►
Description
Lightning Talk: Portworx by Pure Storage - Anthony Hodson (Portworx)
This OpenShift Commons Gathering was held on July 6th, 2022 live in London, England
https://commons.openshift.org
A
Thanks
for
joining
really
great
to
see
everyone
in
person,
of
course,
it's
been
years,
so
I'm
anthony
hudson,
I'm
cloud
native
solution,
architect
at
portworx,
portworx
part
of
pure
storage,
pure
storage,
bolt
port
works
about
a
year
18
months
ago.
It's
before
I
joined
but
yeah.
It
really
completed
their
portfolio
right
storage.
Well,
what's
what
is
portworx
well.
Portworx
is
a
storage
layer
of
kubernetes
okay,
so
we
help
simplify
the
running
of
stateful
apps,
which
opens
up
kubernetes
to
more
workloads.
A
So
if
you
had
a
business
case,
some
of
your
workloads
go
to
kubernetes
cost
saving
velocity.
That
sort
of
thing
imagine
if
the
doors
were
open
for
others,
that's
what
we
do.
So
we
work
with
some
big
enterprise,
clients
and
here's
some.
I
can
talk
about
so
t-mobile.
We
all
know
t-mobile
world
bank,
canada,
kpnr
a
managed
service
provider.
Let's
think
multi-tenant
kubernetes
roads
blocks
make
video
games
big
and
pandemic
kept
kids
quiet,
chg
healthcare
shows
that
we
can
get
the
hipaa
stamps
when
we
need
to
and
so
on
there.
A
A
They
said
well
for
enterprise
storage,
kubernetes,
enterprise,
storage,
outstanding
data
management
capabilities,
unmatched
deployment
possibilities
and
superior
management
features
thanks
very
much
so
and
if
you're
doing
cloud
native
data
storage,
you
know
differentiating
your
apps
that
are
actively
developing
for
cloud.
We're
good
for
that
as
well.
So
here
are
some
of
those
capabilities.
This
clicker
is
working.
A
It's
not
me
very
good,
so
yeah
here's
some
of
the
capabilities,
so
we've
got.
You
know
when
you're
thinking
of
working
with
different
platforms.
You
often
have
to
do
things
twice
right.
I
don't
know
about
you.
I
don't
like
doing
things
twice
and
containers
were
here
to
solve
that,
but
it
seems
that
you
know
the
storage
piece
didn't
really
get
that
memo.
A
So
portworx
includes
modules
for
core
storage,
backup
dr
applications,
data
migration
and
so
on,
and
we'll
talk
about
those
a
bit
more
and,
of
course,
automation
of
your
infrastructure
and
we
work
with
any
distribution.
Of
course
we
work
with
with
red
hat's
openshift,
but
we
work
with
others
as
well,
but
you
probably
think
that's
a
lot
of
icons.
Where
do
I
start?
So
when
I
speak
to
organizations
about
you
know
what
we
can
do
for
them.
A
I
start
with
the
base
responsibility
of
the
operators
right,
someone
created
your
cluster
and
the
workload's
inside
of
it
and
they
use
the
command
to
do
that
right.
I
don't
want
to
break
anyone's
sleep
pattern
at
night,
but
someone
could
destroy
it
with
those
commands
as
well
right.
So
we
need
to
make
sure
if
that
happens,
we
can
restore
right.
A
We
can
get
it
back,
so
you
must
have
a
way
to
get
your
date
back
and
if
this
happens,
you
know
we've
got
to
restore
and
I
put
restore
first
there,
because
you
know
we.
We
only
really
care
about
that
bit.
We
do
that
bit
the
backups.
We
only
really
care
about
the
restoring.
I
think
the
analogy
is
like
cooking
and
eating
right.
You
spend
a
lot
of
time
doing
the
cooking,
but
you
only
really
care
about
the
eating.
We
don't
call
eating
cooking
anyway.
That's
my
hobby
horse,
okay.
A
So
so
when
I
speak
to
kubernetes
about
organizations
about
running
it
in
production,
I
ask:
how
do
you
back
up?
And
I
say
you
know
they
might
say
we
take
care
of
it.
We've
got
stateless
apps,
we
don't
really
mind,
that's
fine.
I
say
all
right.
Here's
my
card
come
back
to
me,
we'll
talk
about
when
you
get
there
or
they
say
you
know
we
back
up
your
snapshots
and
we've
got
a
product.
Maybe
it's
from
an
incumbent
storage
provider
that
sort
of
thing
and
they're
not
usually
built
for
kubernetes.
A
But
you
know
it's
better
to
avoid
these
problems
right,
avoid
doing
things
twice,
make
it
easy
for
yourself.
So
poor
works
have
been
around
working
with
kubernetes
since
2014,
and
you
know
we
grew
up
in
the
same
place
right
silicon
valley
over
there
mountain
view,
and
we
were
around.
We
were
originally
doing
things
with,
not
in
particular
kubernetes,
but
we've
been
around
the
space
for
a
long
time.
A
So
paul
works
has
a
backup
tool
which
simplifies
your
backup
right.
Where
are
we?
Are
we
on
a
different
slide
up
here?
I
think
so.
Let's
talk
you
through
this
one,
so
we
have
appleware
backups.
We
can
auto
discover
clusters.
We
have
a
sas
solution
for
poor
works
as
well.
So
if
you're
in
a
bank,
you
probably
want
to
run
your
own
if
you're
elsewhere
in
the
world,
you
might
want
someone
to
run
this
stuff,
for
it
also
appears
then
in
a
different
blast
radius.
A
So
someone
took
out
your
main
thing:
they're
not
going
to
take
out
our
thing,
but
we
do
our
back.
So
you
know
don't
create
new
users
for
things
use
your
existing
systems,
there
interfacing
with
urad,
very
popular
and
other
oicd
providers,
and
we
also
do
ransomware,
which
is
big
right,
there's
a
lot
of
ransomware
going
on
at
the
moment.
I
was
doing
a
talk
the
other
day
about
it.
Eighty
percent
of
organizations
are
getting
hit
by
a
ransomware,
so
you
look
at
yourself
two
to
the
right
two
to
the
left.
A
One
of
you
will
probably
not
get
hit
by
ransomware
this
year,
but
it's
not
probably
that's
bad
news.
Okay
and
the
average
cost.
If
I
remember
my
stats
correctly
is
1.86
million
dollars
for
ransomware
attack,
so
get
your
data
squared
away,
make
sure
you've
got
it
safe
and
make
sure
you
can
restore
easily
and
portworx
will
help
with
that
right.
A
Here
we
go.
Oh
it's
back
there.
We
go
okay,
so
beyond
the
backup
piece,
enterprises
want
an
ability
to
run
workloads
with
those
non-functional
requirements
that
you
always
have
a
requirement
for
right.
So
is
it
highly
available?
Can
I
make
it
work,
even
if
one
of
the
data
centers
goes
down
right
if
you're
using
cloud
that
might
be
two
availability
zones?
Of
course
you
can't
stretch
storage
between
those
availability
zones
and,
and
that's
a
problem
right
so
portworx
helps
with
that.
A
Can
I
make
it
so
that
I
can
stretch
clusters
between
data
centers
that
are
in
the
same
metro
area?
We
use
the
word
metro
and
allow
you,
when
there's
less
than
10
milliseconds,
to
stretch
cluster
between
them
with
zero,
rpo
and
zero
five
minutes
roughly,
you
don't
generally
want
to
send
things
over
immediately.
You
want
to
work
out
what's
gone
wrong,
but
you
can
script
all
that
you
get
it
over
nice
and
quickly.
I
think
there's
another
one
here
we
go
so
a
bit
more
information
on
that
right.
We
will
scale
your
resources.
A
We
will
scale
your
storage
up
when
it's
needed
as
well,
so
that
you're
not
doing
that
toil
right,
someone's
running
out
of
space
or
you're,
not
having
an
incident
where
someone
has
run
out
of
space
security
right,
we'll
we'll
deal
with
the
secure
change
of
systems
using
an
operator
we
use
an
operator
we'll
know
more
about
operators
now.
So
that's
good!
Thank
you
for
that.
A
A
Databases
are
not
installing
databases
and
not
the
goal
of
the
operative
who
wants
a
database.
They
just
want
to
get
to
the
database
right,
and
we
all
understand
this
because
we've
all
seen
the
rise
of
database
as
a
service
platform
as
a
service
and
that
sort
of
thing.
But
when
you
have
a
cluster,
that's
available
to
you
and
you,
your
principle
of
going
into
kubernetes,
to
keep
with
portability
and
you're
trying
to
avoid
doing
things
twice:
databases
as
a
service.
Isn't
that
appealing,
because
you'd
have
to
code
for
it.
A
Azure
you'd
have
to
code
for
it
in
aws,
maybe
you're
an
ali
cloud,
whatever
right
they're
all
different.
They
might
use
the
same
underlying
systems,
but
wouldn't
it
be
nice
if
you
could
deploy
things
in
the
same
way
within
your
cluster
and
have
that
same
experience
where
your
developers
just
say,
hey,
give
me
a
database
and
they
get
one
right.
You
as
the
operator
can
say:
okay.
Well,
it
needs
to
be
this
size
and
it
needs
to
be
patched
and
actually
all
that
operation
goes
to
portworx
to
look
after.
So
that's
what
we've
done
right.
A
We've
developed
a
portworx
data
services
solution
and
it's
available
to
run
in
your
cluster
and
we've
got
a
really
strong
road
map
to
bring
these
things
together.
I
think
it's
public,
but
we're
we're
going
into
in.
We
hope
to
go
into
very
soon
and
other
stuff
around
microsoft,
towards
the
end
of
the
year.
A
So
there's
you
know
this
is
sort
of
the
architecture.
I
know
you
love
that
term
slide
to
say
you
know.
These
are
all
the
things
we
we
work
with.
I'm
gonna
be
on
a
booth
over
there
for
the
afternoon
and
if
you
want
to
talk
about
any
of
these
pieces
come
and
have
a
chat
and
we'll
work
out
if
we're
the
right
fit
for
you,
and
if
we
are,
let's
make
something
happen,
I
think
that's
there
we
go.
One
more
thing
to
raise.
A
A
lot
of
this
stuff
can
be
done
over
your
lunch
time.
I
call
it
a
lunchtime
poc
particularly
around
the
backup
piece,
so
if
you
want
to
give
the
backup
a
try,
install
it
it's
two
lines
of
code
into
your
cluster
and
I
suggest
next
time
you
need
a
copy
of
your
cluster,
just
run
that
pull
it
out
and
see
how
quickly
the
app's
come
up.
I
was
at
a
client
the
other
day.