking
if
the
first
drink
is
inside
of
the
next
drink,
and
you
can
actually
see
that
in
case
of
wrong
stuff,
we
are
actually
raise
an
exception.
I
thought
it
would
be
just
a
boolean
function,
but
I
know:
okay,
let's.
A
No
III
don't
want
to
rewrite
cat
okay,
so
you
load
this
Python
file.
We
have
our
two
variables
which
are
robot
and
then
we
are
checking
if
should
be
cool
as
strings
okay,
so
we
are
checking
if
robot
is
part
of
the
framework,
if
robot
is
part
of
the
robot
and
if
robot
is
part
of
the
robot,
but
with
a
proper
hello,
lower
and
upper
case
letters,
and
then
we
have
another
name
is
something:
oh
okay,
so
this
is
should
be
equal.
A
A
Okay,
so
we
have
this
robot
and
we
have
the
substring
pure
okay.
So
let's
execute
this:
let's
take
a
look,
so
we
have
check
equal
three
or
we
actually
have
some
details
here.
So
robot
is
not
equal
to
framework.
Robot
is
not
equal
to
lowercase
or
oh,
but
this
one
is
correct.
Check
substring
is
correct
and
then
Jun
is
not
substring
of
Juniper.
A
A
A
Let
me
also
see
what's
going
on
in
the
shed.
That's
one
buck,
I
still
need
to
hunt
down.
There
is
something
going
on
with
HTTP
tunnel
between
the
front
end
and
go
up
proxy.
Sometimes
randomly
disconnects.
The
lesson
still
exists,
so
it
will
reconnect
if
you
refresh,
but
it's
still
annoying
yeah
makes
sense
Oh.
A
A
The
next
day
container
running
back
in
we're
nice,
ok,
cool,
ok,
we
are
finished
use
this
lesson.
This
is
pretty
straightforward.
So
what
did
we
learn
from
this
lesson?
First,
how
you
can
use
the
robot
framework
to
use
the
like
using
built-in
keywords,
in
this
case
the
built
in
keywords
which
should
be
equal,
and
then
we
looked
how
we
can
define
your
own
function
and.
A
A
A
Okay,
so
okay,
we
will
how
to
communicate
vaginas,
use
keywords
from
Jonah's
private
library
and
pass
command-line
arguments
to
robots,
pass
arguments
for
private
functions.
Keywords
you'll,
use
your
chalk,
your
fix
connected
to
a
Linux
machine.
You
around
test
cases
uses
a
robot
on
the
links
machine
and
talk
to
take.
Your
IQ
fix.
A
You
can
install
it
okay.
We
already
had
this
dream
about
PI
easy
before.
Similarly
to
the
example,
in
the
previous
chapter,
we
will
create
a
Python
file.
We
will
define
functions
that
use
PI
C
library
to
connect
our
device,
Fitness
sister
information
clothes
echinacea
to
power
device.
Now,
let's
go
ahead
and
do
that.
I
think
this
should
be
pretty
straightforward
since
29.
A
A
Get
mom
okay,
so
in
this
case
these
two
things
either
get
hostname
or
get
model.
In
this
case
we
will
connect
to
the
device,
fetch
the
details
and
return
the
proper
key
and
then
I,
don't
think
cause
connection
here,
no
I,
don't
think
so.
So
we
actually
pan
connection
return.
Result.
We
open
another
connection,
return!
The
result.
We
don't
you
don't
close
the
connection.
That's
right.
Let's
find
the
Missy.
A
A
Yeah
I
want
to
have
it
here
so
that
I
have
some
kind
of
notes.
After
and
after
these
finishes,
because
I
have
one
note
where
I
put
all
of
my
learned
things
so
that
I
can
quickly
look
it
up
later,
yeah
I
think
all
of
now
it
will
be
a
pain
if
I
will
continue
doing
that.
Oh
it's
right
here!
Actually,
that's
what
I
should
have
done
to
begin.
B
A
A
A
A
A
But
without
all
these
secret
details,
okay,
I,
don't
want
to
retype
this
to
be
honest,
because
I
will
definitely
make
ten
mistakes.
So
it's
just
loop
is
closing,
so
minus
minus
variable
provide
host
the
same
thing:
minus
minus
variable,
user
password
and
then
the
robot
file.
Okay,
let's
just
run
this
okay,
it
will
take
some
time.
B
A
A
A
A
C
A
A
Okay,
so
it's
kind
of
like
you're
creating
function
which
does
some
tests
nested
in
this
case
in
terms
of
robot.
It's
just
a
keyword
using
resource
files.
Variable
files
have
the
same
sort
files
only
difference
being
they
can't
contain
test
cases
table
makes
you
have
to
go
through
this
I
thought.
I
couldn't
just
scroll,
but
unfortunately,
I'll
have
to
give
it
more
thought.
A
So
all
variables
and
user
keywords
can
only
be
used
using
a
file
when
you
say
ever
created
in
order
to
reuse
these
variables
across
multiple
files,
you
create
resource
files,
so
resource
files
started
the
same
era
what
files,
but
without
this
case
is
table.
Also,
the
settings
can
only
contain
the
important
files
in
the
dogs,
the
uses,
separate
yeah.
The
approach
helps
in
keeping
the
test
cases
problem
cluttered
in
you
can
keep
variable
files
as
a
Python
module.
A
All
variables
declared
inside
the
Python
module
can
be
directly
imported
and
things
a
robot
from
the
robot
framework
supports
to
two
times
of
setup
and
teardown
tests
set
up
to
test
teardown
suit,
desktops
to
them.
Okay.
So
this
is
very
similar
to
something
like
contest
in
PI
test
or
the
picture
which
is
limited
to
two
session,
for
example,
keywords
defined
on
their
test
setup
and
teardown
settings
around
before
and
after
every
test
case.
Givers
defined
on
this
suite
desktop
I
always
forget,
is
it
suit
or
suite?
A
A
Consider
scenario
were
in
that
conversation
session
is
initialized
and
terminated
before
an
hour
after
I
read
this
case.
In
this
case,
the
size
of
the
robot
file
can
be
reduced
by
using
test
and
teardown
settings
as
required.
Yes,
so,
for
example,
we
had
this
dream
recently
about
nor
near
respite
test,
and
we
had
kind
of
like
very
similar
stuff
where
we
would
set
up
our
connections
in
the
in
the
contest
so
that
it's
propagated
throughout
through
the
whole
for
all
these
cases,
so
we
don't
open
and
cross
connections
all
the
time.
C
A
C
A
B
A
Okay,
so
we
have
a
multiple
robot
files.
We
have
Gino's
device
which
I
think
it's
the
same
thing
well
the
same
file
as
we
saw
in
the
previous
step.
Yeah
it's
the
same,
and
then
we
have
multiple
robot
files,
which
we
will
try
to
understand.
So
we
must
have
noticed
that
the
new
keywords.
Okay,
first,
let
me
see
the
phone
okay,
so
okay,
so
this
file
loads,
the
Junos
device
py,
and
now
it
has
keyboard
section
where.
A
A
B
B
B
A
A
So
this
is
this
case.
This
would
be
something
like
a
register,
I
guess
non-civil
or
something,
and
then
we
have
all
the
date.
Facts
which
is.
We
are
gathering
facts
again
or
maybe
Locke
is
just
actually
printing
it
on
me.
Oh
I
think
I
know
what
it
is
right.
So
I
okay
look
I
guess
it
will
just
add
it
to
the
log
file
right,
the
invalidate
facts.
A
A
B
A
A
Automation
is
just
the
product
of
our
imagination,
so
I'm
pretty
sure
you're
all
doing
it
manually
and
we
just
start
hype
on
the
Twitter
that
we
should
do.
Network
automation,
but
really
it's
just
an
army
of
people
trying
to
play
console
cables
and
doing
and
building
those
config
files
manually,
as
we
run
those
playbook
scripts
or
whatever.
We
have.
A
Okay,
so
this
file?
Okay,
so
this
file,
I,
guess
it's
rework
of
this
previous
case,
where
this
file
is
referring
to
another
robot
file
as
a
resource.
We
have
test
setup
and
tests
teardown.
So
for
every
test
we
will
do
setup
actions
and
teardown
actions
and
then
test
cases
verify
facts
validate
facts.
Okay,
so
then,
let's
look
on
this
chapter.
Three
resource
file.
A
Ok,
so
then,
and
this
resource
file
contains
the
set
up
actions
in
teardown
actions,
very
set
up
actions.
We
connect
to
the
device
teardown
actions,
we
close
it
device
and
then
validate
facts.
We
are
checking
this
stuff,
ok,
so
this
is
just
just
a
way
to
like
rearrange
it
to
make
it
more
flexible,
and
let's
run
this
robot
for
now,.
A
A
A
But
it's
something
like
this.
It's
so
Oh
easy
for
me
to
get
this
hands-on
knowledge
to
reduce
this
fear
of
this
particular
tool
or
feature
whatever,
so
that
I
get
more
confidence
to
explore
it
on
my
own,
so
I
see
a
lot
of
value
in
something
like
this
I
think
I
think
this
is
just
amazing
that
you
have
this
this
capability.
You
don't
need
to
install
anything.
You
just
open
your
browser.
You
select
something
that
got
your
attention
or
got
your
interest.
You
spend
45
minutes
or
30
or
45
whatever,
and
now
you
were.
A
A
I'm
looking
if
I
want
to
I,
don't
really
want
to
do
Jean
jsj
sniping
napalm,
we
did
stack
storm
sounds
interesting,
not
sure
if
I
really
want
to
do
today.
Guns.
This
one
looks
very
interesting
to
me.
So
the
automating,
the
troubleshooting
chain
and
also
I,
would
like
to
see
stack
storm
as
well,
but
I
would
like
to
see
this
more
I.
Would
first
external
lesson:
ok,
Matt!
A
B
A
But
I
think
what
really
helps
in
this
particular
case
is
that
you
have
this
button
run
so
that
you
don't
have
to
like
you,
don't
necessarily
need
to
use
that
copy/paste
at
all,
though,
for,
like
my
own
notes,
I
would
love
to
like
have
an
artifact
of
my
learning
process
somewhere.
So
it
it's
a
small
issue.
It's
it's
not.
B
A
A
B
A
It's
it's
like
I'm,
just
providing
the
feedback,
since
you
are
on
the
stream,
but
I
don't
think
it's
like
big
deal
or
something
I
still
really
like
the
whole
idea
and
the
execution
so
easy
for
me
to
like
I,
don't
have
to
pay
anything
I
don't
have
to
log
in
any
anywhere
I
just
jump
into
the
lab
using
five
minutes.
I
spent
as
much
time
as
I
want
or
can
to
learn
something
new
right.
That's
just
amazing.
A
A
A
B
B
A
Oh
by
the
way,
I
have
a
question
for
you:
do
you
have
auto
scaling
for
the
for
the
like
notes
and
stuff
so
like
when
there
is
a
big
in
Lin
laughs
for
some
reason
right
many
people
doing
collapse
at
the
same
time,
do
you
how
to
scale
the
notes
in
the
containers,
for
that?
Did
you
have
any
issues
with
that
or
you
have
some
kind
of
like
limit
hard
limit
for
all
of
that.
A
A
B
A
B
A
A
A
A
Okay,
so
again
like
right
now,
I
like
don't
exactly
understand
what
I'm
doing,
but
let's
try
to
figure
it
out
right.
So,
okay,
so
this
is
text
or
debug
mode
run
core
local
I
guess
this.
Is
we
specify
that
we
want
to
run
this
something
afterwards
locally?
I
guess
so
this
would
be
like
local,
hosting
and
civil
I
guess.
A
A
A
A
A
Probably
tour
somewhere
and
then
we
query
requires
this
job.
You
say,
succeeded,
blocks,
parameters
and
then
in
the
result
we
have
the
CD
out.
Okay,
so
you
you
have
some
action,
you
create
a
job
from
it.
You
submit
it
using
HTTP.
Well,
rest
api
right
to
do
post
to
stack
store.
It
creates
this
task,
it
gives
you
back
ID
and
then
you
can
query.
The
idea
of
this
task
can
see
the
result
of
the
section.
A
Yes,
there's
commenting
to
chat
about
the
stats
page
position.
Yeah
I
would
also
agree.
It
was
a
little
bit
hard
yeah
like
if
it
was
on
the
in
the
middle
of
the
page
and
then,
if
I
could
also
check
this
to
change
it.
That
would
be
awesome.
Okay,
cool
now,
actually
I
start
to
getting
some
understanding
of
how
this
textur
works
a
little
bit.
A
A
B
B
B
A
A
Mentor
is
saying
just
wanted
to
say
thank
you
for
your
knowing
your
reads
as
well
enjoyed
watching
the
interviews.
David
fumble
at
Cisco
life
sure
celebrity.
Please
don't
say
that
like
what
great
work
thanks
for
shank
you
know,
thank
you
so
much
for
kind
words
really
appreciate
it.
Please
try,
please
don't
call
me
celebrity,
though,
please
I
know
not
better
than
anyone
else
so
I
like
sharing
the
knowledge
and
things
that
I
know,
but
it
doesn't
yeah
I
I
should
keep
my
echoing
check.
A
A
A
A
C
A
B
A
A
A
Napalm,
okay,
so
get
our
table,
get
config,
get
facts,
get
log
napalm,
pink,
okay,
so
those
are
the
usual
function
that
you
get
after
you
install
napalm
right.
Then
there
are
stuff
waterparks,
okay,
see
that
each
action
is
located
within
that
pack,
which
we
explained
in
the
last
love.
So
a
local
action
in
the
core
park
is
referred
by.
Its
full
name
is
Cora
dot
local.
We
can
pass
this
joystick
to
run,
execute
action.
Okay,.
A
B
A
A
so
what
this
does,
if
you
like,
more
familiar
with
spices,
for
example,
and
probably
for
something
like
this,
you
would
use
celery
in
Python
right,
which
is
tasks,
tasks
you
right,
you
have
a
function
and
then
you
launch
it
from
different
place
with
dot,
dot,
delay
or
dot,
apply
sync,
and
you
get
an
idea
of
the
job
instantly
so
and
the
job
itself
is
being
put
on
the
on
the
queue
right.
So
in
this
case
it
would
be
very
similar
right.
A
B
A
In
this
case,
to
kind
of
like
celery
architecture-
well,
it
will
take
take
on
exams
during
execution,
is
used
to
describe
an
instance
of
running
action
each
time
around
the
nation.
It
gets
an
ID,
be
told
to
run
this
asynchronously
it's
useful,
because
you
will
see
the
next
slide,
but
don't
always
run
actions
on
the
command
line
directly
like
this.
Sometimes
it's
done
for
a
spiral
or
a
workflow.
In
this
case,
retrieving
execution
details
using
AC
to
st
to
execution
get
is
often
often
the
only
way
you
know
action
is
performed.
A
A
B
A
B
B
A
A
B
A
A
B
A
A
B
A
A
B
A
B
A
C
A
A
Okay,
we
ran
these
log
config
config
file,
but
start
to
succeed.
It
and
then
result
load
merge
successful.
If
you
go
to
this
device
and
every
not
around
show
BGP
summary
yeah
I
can
poke
around
the
back
ends,
don't
worry
about
it.
Okay,
then
I
will
just
kill
this
love.
I,
don't
think
it's
killing.
It.
B
A
Up
perfect,
this
was
a
brief
look
at
stick:
storm
actions
using
napalm
puck,
it's
extremely
useful
for
our
purposes
here:
learning
even
driven
Network
automation.
Those
are
many
more
actions:
yeah,
okay,
so
okay,
let's
kind
of
like
do
brief
summary
of
like
this
text
storm
in
this
case,
so
we
saw
like
how
to
run
actions
which,
in
the
world
that
I
know
right.
It's
just
well
in
case
of
ansible.
It
would
be
uncivil
module,
but
it's
just
you
run
something
right.
You
do
something
that's
pretty
much.
A
A
A
To
be
honest,
so
far,
xu
likes
tech
storm,
especially
compared
to
something
like
ansible
this.
This
is
my
like
first
impression
so
far,
but
we'll
see
just
like
real
tasks
are
performed
manually
or
by
another
tool.
Actions
almost
never
run
in
isolation.
They
usually
run
in
parallel
with
each
other
or
in
sequence.
They
usually
run
based
of
some
decision.
A
A
Okay
example
1
action
chain,
those
are
simplest,
but
also
the
least
robust
workflow
option
is
text
or,
if
you
want
to
run
a
sequence
of
actions
with
some
minimal
error,
handling
action
chains
are
probably
sufficient
for
your
purposes.
There
are
the
bare
minimum
or
close
option
instead
saw.
The
Holt
example
for
the
action
chain
has
to
be
a
coaching
in
particular
examples.
B
A
A
A
A
B
A
Check
up
a
couple
of
comments
in
the
chat,
let
me
I
didn't
check
the
salt.
Did
you
see
how
I
change,
how
I
did
not
come
cast
a
spin
come
cast
on
my
view
of
the
stream
keeps
cutting
in
and
out,
if
I
will
join,
if
I
can
put
in
case
not
create
content
test
for
thanks
for
the
valuable
feedback?
Thank
you
so
much
for
joining
Matt
I
really
do
believe.
This
is
something
really
cool,
especially
for
people
who
are
just
starting
and
don't
necessarily
know
all
of
these
tools.
A
I
think
it's
just
an
amazing
resource,
and
it
definitely
so
after
well.
We
will
have
like
a
whole
summary
ten,
but
I
can
tell
you
that
in
relapse
is
definitely
going
to
a
list
of
resources.
I
recommend
to
people
who
are
starting
with
Network
automation,
Network
programmability,
so
thank
you
so
much
for
creating
it.
I
think
it's
just
an
awesome
idea
and
very
very
cool
execution
as
well.
B
A
So
Mistral
is
a
more
powerful
than
action
chain.
Well
action
chain,
excelent
simplicity,
it's
the
simplicity
that
kinda
times
prove
inadequate
for
some
decision-making
power
or
execution
style.
We
need
in
our
workers.
This
case
is
some
of
the
advanced
feature
that
mr.
Auburn
brings
to
the
table.
Improv
useful.
However,
unlike
action
chain,
Mistral,
is
not
a
stack
project
and
out
part
of
these
texts
on
foot
base.
C
A
A
A
A
Okay
and
I
think
something
like
this
is
actually
kind
of
like
the
branching
here
right.
So
I
guess
one
success.
We
execute
task
a
if
the
past
contains
a
well.
If
the
past
is
a
we
execute
B
if
the
past
contains
B
and
C,
if,
if
neither
one
of
those
yeah
I,
think
that's
what
is
happening
so
we
would
say
echo,
so
we
provide
an
input,
a
B
or
something
else.
A
We
will
print
this
well,
we
will
echo
this
right.
We
will
save
the
output
of
printing
to
the
past
variable
and
then
depending
on
which
one
of
those
we
will
say
echo
to
pass
a
to
pass,
be
able
to
pass,
see,
I,
think
I,
get
some
understanding,
so
underscore
dot
is
just
kind
of
like
self
in
Python
kind
of
think,
so
maybe
I
make
zero
sense.
A
B
B
B
B
A
A
B
A
A
A
A
A
A
So
yeah
you
can
trigger
based
on
some
logs
or
well
in
case
of
logs.
You
could
use
napalm
logs
so
that
all
of
these
processing
is
done
done,
centralized
and
trigger
some
actions.
But
if
you
use
something
more
than
just
two
logs
right,
then
you
could
set
up
a
em
script
which
will
do
HTTP
GET
to
some
place
and
will
trigger
workflow
that
you
need
to
trigger
I
mean
the
whole
problem
resolved
like
given
based
automation
is
how
to
get
that
event.
A
A
A
Actually,
I
think
we
I
think
we
could
go
through
this
lab
quickly.
It
seems
like
only
one
scenario,
so
this
lesson
you'll
talk
about
breaking
common
workflow
down
to
smaller
pieces
using
knowledge
began
from
previous
lessons.
We
put
a
small
script
for
each
piece
of
the
work
phone
just
ring
the
small
scripts
together
the
Linux
command
line.
In
order
to
revealed
our
work
loss
scenario,
an
IP
phone
user
casts
call
to
report
problems
to
their
audio
they're,
unable
or
unwilling
to
provide
any
additional
information.
Our
go
over.
A
This
lesson
is
to
find
the
name
of
the
access
switch
and
the
interface
of
their
phone
is
attached.
To
this
first
step
we
have
created
a
small
script
or
three
phi
p
address
of
the
phone
from
IP
pbx,
giving
the
extension
number
of
the
phone.
Before
we
run
the
script,
we
will
have
to
start
the
support
software
located
on
and
C
phone
C
phone
box
run
this
okay
folks,
I
have
no
idea
about
voice
at
all
of
the
phone
is
now
registered
to
the
PBX,
see
this
yourself.
It
enter
in
this
seafoam
lab.
A
A
A
B
A
Okay
and
then
we
parse
xml
from
asterisk,
and
we
get
to
date
how
many
enterprise
customer
getting
IP
address
of
a
software
hardware,
client
such
as
phone
little
machine,
hosted,
chat,
client
or
storage
element
can
be
difficult
time
consuming,
meaning
a
script
such
as
this,
where
you
can
pass
in
no
network
data
such
as
accession
number
3.
If
network
data
such
I
P
address,
can
save
time
and
frustration,
yeah,
ok,.
B
A
A
B
A
Hey
there
were
a
couple
of
questions:
I'm,
sorry,
folks,
I
was
really
focused
on
the
lab,
so
I
didn't
didn't
pay
attention,
though,
if
I
wanted
to
learn
about
knowing
what
one
document
explained,
what
it
is,
how
it
works
and
how
to
use
it.
I've
been
looking
the
github
yeah,
okay,
so
I
would
just
go
to
nor
in
your
Docs
and
so
go
to
nor
in
your
Docs,
and
you
just
go
like
page
by
page
there
they
are
very
small
so
like,
and
we
really
try
doing
our
best
to
have
this
concise
and
up-to-date.
A
So
you
see
like
also
like
one
screen
things
and
then,
if
you
like
go
forward
to
logo
installation,
one
of
the
things
that
we
do
in
our
dogs
is
bu.
We
use
Jupiter
notebooks
directly
as
part
of
our
dogs.
What
this
allows
us
to
do
as
part
of
our
continuous
integration.
Whenever
we
push
any
new
codes
to
our
repo,
we
actually
run
the
our
tutorials
and
verify
the
output
is
the
same,
and
if
the
output
is
not
the
same,
then
the
CI
fails.
A
So
this
is
like
a
great
way
to
like
just
follow
along
with
these
tutorials
rear
and
then
rerun
them
now,
since
you
asked
me-
and
you
could
also
use
the
workshop-
that
I
built
so
I
had
the
Cisco
life
here
of
the
workshop,
which
you
can
clone
and
use
it,
so
it's
deaf
so
actually,
first,
let
me
close
the
dots
dog
dogs
here
and
the
workshop
is
here.
I,
don't
really
have
like
a
lot
of
details
here,
but
what
you
need
to
do
you
have
to
run
Jupiter,
and
then
you
launch
this.
A
This
just
Christ
not
plate
of
the
notebook
and
it
kind
of
walks
through,
like
the
main
features
of
the
whole
framework
like
okay.
What
is
the
inventory?
How
do
you
do
simple
tasks?
How
do
you
do
simple
net?
Miko,
simple
Napalm?
How
do
you
deal
with
connections?
How
do
you
retrieve
different
data?
How
you
change
the
data
filter?
A
B
A
Back
in
business,
so
how
far
we
have
seen
a
few
options
of
doing
works.
Okay,
so
what
we?
What
did
we
see?
We
did
see
packs.
We
did
see
actions
which
you
can
run
in
blocking
mode
or
in
a
sync
mode,
and
we
saw
simple
workflows
where
we
were
changing
tasks
and
we
had
three
different
ways:
you
can
chain
it.
A
A
This
automation
in
a
Thomas
way-
in
other
words
you
don't
want
just
define
what
to
do.
We
want
to
define
to
do
it,
not
Air
Force
tech
storm
to
know
when
a
certain
event
has
taken
place
needs
to
have
some
kind
of
process
of
gathering
the
data
about
other
systems
or
processes
needs
a
mechanism
for
gathering
error
data
so
that
it
can
make
a
decision
about
when
they
take.
A
A
B
A
A
If
you
won't
censor
the
post
external
system
yeah,
so
we
had
the
polling,
we
see
a
lot
more
detail
about
the
sensor
greasing
the
field
artifact
awry,
which
is
an
absolute
pass
to
the
actual
Python
quite
running
the
sensor.
The
content
of
the
sensor
is
a
bit
beyond
the
scope.
Okay,
we
already
did
that.
Let's
see
what
it
does.
A
A
A
Okay,
we
can
be
saved
the
status
about
the
interfaces
to
the
self
device
state
and
then
not
on
every
run.
We
look
on
the
current
details
of
the
interface
and,
if
those
are
different
and
we
trigger
something,
she
is
actually
quite
a
bit
of
code
here.
It's
not
like
it
takes
some
work,
but
okay,
now
it
kind
of
like
makes
sense.
It's
impossible
to
talk
about
sensors
without
DoCoMo
triggers
sensors
would
have
no
point
in
existing
if
it
wasn't
for
triggers.
A
You
think
of
sensor
is
a
bit
of
fairly
self-contained
code
for
constantly
bringing
in
information
about
some
external
entity
or
system
it's
after
this
sensor
to
determine
if
the
information
being
brought
is
actionable.
For
this
instance,
interface
sensor
periodically
pause
our
network
devices
for
interface
information.
If
nothing
changes
fuse
the
respect,
our
device
network
interfaces,
no
meaningful
event
has
taken
place.
However,
if
we
disabled
an
interface
next
time,
our
sensor
pulled
that
interface.
It
would
notice
a
change
and
it's
time
to
notify
stack
storm
that
this
meaningful
event
has
taken
place.
A
Yeah
look
at
this.
We
say
we
will
say
interface
up-down
trigger.
If
down.
Let
me
find
the
if
up
or
if
down,
I,
think
it's
like
a
constant
at
the
top
yeah
it's
right
here.
So
we
have
this
if
up
and
down
and
then
at
some
point
we
trigger
there
and
then
we
also
define
them
in
gamma
the
same
one,
which
is
napalm
interface
down
and
napalm
interface
up,
which
will
also
return
device
in
interface.
Okay,
thanks,
perfect
sense,.
A
You
can
see
the
triggers
have
properties,
oh
just
what
I
heard
I
discussed
so
device
an
interface.
So
how
do
we
know
when
any
man
has
actually
taken
place
and
trigger
has
been
fired?
For
this
view
slightly
different
term
trigger
instance,
it's
a
special
resource
type
using
stack
stone
that
represents
an
instance
of
trigger
it's
kind
of
like
he
went
I
guess
this
very
similar
to
relationship
between
actions
and
executions.
A
A
A
A
Yeah
I
I
have
a
little
bit
like
mixed
feelings
about
this.
What
idea
yeah
like
what
you
could
do
and
I
know
you
can
do
it
on
like
Cisco
and
probably
you
can
do
it
on
juniper
as
well.
Yes,
as
I
described
to
you,
like
I,
am
more
familiar
subscribe.
You
would
create
an
e/m
script
which
looks
for
the
change
directly
on
the
device.
A
It
could
be
like
routing
table
change,
interfacing,
whatever
right,
and
then
it
triggers
a
REST
API,
and
this
REST
API
could
actually
refer
to
likes
text
or
trigger
I'm.
Pretty
sure
you
can
always
you
can
trigger
the
trigger
trigger
trigger
remotely
I.
Think
so
I
mean
it
would
be
pretty
illogical,
so
your
device
once
it
goes
to
HTTP
as
like
10001,
which
is
use
text
or
minced
and
slash
whatever
like
api,
whatever,
whatever
whatever
and
then
it.
A
A
Well,
it
seems
we
keep
the
connection
open.
So
that's
good,
okay,
it's
it's
more
or
less.
Okay.
We
keep
the
at
least
the
connection
open
you
tracking
per
device,
not
state.
So
we
have
like
an
open,
SSH
connection
or
something
but
yeah
I
mean
it's
still
a
lot
right.
You
basically
every
five
seconds
you
go
to
device.
Even
if
you
have
a
connection
already
open
you
go
to
the
device
and
say
show
interfaces
right,
and
then
you
parse
that
data,
and
then
you
make
a
decision
based
on
it.
A
Imagine
if
you're
trying
to
define
more
and
more
of
those
triggers
so
yeah.
There
is
one
idea
in
the
chat
like
couple
it
with
telemetry.
Yeah
I,
agree
with
you,
so
well.
I.
Don't
know
like
how
it's
actually
done
in
stacks
term
right,
but
you
would,
you
would
use
some
telemetry
to
get
data
and
telemetry
was
invented.
All
the
like
protocols
are
for
invent
that
with
a
use
case
of
either
polling
or
on
change
notification.
Those
are
protocols
that
are
designed
to
do
specifically
this
right.
A
A
A
A
A
A
A
True
can
be
seen
here
continue
to
up
to
each
component,
so
on
C.
So
what
is
this
rule?
So
it's
a
Yama
file.
It's
the
clarity
file.
It's
got
to
replace
interface
config,
which
is
for
interface
down
event
from
from
napalm,
since
I
can
fix,
nip
it
to
the
device
to
turn
the
interface
back
up
trigger
napalm
interface
down
criteria,
nothing
okay!
So
when
we
think
when
this
trigger
occurs,
you'll
take
the
action,
napalm
load
config
with
these
parameters
and
oh
by
the
way
you
can
see
that
it
says,
trigger
dot
device.
A
A
B
A
B
A
Action
criteria
trigger
okay,
double
output
showed
us.
Texture
has
ingested
our
new
rule
definition
and
is
ready
to
which
for
new
trigger
instances.
Now
the
rules
in
place,
any
new
instances
of
trigger
will
be
handled
by
this
rule
in
the
name
of
by
Verity.
Damn
I
can't
pronounce
it.
Let's
tinkering
in
event,
on
ivq
of
thanks
to
configure
set
interfaces
here,
for
example,
commit
okay.
It
was
going
run
it
this.
Yes,
you
this
before
we
can
check
instances.
B
A
A
So
we
created
an
action
from
that
trigger
we
here's
what
it
is.
It's
basically
trigger
instance
led
to
do
the
action,
which
is
the
execution
ID
here.
This
is
a
very
handy
tool
for
seeing
our
rule
in
action.
If
you
look
at
the
output
you'll
notice
that
the
trigger
instance
ID
show
ID
of
the
trigger
instance
that
was
handled
handled
by
the
stroke
as
well
as
execution
idea
of
the
execution
of
the
action
execution
that
was
created.
Remember
how
we
retrieved
the
execution
details
asset
to
execution
details
will
get
it
for
us.
A
Yeah
so
now
we
queried
a
stick
to
execution
again
based
on
this,
and
then
we
see
that
it
applied
config
file
on
the
Q
of
X,
2
and
result
was
more
successful,
looks
like
our
load,
X
and
O
configured
good
successfully.
Let's
look
on
interfaces
on
the
m4
on
the
interface
m4,
and
you
see
that
our
interface
is
up
again,
so
to
recap,
actions
and
work
clothes
from
used
to
performing.
This
would
asked
what
about
my
network
censors
triggers
help
us
integrate
with
external
systems.
A
Monitoring
platforms,
network
devices
rules
also
allow
us
to
describe
it
this
than
that
logic.
That
ties
triggers
with
actions
no
more
closed,
so
we
don't
have
to
execute
scripts
ourselves,
ok,
cool,
okay!
So
let
me
I'm
going
to
right
now
to
wrap
up
since
we
are
approaching
almost
three
hours.
So,
first,
let's
wrap
up
about
sex
storm
right,
so
I
have
never
seen
stack
stone
before
in
my
life
and
to
be
honest,
I
I'm,
intrigued,
I
think
it's
an
interesting
system
that
probably
worth
exploring
I
see
it
has
much
more
power
than
unstable.
A
A
So
in
general,
I
I,
like
the
approach
I
I'm,
not
sure,
like
how
efficient
the
system
is
and
stuff
like
that,
because
ansible
super
slow,
for
example,
but
I
am
a
little
bit
in
tricked
I'm,
not
sure
if
I'm
going
to
start
using
it
for
my
projects
yet
but
I
think
it's
worth
trying
actually
yeah.
So
this
was
about
texture.
We
also
had.
A
A
They
are
like
scared
to
change
something
like
they
like.
The
users
are
not
proficient
enough
to
be
able
to
change
like
testing,
and
they
want
to
use
like
English
like
language
to
define
their
tests.
So
I
personally,
I-
don't
have
like
this
since
me
like
when
I
knew,
testing
I
can
use
like
PI
test
or
something.
A
A
It
wasn't
super
impressed
if
you
just
run
one
script
and
it
gets
the
data
from
one
place,
feeds
these
data
processes,
the
output
from
some
database
and
returns
it
to
you,
I
I,
don't
think
it
was
super
impressive.
To
be
honest,
but
you
know
you
can
build
something.
It's
something
interesting
there
I'll
later
now,
let's
go
to
the
feedback
around
the
hole
in
the
relapse.
So
I
already
told
some
of
my
impressions
during
the
stream.
A
A
You
don't
need
anything,
you
don't
need
to
install
anything
locally.
You
don't
need
to
find
a
proper
tutorial
for
your
appropriate
software
version
or
something
you
just
go
to
the
portal.
You
click
on
the
lab,
lab
provision
to
use
in
five
minutes
or
so,
and
you
just
follow
the
directions
to
learn
something
super
easy
they're,
a
couple
of
like
small
technical
things,
but
this
project
was
just
born.
What
like,
four
months
ago,
it's
still
too
early,
and
even
at
this
stage,
I
I,
think
it's
a
great
learning
source
and
I
would
encourage
everyone.
A
Well,
if
you
are,
if
you
are
watching
this
stream,
probably
it
means
that
you
are
somewhat
interested
in
network
automation
or
network
program
ability.
So
if
you
are
one
of
those
people
and
I
hope,
you
are
I,
definitely
recommend
you
going
to
energy
labs
on
every
topic
that
you
don't
think
you
know
enough
like.
A
Yes,
there
are
like
tools
in
this
on
this
list
that
I
have
I
haven't
tried
like
J
snappy
stack,
storm,
robot
I
use
some
of
other
tools,
but
so
I
didn't
do
labs
for
them,
but
this
is
two
that
we
explore
today
haven't
seen
before,
so
it
was
so
easy
to
start
to
get
started
and
it
definitely
reduce
these
barrier
this
entrance
per
year.
For
me,
because
look
if
you
look
on
the
whole
network,
automation
at
the
Bob's,
there
are
so
many
different
projects
and
tools.
A
A
You
spend
30
to
45
minutes,
maybe
even
less
go
through
the
lab
and
you
already
get
some
hands
on
the
hands
on
knowledge,
hands
on
skills,
and
then
it
can
make
a
decision
AHA,
just
something
interesting,
I
see
use
cases
in
my
environment
or
maybe
okay.
This
looks
interesting,
but
I
don't
really
care
about
this
tool.
So
then
you
go
ahead
and
select
something
else
right
and
you
are
feel
much
more
comfortable
because
then
the
next
step
would
be.
You
could
install
this
tool
on
your
local
machine.
A
You
can
install
some
network
devices
and
try
to
replicate
the
same
steps
of
the
lab
put
in
your
environment
and
so
that
you
gain
this
gradual
knowledge
of
this
all
these
different
tools,
because
I
do
think
that
the
whole
net
DevOps
right
now
super
overwhelming
area.
There's
just
so
many
tools,
sometimes
I'm
tired
of
all
these
different
tools
like
unlearn.
This
thing
I
feel
like
it
solves
my
needs
and
I'm
hesitant
to
check
something
other
because,
like
it
will
require
a
lot
of
time
to
learn
this
tool
to
the
same
level.
A
A
A
The
risk
questions
chat
is
this:
introduce
you
to
configuring,
routers
and
switches?
Well,
it
depends
if
you
are
talking
about
automating
configuration
of
the
devices
and
yes,
if
you
are
talking
about
how
to
configure
actually
the
routers
and
switches
in
actual
technologies,
then
no
it's
just
about
Network
automation,
Network
program
and
division.
A
Okay,
folks,
I,
don't
see
any
other
like
comments
or
questions
in
the
chat.
I
would
like
to
say
thank
you
to
Matt
Oswalt
I'm,
sorry,
if
I,
don't,
if
I
pronounced
your
name
incorrectly
I
think
you
did
an
amazing
work
here.
I'm
looking
forward
to
hearing
more
and
to
to
see
more
lessons
in
the
future,
I
do
believe
you're
on
the
right
track,
and
this
is
something
that
was
definitely
missing
in
the
community.
So
thank
you
so
much
for
creating
this
working
on
this.
A
Also,
thank
you
so
much
for
joining
the
stream
today
and
answering
some
of
the
questions
in
the
chat.
It
was
super
cool
so
and
with
that,
I
would
like
to
say
thank
you
to
people
who
watch
the
stream
today.
The
recordings
will
be
posting
posted
sometime
next
week
and
I
will
see
you
next.
Sunday
bye,
everyone.