►
From YouTube: Panel Discussion: Containers in Enterprise Cloud Strategy - Moderated by Anni Lai, Huawei
Description
Want to view more sessions and keep the conversations going? Join us for KubeCon + CloudNativeCon North America in Seattle, December 11 - 13, 2018 (http://bit.ly/KCCNCNA18) or in Shanghai, November 14-15 (http://bit.ly/kccncchina18)
Panel Discussion: Containers in Enterprise Cloud Strategy: Pitfalls, Best Practices, and Predictions - Moderated by Anni Lai, Huawei (Intermediate Skill Level)
Container technology promises to help enterprises modernize legacy apps & create cloud-native apps for today’s fast growing digital economy. But, the container ecosystem, built around open src, is still relatively new & evolving at a breakneck pace. CIOs/CTOs today face the need to develop & implement a container strategy, knowing the risks of deploying a still immature technology into production environments that require 0 downtime. This panel will bring together stakeholders from enterprise IT & open src cloud vendors to discuss the role of open src container technologies in their cloud strategies. Panelists will recount challenges & pitfalls faced in deploying containers in production environments & highlight gaps before containers are fully “enterprise ready", & best practices. Panelists will offer their views on the near future of containers as an enabler of the enterprise cloud.
About Brad
Dr. Brad Topol is an IBM DE leading efforts focused on Open Technologies and Developer Advocacy. Brad is a Kubernetes contributor and serves as a member of the Kubernetes conformance wg.
About Ying
Dr. Ying Xiong is currently responsible for leading an US R&D team and building PaaS on open source technologies for Huawei. Dr. Xiong has been involved in Kubernetes/container technologies and communities for over 3 years, ICT technologies for over 20 years. Prior to Huawei, he was a Cloud Platform Architect/Sr. Dev manager for Centrify, Principal dev manager and principal architect at Microsoft, principal technical architect and sr. dev manager at AT&T
About Nils
Nils designs with his team the Open Telekom Cloud, providing a public cloud offering based on OpenStack. As a strong believer and 20+ year user of Open Source technology, he envisions solutions on top of classic cloud services that both create value (business’ perspective) as well as happiness (devops’ perspective). He is a community outreach evangelist in several Open Source projects, speaks at conferences, and is a tech writer for magazines on container security, cloud strategy, and survival guide in a jungle of ever denser growing tools, techniques, and terminologies. Nils enjoys a good hike in nature and lives in urban Berlin, Germany.
About Brandon
Brandon Philips is helping to shape the future of modern Linux server infrastructure at Red Hat. Brandon joined Red Hat as member of technical staff following the acquisition of CoreOS where he served as CTO. Prior to CoreOS, he worked at Rackspace hacking on cloud monitoring and was a Linux kernel developer at SUSE. As a graduate of Oregon State's Open Source Lab he is passionate about open source technologies.
Join us for KubeCon + CloudNativeCon in Barcelona May 20 - 23, Shanghai June 24 - 26, and San Diego November 18 - 21! Learn more at https://kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy and all of the other CNCF-hosted projects.
A
Good
afternoon,
welcome
to
the
container
enterprise
cloud
strategy
is
so
great
to
see
so
many
of
you
here.
So
so
you
know
you
don't
know.
That
container
is
still
very
hot
right
now
and
from
various
analysts
reports,
CN
CF
user
surveys
and
various
vendors
reports.
We
know
that
the
developer
numbers
is
increasing,
the
customer,
the
user
adoption
number
is
increasing.
The
number
of
coop
Contin
attendees
is
tripling
compared
to
last
year,
so
it's
very
exciting
time
for
us,
but
at
the
same
time,
I
do
think
that
we
as
a
community.
A
We
should
constantly
do
a
temperature
check
on
how
our
users
are
using
the
technologies
that
we're
developing.
This
is
why
I
propose
this
panel.
So
in
this
panel
we
are
going
to
discuss
how
containers
are
being
adopted
in
real
world
production
environments.
We're
gonna,
you
know,
discuss
all
the
pitfalls,
best
practices
learnings
with
everyone
here
and
today,
I'm
very
honored-
that
we
have
such
an
elite
and
diverse
panel
here
and
this
group
of
people.
A
They
have
extensive
experiences
in
not
just
container
technologies,
but
also
called
computing
technologies
and,
more
importantly,
they
can
represent
three
geographical
regions.
North
America
amia
in
Asia
pack,
so
we
can't
hear
all
various
different
perspectives
about
how
their
regions
use
container
technologies
in
production
environment.
So
without
further
ado,
let's
just
start
the
panel
and
panelists.
Thank
you
so
much
for
joining
this
panel
and
my
first
question
to
you
is:
what's
your
name
introduce
yourself
and
also
tell
us
about
your
company,
how
your
company,
the
road
that
your
company
plays
in
the
container
ecosystem?
C
Hi,
my
name
is
sean,
I'm
chief
architect
of
cloud
platforms,
huawei
many
responsible
for
the
container
based
platform
or
services
for
our
public
cloud
or
private
cloud.
Poway
does
have
a
public
cloud,
many
the
bucket
in
China,
and
we
also
have
public
cloud
joined
public
cloud
services
with
the
some
other
telecom
carrier
like
a
German
Toluca
in
the
French
orange,
and
they
joined
public
cloud
services.
So
my
ROI
is
basically
we
have
something
similar
to
a
Samsung
course.
I
see
is
a
cloud
container
engine
similar
to
the
TKE
ata.
C
It
had
a
place
overnight
his
services.
So
we
have
that
service.
That's
am
responsible
for
in
terms
of
the
engagement
we
have
si
n
safe.
In
fact,
highway
was
one
of
the
founding
member
of
the
same
safe.
We
have
been
loved
in
the
same
safe
since
2015
we
ran
now
but
current
in
a
plenty
member
of
the
same
safe.
D
E
So
from
an
IBM
perspective,
we
are
obviously
deeply
involved
and
invested
in
kubernetes
and
CN.
Cf
fact,
my
my
boss
is
the
chair
of
the
CN
CF.
So
we
have
a
lot
of
people
working
upstream
in
a
variety
of
areas.
It's
in
the
lower
level,
areas
of
containers
and
things
like
Service
Catalog
across
the
board.
We
have
folks
contributing
and
from
an
IBM
perspective,
we've
got
an
on-prem
offering
based
on
kubernetes
IBM
cloud.
E
D
Hi
I'm
Ian
I'm
from
Deutsche
Telekom
I'm,
a
cloud
architect
at
open
telecom
cloud.
We
provide
cloud
services
on
several
layers.
We
provide
an
infrastructure
layer
that
is
made
of
OpenStack
and
a
number
of
additional
components,
as
well
as
the
layer
on
top
of
that,
and
that
is
kubernetes
and
we
use
for
that.
A
product
just
mentioned
before
called
CCE.
The
cloud
contain
engine
that
sits
on
top
and
provides
a
managed
kubernetes
service.
Yeah,
ok,.
A
F
D
E
You
so
you
know
we
obviously
provide
for
our
customers.
We
have
enterprise
customers
and
we've
been
working
with
customers
in
retail,
healthcare,
transportation,
financial
technologies
and
I'm,
going
to
kind
of
look
at
this
from
two
different
angles
from
the
customers
that
are
coming
and
wanting
to
run
on
the
public
cloud.
What
they're
really
looking
for
is
the
fact
that
we're
going
to
manage
the
service,
so
they
don't
have
to
worry
about
anything
except
developing
their
application.
Those
same
customers
typically
want
high
availability.
E
E
A
wonderful
component
of
the
CN
CF
and
of
kubernetes
called
the
Service
Catalog
anybody
heard
of
Service
Catalog.
It's
a
wonderful,
wonderful
way
to
take
things
that
are
running
in
containers
and
hook
them
up
with
legacy,
services
and
databases
and
other
services,
where
you're
not
quite
as
comfortable
yet
running
them
as
containers.
And
so
you
can
pull
a
lot
of
things
all
together
using
that
piece.
Okay,.
F
C
F
C
Ibm,
the
huawei
also
has
a
property
cloud
and
the
private
cloud
services
promise
we
deal
with.
We
have
many
use
cases.
One
was
use
case
on
this
year,
which
used
is
one
of
the
early
adopters
of
technically
no
technology,
the
company's
channel
mobile
entertainment
divisions.
They
have
a
platforms
called
Miku
operation,
the
backend
platforms,
we're
responsible
for
process
all
the
authorizations,
the
beans,
the
add
push
input
in
the
s2,
the
cell
phones
for
the
other
games,
they're,
the
people
play
on
the
cell
phone,
so
the
platform
actually
process.
C
25,000
requests
pour
actually
per
seconds
so
that
translate
220,
meaning
only
20
minutes
per
hours.
Now
one
of
the
manager
can
declare
a
platform
manage
300
virtual
machines,
as
promised
case
in
this
case.
One
of
the
problem
they
have
is
the
of
course
resource
utilization,
where
the
average
on
average
is
like
10%,
temporary
tempers
and
the
CPU
usage
for
the
virtual
machines
that
they
have
an
extra
Messer.
So
anyway
ends
so
there
are
some
needs.
C
They
don't
have
end-to-end
automation
process,
a
lot
of
it
is
the
menu
they
manually
curve
and,
of
course,
using
script
by
the
Berman
equilibrium
is
not
enter
in,
so
they
start
to
move
distance
to
the
to
the
containers.
So
the
mag,
with
the
platform
to
the
container
platforms
the
person's
actions
are
observed,
is
that
what
the
same
note
they
can
only
they
character
with
half
of
the
server's.
C
They
have
the
used
to
run
away
aims,
so
they
see
immediately
the
resource
unit,
18,
tava,
well,
half
of
the
server's
they
run
in
containers
so
with
kubernetes
they
can
actually
move
the
the
app
instance
around
around
those
are
those
servers
and
they
character
he
can
deploy
very
quickly
so
that
there's
veneer
kind
of
important
for
them.
One
after
sensitive
winner,
when
you're
working
with
this
customer
is
that
it's
not
just
the
container
right.
It's
not
just
it's,
not
just
the
kubernetes
and
your
install.
There
is
it's.
C
They
wanna
end-to-end
kind
of
solutions
for
the
container
DevOps
to
say
a
deep
applies
to
do.
Mandatories,
ready
to
run
in
the
lifecycle
manager
to
monitoring
that.
So
that's
since
of
an
American
leaders,
not
just
a
one
customer
most
of
a
Cosmo
has
did
this
requirement.
It's
not
just
you
know
install
the
kubernetes.
Then
they
are
you.
Are
there?
No,
it's
not
it's.
They
have
the
planning
and
strategy
into
a
solution.
That's
what
they're
looking
for
so.
B
At
chorus,
we'd
work
with
a
bunch
of
customers,
but
also
spare
the
customer
stories
is
a
small
story
about
an
application.
We
were
running
for
core
OS
calm,
so
we
were
running
this
little
API.
We
we
built
a
little
API
to
manage
that
manage
customer
data
and
we
had
it
set
up.
We
have
CI
CD
around
it,
etc,
as
deploying
automatically
but
kind
of
an
inner.
Interesting
lesson
learned.
This
is
very
practical
lesson
learned.
B
Is
we
never
hooked
up
the
pod
health
checking,
which
actively
checks
that
into
each
individual
instance
of
the
application
is
behaving
correctly
and
unbeknownst
to
us?
There
was
a
bug
in
the
system
and
the
beautiful
thing
about
kubernetes
is.
It
makes
it
really
wonderful
to
run
five
copies
of
the
application,
but
silently
these
applications
are
just
failing
behind
the
scenes,
and
so
without
that
pot
health
checking,
they
weren't
actually
getting
killed
and
restarted,
and
so
just
slowly
the
service
get
got
flakier
in
flakier
and
they
just
stopped
working
completely.
A
So
so,
when
I
asked
you
so
for
a
company
to
enter
into
a
club
native
or
container
strategy,
what
generally
what's
their
biggest
challenge,
is
that
people
processes
number
of
developers
who
know
how
to
code
in
native
environments
or
politics?
And
then
you
know
how
do
generally
people
companies
resolve
that
issue?
E
So
I
was
going
to
look
at
that
question
from
two
different
ways.
You
know
from
from
the
provider
side,
what
were
our
challenges,
just
real,
quick
tip
and
is
providing
private
connectivity
to
our
container
applications
and
and
how
we
address
that.
As
you
know,
a
lot
of
VPN
options
that
we
provide
in
the
cloud
now
from
the
customer
point
of
view
is
really
interesting.
They've
got
to
make
some
decisions
about
how
they
move.
You
know
their
their
applications
to
cloud
native.
E
Obviously,
the
things
that
are
stateful
are
a
little
tough,
but
even
for
the
things
that
they
can't
move,
you
know
we've
seen
well.
Do
you
move
it
lift
and
shift
as
a
big
monolithic
application,
or
do
you
actually
take
the
time
to
take
the
application
and
maybe
break
it
into
a
few
micro
services,
with
the
benefit
of?
If
you
can
bring
it
into
a
few
different
micro
services,
then
those
teams
can
typically
work
independently.
Have
independent
development
schedule
some
more
flexibility,
so
it
really
depends.
A
B
They
took
all
those
little
VMs
and
those
micro
services
and
put
them
in
containers
and
ran
them
on
kubernetes
and
was
able
to
show
like
an
80%
cost
reduction
because
they
simply
were
now
running
one
really
big
VM,
with
all
the
containers
and
micro
services
talking
to
each
other
on
top
of
kubernetes
and
inside
of
this
one
cluster,
and
so
this
was
actually
a
really
good
compelling
way
across
the
business,
because
the
moment
that
they
are,
it
will
show
hey.
Look.
D
D
So
key
is
to
have
the
right
people
and
to
have
the
right
people
and
to
have
the
right
experience
that
these
people
it's
key
to
start
as
early
as
possible,
don't
start
with
while
migrating
your
your
major
key
application,
but
start
with
the
small
ones
with
the
low-hanging
fruit,
I
hate
the
term.
But
it's
it's
suitable
here
so
that
you
can
build
up
some
experience
how
to
design
these
architectures
and
then
everything
else
will
go
on
and
say
on
itself.
I!
Think,
okay,.
C
There
are
we
have
a
few
examples
that
can
show
you
know
the
way
experience
the
challenges
that
we
have
as
customers
are
as
popular
as
kubernetes
today
is
still
complicated
to
some
of
the
customer,
if
not
most
specially
in
the
storage
and
the
network
areas,
most
almost
all
the
enterprise
has
to
own
a
server
infrastructure.
Authority
infrastructure
has
the
you
know:
network
policy,
network
design
so
getting
this
container
platform
the
working
on
top
of
this,
this
storage
in
network
policy.
It's
quite
challenges.
In
fact
they
have
authority
servers
where
we
have
cubic
vo.
Customer
storage.
C
C
If
they
have
the
own
network
policies,
the
users
around
applications
in
IBM,
they
can
access
the
application
with
you
know,
through
the
IP
address
now
then
move
to
the
kubernetes
the
container,
the
part
that
is
dynamic,
you,
don't
you,
can
access
to
it
so
they're
in
kind
all
those
problems
we
had.
There
are
the
problems
we
can
experience
with.
Some
customer
is
a
much
earlier
is
once
and
it's
not
just
the
container
platform,
they
have
the
integrator
weights.
They
they
exist
in
release
pipelines
like
say
ICD
pipelines.
C
They
have
the
indicator,
which
is
this
team
monitoring
systems,
and
if
you
pick
up
this,
you
know
saying
save
that
community
the
same
safe
monitoring
tools,
there's
easy
to
integrate
same-same
did
doing
a
very
good
job,
actually
integrate
all
senses,
but
most
of
the
enterprise
has
their
own
monitoring,
alerting
systems.
You
have
integrates
with
one.
C
Specific
kind
of
a
challenges
we
have
is
one
of
the
customers.
You
know
some
of
the
customers
sterilized
in
all
ways.
One
of
the
ways
they
have
it
doesn't
support
containers.
Oh
it
just
doesn't
support
good
features
of
the
containers,
so
we
have
to
build
our
own
kind
of
containers
to
it
to
adapt
to
that
that
OS,
okay.
A
I'm
gonna
do
a
little
commercial
here,
so
so
you
all
know
that
CN
CF
just
publish
the
the
landscape
document.
In
case
you
don't
know,
you
have
not
seen
it's
at
L
CNC
of
the
I/o
and
if
you've
seen-
and
you
know
that
you
know
the
ecosystem
of
container
is
growing
explosively.
The
good
news
is
great.
We
get
this
huge
momentum.
The
bad
news
is,
it's
really
confusing.
You
don't
know,
what's
really
solid
reliable.
What's
not
so
I
would
like
to
ask
the
panel.
A
E
So
you
know
what
we
found
is
is
the
base
components
of
kubernetes
are
very,
very
solid
so
that
that
base
infrastructure
is
fine
monitoring,
logging,
very
solid
lots
of
options.
The
piece
where
we
need
to
see
more
work
would
be
sort
of
the
security
focused
pieces,
the
authentication
authorization
certificate
management,
image
management.
Those
are
pieces
that,
in
an
enterprise,
I
think
you'd
want
to
see
more
work
being
done.
Okay,.
B
F
D
However,
it's
important
for
for
users
to
find
out
which
components
are
reliable
and
are
stable
and,
for
example,
standard
standardization
and
the
open
source
project
process
is
a
very
important
feature
in
that.
So
projects
that
are
under
the
umbrella
of
this
UNCF,
for
example,
are
kind
of
well
QA
tested,
and
this
is
something
I'd
recommend
for
everybody
to
closely
monitor,
which
component
is
really
developed
and
open-source
principles
so
that
users
don't
get
in
in
just
another
round
of
vendor
lock-in
or
whatever.
Okay.
C
My
experience
always
the
customer
that
are
as
Brad
mentioned
this
corporation
of
kinetics,
is
a
mature
and
stable
in
turn.
You
know
for,
for
example,
the
scheduling,
impartial
management
work
situations
and
we
have
not
heard
major
complaints
or
problems
from
our
customers,
but
the
security,
the
storage
in
network
inside
we
have
I'm
not
say,
that's,
not,
you
know
stable,
but
just
we
have
many
customer.
Ask
the
questions
about
security.
The
storage
I
mentioned
earlier
and
then
they're
working
a
lot
of
Kasim
want
to
integrate
with
the
third-party
storage.
C
Your
third
party
or
the
third
party
network
policies,
hot
anywhere,
is
that
I
think
the
community
can
focus
on
one
there's
much
more
sensible,
right.
I
know
we
are
doing.
The
sensor
is
doing
in
the
right
direction,
with
creators
without
policy
interface
will
create
a
storage
interface
as
long
as
the
read/write
are
rather
ways
right
direction.
They
go,
but
with
those
interface
there's
none
on
as
interface
is
easier
to
add.
Acton
extensible.
Add
third
party
support.
Fourth,
kubernetes
platforms.
A
Good,
so
so
we
all
know
that
the
value
proposition
of
you
know
container
technologies
number
one.
You
know
it
helps
us
launch
the
application
services.
A
lot
faster
number
two
is
to
make
the
infrastructure
resources
a
lot
more
efficient
and
I
think
another
benefit
value
proposition
of
containers
is
container
interoperability
and
so
no
vendor
lock-in.
But
this
is
such
a
broad
topic.
So
I
would
like
to
ask
the
panel
to
clarify
from
your
customer
or
from
users
perspective.
How
do
they
see
container
interoperability?
A
E
E
Interoperability
is
huge.
It's
used
to
our
customers
and
I'm
gonna.
Let
all
my
colleagues
know
my
best
trick
when
I
go
talk
to
my
customers,
because
we
should
all
use
it
and
I'd
say
this,
you
know,
God
forbid,
IBM
doesn't
make
you
happy.
I'm
gonna
feel
terrible,
but
you
can
take
your
stuff
and
you
can
go
to
hallway.
You
can
go
to
Red
Hat,
you
can
go
to
Amazon,
you
go
to.
E
You
know
whoever
is
running
the
the
the
kubernetes
Google
and,
and
so
you
should
feel
very
comfortable
that,
if
you're
picking
this
that
that's
what
your
opportunities
are
going
to
be-
and
it's
also
going
to
enable
you
what
we've
seen
multi-cloud
multi
cloud
is
huge
to
our
customers.
So,
what's
been
very
fortunate,
is
in
kubernetes,
we've
been
doing
a
lot
of
work
early
on
conformance
interoperability,
I,
don't
know
if
you
heard
from
from
Dan,
if
you
mentioned
at
this
time,
but
we
now
have
56
companies
that
are
now
certified
through
our
conformance
program.
E
We
we
had
a
great
model
of
starting
this
early.
While
you
know
if
you
start
interoperability
too
late,
it's
a
it's
like
herding
cats
that
have
run
way
far
away,
it's
very
hard
to
pull
them
in.
So
we
started
early
in
kubernetes.
We
got
everybody
on
board
everybody's
been
wonderful
in
the
community
and
we've
got.
You
know
a
nice
suite
of
certification
tests
and
we
have
more
resources
coming
from
the
CN
CF
to
actually
improve
the
test
suite.
So
a
lot
of
resources
to
improve
the
test,
suite
and
focus
us
on
interoperability.
E
F
B
I
would
say
that
the
the
way
that
we've
been
driving
Sanders
is
essentially
inside
of
kubernetes
holding
the
line.
I
think
this
question
was
would
be
different
like
two
years
ago,
because
there's
all
these
questions
around
like
what
is
the
container
image
format
and
is
it's
swarm
or
kubernetes
or
whatever?
And
luckily
we
that's
all
gone.
It's
cute.
That's
why
there's
4,000
people
here
and
then
the
other
piece
is
just
kubernetes
holding
the
line
against.
B
I
think
one
of
the
lessons
learned
from
OpenStack
was:
don't
let
the
API
be
extended
just
willy-nilly
all
over
the
place,
hold
the
line
and
then
show
vendors
where
they
can
plug
in
and
then
that
allows
and
enables
things
like
the
conformance
tests
and
that
sort
of
thing
to
happen
and
I
think
to
the
credit
of
a
lot
of
folks
in
the
kubernetes
community.
We've
done
a
pretty
good
job
of
that
so
far
and
it's
it's
an
ongoing
discussion
because,
obviously
you
don't
want
to
like
completely
shut
down
any
innovation.
B
That's
coming
out
of
networking
storage
vendors,
but
you
also
don't
want
to
work
your
way
into
a
corner
where
applications
have
to
be
reconnected.
Every
time
they
move
between
a
kubernetes
and
that's
something
that
I
think
is
an
opportunity
that
we're
executing
on
here,
which
was
an
opportunity
that
was
very
similar
to
Linux
15
years
ago,
was
Linux
provide
an
abstraction
across.
Although
Norwin
storage,
vendors,
when
they're
competing
over
hardware
and
kubernetes,
is
providing
a
abstraction
over
networking
storage
vendors
when
they're
all
competing
over
cloud
and
api's.
So.
C
C
So
if
this
means
that
when
the
same
format,
the
same
runtime
environment,
the
same
storages
same
T
Network
and
that's
where
we
need
to
stand
an
eyesore
and
is
interoperable
among
those
those
interface,
so
I
think
the
same,
save
has
now
has
CP.
I
CRI
says
in
the
same
is
a
great
great
kind
of
a
standard
interface
that
hoping
there's
a
product
moving
forward.
Okay,.
F
D
One
single
item
I'd
like
to
get
sorted
out
is,
for
my
personal
opinion,
the
container
run
time
so
don't
get
me
wrong,
I,
really
like
the
variety
we
have
here
and
well
really
big
span
of
different
technologies.
I
I,
just
I
I
thought
I
learned
about
two
or
three
new
runtimes,
just
the
last
couple
of
days,
but
I
thing
especially
for
users.
D
It
is
important
to
get
more
documentation
and
more
material
that
supports
them
in
the
decision
process
which
runtime
to
use
in
which
cases
or
how
to
architect
their
overall
infrastructure,
because
at
the
moment
it's
a
little
bit
well
confusing
at
times.
That's
my
personal
opinion,
but
obviously,
luckily
there
have
been
some
very
insightful
talks
here
on
cue
con.
That
is
the
good
point
and.
A
Good
news
here
great
so
we
are
reaching
the
top
of
our
but
I
have
to
ask
this
question
because
I
think
it's
fun
so
just
to
humor
me
I
like
to
ask
the
panelists
said:
you
know
if
you
can
share
your
crystal
ball
with
all
of
us.
What
do
you
think,
five
years
from
now
containers
how
technology
will
be
I'll?
Give
you
multiple
choice.
A
You
know
number
one
number
one:
it's
like
still
the
same:
lots
of
growth,
lots
of
new
projects,
but
still
confusing
all
right
and
then
coupe
cars
gonna
be
probably
about
10,000
people
number
to
continue
to
our
knowledge
is
so
boring
because
it's
done
everybody's
using
it.
Everything
automated
into
prices
are
using.
It
nobody's
talking
about
it,
but
it's
there
in
every
data
center
number
three,
it's
gone,
not
because
it's
done
is
because
it
got
cannibalize
by
other
technologies.
B
B
C
My
organ
is
the
same
as
the
brand.
Instead,
you
want
the
surveys
I'm,
making
the
surveys
that
I
just
found
a
month
ago.
They
did
a
survey
about
4,000
customers,
enterprise
customers
and
they
be
nice.
These
surveys
are
last
month's
and
in
that
survey
they
asked
well
in
the
survey
actually
say
that
30
percent
of
the
customers,
enterprise
you're
already
using
containers
in
production
and
certain
percent
of
in
the
testing
and
another
30
percent,
not
so
30%
of
evaluation.
And
if
you
addressed
okay,
so
it's
90
percent,
it's
90
percent,
so
so
the
market.
C
This
world
is
already
calling
the
direction
that
we've
mastered
as
user
of
the
containers,
because
they
really
provide
the
benefits
communities,
all
those
in
a
packaging
faster
deployment
upgrade
this
really
provide
the
benefit
to
the
business
as
saying
that
that
whatever
well,
that
was
that
that's
my
opinions
on
that.
So.
E
I
think
what
you'll
see
in
say
five
years
is
from
the
enterprise
side
that
they
will
have
been
able
to
successfully
migrated
many
of
the
applications
that
they
wanted
to
migrate
to
a
cloud
native
to
a
container
kubernetes
based
infrastructure
and
that's
gonna,
be
you
know,
wonderful
and
I.
Think
you're
gonna
see
the
kubernetes
platform
become
more
ubiquitous,
I.
Think
you're
gonna
see
what
converges
on
it
as
deep
learning
machine
learning,
high-performance
computing.
All
of
these
are
gonna,
be
building
on
top
of
this
infrastructure,
and
it's
gonna
be
an
exciting
five
years.
Great.
F
D
Why
not
have
thousands
and
ten
thousands
of
them
and
manage
them?
The
technology
we
might
not
have
have
today
and
I
think
we
see
in
five
years
also
a
different
approach
to
security,
mostly
most
probably
fueled,
by
some
privacy
activities
or
privacy
issues
that
spark
a
new
a
new
feeling
towards
the
security
to
overcome
those
problems.