►
Description
En esta charla, Alejandro nos presenta Civo, una compañía Cloud de Reino Unido (UK), la cual brinda servicios de Kubernetes basados en K3s y servicios IAAS. Alejandro también nos cuenta sobre los retos para Civo del año presente, en el cuál fue creada una plataforma desde 0 para manejar los distintos data centers de Civo alrededor del mundo. También compartirá cómo está organizado el ecosistema de Civo y los beneficios de Kubernetes para los usuarios de Civo.
Al finalizar la sesión, tendremos un espacio de preguntas para que los asistentes puedan preguntar cualquier duda sobre Civo. Esta será una sesión única en español, para que los hispanos conozcan Civo y K3s para sus soluciones diarias.
Contacta a Alejandro en:
- https://www.linkedin.com/in/alejandrojnm
- https://twitter.com/alejandrojnm
A
A
B
B
B
'm
going
to
talk
about
what
we
say
about
things
like
this
and
well,.
Basically,
when
I
spoke
to
Sergio
the
first
time,
I
told
him
that
I
wanted
the
presentation
to
be
about
issues
'night
work',
that
is
to
say
that
people
will
ask
what
more
questions
I
could
say
and
well.
Basically,
let's
see
like
this
witness
archive
was
born
from
a
team
that
already
had
more
than
a
decade
of
experience
in
the
tool
sector
and
had
already
founded
a
hosting
company
in
the
uk
in
2018.
B
B
Quickly,
the
follow
I
still
directed
attention
to
four
months
of
spots,,
a
light
distribution
of
govern
to
base
on
ordinary
and
it
fit
perfectly
with
the
objective
we
were
looking
for
from
the
platform,
that
is,
I
included
myself
at
the
beginning,,
although
it
is
not
always
part
of
the
evaluation
of
I
still
air,,
but
hey,
I
included
myself.
I
was
among
the
first
beto
texts
of
this.
B
B
From
scratch
to
be
able
to
manage
everything
that
the
users
asked
us,
basically
with
three
bots,
if
it
is
only
known
that
the
city
tower
turbinates
us
for
not
to
build
other
formulas-
and
we
decided
to
do
it
like
this-
we
use
cover
net,
no
write.
We
use
stories.
Veronesi
key
that
basically
comes,
are
doctors
of
our
reform
platform.
B
B
We
have
every
service
that
board
word
has
and
we
provide
our
own
hardware,
because
I
have
read
several
conferences,
and
everyone
asks
me
the
same
thing
about
the
train
about
the
group.
It
runs
about
doing
and
we
don't
pass
anything
to
opel
with
the
project.
It
is
a
hardware
created
avenue.
According
to
the
specifications
of
the
platform.
Today
I
am
going
to
present
a
small
video
of
when
we
indicated
a
traffic
and
the
hardware
that
is
currently
used,.
We
have
a
data
center.
B
B
They
will
see
all
the
infrastructure
is
that's
how
it
is
and
that's
how
the
nodes
are
deploying
the
infrastructure
archive
can
take
time.
Well,
if
discounting
the
transport
of
the
vespre,
that
is
the
infrastructure
provider
until
he
says
we
can
stop
for
30
to
40
minutes,
deploy
a
complete
region
already
functional
available
to
provide
service.
B
C
A
A
As
in
a
passive
context,
it
is
like
a
cloud
where
a
service
is
offered.
That
is
a
superlight
horn.
You
can
replace
it
in
Instead
of
using
txiki,,
which
is
and
what
it
is
and
whatever
they
want,,
but
using
that
there
are
three,.
So
that's
like
the
Cíbola
core,.
So
how
do
they
manage
thousands
of
clusters?
How?
Do
they
have
those
regions?
How?
Do
they
have
virtual
ones?,
Well,
I?
Think
we
already
have
a
question
here.
It's
good,?
A
A
B
B
We
offer,
of
course,
as
all
the
cloud,
the
most
expensive,
an
account
1
car
managed
all
the
platform
that
we
have
are
the
nodes
and
they
are
only
four
dollars
per
no2,
depending
on
the
amount
of
resources
that
the
mod
has,.
If
we
did
not
achieve
4
dollars
per
nador
per
month,
well,,
we
will
never
go
to
you,.
I
will
never
continue,.
It
will
make
some
invoices
for
resources
that
you
really.
You
have
not
used
that
is,.
B
C
B
C
B
Providers
and
I
think
that
the
most
important
thing
is
to
go.
Everyone
asks
me
the
same
well,
because
the
difference
I
think
is:
We
focus
a
lot
on
developers,
that
is,
they
are
developing
a
product
for
the
Internet
and
they
don't
have
to
wait.
15
minutes
20
minutes,
because
I
do
ask
everyone
to
set
up
a
node
or
create
a
cluster
for
you
when
we
can
do
it
in
two
minutes.
In
two
minutes
you
tried
your
product,
it
worked
for
us.
You
erased
the
cluster.
You
don't
get
up
again
in
two
minutes.
B
B
B
B
Seen
the
change
yours,
yes,
I,
think
that
is
where
the
advantage
of
5
lies
in
that
6
cd
of
your
power
to
do
it
quickly,,
which
is
what
matters
most
at
this
time,
that
is,
to
be
able
to
test
deploy,
quickly,,
not
waste
time,
because
you
have
a
client
waiting
or
a
user
end
waiting,
and
you
can't
go
to
the
end
customer
or
the
end
user.
You
don't
have
to
wait.
30
minutes.
The
cluster
is
created
to
test
it
and,
after
it
is
approved,
send
the
production.
C
B
B
At
the
beginning
of
next
month,
They
come
with
other
services,,
although
they
are
undergoing
improvements
to
govern
to
serve
Vernet
fence,,
which
are
already
being
worked.
On,.
There
are
things
that
are
on
the
way
and
I
think
there
is
a
public
roadmap,,
but
I
do
n't,
know,
I,
really
don't
remember
where,
anyway,
with
the
platform.
It
has
something
to
do.
We
use
it
a
lot,
which
is
when
you
register,,
you
have
the
option
to
suggest
services,
and
if
this
service
is
voted
for
by
the
community,
Vivo
implements
it.
B
Is
me
yes,
I
worked
on
it
or
started
working
on
it,
but
the
first
I
was
a
bet
tester
to
develop
the
great
massive
forgiveness,
and
then
they
made
me
a
job
interview
and
currently
the
job
itself,
that
is,
I
continue
to
draw
a
lot
from
the
community
and
they
listen
to
the
community.
A
lot.
I
think
that
this.
A
B
The
team
from
the
box
that
already
in
the
previous
conference
brother
of
the
end
of
pop
teams
axis,
but
in
the
team
from
the
box,
he
always
said
no,
no
brock
instant,
is
broken
again,
that
if
it
doesn't
it
will,
if
it
can't
and
so
the
api
and
in
12
we
build
it.
I'm
still
going
to
get
out
of
openstack
the
virtualization
technology
we
use.
Sean
mcvey
is
a
project
of
the
sciences
and
it
is
to
virtualize
virtual
machines
on
q,
vernet.
B
We
create
an
operator
called
pereiro
using
pereroles
and
who
is
the
one
who
manages
the
virtual
machines
and
in
turn,
there
is
a
Quechua
operator
who
manages
the
clusters,
who,
in
turn,
communicates
with
the
virtual
machine
operator
and
commands
to
create
the
entire
virtual
machine
and
commands
to
create
the
images
and
commands
to
create
the
routers.
The
firewalls
everything
a
cluster
needs
to
make
it
work.
B
First,
it
was
me
alone.
We
already
have
more
than
being
left
in
goal
for
at
the
beginning,
I
had
the
task
of
doing
it
in
the
centers
of
the
operators
and
building
them
all
three
types
of
going
in
good
of
an
email
to
make
everything
semi
functional,
because
now
simply
the
base
to
be
able
to
build
the
others
and
I
had
never
had
to
have
one.
So
never-
and
it
was
your
turn
to
good
and
spectacular
the
experience
simply
spectacular
massive
lets.
You
be
creative.
B
B
And
we
are
at
the
point
of
bringing
everything
to
'stable'
and
we
are
going
to
release
the
code
of
the
platform
because
we
believe
that
is
the
way
to
give
back
to
the
community
everything
that
has
given
us
all
the
advice,
all
the
feedback
from
the
community.
But
we
want
to
get
to
a
point
where
it
is
quite
mature
and
quite
this
and
with
all
the
code
to
be
able
to
release
it
and
we
are
already
actively
contributing
to
support
projects
like
now.
B
B
C
B
Well,
yes,
currently,
when
a
user
is
administrative,
the
platform
is
multiple.
No,
the
clusters
are
only
visible
to
that
user
online.
All
the
resources
of
that
user
are
only
visible
to
the
user
in
that
network,
that
is
to
say,
a
user
can
have
many
networks
and
all
the
networks
are
independent,
so
they
don't
see
each
other
even
if
they
belong
to
the
same
user.
It
was
designed
that
way
for
security
reasons.
B
C
A
Here
because
it
caught
my
attention
to
ask
something:
good.
I'm
like,,
as
Alejandro
says,,
not
everything
can
be
counted.
I'm
doing
something
with
Kate
3,,
but
they'll
see
him
there
in
a
few
months,.
They
call
me.
Well,
yes,
I
wanted
to
give
an
opinion
because
it
seems
great
to
me,
I'll
continue,
because
it's
good
that
you
were
like
how
age
was
oriented,,
but
so
that
you
can
use
what
you
include,.
You
have
to
make
modifications
because
it
doesn't
work
like
normal
balances
or
stores.
Then
As.
B
The
pinet
as
a
government
provider,
the
official
using
Quechua
the
doctor
driver
the
site,
wrote
it
using.
That
is
to
say,
if
it
is
oh
I,
don't
know
when
you
will
be
the
default
cluster
is
in
charge.
Is
the
driver?
This
driver
communicates
with
the
zeebo
api
and
when
you
you
send
to
create
a
blind
h,
he
creates
a
volume
inside
vivo
and
it
is
attacked
to
that
virtual
machine.
He
attaches
it
to
the
virtual
machine.
B
Is
the
solution
if
that
is
created
and
the
advanced
solution
to
give
you
an
external
ip
is
also
created?
I
perceive
that
when
you
request
it,,
it
is
to
create
an
English
type,
I
advance
it
automatically,.
It
goes
from
that
driver,
so
to
speak,
in
some
way,
it
will
communicate
with
the
api
and
it
will
request
an
existing
ip
and
they
will
cross
it
off
the
cluster
and
it
will
be
given
to
you
and
it
will
be.
They
are
going
to
deliver
it
to
the
cuber
net
cluster,.
B
We
had
to
build
up
to
the
ipam
to
be
able
to
deliver
the
ips,
because
the
zeebo
public
ip
pools,
that
is,
we
had
to
find
a
solution
to
handle
them
puri
and
how
to
know
what
status
has
been
delivered
and
what
was
not
delivered.
I,
don't
know
we
built
an
operator
that
handles
the
IP
pools
by
regions
from
any
region.
So
every
time
one
asks
for
an
email
or
requests
an
email
in
the
region.
There
is
an
operator
that
is
in
charge
of
searching
a
list
supposedly
delivered
by
Don
Felipe..
B
B
We
are
very
careful
with
that,
because
we
do
not
want
any
zeebo
ip
to
fall
on
the
list
with
that,
that
is,,
so
that,
if
your
not
yet
exploited
to
black
list,
delete
childhood
and
another
user
takes
ip,
and
so
it
is
in
black
list
2.
It
is
the
responsibility
of
zeebo
to
request
to
leave
the
black
lists.
B
C
B
B
And
I
tell
you
from
my
experience,
not
so
much
in
disip
and
if
it
doesn't
come
w
I
know
what
do
you
want
to
do
something
about,
and
it's
believe
I
have
permission
for
this.
This
has
permission
for
this.
It's
all
an
easy
process.
What
I
want
is
a
cluster,
so
they
start
to
complicate
you
not
to
see
if
the
laws
and
security
policies
that
fund
permission
for
the
inclusion
that
if
then
basically
we
both
want,
is
to
simplify
what
a
cluster
wants.
B
C
B
After
that,
they
gave
me
an
interview,,
they
asked
me
if
you
want
to
work,
yes,.
They
gave
me
a
job.
Interview,
I
passed
the
interview
and
I
currently
work
in
Simo,
as
well.
It
happened
to
Day,,
which
is
one
of
the
cinemas.
He
is
the
leader
of
the
team
that
is
coming
to
the
US.
He
started
as
a
beta
tester,.
They
did
an
interview
and
he
currently
works
intensively,
that
is,
many
of
the
people
who
work
decidedly
were
external
sales
and.
B
Can
contribute
whatever
you
want?
If
you
live,
you
can
register
in
sib
or
the
community
in
which
it
is
large.
We
respond
quickly
because
I
think
it
is
another
of
the
premises
to
respond
not
to
let
an
incident
or
error
go
from
just
a
start.
It's
been
open
for
two
months
and
we
don't
respond.
Quickly.
I,
don't
know.
If
Sergio
had
the
pleasure
that
they
responded
quickly.
B
So
the
quick
response
to
say
that
the
feedback
that
the
user
of
this
gives
you
doesn't
work.
and
after
10
minutes,
you
tell
him
it's
fixed
and
he
can
try
it
and
he
tells
you
if
it
works
that
fills
it
up,
because
I've
been
on
this
side
of
the
user
and
now
I'm
on
the
other
side.
And
yes,
yes,
yes,
it
works.
There
is
one
of
the
one
of
The
network
engineer
from
Anais's
team
is
known
to
her..
B
She
was
doing
a
live
streaming
on
YouTube
with
other
people
from
Germany,
and
they
told
her
it's
not
excessive,,
it
doesn't.
Work.
This
is
a
true
story.
That
happened
because
I
stayed
and
I
was
sitting
in
front
of
The
machine
arrives,
writes
to
me
today
in
the
siles,
and
it
doesn't
work,
but
with
the
problem.
Don't
look
what
you
say:
I
tried
it
it's
true,
it
didn't
work
in
the
middle
of
the
streaming,
I
solved
the
problem
and
the
changes,
and
she
told
that
person
the
problem
again
and
The
day.