►
Description
London OpenShift Commons Gathering 2019 Lightning Talk PureStorage
A
Pure
storage,
now
storage,
before
you
start
checking
all
your
emails
on
your
phones,
we're
actually
about
data.
So
I'll
make
this
quite
interesting
a
bit
about
pure
storage.
We
started
off
in
silicon
valley.
We
did
a
billion
dollars
in
revenue
yes
last
year,
which
makes
us
the
fastest
enterprise
it
company
to
be
unicorn
status.
A
How
we're
doing
that
is
disrupting
the
storage
market
by
helping
innovators
build
a
better
world
with
data.
Now
the
observation
that
I'm
sure
you're
all
familiar
with
applications
moving
from
siloed
environment
to
much
more
stateless
scale
out.
Well
we're
seeing
the
same
with
data
in
in
the
industry
that
structured
business
data
is
moving
to
far
more
unstructured
lots
of
different
sources,
lots
of
different
dynamics
and
different.
I
o
patterns,
so
put
that
into
context
of
a
solution
and
where
we
fit.
A
Where
that's
important
to
this
audience.
Is
that
a
lot
of
that's
being
built
in
containerized
environments
and
needs
to
be
managed
at
enterprise
scale
and
we're
working
with
clients
having
compute
platforms,
both
cpu
and
gpu,
just
to
fill
out
the
the
picture
for
you?
A
Now
the
neat
thing
about
it
is:
it
gives
you
a
single,
simple
interface
for
your
data
environments,
whether
you're
edge
core
cloud,
and
it
has
a
single
commercial
model
which
we
could
get
into.
If
you
come
and
see
us
at
the
booth,
a
bit
of
boasting
we've
won
the
ai
best
innovation
two
years
running
at
a
iconics.
A
This
slide
isn't
rendering
correctly,
but
you'll
see
the
the
airy
product
that
we
have
now.
What
it
is
is
a
a
converged
platform
for
ai.
It's
rather
bling
you'll
see
the
photograph
we're
bringing
out
a
grey
pinstripe
version
for
the
uk
market.
A
The
interesting
thing
for
this
audience,
I
think,
is
probably
the
software
stack
so
working
up
from
the
bottom.
We
provide
us
a
scale
out:
object,
storage,
environment,
all
flash,
a
software
data
layer
that
manages
the
data
we
work
with
our
partner
nvidia
for
the
the
gpu
super
computer
layer,
and
then,
on
top
of
that,
we've
got
a
gpu
optimized
container
environment,
which
allows
you
to
run
eight
different
ai
frameworks
and
effectively
a
b
test,
your
algorithms,
when
you're
building
models
for
for
large
scale
data.
A
We
also
have
a
scaling
toolkit
which
allows
you
to
scale
far
more
in
terms
of
of
compute
processing
capability
than
you
would
be
able
to
on
a
public
cloud.
For
example,
one
of
our
customers
is
a
rather
large
social
media
network.
They
use
exactly
this
reference
architecture
to
process
all
of
your
data.
A
So,
let's,
let's
not
worry
about
too
much
about
the
integrations
on
the
left-hand
side,
probably
interesting
for
this
audience,
we've
got
all
of
our
integration
code
through
rest.
Api
is
all
held
on
github
we
integrate
with
ansible
and
most
of
the
automation
tool
sets.
We
have
an
openshift
integration
for
all
of
our
platforms.
A
This
is
particularly
called
pure
service
orchestrator,
so
we
have
volume
plug
into
docker.
We
have
flex
driver
and
plugins
for
kubernetes,
and
that
provides
both
stateless
and
stateful,
so
persistent
storage
for
container
environments
across
block
and
object
and
and
handles
all
of
the
automation
work
for
that
data
layer.
A
The
final
thing
is
that
we
also
partner
with
with
sis
we
have
partnership
with
accenture.
So
those
of
you
that
don't
want
to
necessarily
do
this
for
yourselves,
there's
there's
a
lot
of
complexity
involved.
We
have
some
pocs
running
with
our
partner
accenture
in
in
their
labs
and
they're,
looking
for
clients
to
test
new
ideas
out
with.