►
Description
La era TCloud - Carlos Moreno & Gustavo Rivero (Telefónica Argentina) OSCG 2022
A
A
Good
afternoon
everyone
I
don't
know.
If
you
are
listening
to
me
well
gustavo,
we
are
ready
and
good
with
everything
gustavo.
We
are
ready,
good
good.
We
are
here
to
tell
you
about
the
telephone
case,
the
era
of
claudio
the
keyboard.
As
we
say,
let
me
tell
you
first
A
little
bit,.
What
does
the
keyboard
mean
in
Telefonica
in
Telefonica?.
A
A
Provide
capacities
for
both
Haiti
and
for
the
network,.
It
is
where
we
began
to
make
the
technological
change
and
the
change
of
the
cultural
paradigm
and
where
we
added
Red
Hat
for
that,
so
that,
in
a
way,
it
helps
us
to
implement
openshift
for
everything
that
is
Haiti..
We
are
going
to
tell
you
a
little
about
the
strategy,,
how
it
was,,
how
we
are
organized
and
how
the
evolution
of
our
digital
culture,
was.
A
We,
can
tell
you
about
a
project
that
was
the
one
that,
in
a
certain
way,
propped
up
and
gave
us
the
business
case
to
be
able
to
do
this,
because
yes.,
there
is
no
business
case.
There
is
no
investment
possible,
and,
after
all
this,
you
like
it
I
am
going
to
start
telling
what
tools
were
so
to
say
that
we
chose
openshift,
how
we
are
doing.
We
are
managing
with
red
hat
and
what
other
things
we
have
within
our
type
of
cloud,
which
is
what,
in
a
certain
way,
points
us
in
our
digital
transformation.
A
Basically,
like
many
of
the
companies
since
2018
have
been
making
changes
in
agriculture
working
with
agile
teams
telephone.
We
also
have
all
the
roles
defined.
We
work
in
trains
that
was
very
good
But.
What
was
happening
to
us?
We
reached
a
point
where
the
Telefónica
team
had
both
Haiti
and
the
business
working
together,
creating
ideas,,
but
we
still
worked
in
silos
because
the
person
who
had
to
develop
the
solution
for
us
was
perhaps
outside
the
country,.
A
This
is
the
team
that
all
these
people
have
today
on
the
key,
or
are
they
the
ones
that
moved
on
trains?
As
I
told
you,
we
have
the
train
dique
ar
the
train
of
commerce
of
power,
crm
scm
for
assisted
channels
that
we
use,
and
also
the
tuenti
train,,
which
is
surely
an
application
that
some
of
you
may
be
using
now,.
A
With
that
alone,
we
could
not
make
the
cultural
change,
that
is,
I
was
sitting
at
the
table
together
with
the
commercial
team,
they
defined
solutions,,
but
there
was
still
more
to
go,
and
that
more
was
how
we
took
the
next
step,
and
the
next
step
was
to
get
on
a
voice
style
model,
or
at
least
it
says
itself,
at
least
that
well,
we
had
to
do
for
that.
Among
the
things
I
have
to
do
is
As
long
as
we
continue
with
redhat
starting
to
implement
openshift
on
our
keyboard,.
A
We
had
to
create
teams
because
it
is
important
and
it
was
what
we
were
doing
to
implement
it,
and
also
the
day
after
that
who
was
going
to
be
supporting
how
we
were
going
to
continue
doing
it
and
how
I
am
growing
and
standardizing
the
and
the
tools
that
we
needed
to
continue
being
digital.
That
is
when
we
add
these
four
teams.
There
is
a
team
that
is
a
digital
operation
that
basically,
what
it
attends
to
is
number
1,
because
after
number,
2
and
3,
the
known
error
occurs
on
the
table.
A
That
team
takes
care
of
it
and
the
rest
begins
to
happen
at
the
table.
When.
That
goes
on
to
a
more
detailed
analysis,
to
a
need
for
correction,.
It
goes
to
the
table
or
to
the
train,,
or
rather
to
the
train,,
where
the
boxing
management
team
can
provide
that
solution.
and
Paz
Gustavo
is
the
head
of
that
team,.
He
is
in
charge
of
maintaining
the
openshift,
productively,
maintaining
the
openshift
tools,
and,
in
turn,
being
able
to
change
version
by
version,
each
of
each
of
the
new
versions,.
A
That's
what
we
are
now
410
of
410
now,
but
each
tank
every
three
months
or
every
six
months.
We
have
a
new
version
that
gustavo
is
in
charge
of
homologating
and
re-uploading
gustavo
and
his
team.
We
don't
like
each
other
just
because
there
are
16
people.
I
was
not
going
to
be
able
to
just
fine
and
from
her
a
team
that
is
I
was
in
charge
of
standardizing
what
we
do
in
the
digital
back
end
and
standardizing
how
we
want
the
microservices,
what
microservices
standards
we
want
to
maintain
how
those
microservices
should
achieve
well,.
A
The
life
project
to
tell
is
this.
That
was
the
one
that
leveraged
the
business
case
of
being
able
to
migrate
this
to
a
page
model,
and
it
was
not
to
repatriate
it
is
not
the
app
of
my
movistar
worth
an
application
that,
despite
the
fact
that
we
defined
haiti
and
the
business
that
we
wanted
to
put
inside,
the
app
was
still
being
built
in
a
team
on
the
outside
from
that
table,
and
it
was
deployed
in
a
data
center
in
the
usa.
The
idea
was,
we
have
to
strategically
bring
that
to
argentina.
A
A
I
hope
that
many
of
you
use
it
at
least
half
plus
one
that
are
from
movistar
and
well,
and
the
other
challenge
was
to
be
able
to
improve
on
the
ps
It
was
an
escrow
is
an
application
that
when
we
took
it
to
bring
it
already
had
a
high
level
of
acceptance.
It
was
above
4.3
in
google
stores
and
they
see
it
then
bring
it
to
our
datacenter
changing
them.
We
had
to
redo
it.
A
We
had
to
redo
it
in
many
aspects
to
change
it
so
that
serve
for
our
or
for
our
openshift
and
put
a
high
risk
of
impact
with
the
client,
he
succeeded.
Luckily,
we
achieved
it
today.
He
is
back
at
that
level
quickly.
He
returned
to
that
level
of
acceptance
and
well
and
a
little.
He
is
the
one
who
leveraged
the
investment
To
activate
our
our
our
internal
cloud.
In
Telefónica,
we
went
through
different
stages
to
have
the
infrastructure
to
have
the
teams
the
teams
had
to
be
trained.
A
While
we
implemented
openshift
together
with
red
hat
side
by
side,
our
people
were
learning,
they
did
reskilling
of
what
they
knew
a
lot
about
this.
What
We
did
a
lot
of
this
with
our
own
people,?
We
did
that
they
had
to
do
a
screen,.
They
came
from
working
in
virtual
machines
and
they
would
have
to
start
learning
about
containers.
A
A
B
Just
to
add
and
tell
you
that
the
teamwork,
accompanied
by
the
gates
team,
the
team
is
already
present.
We
started
this
project
with
version
4.6
of
openshift
until
reaching
version.
410
Today
in
production
I
think
that
if
there
is
below
what
I
am
going
to
tell
you
is
our
experience,
how
we
focus
on
automating
the
developer
experience,
what
selection
of
tools
we
made
I
am
going
to
go
into
a
lot
of
detail,
but
what
we
achieved
to
reduce
time-
and
here
Don't
be
scared
by
the
architecture,.
B
There
is
our
current
architecture,
and
today
we
have
only
one
node,,
the
TVs,
various
metal,.
We
also
have
our
virtualized
low
environments,,
but
what
I
wanted
to
show
you
was
our
selection
of
bars,
products.,
In
principle,.
The
layer
for
me
is
made
up
of
for
three
that
we
are
using
the
king
heart
single
sign-on.
We
are
using
openshift
by
line,
we
add
m
the
broker
casca.
B
In
addition
to
this,
we
are
working
on
the
deployment
of
dressed
colleagues
with
the
red
team
that
we
use
the
observability
offered
by
the
path.
The
pencil
platform
We
also
went
to
the
catalog
of
services
or
the
catalog
of
images
that
Redhat
offers,
in
this
case
the
technologies-
you
don't
follow
and
rooms
where
the
images
are
already
sanitized
are
cured,,
so
we
didn't
have
to
think
about
setting
up
a
pan
to
setting
up
golden
and
may
start
working
on
scans.
We
needed
to
reduce
time.
B
Finally,
we
made
the
decision
to
set
up
a
rancher
cluster
to
outsource
those
tools
where
we
did
not
want
to
generate
an
overload,
the
principe
platform
among
them.
We
chose
the
gamelab.
That
today
is
our
core,
let's
say
a
repository
and
we
need
it
in
high
availability.
This
ranch
he
is
working
active
passive.
B
B
B
B
B
Then
we
added
all
the
technological
state,
I
showed
them
the
hour.
Metrics,
then
included
in
the
pailón
code
and
from
there
extract
the
information
up
to
Today.
We
reach
the
milestone
model
where
we
have
fully
automated
semantic
portioning
and
we
also
deploy
it
in
different
sites.
We
are
preparing
for
site
2,,
which
is
another
cluster,
and
another
2
centers.
B
Well,
first
of
all,
I
am
going
to
tell
you
what
the
developer's
experience
is
like
today
with
this
type
of
pay
line.
What
we
are
always
looking
for
is
to
automate.
So
we
just
want
to
think
about
the
developer.
Making
add
add,
request,
execute
the
open
plan.
If
pay
line
does
all
the
work
in
different
tasks
generates
an
image?
The
image
will
be
versioned,
it
would
be
documented
in
libya,
which
is
our
case
of
our
tool.
We
are
using
as
a
repository
and
you
go
and
chirp
everything
automatically
we
are
using
conventional
commit.
B
This
is
like
a
natural
language
game
that
we
use
a
robot
within
the
pai
line
and
the
version
is
automatically
taken
care
of
in
the
deployment
part
we
separate
into
the
infrastructure
as
code
we
standardize
the
diploma
and
every
time
there
is
a
change.
We
want
to
take
an
image
to
a
different
site.
We
are
not
preparing
to
activate
the
second
site
through
the
argos
application
set
and
we
can
automatically
add
and
deploy
those
applications.
B
A
little
show
you
the
case
of
a
network
application.
In
this
sad
case,
we
were
with
a
process
where
we
had
to
request
development
to
load
in
a
document,
everything
that
was
needed
to
generate
an
app
and
validate
it
validate.
So
this
took
days
the
document
would
go
backwards.
We
could
start
using
it
ofencia
line.
B
A
In
summary,
I,
basically
liked
what
Gustavo
said
and
what,
when
it
is
a
bit
like
the
transformation
was
what
he
just
said.
2
about
the
deployments
think
that,
despite
building
solutions
in
our
agile
teams,
when
we
passed
the
request
to
the
area
that
had
to
produce
the
development
and
deploy,
sometimes
we
had
to
plan
it
for
15
days
or
20
days
and
nowadays,
totally
autonomous.
And
since
we
went
through
many
of
the
cases
that
they
told
before,
like
we
need
you
to
make
me
a
ticket
to
pass
the
production
part
of
the
transformation
we
achieved.
A
And
today
that
happens
and
runs
directly
and
each
team
is
autonomous.
In
that
aspect,
we
had
to
put
together
a
team
during
the
process,
as
I
say,
the
team
that,
along
with
bars,
deployed
the
path,
the
openshift,
the
team,
since
also
He,
was
in
charge
of
who
today
manages
all
the
infrastructure
and
the
Haitian
operations
team
in
zinc.
We
had
to
improve
some
things
in
the
discipline
we
already
came
with
with
disciplines.
A
We
had
the
discipline
of
testing
and
the
discipline
of
boxing
in
the
case
of
text
in
we
had
to
change
some
things
to
improve
the
automation
performance
automaton.
When
you
start
to
manage
your
own
development,
you
start
to
have
growth
in
this
type
of
thing.
You
have
to
use
new
tools.
Well,
all
that
happened
in
the
meantime.
We
implemented
a
pay
it
and
let
openshift
working,
ok,
good
next
challenges
or
multiple
pepsi
esther
with
tunnel
to
gustavo.
We
are
going.
A
B
Are
going
to
explain
it
in
the
laboratories?
What
we
are
looking
for
is
to
have
service
resilience
today,
only
we
have
a
site
is
going
to
activate
the
second
site.
We
do
not
want
to
have
idle
resources,
so
we
are
going
to
an
active
active
deployment
and
to
be
able
to
bring
the
service
month
to
production
and
to
be
able
to
invoke
services.
If
a
cluster
is
affected
or
a
site
is
affected,
it
will
continue
to
give
services
with
the
capacity
of
the
other
site.
A
We
are
going
to
continue
implementing
disciplines
communities.
We
do
not
have
communities
where,
within
telefonica,
people
join
the
community
and
begin
to
contribute
ideas
for
improvement
for
the
new
versions
of
paila.
In
many
of
the
versions
of
paila
that
the
The
peace
team
is
getting
out
of
the
community,
meetings.
I
need
this..
We
don't
need
this,
since
it
improves
my
process
and
we
see
to
standardize
it
and
add
it
to
the
other
challenge
of
adding
security.
There
is
nothing
in
telephony
that
can
go
to
production,.
Yes.
It
does
not
pass
through
the
hands
of
security.
A
Today,
that
is
a
text
and
a
manual
today.
It
is
that
part
in
which
we
still
do
not
have
automated.
Today.
It
is
a
test,
a
test
and
manual
that
there
comes
a
time
when
it
reaches
the
integration
environment,
security
comes
and
does
its
tests
and
unless
As
security,,
it
is
part
of
the
tables
and
participates
with
members
with
members
inside
the
trains
that
say
this
will
not
need
security,
because
there
was
already
what
it
is
and
it
does
not
need
to
be
there
if
it
can
run
directly
if
it
does
not
require
manual.
A
Testing,,
which
is
what
we
are
to
change
and
we
are
beginning
to
add
telephone
services.
We
want
to
become
good
and
continue
democratizing
the
metrics
just
blooded.
Now
our
metrics
are
automated,
but
we
have
to
get
the
metrics
to
reach
everyone,
all
the
teams
that
use
them,
because
it
is
a
cultural
change.
Today
they
are
concerned
for
taking
development
and
they
don't
look
at
the
metrics,.
A
I
usually
call
that
democratizing
and
we
have
to
make
it
democratized
and
use
something
else
that
we
are
also
doing,
that
we
have
in
the
non-productive
environments
of
the
legacies,,
which
is
to
automate
the
self-
remediation
of
the
environments.
We
have
tools
named
gustavo
tools
that
today
would
allow
us,
in
some
way
to
automate
and
self-remediate
known
situations
in
cloud
environments
in
the
cloud
in
production
or
in
non-production
environments,
and
it
is
part
of
what
we
have
to
do
as
challenges
for
2023,
well,.