►
From YouTube: Cloud Foundry CAB Call recording June 2022
Description
The agenda for the call is as follows:
‣ Korifi updates
‣ Korifi Architecture discussion
‣ TOC Election results
‣ Cloud Foundry Event
A
We
have
a
bunch
of
very
interesting
updates
from
the
community
so
for
those
who've
been
following
closely.
It's
been
a
year
since
we
elected
the
inaugural
technical
oversight
committee
and
we've
had
the
opportunity
to
conduct
the
first
election
to
fill
in
seats
that
we
are
now
starting
to
rotate
and.
A
A
We
don't
have
a
whole
bunch
of
updates
yet,
but
just
want
to
remind
folks
about
dates
and
if
you're
wondering
about
logistics
and
things
like
that,
we
can
chat
about
that
and
then
finally,
I
guess
the
meat
of
the
call
is
meant
to
focus
on
some
kurifi
architecture,
documents
that
were
prepared
by
the
community
in
the
past
week
or
so
and
generally
have
a
discussion
about
the
karifi
project
as
it
stands.
A
B
Yeah
I'll
be
posting
to
cf
dev
shortly,
and
I've
got
a
pull
request
and
update
the
results
to
the
community
repo,
but
the
results
for
the
election
we're
going
to
see
reuben
koster
from
vmware
bajan
belly
from
sap
and
andy
hunt
from
cloud
dot,
gov
dot,
uk
and
so
we're
very
excited
to
have
them
join
and
that
keeps
about
the
same
balance
as
the
as
the
exact
same
balance
as
we
had
previously.
B
Eric
malm
from
vmware
and
stephen
marker
from
sap
will
continue
on
for
another
year,
they're
still
on
the
on
board,
and
their
seats
will
be
up
for
election
one
year
from
now.
So
yeah
I'll
share
that
more
widely
with
the
community
later
today,
but
just
figured
we'd
give
you
an
update
on
that
now
so
congrats
to
everyone
who's
now
on
the
tlc.
A
Yeah
we'll
ratify
all
of
this
during
the
next
eoc
call.
I
think
we
are
also
due
to
elect
a
new
toc
chair,
which
I
guess
will
also
happen
in
the
coming
weeks
and
yeah
once
that's
set.
I
think
the
new
toc
will
begin
their
one
year
term
well
before
july.
I
guess,
but
yeah
remains
to
be
seen.
A
Okay,
next
sort
of
quick
update,
a
reminder
about
the
event
that
we're
scheduling
so
just
some
points
to
recall.
First
of
all,
it's
going
to
be
a
one
day,
single
track
event
co-located
with
kubecon
north
america,
so
that's
either
the
24th
or
the
25th
of
october
2022..
A
So
we
have
either
the
monday
or
the
tuesday
we
don't
know
which
yet,
but
I
guess
we
will
soon,
and
so
it's
going
to
be
a
one-day
event
like
I
mentioned.
A
Obviously
that
also
means
that
we
have
to
cram
in
like
a
lot
of
different
tracks
and
different
segments
to
the
event
which
we're
all
in
sort
of
the
planning
phase
right
now.
So
as
soon
as
we
know
the
date
and
as
soon
as
we
nail
down
some
of
some
other
logistics
about
the
event
itself,
we'll
be
sending
out
notifications
about
cfps
about
getting
some
papers
and
submissions
in
from
the
community.
A
A
A
So
the
venue
is
at
huntington
place,
which
is
in
downtown
detroit,
there's
a
bunch
of
different
hotels
here,
a
lot
of
which
are
actually
booked
fully.
So,
if
you're
looking
for
places,
I
would
encourage
you
to
do
it
sooner
rather
than
later.
A
A
We
are
not
sure
yet,
whether
it's
going
to
be
on
the
monday
or
the
tuesday
preceding
kubecon.
So-
and
you
know
we're
speaking
to
the
lf
events
team
and
the
cncf
events
team-
to
have
all
of
this
information-
and
you
know
it's
it'll-
hopefully
be
here
soon,
but
in
either
ways.
A
B
I
know
we
should
know
great
confirmation
and
get
a
date
in
the
next
couple
of
weeks,
certainly
by
the
first
week
of
july,
so
really
look
forward
to
that.
It's
hard
to
share
a
ton
of
details
until
you
know
the
day
of
but
yeah,
I
would
recommend.
I
mean
I'd
love
to
see
if
any
people
come
as
can
so
I
would
recommend
you
know,
checking
out
hotels
sooner
rather
than
later.
B
I
know
that
sunday
has
booked
up
in
some
cases
and
since
we
don't
know
whether
we're
on
monday
or
tuesday
it'd
be
great
to
get
hotels
booked
as
soon
as
possible,
but
we'll
share
a
lot
more
about
this
in
the
next
two
weeks
or
so
with
the
community.
B
A
Cool,
so
any
questions,
participants
on
the
call
have
about
elections,
election
results,
event,
event,
day,
etc.
A
A
What
what
query
fee,
what
the
query
fee
team
has
been
up
to
in
the
past
week
or
two?
I
guess
so
about
a
month
ago,
we
released
an
important
milestone,
the
0.1.0
tag,
so
we.
A
We
sort
of
labeled
this
as
our
beta
release
and
we
put
some
pressers
out.
We
had
a
lot
of
social
media
information
out
about
this
and
we
ran
like
a
bunch
of
different
efforts
to
create
some
awareness
about
the
tool.
Now,
since
then,
a
few
things
have
changed
I'll
go
into
some
of
the
details
during
this
call
about
what
are
some
of
the
sort
of
big
changes.
A
Obviously,
our
issue
backlog
has
been
pretty
active,
which
is
great
again.
The
kurifi
team
meets
every
other
tuesday
for
a
community
call.
That's
focused
on
getting
cloud
foundry
to
work
with
kubernetes.
A
A
In
you
know
as
much
detail
as
possible
and
to
let
folks
know
about
what
exactly
it
is
that.
A
This
you
know,
sort
of
is
like
the
main
diagram
that
represents
the
tool
itself.
So,
as
you
can
see
here,
it
illustrates
that
there's
you
know
a
client
that
actually
connects
to
the
service
and
then
this
service
is
basically
triggering
two
kinds
of
workloads.
So
there's
a
run
workload,
but
before
that
I
guess
there's
a
build
workload
which
actually
takes
the
code.
A
Does
the
conversion
into
an
image
puts
the
image
on
our
container
registry
it's
here
and
then
the
run
workload.
You
know
grabs
it
from
here
and
puts
it
on
the
container
runtime.
So
the
community
is
hard
at
work
documenting
all
of
this
information
and
building.
You
know
some
images
and
some
resources
that
people
can
make
use
of
to
understand
this
better.
So.
A
The
reason
that
I
wanted
to
bring
this
up
during
the
call
today
is
twofold,
a
obviously
to
communicate
the
architecture
of
gurifi
itself
and
b
to
see
if
you
know
there
are
folks
who
are
listening
here,
who
want
to
jump
in
on
some
non-code
contributions
and
learn
this
better
document
it
better
or
help
improve
whatever
documentation
and
that
we
have.
You
know
I
just
wanted
to
welcome
such
contributions
as
well.
A
So,
like
I
mentioned
the
core
components,
so
I'm
going
to
read
off
of
the
screen
for
a
bit
here,
so
the
and
and
again
feel
free
to
interrupt
with
any
questions
that
you
might
have
or
any
points
that
you
might
want
to
add
or
any
corrections
that
you
might
want
to
make.
A
Now
what
has
changed
is
the
interaction
to
these
can
be
made
directly
through
cube
cutting
or
it
can
also
happen
via
the
cf
cli,
so
that,
I
think,
is
one
significant
change
that
has
happened
and
basically,
each
of
the
different
you
know,
segments
of
what
cloud
foundry
provides
is
a
custom
resource
that
runs
on
a
cube
cluster.
A
The
architecture
basically
identifies
four
resources
here
that
are
of
you
know
more
significance
than
the
rest.
I
guess
so.
The
first
one
is
the
api
deployment
crd.
So
I
have
also
opened
like
a
canines
here
that
is
sitting
over
a
query
fee
somewhere
in
the
back,
so
some
of
these
things
you'll
be
able
to
actually
visually
see
and
verify
as
we
talk
about
them,
and
so
this
is
like
the
set
of
pods
that
are
running
as
part
of
a
curry
fee
deployment.
A
Then
next
up
so
the
kurifi
api
is
basically
the
interface
to
all
of
the
crds,
and
it's
the
go
between
for
any
of
the
clies
that
are
connecting
the
query
fee
api,
which
basically
enable
the
different
workflows.
A
A
The
next
sort
of
big
piece
is
the
build
and
the
run
segments
of
the
whole
curry
fee
deployment,
and
both
of
these,
I
guess
anybody
who
has
some
idea-
I
guess
of
projects
such
as
k
native,
will
obviously
see
the
congruence
here
and
understand
that
there
are
very
similar
sort
of
steps
that
the
query
fee
team
takes
as
well,
so
the
build
workload
obviously
takes
takes
care
of
the
build
process.
Where
you
know
the.
A
I
think
there's
also
a
role
that
irene
plays
here
again
for
the
moment,
and
that
has
to
be
swapped
out
for
like
the
build
workload
and
the
run
workload,
resources,
and
that
will
happen
soon.
I
guess.
A
I
think
the
authors
have
also
emphasized
how
there
is
a
lot
of
extensibility
here
and
we'll
there's
some
documents
that
I'll
show
about
these
being
very
explicit
as
well.
A
Some
of
the
dependencies
as
they
are
defined
are
some
of
the
other
sort
of
tools
or
sub
components
that
are
used
in
order
to.
A
Design
and
define
query
fee
in
general,
and
so
certs
are
one
there's
a
metric
server
and
I
hope
to
do
a
deep
dive
into
how
the
metric
server
can
be
used
with
different
apps
in
a
session
later
this
year,
but
there's
obviously
a
kubernetes
metric
server
that
is
also
deployed
as
part
of
curryfee
that
can
be
made
use
of
kpac
is,
I
think,
a
very.
A
How
can
I
say
it's
it's
very
intrinsic
to
this
community
and
it
builds
on
the
cloud
native
buildback
specification
in
order
to
export
that
oci
compatible
image.
A
Contour
is
the
ingress,
I
think
the
cf
for
kate's
project,
which
was
there
previously
made
use
of
istio.
I
think
a
lot
of
research
has
gone
into
why
car
and
consequently
envoy,
is
a
much
more
appropriate
solution
and
if
folks
are
interested
in
reading
a
little
bit
about
that,
I'm
happy
to
show
where
they
can
and
finally,
service
bindings
are
taken
from
the
kubernetes
project
as
it
is,
and
it's
basically
with
some
of
the
services
that
are
needed,
and
things
like
that.
A
A
So
is
is
a
slightly
more
in-depth
discussion
about
three
areas,
I'll
start
from
the
bottom,
because
it
represents
like
a
rather
significant
deviation
from
before
so
cloud.
Foundry
on
vms,
typically
used
droplets
and
they've
moved
away
from
these
droplets
for
corifi,
and
instead,
like
I've
mentioned.
I
guess
a
few
times
on
the
call
before
a
oci.
A
Compatible
container
image
is
what
is
exported.
This
oci
image
is
put
on
a
container
registry,
which
I
think
is
again
something
that
folks
are
quite
used
to
as
a
paradigm
when
using
docker
it
was
like
docker
hub
is
a
container
registry
and
you
could
download
images
from
there
and
then
use
them
for
deployment.
So
it's
maintaining
sort
of
that
same
workflow
and.
A
A
A
Piece
of
this
whole
architecture
is
the
way
multi-tenancy
is
being
implemented.
So
for
the
first
version,
a
significant
amount
of
effort
went
into
creating
I
using
hierarchical
namespaces,
which
is
again
a
kubernetes
sig
project,
so
hnc,
which
stands
for
hierarchical
name.
Space
controller
is
the
name
of
the
project
and
I
guess
hierarchical
name
spaces
with
what
are
known
as
sub
name.
Space
anchors
were
made.
Use
of
for
this.
A
So
I'll
go
into
some
details
now,
because
there's
no
sort
of
argo
space
in
this-
it's
not
visible,
but
that's
essentially,
where
I
expect
it
to
be
so
sub
name.
Space
anchors
made
it
possible
to
implement
the
cf
definition
of
an
org
spaces
within
the
orgs
and
so
on,
because
kubernetes
has
a
flat
sort
of
notion
of
name
spaces.
They
do
not
allow
nesting,
name
spaces,
and
so,
whereas
you
know
cloud
foundry
at
the
first
instance
of
the
first
step
itself
requires
an
org
and
a
space
within
the
arm.
A
A
It's
been
documented
in
significant
detail
here,
so
there
is
a
spike,
I
guess
much
more
than
a
spike
right
now
that
has
gone
into
creating
a
permissions
model
that
mimics
the
cloud
foundry
permissions
model
and
to
make
use
of
a
native
sort
of
cf
ish
controller
in
order
to
implement
these,
as
opposed
to
going
the
hierarchical
name
spaces
way.
So
this
document
has
in
detail
what
the
limitations
were.
A
So,
if
folks
are
willing
to,
if
folks
require
more
information,
you're
more
than
welcome
to
take
a
look
at
this,
so
I
guess
the
goals
for
this
migration
are
to
produce
something:
that's
more
lightweight
and
reliable,
while
also
reducing
the
dependency
on
something
that
you
know.
Another
open
source
team
is
working
with,
and
I
guess
the
third
rather
more
important
thing
here
is
about
authentication,
and
so
this
is
all.
A
This
also
represents
a
significant
deviation
from
how
cfcli
for
vms
actually
work,
so
corifi
makes
use
of
native
kubernetes
authentication,
so
anytime,
you
sort
of
want
to
log
in.
I
guess
I
could
demo
that,
but
anyhow
I
every
time
you
want
to
log
in
it
makes
use
of
the
same
interface
that
you
would
when
you
use
like
cubectl,
so
it
will
recycle
the
authentication.
A
From
the
kubernetes
cluster,
whatever
it
is
that
you're
using-
and
so
I
think
this
diagram
does
a
good
job
of
explaining
what
the
overview
of
the
auth
looks
like
and
there's
some
more
information.
But
if,
if
folks
want
to
again
learn
more
about
this,
so
but
then
what
I
wanted
to
highlight
is
it
it's
a
rather
significant
deviation
from
the
way
cloud.
Foundry
did
it,
and
even
from
the
way
that
cf
for
kids
was
doing
auth.
A
Let's
see
the
architecture,
docs
also
provide
like
a
detailed
mapping
between
some
of
the
cloud
foundry
concepts
and
what
is
the
kubernetes
analog
or
equivalent
for
some
of
these?
So
you
know
if
folks
want
to
create
a
mind
map
of
here's.
A
You
know
what
here
are
you
know
on
the
left,
some
of
the
concepts
that
we
are
used
to
and
on
the
right
here
are
like
some
of
the
things
that
are
new.
You
know
feel
free
to
peruse
this,
and,
like
I
mentioned,
a
lot
of
the
architectural
decisions
have
come
from
information.
That's
here,
for
example,.
A
Using
image
formats
is
available
here,
and
one
of
the
reasons
I
like
to
refer
to
these
is
they
always
provide
like
a
context
from
the
cf
for
vm's
world.
So,
if
you're
used
to
like
bosch
based
stuff,
you
can
get
the
context
and
the
delta
is
also
highlighted.
Now,
like
I
mentioned,
here's
a
another
document
that
provides
the
reasons
for
why
the
team
picked
contour
and
envoy,
as
opposed
to
any
of
the
others
that
were
here.
So
all
of
these
are
also
documented
again
linked
from
the
main
sort
of
architecture.
A
Dock
for
upcoming
changes,
there's
a
link
to
proposal
documents,
so
you
can,
if
you
would
like
to
sort
of
review
what's
going
on
in
query
field.
I
guess
these
are
super
useful
in
terms
of
learning
about
you
know,
what's
to
come.
A
That
is
that's
all
I
had
for
show
and
tell
if
folks
have
any
questions
comments,
feedback,
I'm
happy
to
take
them
on
there's,
also
like
a
bunch
of
other
resources
that
are
in
the
works
and
as
we
start
to
release
these
I'll
share
links
to
them
on
the
slack
as
appropriate,
but
yeah.
That's
it
for
now,
thanks
for
listening
and
I'm
happy
to
take
any
questions.
A
Time
frameworks,
these
are
all
on
github
people
can
check
them
out
I'll
drop
them
into
the
cab
slack
or
I'll.
Send
them
to
you
on
slack
as
well.
Okay,
yeah.
B
A
That's
all
I
had.
I
guess
we
come
to
a
close
thanks.
Everyone
for
attending
today
and
see
you
in
next
month's
call.