►
Description
KBE Insider interviews Jim Wittermans, Chapter Area Lead at ABN AMRO Clearing Bank while in Amsterdam for KubeCon Europe. We discussed various challenges and myths in the transition to cloud, such as vendor lock-in, replatforming, and more. Jim also shares how his team at ABN AMRO Clearing Bank solves these problems by providing development support services.
Watch to gain Jim’s real-world insights on transitioning to open source technologies at enterprise scale!
A
B
B
A
B
B
Yeah,
so
we
do
it
by
by
technology
now.
So
that
also
means
that,
because
we
have
a
global
presence
from
an
ABM
Brooklyn
perspective,
is
that
even
though
my
maybe
the
python
developers
might
be
in
Sydney
Chicago
Amsterdam
yeah
that
even
they
also
communicate
with
each
other
and
learn
from
the
best
practices
amongst.
B
B
B
A
Helps
a
lot
that
is
one
of
the
things
that
I
continuously
rant
about
with
the
tech
hiring
Market
is
like
you
know:
do
you
really
need
to
find
somebody
who
specifically,
has
you
know
five
years
of
experience
with
python
with
this
particular
module?
Or
can
you
find
somebody
who
knows
how
to
program
for
a
bunch
of
years,
and
you
can
just
teach
them
python
in
that
particular
module
exactly
exactly
yeah
couldn't
agree
more
yeah
yeah?
It's!
It
really
kind
of
drives
me
crazy,
because
it's
like
I
know
how
to
do
a
for
loop.
B
A
A
A
A
So
so
you
I
mean
so
you
work
at
a
you
know
a
clearing
right,
so
a
bank
and
but
what
is
it
about
kind
of
the
open
source
world
or
what
you
know?
How
involved
have
you
been
in
open
source?
I
mean
you
use
Python,
so
you
probably
have
some
exposure,
but
where
does
that
fit
into
your
kind
of
professional
or
even
non-professional
life?
So.
B
B
Also
because
the
sort
of
the
the
intelligence
of
the
crowd
is
always
so
much
larger
than
an
individual
could
ever
do
yeah
100
exactly
so
I
I'd
like
to
contribute
I
did
a
bit
from
a
Google
perspective
in
my
early
years.
Also
from
a
python
perspective,
I've
contributed
a
bit
and
I
must
admit.
In
my
new
role,
I've
tried
to
develop
bits
from
time
to
time.
I
always
told
myself,
I
want
to
keep
it
in
development.
One
day
a
week
right
right,
how's
that
going
for.
A
A
A
Professor
right
like
as
in
you
know,
I'm
still
learning
how
to
do
all
that
stuff
and
I
didn't
have
a
lot
of
the
exposure
that
I
think
a
lot
of
you
know
kind
of
professors
do
going
up
through
the
normal
path
of
like
you
know,
because
I
was
never
a
TA
anywhere
yeah.
You
know.
So
it's
like
it's
all
been
kind
of
self-taught.
You
know
yeah
that
you
know
I've
done
a
lot
of
like
conferences
and
workshops
and
all
that
jazz
but
I
haven't
you
know.
A
B
Oh,
that's,
nice,
that's
cool,
yeah,
I!
Think
it's
very
from
a
leadership
perspective
right
I
know
you
talked
about
Open
Source
before,
but
I
think
it's
really
hard
to
find
people
who
actually
know
how
to
coach
a
bit
and
I
like
to
explain
stuff
and
also
are
quite
technically
Advanced
still
and
I
know.
I
now
know
why?
Because
as
soon
as
you
sort
of
reach,
let's
say
one
level
higher
it's
you're
consumed
by
other
things
and
development
right.
A
Right
or
the
one
of
the
things
I,
you
say
too
is
like
I.
You
know,
I
had
a
cube
for
a
long
time
in
in
the
various
jobs
I
had
and
eventually
I
was
promoted
to
like
a
management
level
or
whatever
and
I
finally
got
an
office
with
a
door
and
I
found
out
what
why
the
door
is
terrible
because
it
means
you
have
to
have
closed-door
conversations.
You
know
about,
you
know
someone's
under
performance
or
you
know,
you
know
things
like
that,
and
it's
like
these
are
not
I.
Don't
want.
A
A
B
B
But
the
kubernetes,
of
course,
was
renovated
disruption
from
a
technology
perspective,
and
you
see
that
multiple
companies,
including
ours,
in
all
honesty,
are
still
making
their
transition
to
wash
containers
and
we're
doing
quite
well.
Now,
that's
which
is
nice
yeah,
that
is
nice,
but
that's
one
of
those
I
say
big
disruption,
categories
that
were.
A
B
100
score
so
far,
no
but
I
think
the
from
a
cloud
native
perspective
and
I
think
that's
both
from
a
private
and
public
Cloud
perspective.
The
the
cloud
native
part
is
that,
if
you're
going
to
use
native
Services,
regardless
of
the
platform
that
you're
using
you
actually
get
lost
out
of
it
and
I,
think
there's
always
a
discussion
which
is
a
valid
one.
Of
course,.
B
Cool
okay,
that's
fine!
No,
but
still
I!
Think
the
if
you're
looking
at
the
vendor,
looking
for
only
when
you're
doing
Cloud
native
stuff,
I
think
you're
lying
to
yourself,
because
vendor
login
is
not
something
that
only
exists
when
you
do
Cloud
native,
it
might
be
a
bit
more
apparent
because
the
focus
is
there
right.
But
if
you
have
a
complete
landscape
of,
let's
say
a
Linux
servers
VMS
that
are
hosted
by
maybe
even
external
Contracting
Parts
in
external
vendor
and
then
trying
to
move
to
a
new
technology.
B
A
Yeah
one
of
the
things
you
kind
of
are
putting
it
nicely
like
it's
more
obvious
like
when,
when
I
was
in
some
of
the
Consulting
I
was
doing.
We
were
first
trying
to
get
people
to
move
to
the
cloud
and
they
would
talk
about
like
how
concerned
they
were
about
the
security
of
the
access
into
their
environments,
and
you
know
and
and
I'm,
and
so
then
I
would
essentially
proceed
into
a
discussion
of
how
many
nightly
batches.
Do
you
run
with
how
many
different
partners
that
all
have
custom
firewall
punches?
A
You
know
and
and
I
think
that
the
what's
nice
about
those
native
Services
is,
is
your
I
think
you're
you're
more
clearly
making
a
choice
to
choose
that
vendor
or
whatever?
Then
you
did
sometimes
in
the
past,
where
you
were
just
kind
of
firing
off
this
API
over
here
and
nobody
really
knew
what
it
was
and
so
I
think
it's
a
little.
It's
a
little
simpler
to
have
a
like.
You
said
visibility
into
well:
where
is
our
vendor
lock
exactly.
B
Exactly
that
abstraction
layer
goes
up
and
up
and
up
even
a
bit
further
by
time,
so
I'm
also
trying
in
our
organization
of
course
right,
there's,
always
a
sort
of
stigma
on
a
bank,
because
the
bank
is
old
and
I
won't
say
that
we
don't
have
Legacy
systems
in
our
in
our
case.
But
I
do
see
that
we
we're
moving
towards
a
more
native
approach,
both
from
a
public
perspective
and
and
private
perspective,
and.
B
A
Well,
I
think
it's
particularly
difficult
because
you
you
can
develop
systems
in
the
cloud
that
are
very
traditional.
It's
just
you
know
and
and
it's
super
easy
to
fall
into
that
trap
and
you
really
aren't
taking
advantage
of
of
what
it
means
to
to
have
that
you
know
kind
of
distributed
infrastructure.
B
Exactly
and
I
think
the
the
mistake
that
we
tend
to
make.
Sometimes,
of
course
you
probably
know
the
7R
sort
of
AWS
as
well.
Let's
say
re-platform
refactor,
that
sort
of
stuff
and
what
we
sometimes
make
as
a
mistake
is
that
we
just
gives
our
current
application.
We
throw
it
all
in
the
same
container
and
then
we
hosted
container
rights,
and
then
we
say
no
now
we're
actually
Cloud
native.
A
Although
I
will
say
my
guidance
for
a
long
time
to
people
was,
if
you,
if
you
were
trying
to
bring
something
you
know
to
a
container
or
whatever
is
actually
to
do
that
to
start
right,
yeah
and
then
kind
of
start
to
pair
off
the
services.
You
know,
but
don't
let
the
the
you
know
you
have
to
re-architect
the
whole
thing
gate
you
from
even
taking
the
first
step,
which
can
be.
You
know
a
little
bit
scary,
because
all
the
documentation
tells
you
well.
A
You
can't
run
more
than
one
application
per
container
and
all
this
other
stuff
and
it's
like
you're,
better
off
kind
of
starting
to
think
container.
Even
if
you're
not
quite
doing
it
right
exactly,
you
know
as
quickly
as
possible
than
you
are
to
to
wait.
You
know
exactly
yeah,
it
was
actually
I
mean
in
a
lot
of
ways.
That's
why
the
big
push
behind
like
service
oriented
architecture,
kind
of
didn't,
take
off
very
well,
so
it
was
all
top
down.
B
Yeah,
but
it
makes
a
difference.
I
think
you
touched,
you
touched
up
on
the
the
microservices
I
think,
there's
always
as
it
is
with
everything
right.
There's
a
balance
between
where
your
predictability
or
productivity
is,
if
you're
talking
about
microservices,
then
if
you
peel
off
very
specific
services
from
an
existing
application,
that's
also
a
great
way
to
start
building
and
containerized
application,
because
it
also
makes
it
a
lot
smaller
and
easier
to
start
with
right.
That's
the
right
right!
You.
A
B
The
interesting
part
comes
yeah
yeah,
because
we've
we've
been
using
it
for
six
or
seven
years.
That's
when
we
first
started
sort
of
the
talks
about
going
towards
openshift,
because
hosting
our
own
kubernetes
platform
again
is
not
something
that's
our
in
our
core
business.
So
why
should
we
do
it
ourselves?
B
Then
it's
better
to
let
someone
manage
the
platform
or
the
minister
cluster
that
well
I
have
have
proven
experience
with
it,
but
still
not
nowadays,
and
that's
I
think
our
biggest
challenge
we
have,
let's
say
about
10
to
15
percent
of
our
applications
have
moved
and
we
do
see,
let's
say
see
the
exponential
growth
now
the
platform
consumability
is
is
going
up.
Yeah
yeah
yeah,
but
it's
been
well
a
tough
journey
to
get
to
get
to
move
over
right,
because
also
it's
not
only
from
a
tech
perspective.
If.
B
A
Yeah
I
mean,
and
it's
it's
always
so
tough
right
to
write
a
you
know,
a
cost-saving
story.
You
know,
and
you
know,
because
most
of
the
time
it's
like
you
know,
I'd
rather
I'd
rather
operate
a.
You
know
a
new
feature
story.
You
know
or
or
sell,
particularly
for
those
kinds
of
people.
So
you
know
I,
think
most
of
my
work
has
always
been
around
trying
to
make
sure
that
you
know
in
consulting
or
whatever
it's.
B
Yeah,
you
know
that's
the
perfect
way
of
doing
it.
It's
really
true,
yeah,
yeah,
I,
think
I
mean
we're
working
for
a
bank.
Obviously,
so
what
always
works
great
and
it's
not
even
a
lie,
to
be
honest,
actually
very
true
as
well.
Your
risk
appetites
also
right
it's
much
better
than
it
was
before.
If
you
actually
do
make
that
step
right,
yeah
and
that's
because
well,
banks
are
quite
risk-averse,
which
is
a
good
thing
right.
A
B
Yeah
but
I
think
that's
actually
an
interesting
point
because
being
sort
of
risk-averse
and
Innovation
usually
doesn't
go
together.
That
well
yeah,
so
I
think
even
well.
I'm,
not
generally
a
fan
of
doing
forced
proof
of
Concepts
and
forcing
people
into
a
room
to
do
something
new
right,
because
we
have
to
do
it
right.
A
B
In
these
kinds
of
Concepts
and
these
kinds
of
environments,
it
does
sometimes
work
to
really
get
people
in
the
room
and
do
something
completely
isolated
from
what
we
do
on
a
day-to-day
basis,
just
to
get
them
in
touch
with
the
new
technology,
because
you
can
get
lost
in
day-to-day
business.
If
you
don't
do
that
from
time
to
time,
right.
A
Sometimes
well
sometimes
I
don't
think
you
necessarily
even
know
you're
in
the
box
right.
You
know
it's
like.
Oh
there's
walls
out
there.
You
know
until
you
see
somebody
who's
doing
something
completely
different
yeah
and
it
really
does
change
your
mind
yeah.
You
know
and
I
think
and
a
big
part
of
that
kind
of
cloud
native
adoption
really
whatever's
like
you,
it's
fundamentally
architecturally
different
yeah
and
if
you're
not
seeing
that
part
of
it
you're,
not
really,
you
often
are
not
going
to
be
able
to
take
advantage
exactly.
B
Yeah
yeah,
and
that
makes
it
difficult
right
because,
usually,
if
you,
if
you
take
the
decision
makers
in
the
company,
it's
even
harder
for
them
to
explain
why
this
fundamentally
different
and
if
the
development
teams
don't
get
that
new
Concepts
either
for
them.
They
won't
explain
it
to
the
decision
makers
either
right.
It
makes
it
really
hard
to
make.
A
That
movement
going
right.
So
what
have
you
found
you
know
so
you
said
recently
you've
really
been
seeing
adoption
curve
kind
of
go
up.
What
what
do
you
think
has
been
that
driver?
Is
it?
Are
you
doing
something
different?
Is
it
Snowball
Effect?
Is
it
you
know
what
what
kind
of
hints
do
you
have
for
people
listening
to
the
show
for
them
to
be
able
to
get
other
people
to
adopt.
B
Things
so
I
think
it's
it's
drivers,
I
think
one
of
them
is
the
the
service
oriented
part
from
a
platform
perspective.
I.
Think
the
one
of
the
pitfalls
that
we
had
in
the
past
is
that
the
the
platform
team
is
actually
making
sure
that
the
developers
have
the
ability
to
develop
on
openshifts,
for
example
or
AWS
whatsoever,
that
they
actually
are
developing
the
platform
for
our
developers
to
work
on
and
because
they're
technicians
themselves
as
well.
B
They
also
like
to
make
the
platform
as
perfect
for
themselves
to
work
with,
but
sometimes
forget
the
customer
on
on
the
lra,
yeah
yeah,
and
that's
one
thing:
that's
really
going
well
recently,
also
because
of
the
communication
between
both
parties
and
I
think
the
second
part
is
related,
but
it
has
to
do
with
the
consumability
of
the
platform
yeah.
So
what
you
see
happening?
A
B
B
So
we
have,
of
course,
our
own
internal
gear,
repos
yep,
but
there
are
consumable
Docker
files,
Jenkins
files,
openshift
template,
yamos,
all
that
sort
of
stuff,
but
already
with
the
the
requirements
for
the
platform
team,
the
security
team
Governors
compliance
architecture
baked
in
right.
So
that
means
they're
pre-approved
they're
consumable
and
it
works
in
our
environments,
which
means
that
they
don't
have
to
start
all
the
way
from
scratch.
And
that's
something
you
see
that
the
adoption
is
is
really
going
going.
Well
so
far,
yeah.
A
Yeah
I
I'm
wholeheartedly
in
agreement.
One
of
the
things
that
I
actually
think
is
interesting.
That
I've
seen
a
few
organizations
do
is
also
do
rotation,
so
you
actually
have
some
of
the
you
know
normal
developer
teams.
You
know
one
of
their
members
joins
the
developer
enablement
team
for
a
few
months
you
know,
or
a
year
or
whatever,
and
then
you
know
kind
of
rotating
back
through
it's
kind
of
similar
to
the
model
of
you
know.
B
A
B
A
B
B
B
A
A
Almost
very
Noble
yeah
I
am
well
there's
something,
but
it's
certainly
a
lot
of
fun.
Yeah
Volkswagen
Bug,
oh
wow,
old
school,
like
that's
like
a
73
72.
It
looks
like
a
nice
color
as
well.
Yeah
I
had
I
had
one
that
was
like
dark
blue,
but
it
was
a
75,
so
it
was
like
a
Super
Beetle
yeah,
but
then
Chrome
everywhere
on
it.
It
was
gorgeous
first
time,
I
turned
it
on,
though
the
entire
steering
column
lit
on
fire,
so
yeah.
A
A
One
right
yeah
and
it's
a
very
straightforward
engine.
So
as
you
can
you
know,
basically
anybody
can
do
it
like
I
couldn't
well,
there
was
a.
There
was
a
great
book
in
my
my
father
actually
had
a
bit
of
a
you
know,
mechanical
bent,
and
so
he
would
guide
me
when
I
got
really
stuck,
but
for
the
most
part
I.
You
know
there
was
the.
What
was
it
the
Zen?
No,
it
was.
A
There
was
the
Zen
In
The
Art
of
Motorcycle
Maintenance,
but
there
was
like
a
similar
book
for
Volkswagen
bugs
yeah
and
yeah.
I,
basically
told
you
how
the
entire
car
work,
so
you
can
break
it
out
and
you
know
start
tinkering
with
stuff.
I
mean
the
worst
you
could
do
is
like
make
it
not
run
yeah,
and
so
it
wasn't
running
anyway.
It
doesn't
make
any
worse.
B
A
A
Gonna
have
any
idea
how
cars
work,
no,
no
I,
just
don't
think.
We
know
how
it
would
work
anymore,
like
I.
I,
didn't
know
this
that.
So,
even
if
you
have
a
gas
powered
engine
yeah
a
lot
of
the
time,
the
whatever
like
the
soundproofing
yeah,
is
so
good
that
you
actually
can't
hear
the
engine,
and
so
the
engine
noise
you
hear
inside
the
cab
is
actually
manufactured.
You're,
kidding,
yeah,
really
or
or
it'll
have
like
a
pipe
basically.