►
Description
In this talk from the OpenShift partner theater at Red Hat Summit 2018, Anthony Farinha and Mike Wiederhold from Couchbase demo the Couchbase Operator in Kubernetes.
A
Great
well
good
morning,
everyone
I'm
Anthony
Farina
from
Couchbase
I,
lead
our
technology
partnerships
and
solution
partner
relationships
at
the
company.
We're
happy
to
be
here
at
Red
Hat
summit,
as
some
of
you
noticed
in
the
keynotes
the
day,
one
keynotes,
both
from
some
customers
joint
customers
like
Amadeus,
as
well
as
some
of
the
open
shift
team.
We've
demonstrated
really
some
exciting
innovation
in
the
open
shift
in
kubernetes
ecosystem.
A
My
colleague,
Mike
Weir
hold,
who
manages
the
engineering
team,
will
co-present
with
me
and
we'll
do
we'll
do
a
demo
and
an
explanation
of
what
we've
done
with
operators
in
kubernetes,
but
before
we
do
that
I'll
jump
in
just
to
provide
a
little
bit
of
a
level
set
of
Couchbase
who
we
are,
why
we
exist
and
why
we're
growing
and
why
our
customers
are
really
investing
with
us.
We're
a
venture,
backed
private
company
really
with
the
who's
who
list
of
investors
and
board
members,
including
me,
Mayfield
and
Excel.
A
One
of
our
board
members
is
Don
Chamberlain,
who
really
Co
invented
sequel
at
IBM.
Even
Jeff
Epstein
ex
CFO
of
Oracle
is
on
our
board.
So
we
are
an
enterprise,
no
sequel
data
platform,
with
a
tremendous
growth
curve
ahead,
as
illustrated
by
our
customer
deployment,
so
we're
a
horizontal
data
play.
We
have
use
cases
that
fit
across
different
verticals,
and
some
of
our
customers
are
here
in
the
whether
it's
financial
health
care
travel,
retail
e-commerce,
a
lot
of
use
cases
that
we'd
be
happy
to
step
through
in
more
detail.
But
why
is
this
happening?
A
Why
our
customers
really
adopting
Couchbase?
Everyone
talks
about
digital
transformation?
What
are
the
pressures
existential
threats
for
all
of
these
enterprises
to
be
more
innovative,
respond
with
agility
and
really
show
value
and
customer
loyalty
with
really
differentiated
offerings?
And
the
database
is
one
of
the
key
impediments.
The
traditional
legacy
relational
databases
have
proven
to
be
one
of
the
impediments
for
making
these
digital
transformation
project
successful,
which
is
why
we've
really
had
a
lot
of
traction
in
the
market.
A
Oltp,
based
and
OLAP
systems
have
their
limitations,
but
specifically
when
it
comes
to
schema
rigidity,
flexibility,
the
ability
to
deploy
at
pace
and
to
really
innovate,
and
what
a
lot
of
people
are
doing
is
trying
to
layer
in
some
point
solutions
to
get
you
performance
at
scale
to
give
you
a
geo,
replication,
resiliency.
So
there's
a
lot
of
point.
Solutions
that
come
onto
the
market
to
address
these
things
adds
complexity,
adds
cost
and
Couchbase.
We've
created
a
data
platform
which
we
term
the
engagement
database.
A
The
engagement
database
complements
your
OLTP
and
your
OLAP
investments,
whether
you
have
a
Hadoop
data
Lake.
We
have
connectivity,
but
this
is
really
as
Amadeus
Illustrated
yesterday.
In
the
keynote
we
play
in
customer
engagement,
applications
so
interactions
mobile,
IOT
likes
tweets,
browse
personalization
at
the
edge
is
what
we
define
as
an
engagement
data
engagement
database
use
case
and
to
have
an
engagement
database.
It
has
to
be
cloud
native
multi-cloud
portable,
with
those
applications
has
to
be
mobile
first
and
really
offline.
A
First,
even
if
you
don't
have
connectivity
that
app
always
has
to
work
and
be
responsive,
we
have
sequel
for
JSON,
so
we're
both
a
key-value
and
a
document
store.
So
we've
got
all
the
built-in
enterprise
capabilities
and
we're
really
built
for
performance
at
scale,
which
is
really
a
unique
differentiator
for
us
and
with
all
the
security
features
that
are
expected
in
an
enterprise
data
platform.
Just
to
illustrate
what
we've
built
here,
we're
memory.
A
First,
we
have
a
list
and
with
the
persistent
layer
we
can
scale
horizontally,
we're
a
scale
out
distributed
database
and
the
data
platform.
In
addition
to
having
replication
and
indexing
capabilities,
we've
iterated
the
platform
to
add
new
capabilities.
Mobile
analytics
for
complex
query:
query
for
for
operational
queries
with
with
nickel
and
1q
lrj
sequel
for
json
and
full
text
search,
inverted
index
capabilities.
So
we
can
flatten
the
stack
with
your
deployments
and
take
out
a
lot
of
significant
redundancy
and
and
cost
and
give
you
high
performance
at
scale.
A
Are
the
customer
problems
that
we
typically
solve
and
we'd
be
happy
to
discuss
at
the
at
the
booth?
Are
around
three
vectors?
How
do
you
improve
customer
experience?
I
got
a
I
got
a
roll
out.
A
new
app
I
got
a
change.
My
e-commerce
I
have
to
change
how
I
engage
my
customers,
those
are
CMO
CBO
CIO.
How
do
I
use
the
data
at
the
edge
for
better
monetization
and
customer
loyalty,
faster
time-to-market?
Typically
a
DBA
or
data
governance,
or
data
modeling
thing
where
the
rigidity
of
a
relational
database
is
holding
you
back
from
innovating.
A
So
we
we
address
those
use
cases
and,
of
course,
in
the
IT
layer,
people
that
are
struggling
with
Oracle
elas
or
a
traditional
database
cost
or
your
scaling
up
Exadata
or
you
have
kind
of
a
lot
of
technical
debt.
A
mainframe
that
you
want
to
offload
those
IT
and
cost
reduction
plays
are
a
good
fit
for
for
Couchbase.
A
What
do
we
announce
this
week?
So
we
have
a
great
relationship
with
Red
Hat.
Well,
we
announced,
or
a
Red
Hat
demonstrated
in
the
keynote
yesterday-
was
our
investment
in
the
Couchbase
autonomous
operator
in
kubernetes,
so
we've
extended
those
api's
and
we're
the
first,
no
sequel
player
to
really
have
automated
operations
in
incur
Burnette
ease
on
openshift,
so
no
downtime
elastic
scaling
and
you
can
really
run
with
confidence,
business,
critical
applications
on
open
shift
across
clouds,
no
vendor
lock-in
with
geo
replication
built-in.
A
A
But
our
solution
fundamentally
solves
some
of
these
core
problems:
geo
replication
at
scale
with
fast
in
memory
performance,
centralized
management
we're
the
easiest
database
to
to
to
manage,
especially
with
this
autonomous
operator,
auto
provision,
Auto
recovery
for
business
critical
applications,
and
we
do
it
with
open
shift
across
any
public
cloud.
You
can
run
Couchbase
in
any
public
cloud
through
the
marketplace
we
run
on
bare
metal.
You
can
do
it
in
private
cloud.
B
So
an
operator
basically
something
that
allows
us
to
encode
exactly
what
a
human
would
do
when
they
were
managing
a
cluster
in
Couchbase,
but
for
all
that
to
happen
automatically
and
so
we're
going
to
show
you
today
is
a
demo
of
some
of
those
things
running
and
so
today,
what
that
means
is
that
you
can
just
based
on
filling
out
a
simple
configuration.
You
can
automatically
deploy
a
Couchbase
cluster
and
you
can
also
handle
failure
recoveries.
B
So
if
you
have
nodes
that
go
down,
you
can,
if
it
happens
at
three
o'clock
in
the
morning,
it'll
just
be
taken
care
of
automatically
by
the
operator.
So
that's
what
we
can
do
today,
but
tomorrow
we'll
be
able
to
do
automatic
upgrades,
automated
backup
and
restores,
and
we
our
goal
is
to
automate
every
task
that
can
be
that
a
human
would
do
when
they're
managing
their
clusters,
and
so,
if,
if
you're
running
your
operator
and
open
shift,
then
the
first
thing
that
you
need
to
do
is
install
your
operator.
B
You
just
run
a
simple
command
that
says,
create
the
operator
and
this
will
install
a
custom
resource
which
tells
kubernetes
about
exactly
how
you
define
a
Couchbase
cluster
and
will
install
the
software
to
manage
your
Couchbase
clusters,
and
so,
if
I
jump
over
here
to
the
OpenShift
web
console,
we
can
now
see
that
we
have
a
deployment.
That's
running
the
Couchbase
operator.
We
put
this
in
a
deployment
because,
if
your
operator
ever
crashes,
you
want
it
to
come
back
up
so
that
you
it
can
keep
managing
your
clusters.
B
B
B
This
is
as
simple
as
running
another
command,
which
just
says
create
the
Couchbase
cluster
and,
if
I
jump
back
over
here
after
I
do
that
to
the
open
shift
administration
console,
we
can
see
our
first
pod
has
already
come
up
and
I've
created
a
route
in
advanced
so
that
we
can
get
to
the
Couchbase
UI.
So
this
usually
takes
a
second
before
the
route
will
recognize
the
pod.
B
Okay,
so
I
just
created
a
separate
way
to
get
into
the
admin
console
here,
but
now
you
can
see
that
the
entire
cluster
is
already
balanced
and
we
have
the
one
bucket
that
we
created.
So
one
of
the
really
interesting
things
that
the
operator
also
takes
care
of
is,
if
you
have
a
node
crash,
and
so
if
I
come
back
to
the
service
tab
and
then
I
just
go
and
I
say:
I
want
to
delete,
see
be
example:
zero,
zero,
zero
two.
B
This
would
simulate
a
failure
of
a
node,
and
so,
if
I
come
back,
you
can
already
see
that
the
node
is
down
in
the
Couchbase
cluster.
So
we
have
our
auto
failover
timeout
set
to
30
seconds,
and
so
after
30
seconds,
this
node
will
be
auto
failed
over.
And
what
that
means
is
that
so
right
now,
a
third
of
the
data
is
actually
unavailable
in
the
cluster.
But
once
the
node
is
auto
failed
over,
we
have
replicas
on
the
other
nodes,
and
so
those
replicas
will
be
activated.
B
So
your
applications
can
continue
to
access
the
data.
Once
that
happens,
we
will
add
in
another
Couchbase
pod,
and
then
we
will
rebalance
all
the
data
so
that
we
get
right
back
to
the
state
that
we
want
that
we
were
expecting
to
be
in.
So
we
can
see
that
the
the
node
has
just
been
auto
failed
over
and
we
should
see
another
node
get
added
into
the
cluster.
Any
second.
B
All
right,
I
just
kicked
that
off
myself
there
so
yeah,
so
those
are
the
two
main
operations
that
were
able
to
do
with
the
couch
bass
operator
and
we
also
support
other
things
such
as
auto
scaling.
So
if
you
want
to
scale
up
your
cluster,
it's
as
simple
as
coming
back
to
this
configuration
changing
three
to
four
and
that
will
automatically
have
the
operator.
Add
a
new
note
for
you.
A
All
right
thanks,
Mike
and
thanks
for
folks,
if
you
want
to
learn
more
about
the
Couchbase
operator,
we're
in
booth,
342
right
behind
you
here.
I
also
want
to
call
out
some
of
the
resources
that
are
available.
We
did
post
at
cube
con
a
nice
blog
post
on
how
to
use
the
operator
for
kubernetes.
We
are
in
a
technical
preview
and
in
beta
in
collaboration
with
Red
Hat.
So
some
of
our
customers
are
experimenting
with
the
operator
and
you
can
hit
that
link
to
to
learn
more
there's.
A
A
We
we
have
a
technical
workshop
today
in
New
York
and
our
main
show
where
UPS
United,
Airlines
and
a
number
of
customers
are
describing
how
they
use
Couchbase
and
Red
Hat
will
be
a
platinum
sponsor
for
future
Couchbase
events,
including
coming
up
in
London
next
month,
and
our
big
show
in
Silicon
Valley
later
in
the
fall.
So
we
look
forward
to
engaging
with
you
and
appreciate
your
time
here.
Thank
you
very
much.