►
Description
Cloud Seed is an open-source program lead by GitLab Incubation Engineering in collaboration with Google Cloud.
Deploying web application (and related workloads) from GitLab to major cloud providers should be trivial.
Cloud Seed makes it ridiculously simple and intuitive to consume appropriate Google Cloud services within GitLab.
00:00 Intro
00:53 Recap - What's shipped thus far
01:46 Databases proof-of-concept
02:39 Demo 01 database tab and services
03:44 Demo 02 create database instance
04:28 Demo 02 support versions and machine types
05:09 Demo 03 instance creation result
06:22 Closing thoughts
07:13 Google app verification update
08:08 Real closing thoughts
A
A
A
A
We
want
to
support
the
three
big
post,
three
big
relational
date-
databases
that
is
postgres
mysql
and
sql
server,
and
the
use
cases
that
we
like
to
address
are
app
modernization
and
cloud
migration.
That
is
moving
a
traditional
app
to
the
cloud.
Relational
databases
are
still
widely
used
among
enterprise
and
open
source
applications.
We
want
to
support
those
and
they
are
an
excellent
solution
for
several
types
of
modern
application.
Develop
development
needs
as
well.
So
these
are
the
use
cases
we
want
to
support
all
right.
A
Let's
begin
with
the
first
demo,
we're
going
to
look
at
the
database,
tab
and
list
of
database
services.
If
you
remember,
we
had
introduced
a
google
cloud
page
under
the
project
site
menu
which
had
the
configuration
tab
for
service
accounts
where
you
can
create
service
accounts,
and
you
could
configure
your
deployment
regions
as
well.
A
A
Now
we've
introduced
a
new
databases,
tab
and
you
can
see
the
list
of
database
services
available
cloud
sql
for
postgres
cloud
sql
for
mysql
and
cloud
sql
for
sql
server
allows
you
to
provision
fully
managed
relational
databases
for
these
three
popular
and
well
used
database
technologies
and
then
there's
a
bit
of
road
mapping
here,
where
future
database
services
being
provided
by
gcp
could
also
be
added.
A
I
grant
permissions
to
gitlab,
and
now
I
see
this
form
where
I
can
select
my
gcp
project
I
can
select
which
environment
I
want
to
create
a
database
for
in
this.
Let's
say
I
want
to
create
it
for
my
master
environment
and
I
can
select
machine
types
and
I
can
select
database
versions
press
the
button
and
that's
it
shedding
a
bit
more
light
on
what's
possible,
and
here
are
the
supported
database
versions
and
machine
types.
Postgres
support
is
all
the
way
from
9.6
to
10,
11,
12,
13
and
14..
Mysql
is
the
latest.
A
Based
on
your
needs,
you
can
select
the
machine,
the
number
of
cpus
and
the
storage
or
the
memory
available.
You
can
select
these
options.
The
create
instance
form
when
successful
you'll
see
a
message
of
this
kind.
It
says
cloud
sql
instance
creation,
request
successful,
and
it
says
that
the
expected
resolution
time
is
a
few
minutes.
A
Provisioning
sql
databases
on
gcp
is
an
asynchronous
operation.
It
takes
two
to
three
minutes
and
therefore
the
information
is
presented
in
a
way
where
you
are
informed
that
the
resolution
time
is
a
few
minutes.
In
this
case.
What
happens
is
you
will
not
see
the
result
of
your
action
until
the
operation
in
the
back
end
is
complete
once
the
operation
is
complete,
you
switch
to
the
databases,
tab
and
you
scroll
down
and
you'll
see
a
list
of
instances
available
for
your
project.
A
These
instance
details
connection
details
are
stored
as
ci
wires.
This
is
an
approach
similar
to
what
we
do
with
auto
devops
and
what
we
do
for
the
cloud
run
deployments
as
well,
and
you
can
see
the
various
credentials
are
stored
as
part
of
your
ci
pipeline
environment
variables.
A
What
you
saw
so
far
is
a
proof
of
concept.
It
will
start
to
ship
over
the
next
weeks
and
months
as
I
break
down
this
proof-of-concept
branch
into
smaller,
mrs
that
get
merged
into
the
main
code
base.
Mind
you.
This
needs
to
be
hardened
and
needs
to
go
through
the
security
reviews,
the
list
of
ci
wires
that
are
shared.
I
think
there
are
too
many
of
them.
We
need
to
strip
down
the
number
of
ci
wires
available
to
the
bare
minimum
that
an
application
needs
to
connect
to
a
database.
A
The
original
request
was
dated
april
26th,
but
now
we're
on
top
of
things,
we've
taken
care
of
all
the
all
the
all
the
tweaks
we
had
to
make
on
our
end,
and
now
the
ball
is
in
google
skirt
and
we'll
keep
everybody
informed
in
form
of
these
video
updates
on
what's
going
on
over
there,
the
google
app
verification
step
was
required
because
gitlab
now
becomes
the
manager
of
your
cloud
resources
and
services.
This
is
an
additional
scope.
Thus
we
needed
to
go
through
the
review
process.