►
Description
Agenda available here: https://docs.google.com/document/d/1SCOlAquyUmNM-AQnekCOXiwhLs6gveTxAcduvDcW_xI/edit?usp=sharing
A
Hi,
everyone
welcome
to
the
cloud
foundry
community
advisory
board
meeting
for
october
28
2020.
we're
going
to
skip
over
the
cff,
highlights
and
updates
the
cloud.
Foundry
foundation
staff
worked
really
hard
to
give
us
a
great
summit
virtual
summit
last
week.
I'd
like
to
thank
them
all
for
for
that
work.
Some
great
presentations
they're.
A
A
I
thought
I
saw
eric
mom
on
our
attendee
list.
Eric
are
you
here,
and
would
you
like
to
go
over
the
pmc
updates
for
runtime.
B
Yeah,
I'm
happy
to
got
a
got
a
nice
set
of
updates
here,
so
the
cli
team
has
released
what
I
think
is
intended
to
be
the
final
version
of
the
v6
cli
with
a
couple
of
fixes
in
it.
So
that's
6.53
and
so
they'll
be
focusing
on
the
v7
versions.
After
that
and
then
of
course,
we've
got
release
integration
with
the
release
of
cf
gates
version
1.0
around
summit.
So
congratulations
to
all
of
them
on
that
milestone
and
then
not
to
be
outdone.
B
Qcf
has
some
other
significant
releases
as
minor
updates
in
v2.
I
know
they've
gotten
some
really
nice
scaling
configuration
in
to
2.5
and
some
more
configurability
around
the
irony
service
account
in
two
secs
for
some
of
the
component
teams.
Cappy's
finished
up
their
work
to
move
the
application
bit
packages
into
the
oci
image
registry
when
they're
using
kpac
and
cf
for
gates-
and
I
know
they've-
been
working
on
some
scaling
improvements,
especially
with
some
of
the
feedback
that
we've
gotten
from
the
sap
folks
that
have
been
testing
cf
for
gates
at
scale.
B
Irini
they've
just
started
some
work
around
spiking
on
sidecar
support,
but
I
think
they're
also
thinking
a
little
bit
more
about
various
other
extensibility
mechanisms
and
how
to
support
those
networking
has
been
doing
some
work
to
add
optional
support
for
other
ingress
controllers
and
mechanisms
on
kubernetes,
starting
with
contour.
B
So
we're
thinking
that
this
is
looking
like
a
potentially
lighter
weight
option
for
ingress
than
istio,
which
is
very
feature-rich,
but
also
heavy
weight,
sometimes
a
little
difficult
to
wrangle.
So
this
is
probably
going
to
be
more
attractive
for
smaller
or
test
installations
of
cf.
For
case,
then
they're
also
starting
to
look
into
some
of
the
app
to
app
features
and
how
they
can
start
supporting
those
on
top
of
kubernetes
and
then
login
metrics
they've
continued
to
work
on
log,
cache,
performance
and
scalability.
B
I
think
they've
delivered
some
improvements
there,
but
they're
still
hard
at
work
there
and
jesse
weaver.
The
project
lead
for
that
team
has
been
putting
together
a
github
project
on
the
staff
gates,
logging
repo
to
reflect
some
of
the
roadmap.
So
thanks
those.
A
Great
now
I
did
want
to
ask
about
the
cli
v6
release.
I
know
a
lot
of
us
still
rely
on
the
v2
api
for,
for
many
reasons,
do
they
still
intend
on
releasing
patches
for
that
for
the
foreseeable
future.
B
I
think
so
yeah
I
I
assume
that
if
there's
any
security
issues
or
significant
bugs
to
correct,
then
they
would
have
patch
releases
for
that.
A
B
A
You
have
we
got
anyone
from
the
bosch
team
who
can
give
us
a
quick
update.
A
A
The
big
news
there
amongst
a
lot
of
other
improvements
and
extension
point
refactoring
was
the
donation
or
the
contribution
of
the
kubernetes
and
helm
support
stratos
in
susa
stratos
to
the
upstream
stratos
for
two.
So
now
everyone
get
has
access
to
all
of
the
kubernetes
features
and
functionality,
as
well
as
a
nice
web,
helm
client
for
installing
things
to
those
kubernetes
kubernetes
endpoints
that
you
connect
to
just
like
the
cfn
points
you
can
specify
multiple
ones
and
and
deploy
things
to
to
more
than
one
kubernetes
endpoint.
A
I
don't
want
to
attempt
to
summarize
the
the
multi-apps
update.
I
got
suffice
to
say
that
there
were
a
couple
of
new
features.
A
Multiple,
you
can
specify
multiple
instances
of
one
on
the
same
mta
in
a
single
cloud,
foundry
space
you
can
use
the
service
broker,
parameter
to
specify
which
broker
to
use
for
service
creation.
That's
for
if
you've
got
duplicate
service
offerings
from
two
different
brokers
and
then
there's
a
whole
bunch
of
changes
and
bug
fixes
that
are
in
the
notes
from
the
last
cf
extensions
meeting,
which
of
course
are
are
on
the
in
the
github
repository
with
all
the
other
things.
A
I
will
amend
these
notes
if
I
get
any
late
additions
to
the
the
extensions
pmc
updates
from
all
around,
but
the
real
reason
we're
we're
meeting
today,
because
we're
of
course,
one
week
out
from
our
normal
cadence,
we
wanted
to
to
give
room
to
the
diversity
conversation
last
week
at
cf
summit.
We
didn't
want
to
collide
with
that,
but
we
rescheduled,
because
there
were
a
couple
of
community
projects
that
people
wanted
to
present,
one
of
which
I
don't
think
was
presented
at
summit
and
the
other
one
that
we
wanted
a
do-over.
C
A
C
Sure
I'll
take
it
away,
I
don't
have
slides.
I
really
have
barely
have
a
demo.
My
demo
infrastructure
blew
up
this
morning.
I
think
I've
managed
to
to
bring
it
back
from
the
brink,
because
it
is
a
data
protection
solution
which
is
what
it
specializes
in
so
cf
protect
is
a
merge
integration
of
shield
cloud.io,
which
is
a
sas
backup
solution
based
on
shield
open
source
that
stark
and
wayne
launched
back
in.
I
think
august,
which
gives
you
on-demand
access
to
a
shield
cloud
outside
of
your
infrastructure
for
a
monthly
fee.
C
The
the
goal
in
putting
this
project
together
was
to
make
data
protection
inside
of
cloud
foundry
as
simple
as
application
deployment.
Inside
of
cloud
foundry,
the
big
strength
we
see
with
cloud
foundry
is
cf
push,
not
a
big
surprise
to
anybody
on
this
call,
but
cf
push
being
able
to
go
from
code
to
running
app
is
still
pretty
groundbreaking,
even
in
this
day
and
age
of
kubernetes
and
and
containerization
everywhere,
we'd
like
to
do
the
same
for
data
protection,
so
I'm
going
to
go
ahead
and
share
my
screen.
C
Yes,
indeed
excellent,
so
I'm
going
to
start
with
I'm
using
pws.
We
actually
started
this
project
because
we
got
the
eol
announcements
for
pws
and
we
needed
to
get
our
stuff
off
of
pws
and
the
easiest
way
we
could
do.
That
was
shield,
so
I
have
in
in
our
pws
instance.
I
have
this
app
that
I
wrote
called
to
do
it's
just
a
small
to
do
application.
C
C
It's
very
very
minimalistic
and
what
we're
gonna
do
is
we're
gonna
run
a
protection
against
it,
so
I
also
have,
as
you
can
see,
look
at
the
services.
I
went
ahead
and
I
got
a
clear
db:
the
free
plan
out
of
pws.
This
is
just
a
mysql,
app
direct
instance.
It's
just
out
in
the
cloud
somewhere
and
if
I
look
at
that
service
I
could
go
out
and
I
can
use
their
dashboard.
I
can
run
their
backups.
C
I
can
also
look
at
the
bindings
for
the
app,
so
I
can
look
at
the
environment
variables
for
my
to-do
app
which
that
service
is
bound
to,
and
I
could
find
my
credentials
right.
My
host
name,
the
database
name,
what
user
to
log
in
I
could
set
all
this
stuff
up
myself
in
shield
or
in
any
other
backup
solution,
but
I'm
lazy,
and
what
I
want
to
do
is
I
want
to
have
cf
do
that
for
me,
real,
quick,
I'm
going
to
flip
over
to
this
tmux
pane,
which
shows
me
my
shield
instance.
C
So
we're
going
to
start
with
cf
protect
inside
my
environment.
I
have
a
couple
of
environment
variables
that
tell
the
cfcli
plug
or
the
cf
protect
plug-in
for
the
cfcli,
how
to
talk
to
shield
and
how
I
want
my
backup
for
this
app
to
be
done
as
far
as
what
agent,
but,
for
the
most
part
we
run
cf
protect.
We
remember
to
put
the
app
name
in,
and
the
plugin
looks
through
and
says:
hey
I
found
this
to
do.
C
Db
looks
like
it
might
be
my
sequel
and
it
pulls
that
off
by
looking
at
a
combination
of
tags
and
the
instance
binding
credentials.
So
my
sql
in
pws
presents
a
specific
way,
a
separate
set
of
tags.
It's
actually
tagged
mysql
and
a
specific
set
of
credentials
that
we
can
pull
from.
So
here
it's
pulling
the
cleardb.com
hostname,
the
port,
the
database
and
the
username
and
password.
C
It
then
goes
out
to
shield
and
it
creates
a
new
target
named
after
org
space,
app
and
service
name,
and
then
it
creates
a
daily
job
that
is
paused
and
if
I
flip
back
over
to
the
browser,
you'll
see
that
this
is
now
extent
make
zoom
a
little
smaller
there
we
go.
We
have
this
thing
it
came
in.
C
It
is
paused,
so
one
of
the
things
we
did
in
the
design
for
cf
protect
is
we
we
create
a
daily
job,
but
we
don't
turn
it
on
so
that
you
have
a
final
review
step
to
go
hey
this
is
you
know
I
want
this
to
run
daily
at
4am,
or
I
don't,
and
I
want
to
go
in
and
change
it.
C
But
all
my
stuff
is
here
right
all
of
the
things
that
I
need
to
figure
out.
I
picked
a
specific
agent
for
this,
so
we're
going
to
go
ahead
and
we're
going
to
manually
run
this
this
backup,
because
I
don't
want
to
unpause
it
and
we
can
look
through
the
tag,
the
task
log
it
it
connected
out
to
there.
It
complained
about
password
on
command
line,
but
it
did
go
off
and
build
an
archive,
and
that
archive
shows
up
right
here
with
this.
C
C
C
So
if
I
do
a
show
databases,
that's
just
the
stock
standard,
mariadb
loadout,
my
sequel,
the
performance
scheme
and
the
info
schema.
The
configuration
for
this
is
already
in
shield.
So
if
I
go
look
here
at
targets,
I
have
a
target
called
migration
target.
That
is
that
and
it's
running
under
a
different
agent,
so
we're
gonna
grab
that
archive
and
we're
gonna
say,
restore
the
archive
and
I'm
gonna
cheat
and
I'm
gonna
steal.
The
uuid
from
here
we're
gonna,
say:
restore
that
archive
to
the
migration
target.
C
C
Database,
which
is
the
database
that
clear
db
set
up
for
us,
so
I
can
pull
this.
I
can
run
reports
against
this.
I
can
actually
move
the
to
do
app
off
of
pws
either
onto
another
cloud
foundry
by
restoring
to
a
similarly
cf-protected
database
instance
in
a
different
cloud
foundry
I
can
move
it
to
kubernetes.
I
can
run
it
in
docker
compose.
I
can
do
a
whole
bunch
with
it
and-
and
I
think
we
think
that
that
was
a
fairly
painless
setup
for
doing
data
protection.
C
And
that's
that's
kind
of
it.
It's
data
protection
is
really
hard
to
demo
because
if
it
works,
nothing,
it
works.
Catastrophic
happens
right.
D
Case
in
point,
I
just
want
to
re-emphasize
or
emphasize
that
when
we've
gotten
noticed
that
pws
was
end
of
lifeing,
we
we
stood
up
our
own
cloud,
foundry
elsewhere
and
we
literally
used
cf
protect
to
migrate
the
data
services.
That's
right,
jane's
right.
C
Yeah
yeah,
we
used
this
to
move
our
own
stuff
because
we
needed
to
right
now
the
state
of
the
project
it
is
on
github
github.com.
Let
me
actually
go
share
screen
and
we'll
go
look
at
github.
It
is
out
here
in
the
shield
cloud
org
cdf
protect.
You
can
download
this.
The
readme
has
all
the
instructions
for
installing
this.
We
actually
are
running
our
own
cli
plug-in
repository
this
one
that
you
can
use
to
install
the
official
way.
C
If
you
don't
want
to
build
software
on
your
own
machine,
it's
a
fairly
rudimentary
version
of
the
plugin
right
now
we're
only
supporting
my
sql
and
postgres,
and
they
only
work
if
these
are
the
credentials
and
the
tags
we're
planning,
because
this
is
heuristic
right.
It's
looking
at
the
the
vcap
services
buying
the
environment,
variable
and
kind
of
picking
and
choosing
what
makes
sense
to
back
up
and
how
to
back
it
up.
C
So,
if
there's
anyone
out
there
that
wants
to
give
this
a
go,
I
strongly
recommend
trying
it
in
a
non-production
environment.
First,
as
I
do
with
all
data
protection
software,
and
if
there
are
services
that
you
have
access
to
that,
you
would
you
know,
we
know
how
to
back
them
up,
but
the
cf
protect
plug-in.
Doesn't
please
feel
free
to
open
a
github
issue
with
details
about
the
service.
The
cf
protect
plug-in
only
gets
better,
as
we
add
support
for
additional
catalog
services.
A
I
have
a
question:
is
there
a
server-side
component
to
shield
that
runs
in
cloud
foundry.
C
No,
not
for
the
cf
protect
stuff.
The
shield
cloud
is
is
shieldcloud.io
is
a
sas
offering
okay,
so
you
would
you
sign
up
for
that
and
then
it
uses
that
as
the
endpoints
I.
A
Was
just
wondering
about
cloud
foundries
that
that
don't
have
public
endpoints
for
the
databases
that
you're
you're?
Yes,.
C
So
that's
where
the
agent
comes
in,
I
I
kind
of
cheated.
There
is
another
component
that
you
can
run
inside
your
infrastructure
that
provides
your
agent
implementation
because,
what's
out
on
the
public
cloud
is
really
the
orchestrator
and
anywhere
you
can
run
a
docker
container.
You
can,
or
or
you
can
run
just
a
straight
like
we
support
ubuntu,
vms
and
centos,
vms
and
just
straight
up
installation.
C
I
don't
believe
we
have
a
bosch
release,
but
we
do
have
the
docker
image
that
you
can
run
on
top
of
containers,
bosch
release
and
that
spins
up
a
little
piece
of
software.
That
actually
ssh
is
outward
from
your
infrastructure,
so
it
transits
all
the
nat
gateways
and
firewalls
and
everything
else
and
goes
out
to
a
public
endpoint
and
then
from
there
carries
out
the
command
says.
Oh
it's
time
to
do
this
backup,
okay,
great
and
then
works
that
out
to
the
public
storage
and
all
that
stuff
could.
A
You
run
that
container
with
docker,
diego
or
diego
docker
feature.
C
You
probably
could
I
wouldn't
have
a
port.
I
don't
know
if
diego
would
be
problematic
from
this
isn't
really
alive,
because
it
doesn't
listen
for
any
traffic.
It's
an
outbound
only
thing,
but
it
is
a
native
oci
image,
that's
available
on
docker
hub
and
there's
a
couple
different
flavors,
depending
on
what
you
want
to
back
up
with
it.
C
So
there's
straight
shield
cloud:
we've
actually
got
a
blog
post
that
I'll
be
pushing
out
today
to
the
showcase.io
website
that
details,
how
the
docker
image
is
put
together
and
how
one
might
use,
extend
and
otherwise
integrate
it
into
your
environment.
So.
A
Cool
this
is
really
really
awesome.
Any
more
some
questions.
I
think
this
is
long
long
needed.
I,
there
are
probably
things
in
bosch,
backup
and
restore
that
that
do
similar
stuff,
but
this
is
a
great
way
for
end
users
to
be
able
to
get
at
their
data
in
an
organized
way.
Any
plans
for
expanding
the
number
of
services
that
you
support.
C
Oh
100,
we
just
have
to
work
our
way
through
the
catalog,
as
wayne
pointed
out.
This
came
about
because
we
needed
to
do
some
data
migration
yeah
and
we
already
had
all
the
data
pipes
bits
in
shields.
So
we
said
hey
this.
This
could
work.
We
have
all
the
parts
now,
let's
do
this,
but
yeah
as
we
as
we
work
through
our
own
services,
layer
and
and
other
clouds
as
they
go
eol
I
think
swisscom
was
one
we
were
trying
to
get
into
to
see
if
we
could
play
around
there.
C
Definitely
it's
mostly
about
understanding
the
tags
and
how
to
wire
the
because
the
app
already
has
access
to
all
of
its
data
right.
Otherwise
the
service
doesn't
work.
So
all
we
need
to
do
is
get
shield
the
access
in
the
shield
format
for
the
given
plugin
and
it
should
be
able
to
do
what
it
needs
to
do
so.
C
In
the
cf
protect
one,
yes
and,
as
we
add
more
we'll
add
more
there,
it
is
important
to
point
out
that
shield
project
and
github
and
shield
project
dot
io
is
the
open
core
base
and
shield
cloud
and
github
and
shield
cloud
that
I
o
is
the
sas
offering
right
right
now.
Cf
protect
only
works
with
shield
cloud.
That's
only
because
that's
what
we
have,
what
we
had
access
to
and
and
a
desire
to
move
quickly,
there's
no
reason
it
can't
do
other
shields.
C
A
So
I
see
you've
got
file
system,
postgres,
mysql
and
then
some
things
on
the
presumably
on
the
roadmap.
Extra
backup,
cassandra
console
xcd.
C
C
C
A
Awesome
that
is
so
cool
all
right.
That's
gonna
be
a
hard
act
to
follow,
but
we
do
have
thanks
so
much
james
for
presenting
again
any
time,
always
good
content.
If
I
can
just
take
a
look
at
my
agenda
today,
any
questions
I
asked
a
few
questions.
There
could
be
more
out
there.
D
James
is
this:
calls
can
be
used
to
migrate
from
from
cf
whatever
to
cf
to
kate's
and
cf
the
pcf
and
stuff
like
that
as
well.
C
Yep,
in
fact,
I've
got
a
whole
series
of
content
talking
about
using
cf,
protect
in
vm
based
cloud
foundry
to
migrate
to
apps
migrate.
The
data
for
the
apps
that
you've
already
see
I've
pushed
over
to
your
new
shiny
cf
for
case
we're
trying
to
make
the
data
portable
in
the
same
way
that,
right
now,
the
application
code
is
portable,
not
to
necessarily
make
the
state
state
list.
But
as
close
as
we
can
get.
E
Yeah
and
I'll
just
a
half
a
question,
but
it
should
be
possible
for
people
who
have
data
right
now
to
point
shield
and
shield
cloud
at
their
data
with
the
shield,
ui
and
api
yeah.
Just
to
you
know,
do
a
practice
run
there,
while
the
the
plug-in
the
cf
protect
plug-in,
is
being
updated
to
understand
the
v-cap
stuff.
E
A
I
see
that
I've
been
corrected
on
the
chat.
It's
shield
cloud
that
I
o
not
shield
project
that
I
o
is
that
which
url
would
you
prefer
people.
A
Okay,
got
it
got
it.
Okay.
Is
there
any
other
questions,
if
not
kisavan
and
pavan
from
sap?
If
they
are
here.
A
Sap
rocks
they
certainly
do
but
they're
not
here,
maybe
there's
a
little
problem
with
communication.
They
were
going
to
present
represent
the
our
scalability
journey
with
cf
for
k-8s,
but
there.
E
A
Time
jump
they
might
not
be,
they
might
not
be
there.
Let
me
just
ping
them
on
the
on
the
slack.
In
the
meantime,
chris
can
provide
any
updates
he
wants
from
from
the
cf
foundation,
because
I
see
that
he's
joined.
A
Chris
clark,
would
you
like
to
present.
D
Anything
I'll,
say
hi,
thank
you
all
for
for
the
great
work
that
all
led
to
summit
last
week.
That
was,
I
think,
as
fun
as
a
virtual
event
in
my
garage
at
three
in
the
morning
can
be-
and
you
know
some
issues
with
the
with
the
with
the
site,
but
you
know
we're
soliciting
feedback
currently
from
all
that,
and
you
know
certainly
hope
that
we
can
all
get
together
in
person
sometime
next
year,
but
yeah.
D
That's
that
I
don't
actually
have
any
big
updates,
I'm
actually
up
in
gernville
right
now,
at
the
end
of
a
I
was
off
yesterday
I'll
be
working
today
when
I
get
back
home
but
yeah,
that's
all
james.
It's
a
great
great
demo.
We're
excited
to
see
the
shield
project
come
to
fruition.
A
A
Thanks
we,
we
did
thank
the
the
cloud
foundry
foundation
staff.
It
was
a
very
very
early
morning
for
people.
I
will
confess
that
I
did
come
late
later
in
the
in
the
morning
and
and
then
catch
up
with
the
recordings
yeah
the
recordings.
I
did
mention
that
the
recordings
now
look
like
they're
on
on
youtube,
so
we
can
should
we
share
them
out
with
our
twitter
followers
yeah.
D
I
think
that's
good
to
go.
I
think
they,
I
think
they're.
I
think
it
went
live
on
youtube
sometime
late
yesterday,
so
so
they're
all
on
there
and
you
can
catch
up
with
all
of
the
breakout
sessions
and
whatnot
that
you
missed
during
summit
and
yeah
tweet
away
and
yeah.
A
Kevin
joined
any
updates
from
bosch.
A
E
I'm
just
looking
if
anyone
from
the
seville
office
there,
actually
the
bosch
team
is
moved
mostly
to
the
seville
office.
The
only
update
I
can
give
I
was
hoping
diego
was
on,
but
you
know
I
think,
you've
been
hearing
some
of
the
rumblings.
We
do
have
some
of
the
bionic
stem
cell
stuff
progressing.
E
A
That's
fine,
yep,
bionic
stem
cell
stuff
is
progressing.
I
added
some
notes
is
anyone
from
sap
here
yet
to.
A
Demo,
their
stuff-
I
will
point
out
that
now,
especially
since
we
just
made
note
that
the
cf
summit
talks
are
available
online.
The
scalability
journey
with
cf
for
k-8s
is
actually
one
of
the
the
talks
from
from
the
summit
so
I'll.
If
I
can
find
the,
if
someone
can
find
the
youtube
link,
you
could
we
could
just
refer
people
to
it.
D
I
do
have
a
quick
question,
I'm
wondering
if
anybody
has
looked
into
what
scaling
for
cube.
Cf
looks
like
as
far
as
can
I
run
50
000
application
instances
on
this
thing.
If
there's
anything
out
there
about
that.
A
We
owe
that
to
the
community,
and
I
know
that
ibm
has
been
doing
a
lot
of
scalability
testing.
Of
course,
qcf
is
a
big
part
of
their
plans,
sure
and
they
are
running
it
at
larger.
They
they
have
instances
that
are
running
at
larger
scale.
I
have
somebody
on
the
cube,
cf
team
with
a
with
an
item
for
that,
but
I
don't
know
the
status.
So
why
don't?
I
take
an
note
to
provide
an
update
for
that
in
the
next,
probably
in
the
next
app
runtime
emc
meeting
great.
D
That
might
be
a
little
more
time,
sensitive
question
for
for
research
purposes,
for
clients.
Based
on
that
question,
can
we
like
maybe
contact
you
out
of
band
for
that
or
sure
you.
A
Know
what
I
would
say,
cube
cf
cube
cf
dev
on
on
the
cloud
foundry
slack,
because
that
work
might
be
done
and
I'm
just
not
aware
of
the
numbers.
D
A
All
right
there
aren't
any
there
aren't
any
that
you're
aware
of
okay.
Well,
let's
wrap
it
up.
Then
we
blazed
through
it
pretty
quickly
this
week.
I
guess
the
next
one
chris
is
going
to
be
in
three
weeks,
so
we're
back
to
our
cadence
yeah.
That
sounds
right
and
please
feel
free
to
ping
me
on
slack
or
in
the
cab
channel,
we'll
just
discuss
it.
We'll
try
and
get
a
couple
more
presentations
for
for
for
next
time.
A
Maybe
I'll
have
those
numbers
for
for
qcf
and
maybe
show
off
some
of
the
things
in
the
in
the
latest.
Qcf
get
a
demo
there
and,
of
course,
stratos
is
always
fun
to
take
a
look
at,
but
yeah
until
that
time,.