►
From YouTube: London OpenShift Common Gathering 2019 Lightning Talk CrunchyData PostgreSQL-as-a-Service OpenShift
Description
London OpenShift Commons Gathering 2019 Lightning Talk CrunchyData
PostgreSQL Operator
A
All
right
so
I'm
Joe
Conway,
the
VP
of
engineering
for
crunchy
data,
if
you
take
just
one
thing
away
from
this
brief
talk
that
I
have.
If
you're
thinking
about
Postgres
micro-services,
open
shift
and
security,
you
want
to
see.
Crunchy
crunchy
has
led
the
market
with
Postgres
when
it
comes
to
security.
We
are
the
author
of
security
benchmarks
for
the
US
government
for
the
CI,
a
so
Center
for
Internet
Security.
A
We've
got
a
Common
Criteria
certified
version
of
Postgres,
which
essentially
means
extensive
testing
to
validate
the
security,
as
well
as
the
fact
that
we
literally
look
at
every
patch
that
gets
committed
to
post
grass
or
post
GIS
or
several
other
components
that
are
part
of
that
suite
of
tools.
So
when
it
comes
to
security,
we've
got
that
in
spades.
We've
also
got
and
what
I'm
going
to
spend
most
of
the
time
talking
about
for
the
rest
of
the
slides.
A
We've
got
not
only
a
very
robust
containerized
solution
of
postcodes,
not
just
standing
up
a
database
but
standing
up
an
entire
H,
a
cluster
with
integrated
backups
and
integrated
monitoring,
and
basically
all
the
things
that
you'd
want
to
run
a
database
as
an
enterprise.
We've
got
containers
for
that.
But,
more
importantly,
we
also
have
an
operator
that
allows
you
to
orchestrate
many
many
clusters
of
databases.
A
A
So
briefly,
an
operator
I
assume
most
of
the
people
here
probably
know
what
this
is.
But
an
operator
is
basically
an
application
controller
that
lets
you
orchestrate
nuances
and
it
especially
with
stateful
services
in
kubernetes
or
an
open
shift
environment
and
basically
our
operator,
the
crunchy
Postgres
operators.
Short
name
is
P,
go
P
geo.
It
can
be
accessed
via
REST,
API
and
I'll.
There's
a
diagram
on
that
a
little
bit
later.
A
Why
would
you
use
it?
Well?
If
you've
got
complex
operations,
would
you
typically
do
with
a
database?
This
reduces
those
complex
operations
down
to
one
line
commands
it
lets.
You
do
many
different
customizations
to
various
database
clusters
that
you
might
need
to
work
with,
and
you
can
still
use
the
same
workflow
as
I
mentioned,
we
have
a
simple
CLI,
that's
available
command
line
interface,
but
because
that
interface
is
actually
built
on
top
of
a
REST
API.
A
We
also
have
a
GUI
that
we
can
lay
on
top
of
that
and
later
on,
maybe
further
integrations
with
things
like
the
open
shift
GUI
it
allows
you
to
manage
at
scale,
so
you
can
manage
potentially
thousands
of
different
database
clusters
and
we
have
customers
that
are
actually
doing
this
at
this
kind
of
scale.
Now,
both
in
the
commercial
world
and
in
the
government
and
as
far
as
again
security.
By
having
the
operator,
you
don't
have
to
give
full
access
to
your
OpenShift
cluster.
A
The
value
of
open
shift
for
managing
Postgres
instances
basically
goes
up
exponentially
as
the
number
of
instances
of
the
database
that
increase
and
what
what
this
operator
basically
boils
down
to
is.
We
are
Postgres
experts.
Our
company
actually
is
responsible
for
about
a
third
of
the
code
that
is
in
the
upstream
Postgres
database.
I'm
a
committer
we've
got
several
other
committers
on
staff.
We've
got
the
person
who's
founded
the
open,
the
post,
GIS
spatial
extender
for
Postgres
is
also
our
staff,
so
we've
got
extensive
expertise
in
this.
A
So
this
is,
it's
been
GA
since
2017.
So
as
far
as
operators
go,
it's
actually
been
around
for
a
while
and
it's
it's
very
robust
and
it's
it's
been
used
in
production
and
it
continues
to
get
developed
based
on
direct
feedback
from
users.
So
importantly,
the
features
that
are
in
there
have
been
driven
by
real
users
with
real
production
needs.
A
You
can
define
resources,
and
it
also
does
some
some
smart
deployments
in
terms
of
using
node
affinity
to
allow
you
to
specify
where
things
go.
Here's
some
examples,
pgo
create
London,
will
just
create
a
pod.
It's
basically
an
H,
a
database
cluster,
one
symbol:
simple
command,
set
up
increase
the
replica
count,
set
up
your
backups
and
then
schedule
a
full
backup
at
midnight
and
differential
backups
every
six
hours
during
the
day.
It's
simple
to
do
on
the
right
here.
You
can
see
that
we've
got
the
command
line.
A
Real
brief,
not
enough
time
in
five
minutes
to
go
into
this,
but
this
is
the
type
of
cluster
you
can
deploy
with
these
simple
commands:
high
availability
cluster
across
all
of
your
open
shift,
and
this
is
where
it's
available
the
operators
available.
As
I
said
before,
github
crunchy
data
Postgres
operator,
we've
got
a
table
in
the
other
room.
Please
stop
by
and
see
us.