►
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
You
want
to
record
already
recording
great
well
welcome
everybody
to
the
february
16th
cap
call
I'm
filling
in
for
troy
today.
A
couple
of
minor
updates
from
the
cff
we've
got
a
exciting,
hands-on
lab
next
week
from
stark
and
wayne
james
hunt
is
gonna,
be
leading
us
through
how
to
how
to
cf
protect
and
how
to
shield
your
data.
I've
already
gone
through
that
lab.
It's
pretty
pretty
interesting,
so
you
can
check
that
out
next
tuesday
at
10
a.m.
Pacific.
A
There
I
don't
have
any
real
details
to
share,
but
I
will
say
that
there
is
a
tentative
plans,
are
shaping
up
for
a
virtual
event,
sometime
in
june
or
july,
have
more
details
next
month
or
the
month
after
the
latest,
but
that
that's
dates
are
being
floated
around
and
and
yeah
should
have
a
little
bit
more
information
than
that
non-information.
I
just
gave
you
in
another
month
or
two
also.
The
technical
governance
meetings
on
friday
mornings
seem
making
pretty
good
progress.
A
That's
in
the
github
community
github
cloud
foundry
slash
community.
If
you
wanna
check
into
that
a
little
bit,
eric
might
also
be
able
to
add
some
context
there.
That's
what
I've
got
from
cff
and
pmc
updates
eric.
You
have
some
app
runtime
updates
for
us.
B
Yeah
sure
I've
got
a
few
highlights
to
share
there
so
release
integration
continued
with
regular
releases
of
cf
deployment.
I
know
they
had
a
bump
to
major
version
16
because
of
interface
change
related
to
some
stem
cells,
stem
cell
specification
on
some
of
the
add-ons
and
then
they've
also
released
a
major
version
of
cf
for
gates
again
because
of
a
minor
but
breaking
change
in
the
interface
there
and
qcf
has
also
continued
their
regular
releases.
B
I
know
I
think
their
latest
is
217.12
with
some
bump
dependencies,
and
then
I
know
that
clock
controller,
team
and
services
api
have
continued
marching
through
those
remaining
b3
endpoints.
I
think
they
might
be
done
with
services
at
this
point,
but
I
don't
want
to
make
any
promises
without
confirmation.
So
maybe
we
can
get
a
confirmation
update
from
them
pretty
soon
and
then
irene
is
doing
a
bunch
of
really
interesting
work.
B
They've
released
another
major
version
of
their
component
release
where
they've
moved
entirely
to
config
maps
for
configuration
instead
of
templates.
So
that's
resolving
any
of
the
templating
controversy
that
we've
had
at
that
component
later.
B
They
also
did
a
little
spike
around
taking
the
existing
cf,
build
packs
and
shimming
them
in
as
cloud
native
build
packs,
so
that
kpac
could
run
them
with
fewer
incompatibilities
from
the
potato,
build
packs,
and
then
they
they
also
had
an
inception
on
the
team
and
are
looking
at
taking
some
of
the
initial
work
that
they
did
with
a
crd
representation
of
their
long-running
process
entity
and
using
that
more
consistently
throughout
the
internal
apis.
So
I
think
they're
planning
to
move
forward
with
that
in
the
next
few
weeks.
C
B
I
know
that
there
are
some
discussions
that
have
been
going
on
with
with
vmware
and
sap
and
ibm
about
getting
that
moving
forward
more
definitively.
So
some
of
that
is
still
in
flux,
but
I
think
I
think,
we'll
be
seeing
more
forward
movement
on
that
pretty
soon.
D
Eric
I
have
one
question
with
v3
now
basically
finished:
will
there
be
an
announcement
when
the
v2
clock
start
starting
to
tick.
A
All
right
do
we
have
anyone
from
the
bosch
pmc.
A
Okay,
I
do
know
there
was
another.
There
was
a
bosch
release
just
a
week
ago.
271.5
bosch
is
largely
in
maintenance,
but
you
know
work
is
still
being
done.
So
leave
it
at
that.
Anyone
from
extensions
richard
I
see
there
any
updates
on
stratos.
E
No,
I
was
just
here
to
to
to
witness
basically
stuff.
I.
A
No
problem
all
right:
if
there's
any
other
agenda
items,
people
want
to
chat
now
would
be
the
time
we
do
have
guillain-barre
from
orange
here
to
present
a
proposal
on
cf
marketplace.
So
any
other
anything
else.
Anyone
want
to
talk
about
before
we
jump
into
that.
A
Okay,
guillaume,
take
it
away!
Looking
forward
to
this
presentation,
it
looks
pretty
interesting.
G
F
Should
see
slides,
yes,
so
yeah,
I
was
keen
to
share
some
of
our
work
and
yes
or
feeling
that
marketplace
is
still
a
key
differentiator
for
cloud
foundry
and
share
some
of
our
use
cases
around
this.
F
And
so
I
I
split
this
presentation
into
three
parts:
maybe
just
a
status
about
the
usual
primary
use
case
for
cloud
foundry
marketplace
which
is
consuming
manage
manage
service
from
application
workloads
and
comparing
making
a
status
on
their
strength
and
limitation.
F
One
is
using
cloud
foundry
to
implement
a
way
to
share
marketplace
among
multiple
platforms
that
quite
similar
to
the
asap
perry,
pre-service
manager
or
to
vmware
ism
independent
service
manager
with
a
different
approach
and
using
cloud
foundry
for
doing
that.
Most
of
the
platform
record
system
and
another
use
case,
is
to
use
cloud
foundry
as
a
unified
onboarding
tool.
F
So
this
is
more
of
out
and
screenshots
for
now
and
not
yet
implementation,
but
I'm
looking
forward
to
touring
feedback
and
and
and
whether
that's
a
fits
similar
needs
and
the
conclusion
discussion
I'd
like
to
to
to
have
with
you
guys
is
how
is
the
cloud
foundry
marketplace
going
in
terms
of
new
features?
F
How
can
this
be
fueled
by
serving
enterprise
use
cases
in
this
conclusion?
So,
let's
start
with
the
the
first
part,
and
so
yes
for
the
primary
use
case
of
consuming
manage
service
for
map
workloads.
F
Just
to
remind
us
for
us
of
the
different
use
case
that
different
steps
in
the
user
experience
so
discovering
discovering
offers
service
definition,
the
service
plan
and
their
metadata,
such
as
costs,
support,
rails
and
so
forth,
and
as
well.
Managing
service
plan
visibility
within
orgs
and
then
provision
the
service,
usually
with
params
being
able
to
access
a
web
dashboard
for
those
managed
service
using.
F
F
The
ability
for
the
service
user
to
upgrade
a
plan
to
do
maintenance
upgrade
that
we
call
maintenance
info
and
to
share
service
instance
within
these
platforms
among
multiple
space,
and
then
there
is
bind
to
to
access
this.
The
service
instance
by
the
by
the
application,
usually
credential,
but
there
was
root,
and
this
lock
and
by
then
provision
and
to
me
the
ux,
is
just
it's
just
great.
It's
simple!
F
It's
consistent
across
services
and
that
made
the
great
success
within
your
range
and-
and
I
think,
as
well
for
in
the
community
on
the
limitations,
the
things
that
are
still
waiting
from
my
point
of
view.
The
ability
to
send
annotations
that
are
set
on
crosstalk
entities
service
instance
bind
apps
space
and
orgs
to
service
brokers.
F
So
they
can
make
useful
stuff,
with
those
annotations,
a
first
class
ux
to
consume
service
logs
and
service
metrics,
also
first
class
ux
to
perform
backup
and
restore
being
able
to
do
service
composition
and
that
there
are
some
aged
cages
and
that
that
that
do
need
this
composition,
and
maybe
things
like
being
able
to
stop
and
start
iber
net
and
resume
service,
especially
for
cost
saving,
would
be
quite
important.
But
overall,
the
primary
use
case,
I
think
for
me-
is
still
very
a
very
important
strength
to
the
platform
ecosystem.
F
So
what
are
the
alternatives?
From
my
point
of
view
in
the
open
source
community?
The
first
one
is
a
first
class
to
be
on
a
discussion
resources
with
two
type
of
packaging,
either
operator
styles.
Packaging
such
as
operator
live
my
life
cycle
management
or
lm
or
maybe
unsharp
packaging.
F
Another
one
might
be
tougher
modules,
mileage
service,
most
managed
service
are
include
on
public
cloud
after
for
modules,
and
they
can
be
maybe
wrapped
as
a
customer
resource
and
there's
an
interesting
project
that
that
provides
a
terraform
operator
for
hotel
for
modules
and
another
one
is
consume
service.
Catalog
consumed
osb
marketplace
through
kubernetes
client
kubernetes
service
catalog,
either
through
custom
resource
or
it's
hcli,
service,
catalog,
cli
or
web
ui.
F
F
There
used
to
be
cube,
apps,
but
it
was
removed,
and
so
we
were
missing.
We
were
discussing
with
the
stratos
team
as
well
on
the
opportunity
to
provide
one,
and
so
we
experimented
with
generating
m
chart
facade
automatically
for
each
service
broker
to
consume,
so
I'll
be
sharing
some
of
this
poc
now
so
yeah.
The
goal
is
to
provide
a
web
ux
to
consume
the
usb
api
offerings
in
kubernetes
and
the
use
case.
F
I've
prototyped
so
far
is
discovery,
provisioning
and
I
have
still
some
some
work
done
to
be
done
with
for
binding
and
update
and
so
discovery.
They
would
be
learning
more
about
the
availability
offers
the
metadata
and
the
provisioning
will
will
get
completion
and
for
the
available
service
plan,
completion
and
waveform
for
service
parameters
and
as
well,
a
way
to
access,
dashboard,
urls
and
so
the
way
I've
prototyped.
This
so
far,
is
to
to
generate
the
same
chart
from
the
usb
catalog
dynamically
and
the
generated
n-sharp
provision.
F
F
F
When
we
look
at
a
an
n
chart,
this
readme
is
automatically
generated
with
the
icon
and
the
metadata
from
the
osb
cataloging
points,
including
the
service
plans,
the
providers,
the
support
tutorials
and
the
service
plan
is
interesting
to
get
the
pricing
which
is
available,
the
additional
details,
bullets
and
as
well.
I
mapped
the
maintenance
info
to
the
service
plan
as
well
to
display
and
when
the
user
has
finished
reading
the
readme
it
can
go
through
the
generated
forms
for
selecting
the
values
in
the
m
chart
and
one
is
a
service
plan.
F
The
readme
points
to
full
details
and
if
there
would
be
additional
parameters,
the
json
schema
available
in
the
service
in
the
catalog
endpoint
of
osb
would
be
used
to
generate
forms
as
well
to
prompt
the
user
to
select
to
to
fill
in
the
required
service
instance
parameters
and
eventually
there
would
be
service
binding
in
the
same
screen
to
allow
the
user
to
to
create
one
or
more
bindings.
G
F
Currently,
this
requires
at
least
in
ranger,
and
I
need
to
check
in
the
other
uis.
This
requires
upgrading
gm
chart
because
the
release
note
template
is
rendered
at
the
installation
time
at
that,
an
installation
time.
The
service
instance
resource
is
not
yet
populated,
so
the
lookup
into
kubernetes
inventories
is
still
missing,
and
so,
if
we
look
at
the
resources
that
this
m
chart
has
instantiated,
it's
just
a
service
instance
from
service
catalog
and
eventually
there
would
be
one
or
more
service
binding,
so
yeah.
F
We
we
plan
to
go
deeper
and
to
complete
our
automation
on
this
generation
and,
yes,
we'd,
be
happy
to
get
feedback
and
suggestions
on
that
on
the
repo
or
any
question
afterwards,
and
so
I
tried
so
getting
back
to
comparison
about
alternatives.
F
I
was
listing
different
alternatives
for
to
cloud
foundry
marketplace
for
different
use
cases,
so
I
I
tried
to
think
how
do
they
compare
really
for
different
those
different
use
cases,
and
I
won't
go
into
each
cell
because
it
would
take
too
long
just
to
give
an
idea
of
a
color
map.
F
The
cloud
foundry
marketplace
is
mostly
green
with,
I
think
great
ux.
For
for
all
of
the
supported
use
cases,
there
are
some
stuff
that
are
not
yet
supported.
The
concept
of
I
was
showing
is
using
the
service
catalog
custom
resources
using
the
help
facade
such
as
usb
to
helm,
poc.
F
I
was
showing
with
service
catalog
ui,
and
you
we
see
that
I
think
there's
many
yellow
and
some
red
and
the
other
approaches
I
think,
are
okay
for
some
and
and
are
really
good
for
some
others,
such
as
service
composition,
but
have
also
as
of
also
weaknesses
for
others.
F
So
yeah
my
conclusion,
my
personal
feeling
that
yes
cloud
from
the
marketplace
for
its
primary
use
case
of
consuming
managed
service
from
application
workloads
is
still
unrevolt
and
I'd
like
to
share
with
you
beyond
this
primary
use
case.
I'd
like
to
share
some
other
use
case
that
were
valuable
to
ranch
and
I'd
like
to
know
whether
they're
valuable
to
the
to
the
community
and
to
the
enterprise
market.
So
there's
two
of
them.
F
I
was
mentioning
sharing
marketplace
among
multiple
platforms,
so
sharing
a
marketplace
between
different
cloud,
foundry
installations,
different
kubernetes
clusters,
potentially
provision
on
demand-
and
I
estimate
so
we
we
worked
on
this
osb
cmdb
project,
which
is
similar
in
scope
and
then
the
asap
periple
service
manager
and
and
vmware
ism.
F
We,
you
have
a
full
slide
back,
which
I
think
is
50
slides
over
there,
which
I
was
planning
to
trying
to
share
at
the
last
two
platform
resumes.
But
my
talk
proposal
were
were
not
accepted,
unfortunately,
so
I'll
try
again,
but
basically
the
idea
is
to
use
cloud
foundry
as
a
database
and
a
specialized
database.
The
database
that
manage
service
instance
service,
binding
and
service,
offering
so
I'll
detail
that,
and
the
second
use
case
is
which
is
more
faults.
It's
not
just
a
project.
It's
it's!
F
It's
using
cloud
foundry
as
an
identity
and
access
management
component
of
the
shelf
to
support
a
use
case
that
we
have,
which
is
to
to
provide
a
unified
onboarding
for
teams
that
need
to
access
multiple
cloud
offerings,
and
sometimes
we
have
apps.
That
needs
not
only
cloud
foundry
but
as
well
kubernetes
and
maybe
is
and
some
other
platforms
and
instead
of
having
them,
declare
the
members
and
do
their
onboarding
again
and
again
in
in
web
uis
that
are
costly
to
develop
and
maintain.
We
are
thinking.
F
Maybe
we
can
use
cloud
foundry
for
that
and
basically
to
map
cloud
foundry.
Im
structure
workspace
with
the
members
and
services
to
map
that
into
using
apis
in
into
onboardings
of
kubernetes
is
offers,
and
even
cloud
foundry.
F
So
maybe,
let's
start
with
the
first
one,
the
usb
mdb
sharing
marketplace
among
multiple
platforms.
F
F
We
have
a
marketplace
which
might
be
on-prem
or
maybe
public,
and
as
we
grow
a
number
of
platforms,
we
could
have
multiple
cloud-foundry
platforms
deployments,
maybe
because
of
geography,
maybe
because
of
security
isolation
of
our
legal
restrictions,
as
we
grow
number
of
platform
rate
foundations
and
recumbentless
instances.
Maybe
on
demands
as
well
as
is,
is
tonight
it's
hard
to
keep
a
consistent
consumption
by
all
those
clients
of
the
marketplace.
So
we
need
some
kind
of
international
inventory
and
control
in
between
that.
We
can
act
upon
so
in
in
this
next
slide.
F
Why?
Why
do
we
need
an
inventory
when
we
add
a
new
service
broker?
We
want
it
to
be
consumable
from
all
our
client
platforms,
so
we
need
a
consistent
marketplace
among
those
different
platforms
and
from
a
service
provider
point
of
view
as
well.
We
want
to
have
a
consistent
and
comprehensive
view
of
consumption
by
all
all
the
client
platforms
and,
more
generally,
we
want
some
kind
of
a
command
operability
for
all
clients
and
brokers.
So
we
need
a
common
inventory.
F
With
the
access
be
able
to
access
the
usage
history,
we
need
to
be
able
to
trust
calls
from
client
platforms
from
consumer
users
for
apps
and
to
store
those
metadata
and
avoid
implementing
this
logic
in
each
service
broker
to
be
able
to
log
usb
api
calls
received,
and,
in
addition,
we
need
control.
So
one
is
to
unify
the
usb
calls
received
by
service
brokers
from
different
platforms,
there's
different
profile
into
the
usb
api
and
the
cloudform
repo
file.
Is
this
thing
from
kubernetes
profile?
F
In
particular,
the
kubernetes
has
the
specificity
of
in
its
reconciliation,
loop
implementation
of
applying
an
eventual
consistency,
pattern
which
is
costly
to
implement
in
service
broker.
They
need
to
handle
concurrent
calls
and
determine
which
one
has
made
has
been
remarried.
First,
so
osb
shield
service
broker
authors
from
this
implementation
service
broker
office
only
received
cloud,
foundry
profile
calls
and
the
kubernetes
forms
would
be
automatically
translated
into
croton
recalls.
F
Service
providers
need
to
control
which
client
is
consuming
the
service
plan,
so
the
service
plan,
visibility,
use
case,
set
limit
and
quotas
when
they
run
out
of
capacity
for
examples,
and
this
is
as
well
a
good
a
good
area.
We
think
to
implement
future
features
that
we
are
thinking
of
implemented
and
things
like
doing.
Soft
delete
and
undelete
of
service
instance
in
central
place
and
to
do
service
instance
and
service
binding
sharing,
but
between
distant
client
platforms,
for
example.
F
Share
kafka
cluster
between
the
kubernetes
instance
and
a
cloud
foundry
instance
to
be
able
to
share
messages
and
potentially
doing
recharging
as
well
and
stuff
like
that.
F
F
We
have
service
providers,
so
service
provider
register
into
the
central
component,
their
new
broker.
They
manage
the
service
offering
visibility,
they
manage
their
quotas
and
they
can
as
well
look
up
the
content
of
the
inventory
list.
Service
instance
and
service
binding
that
were
provisioned
access,
the
metadata
to
know
which
platform
it
was
which
client
which
tenant
it
was
within
it,
it
can
be
a
namespace
can
be
a
platfordry
or
arkspace
application
instance.
Id.
F
You
might
want
to
get
analytics
dashboards
about
consumption
of
its
services,
maybe
alerts
if
there
is
a
loop,
an
uncontrolled
loop
that
created
a
thousand
dedicated
cluster
in
his
in
a
faulty
ci
cd
and
maybe
as
well
receive
alerts
when
there
is
something
wrong
happening.
F
So
these
central
components
come
with
its
cli,
it's
web
ui,
it's
analytics
web
ui,
it's
monitoring
analytics,
so
a
whole
lot
of
ecosystem
and
basically
it
receive
usb
calls
from
client
platforms
and
translate
that
into
usb
calls
for
brokers,
and
it
gets
an
inventory
of
everything
that
was
provisioned
and
it
might
use
this
inventory
to
export
it
to
corporate
smdb
or
somewhere
else,
and
when
we
were
thinking
about
implementing
this
because
perry
play
was
not
shooting
on
it,
and
I
have
an
issue,
a
public
issue
that
describes
a
reason
on
the
on
page
of
the
repo
and
the
ism
was
made
a
close
source
when
we
were
considering
how
to
implement.
F
We
are
thinking
well.
Do
we
start
with
just
a
plain
database
for
managing
service
instance
service,
binding
service
offering,
and
we
have
to
re-implement
all
this
tooling
ecosystem
tooling.
Or
can
we
use
a
specialized
database
and
we
are
thinking
what
cloud
foundry
is
a
specialized
database?
It
manages
service
instance,
service,
keys
service
offering
and
service
plan.
F
Can
we
just
use
cloud
foundry
as
a
database,
and
then
we
will
be
able
to
use
its
tooling
cli
its
web
ui,
the
existing
analytics
that
we
have
database
analytics
for
creating
the
cloud
foundry
content,
as
well
as
the
promoters
exporters
and
the
graphene
dashboards
for
monitoring
services?
Can
we
use
all
this
ecosystem
to
access
the
specialized
database
and
maintain
our
inventory
of
services?
F
And
so
basically
that's
that's
what
we
we
do.
The
usb
service
broker
translate
usb
calls
into
cloud
foundry
service
api
calls
from
copy
and
then
cloud
foundry
cloud
controller.
Translate
those
cappy
calls
into
sb
calls
to
service
brokers,
so
there's
basically
a
thin
layer
of
translation
between
usb
calls
and
marketplace
calls.
So
let's
look
at
that
a
bit
more
details
when
a
client
platform
asks
for
the
inventory,
the
catalog
endpoint
the
broker.
F
I
just
look
up
the
marketplace
into
the
inventory
into
the
database
into
cloud
foundry,
which
was
provisioned
previously
using
the
native
platform
resource
cf
create
service
broker.
F
We
choose
to
have
one
org
pair,
osb
client.
So
in
the
example
I
had
two
claim
platforms,
cf1
cf2
and
two
kubernetes
platforms,
so
I
would
have
four
orgs
in
each
org.
We
have
one
space
per
service
definition.
So
in
this
example,
I
have
cassandra
my
sequel
and
this
dummy
space
and
then
for
each
service
instance
that
are
created
by
requested
by
user.
We
have
a
that.
We
call
broker
service
instance.
We
have
a
backing
service
instance
there
and
then
for
each
service
key
one
for
each
service,
binding
requested
by
the
user.
F
We
have
a
service
key,
so
yeah
there's
more
details
in
the
slide
deck
of
the
using
of
the
community
ecosystem,
so
stratos
to
to
look
at
the
content
of
the
inventory
and
the
promoters,
dashboards,
graphene
dashboards,
to
to
look
at
the
rate
of
consumption
of
marketplace
alerts
when
a
service
instance
is
failing,
ccli
and
cli
script
to
to
do
some
analytics
on
the
content
of
the
database.
F
F
So
we
we
are
happy
to
to
move
out
the
repo
if
it
makes
sense
and
if
there
is
interest-
and
so
I'd
like
to
go
to
now
to
go
to
second
part
beyond
the
primary
use
case
of
consuming
managed
service
by
my
application
workloads.
F
So
in
this
other
usage
of
a
secondary
usage
of
cloud
foundry,
we
use
cloud
foundry
as
an
ia,
and
so
this
is
an
overview
of
how
the
user
experience
for
projects
that
need
to
consume
multiple
platforms,
maybe
to
onboard
on
different
cloud,
foundry
platform
deployments
and
maybe
as
well
on
kubernetes,
because
the
application
is
not
eligible
only
on
the
on
cloud
foundry
it
needs
to.
You
need
some
kubernetes
support
as
well.
Maybe
some
is
support
as
well.
F
We
have
some
applications
that
that
need
multiple
platforms
to
fulfill,
that
diverse
needs,
and
they
don't
want
to
to
declare
their
numbers
into
each
platform
and
to
do
their
onboarding
multiple
times.
So
in
this
scenario,
they
would
create
a
project,
declare
their
members
and
then
the
project
members
would
be
automatically
available
when
they
ask
for
platform
movement
of
kubernetes
cloud
foundry
or
is
when
they
create
a
platform
tenant.
F
The
project
members
will
be
automatically
created
and
updated,
and
then,
from
each
platform
they
can
consume
the
marketplace,
as
they
usually
do
still
having
the
centralized
cmdb
keeping
an
inventory.
So
those
platform
can
be
public
or
private.
We
keep
a
central,
unified
inventory
of
our
all
project
subscription
to
platforms
and
as
well,
are
all
project
subscriptions
to
service
provider
being
public
or
private.
F
So
let's
look
at
how
how
this
how
this
would
appear
using
stratos
and
stratos
in
cloud
foundry
community
project,
so
the
project
manager.
F
F
I
can
declare
an
experiment.
If
he's
declaring
an
experiment,
then
he
can
invite
teammates
to
share
his
clocks
of
script
or
it
can
declare
a
real
project,
and
so,
in
this
case
this
is
subject
to
internal
recharging.
You
need
to
provide
a
recharging
code,
then
there's
maybe
unlimited
number
of
of
members,
and
there
is
maybe
the
default
pattern
of
separating
fraud,
non-profit
they
have
mqa
members
and
maybe
even
might
think
of
having
different
leases
duration
leases
for
subscription.
F
Physical
physical
compute
resources
and
this
create
environmental
issues
and
financial
costs,
so
adding
automated
lists
for
the
different
project
would
be
would
be
nice
as
well,
so
the
user
end
up
into
an
org.
So
let's
use
cloud
foundry
thinking.
Can
we
use
cloud
foundry
im
as
a
way
to
declass
this
project,
and
so
the
project
manager
or
the
project
members
adapts
into
a
default
arc
of
a
space
by
its
name?
F
That's
my
internal
name
and
then
it
choose
which
plan
so
in
this
case
declares
an
official
project,
the
name
of
each
project
and
potentially
service
parameters
such
as
recharging
codes
and
other
stuff,
and
so
that's
the
service
instance
in
the
marketplace.
So
that's
the
way
declared
this
project,
so
the
dashboard
brings
the
user
to.
F
To
the
org
created
had
selected
that's
a
real
project,
so
it
has
created.
The
service
instance
has
created
into
a
cloud
foundry
an
org
and
multiple
space,
one
space
for
development
teams,
another
space
for
qa,
and
also
space
for
prod
and
using
cloud
foundry
ui
using
stratos.
He
would
then
add
users
its
project
members
to
the
different
space
and
basically
provision
the
aim.
F
F
And
so
it
might
as
well
choose
in
addition
to
the
official
project,
to
work
on
an
experiment,
so
we
create
a
second
service
instance
for
an
experiment,
he's
working
on
as
a
side
project
and
maybe
another
one
for
self
training.
F
So
basically
the
yeah
using
the
marketplace
provision
provision
into
into
cloud
foundry
the
different
elements
that
are
then
used
to
to
onboard
users.
So
we've
seen
the
first
part
creation
of
project
setup
on
members.
F
Now
we
look
at
how
to
how
to
create
platforms
into
the
different
tenants,
so
the
user
needs
now
to
provision
the
cloud
from
retirement
on
organo
space
or
provision.
A
kubernetes
settlement
might
be
on-prem
or
public
or
closer
or
last
term.
So
how
does
it
look
so,
within
its
organ
space,
any
space
member
can
now
have
access
to
a
list
of
services
service
offerings
which
corresponds
to
the
different
service
cloud
service
offerings.
So
in
this
case
we
have
a
cloud
foundry
organization,
that's
our
internal
name
for
cloud
foundry
instances.
F
F
In
our
case,
we
have
multiple
cloud
foundry
deployments,
one
player
az
for
specific
reasons,
internal
reasons
and
the
user
can
choose
to
onboard
on
1az,
2az
or
multiple
easy
already
available,
and
then
it
choose
for
which
team
this
subscription
is
for
for
the
dev
team,
create
team
or
pro
team,
so
in
which,
in
which
space,
if
he's
a
member
of
all
of
them
and
enter
so
the
service
broker.
That
supports
this
service,
offering
would
provision
into
cloud
foundry
and
the
proper
org
and
and
the
the
user
associated
to
that.
F
So
the
way
we're
thinking
of
implementing
this
is
using
the
cloud
service
broker,
which
was
contributed
by
google
and.
C
F
Vmware
and
the
cloud
from
the
community,
and
which
is
awesome
and
to
use
the
clothing
ray
telephone
provider
as
well.
I
provide
some
details,
so
the
way
it
would
work
is
that
the
service
broker
would
look
up
where
it
was
called
from
and
look
up
the
tip
numbers
into
the
the
calling
space
and
provision
those
users
into
the
created.org
and.
G
F
Of
multiple
subscription
made
so
user,
they
provision
multiple
platform,
resubscription
multiple
kubernetes
subscription
and
they
can
log
into
each
one
into
a
central
space.
F
So
that
was
the
part
about
oddboarding
and
then
once
they're
on
board
from
each
platform,
they
can
consume
the
marketplace
using
their
native
platform,
ui.
So
cloud
foundry
marketplace
for
cluttery
and
service
catalog,
for
example,
into
kubernetes,
still
keeping
a
centralized
inventory
and
control
for
all
activities.
F
So
on
the
left,
we
have
the
opposition
platforms,
so
it
might
be
cloud
foundry,
kubernetes
or
is
so.
Let's
that's
easy
for
for
cloud
foundry.
They
consume
the
marketplace
from
from
stratos
or
from
the
ui
and
then
cloud
foundry
has
a
nosb
client
which
goes
to
the
ocmdb
and
then
from
this
they
can
consume
service
broker
for
data
services.
F
So
we
see
similar
ways
for
kubernetes.
I
was
mentioning
with.
G
F
F
We
are
also
considering
generating
automatically
in
the
same
way
that
osb2m
vr
tiles
and
maybe
always
open
stack
horizon
ui
for
having
native
marketplace
consumption
into
into
v-ray
and
to
into
openstack
for
the
first
on-boarding
case
that
was
showing
creating
level
project
and
then
onboarding
platforms.
F
The
user
would
end
up
into
stratos
discover
service
offering
and
this
service
offering
would
be
supported
by
the
different
on
on
service
on
onboard
service.
Onboarding
service
broker,
sorry,
and
so
we
are
considering
using
the
cloud
service
broker
with
the
telephone
service
provider
and
the
cloud
service
broker
supports
this
concept
of
broker
pack,
that's
the
name
for
calling
terraform
modules,
and
so
basically,
there
was
one
module
that
would
instantiate
an
org
and
a
dev
qa
and
plot
space.
F
In
the
example
for
the
service
plan
project
that
was,
I
was
showing
and
and
this
terraform
module
should
also
use
data
sources.
Data
source
is
a
way
to
look
up
an
existing
object
into
a
real
world
and
the
data
source
would
just
look
up.
F
The
the
context
of
the
osb
calls
the
context
being
which
space
apps
and
org
the
service
instance
was
created.
G
F
F
I'd
like
to
have
with
you
guys
and
to
my
my
feeling
is
that
platform
remarketplace
and
the
usb
ecosystem
provides
strong
value
to
range
and
and
I'd
like
to
mention
that
it's
also
the
case
to
the
enterprise
market
and
there's
still
quite
a
bit
of
work
on
those
components
to
to
support
advanced
use
case
and
to
source
supports
enterprise
use
case.
So
I
was
mentioning,
for
example,
the
sending
annotations
to
service
brokers.
F
There
is
maintenance
into
stratos
for
enhanced
experience
of
the
platform
remarketplace
as
well
as
v3.
The
cf
java
client
is
great.
Is
that
a
great
asset
that
allows
to
develop
service
broker
that
that
talks
to
plot
hundreds,
such
as
for
usb
same
db?
Vitreous
report
is
being
contributed
as
well,
and
thanks
again
for
from
vmware
and
sap
for
those
great
contributions
cloud
service
broker
is
is
awesome.
There
are
many
as
well
opportunities
to
serve
this
enterprise
use
case
with
additional
use.
F
Cases
such
as
support
for
dashboards,
currently
cloud
service
broker
doesn't
support
dashboards,
and
I,
I
think
I'll
share.
F
Another
presentation
detailing
detailing
one
use
case
I
was
missing,
is
once
a
user
created
a
data
subscription,
say
a
gcp,
a
cloud
service,
a
cloud
sql
service
instance
and
that
he
needs
to
access
gcp
console
to
access
log,
metrics
alerts
or
backups
currently
from
cloud
foundry
identity.
It
has.
G
F
G
F
F
Some
of
that
yeah,
this
ecosystem
is
strong
as
well
because
of
the
kubernetes
client
service
catalog.
Maybe.
F
So
I'm
I'd
like
to
get
feedback
on
whether
this
usb20
is
interested
interesting
to
the
community
and
as
well.
It's
strong
because
of
the
service
brokers
that
that
provides
a
way
to
provision
kubernetes
resources.
F
So
there's
a
sap
service
service
fabric
broker,
which
I
do
need
to
test,
there's.
G
F
Cloud
service
broker
that
looks
very
promising
as
well
using
telephone
providers
kubernetes
alpha
enables
custom
resources
to
be
provisioned
still
alpha,
as
its
name
indicates.
F
So
we
do
need
to
test
that
and
there
is
the
kubernetes
provider,
and
there
was
a
great
example
contributed
by
the
cloud.gov
team,
gsa
eks
brokeback
example,
which
is
very
promising,
so
the
end
provider
as
well-
and
there
was
an
issue
provided
by
gsa
team
as
well
on
that
there's
many
brokers
that
look
supremizing
as
well,
maybe
others,
and
so
yes,
the
discussion
and
question
I
was
keen
to
to
receive
was
to
those
use
cases
that
I
was
sharing,
that
they
would
also
apply
to
us
to
enterprise
customers.
F
Yeah
advanced
use
cases,
and
would
that
be
a
way
as
well
to
help
maintaining
this
ecosystem
that,
I
think,
is
still
differentiating
from
the
ecosystem,
but
gradually
is
getting
competition
and
so
is
is
probably
threatened
by
by
investment
made
into
into
the
kubernetes
ecosystem.
So,
yes,
that's
the
discussion.
Maybe
we
can
move
to
more
discussion.
A
A
Well
got
a
fairly
small
audience
here
today,
but
I
think
the
very
least
you
should
post
this
in
cf
dev,
trying
to
reach
out
to
my
community
and
and
see
what
kind
of
response
that
gets
and
yeah.
We
can
talk
how
to
move
forward
with
this
but
yeah.
It
looks
really
interesting.
I'm
also
very
excited
about
the
cloud
service
broker
project.
So
yeah
is,
I
guess,
yeah
cf
dev
would
probably
be
the
best
place
to
move
forward
with
this
slide
deck.
A
If
you
can
post
that
in
the
cab
channel
and
or
cf
dev,
I
think
that
would
be
useful,
but
yeah
thanks.
So
much.
F
Yes,
norm
say
that
the
ec
enterprise
usage
same
from
wing
and
yes
I'll,
follow
up
as
well
with
with
troy
on
the
pmc
extension
for
incubation.
If
there
is
interest
in
incubating
your
project
and
yes,
if
there
is
I'm
happy
to
to
get
offline
feedback,
maybe
from
service
teams
cloud
service
teams
as
well.
G
Yeah,
I
just
wanted
to
add
chris
that
I
was
very
interested
in
the
cloud
service
broker
aspect
as
well
and
see
if
there's
any
overlap
there
or,
if
there's
some
lessons
that
we
can
pull
out
of
this,
and
thank
you
for
the
presentation,
yeah
I'll
I'll
reach
out
to
you,
probably
outside
of
the
call
and
get
some
more
info
on
those
aspects
that
I
I
just
mentioned.
A
All
right:
well,
I
think
that
about
concludes
it
we'll
give
everyone
a
few
minutes
back,
but
thanks
everyone
for
joining
and
yeah
guillaume,
thanks
again
and
we'll
all
talk
next
month.