►
From YouTube: Cloud Foundry CAB Call Feb 16 2022
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
Hello,
everyone
welcome
to
the
february
edition
of
the
cloud
foundry,
capcom
and
yeah.
It's
it's
so
nice
to
have
everyone
here
today,
so
I
did
send
out
some
information
earlier
about
this
call
going
to
focus
on
a
particular
aspect
of
jammy,
jellyfish
updates
as
the
new
cloud
foundry
linux
operating
system
and
so
joining
us
are
ruben,
benjamin
and
ramon,
who
I
know,
have
been
hard
at
work
at
making
this
possible.
So
big
thanks
to
the
three
of
you
for
taking
some
time
out
and
helping
educate
the
community
about.
A
What's
going
on
so
I'll,
just
give
a
little
bit
of
background
and
then
switch
it
over
to
reuben
and
the
others.
And
of
course
this
is
more
effective.
If
you
know
folks
participate,
ask
your
questions.
What
have
you
so
it's
a
good
opportunity
to
talk
to
the
contributors
directly
and
yeah,
just
you
know
make
the
best
of
it.
So
by
way
of
background,
I
guess
the
cf
community
has
done
os
migrations
in
the
past.
A
Obviously,
and
after
the
typical
life
cycle
of
a
canonical
release,
I
guess
the
community
gets
together
and
engages
in
a
migration.
Now
the
migration
is
non-trivial,
both
in
terms
of
efforts
and
expense,
and
so
fortunately
we
only
have
to
do
this
once
in
a
while
now
as
a
community,
the
sort
of
stewards
of
the
project
have
decided
that,
instead
of
moving
from
the
bionic
lts
to
the
focal
lts,
we're
just
going
to
take
a
jump
to
the
next
one.
A
So
this
is
going
to
be
1804
to
22
o4
and
we're
skipping
2004
in
the
middle,
and
the
reason
is
we
can
just
extend
support
and
keep
working
with
it
for
longer,
so
just
very
basic
operational
convenience.
A
So
that's
really
the
background
to
this,
and
so
some,
I
guess
alpha
or
pre-alpha
versions
of
jammy
have
been
made
available,
and
so
we
are
making
a
jammy
version
available
as
a
stem
cell
and
then
in
parallel,
also
working
on
the
cf
linux
file
system
and
then
making
it
available.
A
B
Yeah
sure
thanks
ram
yeah,
you
have
to
explain
most
of
it
so
of
the
reasoning
behind
why
we
are
skipping
right.
We
we're
skipping
the
vocal
release
because
we
just
want
to
minimize
the
effort
right.
It's
an
expensive
thing
to
do
so
by
skipping
or
moving
ahead
or
skipping
a
version
we
would
just
buy
more
time
or
of
of
an
lcs
window.
Basically,.
B
B
So
we
started
at
work
and
then
we
got
a
pipeline
going
conquest
pipeline.
So
that
was
all
the
templating
or
the
the
basically
the
naming
stuff
that
was
done,
and
then
we
found
out
that
they
actually
already
started.
Publishing
death
builds
of
the
jammy
stem
or
the
the
to
be
gemmy
image,
or
so
there
was
a
docker
image
being
built
built
with
the
gemmy
stem
cell
or
jami
tag.
B
B
Eventually,
it
has
been
like
playing
catch
up,
basically
dim
bomb,
something
upstream
again,
and
then
we
have
to
change
our
package
definitions
so
that
we
know
that,
like
okay,
this
lip
is
now
at
this
version
or
something
right
because,
but
the
nice
thing
is
of
this
approach
is
that
we
are
keeping
up
right.
It's
not
that
it's
smaller
changes
because
they
break
something
and
then
we
immediately
have
a
red
pipeline
and
we
could
just
fix
one
thing
instead
of
like
having
to
do
it
all
at
once.
B
So,
by
keeping
up
with
it,
we
were
confident
enough
to
publish
an
alpha
release
and
the
reason
why
we
did
this-
I
mean
we
are
ahead
of
schedule,
this,
the
rfc
that
has
been
scheduled
or
published,
because
I
mean
that
says
that
this
work
that
has
already
been
done
right
should
have
been
starting
march
next
month,
but
yeah.
We
already
have
an
alpha
and
that
just
allows
us
to
by
publishing
alpha.
We
we
hope
to
have
other
teams
to
that.
B
B
So
you
need
to
bump
to
3.1.
So,
for
example,
cloud
controller
and
those
types
of
things
need
to
bump
their
ruby
version,
and
that's
also
true
for
the
bosch
director
and
all
the
cpi's.
B
So
that
is
work
that
has
been
going
on
as
well.
We
were
luckily
already
on
for
the
director
on
ruby
3.0,
so
the
bump
to
3.1
was
not
so
big.
B
The
cpi's
are
mostly
on
2.7
2.6,
so
they
need
to
make
a
bigger
bump.
Also,
the
code
base
are
smaller,
so
yeah
yeah
that
that's
a
bit
like
what
it
that's.
Basically
the
status
we
have
an
alpha.
Other
teams
are
aware:
it's
we.
We
are
ahead
of
schedule
and
I
think
we're
not
far
away
from
feature
freeze.
B
B
And
by
then,
we
all
also
hope
to
have
the
director,
the
boss
director
be
compatible
with
jamie,
so
that
should
allow
us
during
the
beta
to
have
at
least
maybe
go
in
some
deaf
environment.
So
that's
what
we
we
hope
to
at
that
point.
We
hope
that
the
community
starts
maybe
rolling
gemmy
stem
cells,
at
least
out
for
the
director
in
their
yeah,
in
some
of
their
pipelines,
maybe
not
to
production,
but
at
least
help
us
get
our
feedback
on
this.
C
From
your
experience,
I,
I
know
you're
doing
version
bumps
of
various
different
packages
and
stuff
like
that.
Are
there
new
packages
that
are
being
added
because
jimmy
jellyfish
that
we
need
to
be
concerned
with,
or
are
there
packages
being
removed
because
they're
not
going
to
be
supported
on
jimmy
jellyfish
that
you're
seeing.
D
Yeah
all
the
compiler
packages,
basically,
everything
related
to
compiling,
has
major
bumps
so
yeah,
and
it's
also,
we
have
a
lot
more.
The
they
have
done
a
lot
of
package
segregation.
It's
like
a
lot
of
things,
have
been
pulled
into
small
packages,
for
I
don't
know,
maybe
for
better
upgrading
updating
in
the
future.
D
But
yeah
the
the
biggest
one
is
the
gcc
that
we
see
and
the
openssl
and
the
ruby
one
related
to
that.
With
the
other
things
we
haven't
seen
that
much
issues
yet
but
yeah
we
to
be
honest.
We
also
haven't
fully
fully
tested
it
with
a
lot
of
applications.
A
lot
of
other
software.
B
D
B
No,
I
mean
so
from
the
first
from
the
first
time
that
that
was
tried
right.
There
were
some
issues
that
we
found
and
those
were
raised
up
upstream
with
the,
for
example,
the
I
think
aj,
proxy
or
nginx
does
have
compilation
issues
on
gemmy.
B
If
you,
if
you
search
for
for
jamie
and
the
cloud
from
new
york,
you
see
that
there's
a
lot
of
upstream
issues
already
outside
of
the
bosch
scope.
So
what
has
been
work
has
been
going
on.
B
D
C
B
Did
do
an
experiment
with
so
there's
this
thing
with
the
there's,
an
open,
ssl
gem
inside
that
is
basically
embedded
in
in
the
ruby
source
code,
but
it
is
a
separate
gem
and
that
gem
itself
publishes
that
it's
compatible
with
2.6
ruby
2.6,
so
it
so.
I've
done
an
extra
successful
experiment
where
I
took
like
the
2.7,
codebase
and
backported,
or
just
took
that
codebase
or
the
gem
source
and
put
that
in
the
source
like
because
we
compiled
ruby
from
source
in
the
bosch
release.
B
I
I
was
able
to
get
a
a
2.7
working
that
that
compiles
on
on
the
gemi
stem
cell,
so
maybe
the
buildback
buildback
people
want
to
look
into
something
like
that.
Because
I
mean
for
the
bosch
releases
we
can
say
yeah.
You
just
need
to
bump
to
3.1,
but
if
you
have
to,
if
you
have
that
requirement
for
all
your
apps,
that
will
be
a
bigger
thing
right,
because
that
that's
fine.
E
But
on
that
subject
about
bigpacks
I
would
like
to
mention:
we
will
be
running
ubuntu
18
containers
on
top
of
ubuntu
22
kernel.
E
So
this
is
what
I
call
the
frankenstein
problem,
but
it's
it's
still
ubuntu
and
the
problem.
The
frankenstein
problem
is
to
run
a
g
ellipse
c.
That
is
not
the
same
and
not
granted
to
work
with
the
camera.
E
So
inside
the
container
you
are
relying
on
the
jlpc,
but
the
kernel
is
another
version
that
the
jlipsy
was
supposed
to
work
with
because
they
have
four
years
basically,
but
ubuntu
on
ubuntu
is
not
that
a
big
deal.
I
mean
you
have
bigger
risk
if
you're
running-
let's
say
red
hat
on
top
of
ubuntu
until
now,
and
the
kernel
is
still
very
conservative,
so
it
shouldn't
be
a
big
deal,
but
still
there
can
be
some
risks
about
this
frankenstein
problem
and
yeah.
Most
people
don't
care
about
what
piano
they
have
below
there.
E
B
The
the
idea
is
to
still
do
like
dual
stack
support,
but
I
don't
know
that
still
needs
to
be
proven
out
right
if
that,
hopefully.
E
It
will
doing
doing
a
new
stack
is
a
big
deal,
because
it's
required
to
rebuild
every
binary
dependency
and
put
that
on
online
for
the
k
for
the
the
bill
packs
to
to
draw
them
when
necessary.
E
B
It
is
in
scope
right.
The
rfc
says
that
we're
gonna
create
a
cf
linux
ls4
root
fs,
so
it
is
going
to
be
built.
It's
just
a
question
of.
Can
we
get
the
current
root
of
s
cf
linux
fs43?
If
that
can
still
work
on
the
gemi
stem
cell
right?
If
that
it
should
right,
but
I
don't
know
we
that
that
needs
to
be
proven
out
like
it.
It
worked
last
time
that
we
got
it
working,
but
I
don't
know
if
that
it
required
some.
B
Some
things
with
star,
I
think
was
I
don't
know
there
were
some
binaries
incompatible
or
something,
but
I
mean
probably
we
can
get
that
to
work,
but
we
are
not
there
yet
right
that
that
will
that
will
be
the
that's
why
we
have
a
year,
basically,
where
we
that
that
we
want
to
get
to
figure
all
these
things
out
and
get
that
working.
C
The
only
other
question
I
got
is
that
that
I
put
inside
that
I
commented
in
inside.
The
proposal
is
the
time
frame,
because
the
ending
of
everything
occurs
at
the
ending
of
the
bionic
stem
cell,
and
the
sooner
you
can,
if,
if
it's
possible,
to
get
stuff
done
before
the
end,
because
we
had
problems
this
time
around
right
going
from
trustee
to
bionic,
because
truster
was
already
already
expired.
When
bionic
wasn't
completely
done,
so
it
caused
frustrations
in
the
community.
C
C
E
C
Yeah,
so
it's
great,
so
I
appreciate
all
the
hard
work
reuben
and
your
team
and
and
sap
and
and
ramon
has
done
on
on
the
project.
So
so
thank
you.
B
C
B
So,
that's
I
don't
know
but
yeah.
So
that's
why
the
the
ubuntu
esm
thing
can
be
interesting
right.
That's
how
you
get
around
this
problem,
but
it
costs
money
so
yeah,
but
I
mean
like
this
is
all
we
we
just
have
that
year,
but
right
a
year,
that's
the
window.
We
have
where
all
this
stuff
needs
to
happen
because
of
how
canonical
set
things
up
so
yeah.
E
I
have
a
question.
It
may
be
nice
that
on
bosch
io
we
distinguish
releases
that
have
successfully
been
tested
on
the
new
stem
cell.
Maybe
some
kind
of
compatibility
grid.
E
Good
question
but
yeah
this
is
to
be
worked
out
just
just
from
the
idea.
B
Thanks
nice
idea,
I
don't
know
how
to
do
it,
but
yeah.
We
can
think
about
it.
Yeah
the
problem
is
brush.
Io
is
not
compiling
anything
right.
It's
just
taking
a
release
from
somewhere
and
putting
I.
E
D
The
only
thing
we
can
solve
this
is
that
the
actual
release
maintainers
are
going
to
update
like
we
like
how
we
add
how
you
add,
a
new
release
to
bosch.
Io
currently
is
just
by
adding
a
small
text
file
and
you
basically
can
add
another
table
in
there,
but
then
the
release,
the
the
release
engineer,
should
maintain.
This
is
going
to
be
a
lot
of
vrs
again.
E
Yeah
sure,
but
it's
been
a
long
time
that
I've
been
thinking
that
bush
release
could
at
least
declare
with
which
stem
cells
they
are
compatible,
and
it
was
already
a
problem
with
when
red
hat
stem
cells
were
still
maintained.
E
C
Benjamin,
I
wonder
if
this
is
really
an
issue
that
rom
can
actually
help
resolve
by
asking
the
stakeholders.
What
have
we
tested
what's
going
on
and
maybe
give
a
a
a
bi-monthly
every
every
two
months
report
saying
here's
the
status
on
jimmy
jellyfish
upgrades
right.
This
is
what's
been
tested.
This
is
what
works
and
it's
kind
of
it's
more
like
a.
It
would
be
cool.
C
C
Maybe
that's
the
way
to
handle
it.
This
way,
just
putting
the
effort
and
trying
to
record
everything
and
what
works
and
what
doesn't
on
the
on
the
stakeholders
is
more
work
than
just
asking
the
questions.
What
have
you
tested
and
then
they
just
you
know
and
rob,
can
write
the
the
article
or
or
maybe
chris
writes
the
article
or
or
we
write
the
article.
It
doesn't
really
matter,
but
the
point
would
be
kind
of
like
a
a
bi-monthly
upgrade
update.
How
are
we
progressing
along
the
chain?
Are
we
are
we
meeting?
C
C
F
I
could
imagine
that
over
time,
the
cf
deployment
pipelines
once
they
are
set
up
can
bring
at
least
some
data
for
the
let's
say,
most
common
bosch
releases
that
are
used
as
part
of
cf
deployment.
F
I
mean
today.
I
think
they
still
run
against
xenia
right,
there's
one
pipeline
somewhere
well,
hidden
that
runs
against
bionic
and
yeah.
There's
nothing
yet
officially
that
dries
out
jeremy,
but
for
the
future
I
could
imagine
to
drop
senile
and
just
to
validate
in
standards,
bionic
stem
cell
and
add
a
new
one
that
hopefully
colors
green
over
time
with
jeremy.
B
Yeah,
I
think
that
that's
the
main
thing
people
want
to
are
interested
in,
like
is
cf
compatible
right.
All
the
other
releases
yeah
are
nice,
but
I
mean
that
that's
not
really
hurting
anybody.
People
can
pr
and
stuff
like
that,
but
this
the
main
cf
thing:
that's
what
people
want
to
know.
C
And
then
the
the
stack
itself,
this
you
know
the
linux
fs4
stack,
but
that's.
B
E
A
Yeah,
I
think,
if
manual
updates
for
like
the
first
few
cycles
are
the
best
way
to
do
it.
You
know
this
I'm
more
than
happy
to
step
in
and
get
updates,
but,
like
stefan
said,
if
there's
some
way
of
baking
it
into
like
a
cf
deployment
thing,
I
think
that
would
be
super
cool
and
super
useful
as
well
and
so
much
more
scalable.
B
Obviously
so
another
way
to
keep
an
eye
on
this
or
track
of
this
is
to
attend
the
working
group
meeting
right
after
the
deployments
working
group
meeting,
I
mean
that
that
working
group
is
still
being
formed,
but
I
assume
that
there
will
be
a
lot
of
information
there.
F
Yeah,
it
would
take
a
little
bit
of
time
under
this
working
group.
It
was
just
approved
yesterday,
so
we
take
a
little
bit
and
I
know
the
work
will
bleed
pretty
well
and
he.
E
A
G
A
G
A
quick
question:
I
am
a
fairly
new
cloud:
foundry
admin
for
lack
of
a
better
word.
I've
taken
on
a
project
at
the
company
that
I
work
for,
and
we're
kind
of
stuck
right
now
on,
and
maybe
I'm
wrong.
It's
an
assumption
that
I've
made
we're
still
using
the
xenial
stem
cell
and
our
security
people
are
asking
us
hey.
When
are
you
guys
going
to
be
moving
so
that
we
can
get?
G
G
But
I
don't
want
to
go
through
all
of
that
effort
which
is
going
to
be
for
me.
I'm
not
a
very
strong
developer
a
year
or
two
worth
of
work
just
to
have
you
guys,
come
in
and
drop
jammie
jellyfish
and
and
be
right
where
matt
so
is
there
some
guidelines
that
maybe
I
can
go
back
to
our
security
department
and
say
hey?
This
is
what
we
can
do
as
kind
of
an
interim
to
getting
off
of
xenial
and
onto
something
that
is
supported.
D
E
G
Correct
yeah
we're
just
right
we're
just
using
bosh
to
deploy
cloud
foundry
again,
because
I'm
kind
of.
G
E
G
I
guess
there
was
one
component
that
I
that
I
read
about
was
on.
We
use
nfs
mounter
or
is
it
nfs
volume
I
think,
is
the
package
and
for
a
while
there
they
had
a
blocking
issue
with
bionic,
and
I
think
that
that
was
resolved,
and
I
think
that
was
the
last
piece
that
we
were
waiting
on
before
we
started
because
we
have.
G
I
think
we
have
like
three
dev
lanes
that
we
can
test
this
in,
and
is
it
just
going
in
using
the
ops
files
or
do
we
just
go
into
the
manifest
and
just
change
the
default
from
xenial
to
bionic?
D
Deployment
files
yeah
yeah,
because
then
you
know
that
everything
is
changed
correctly
and
you
need
to,
of
course
change
your
cl.
You
probably
need
to
update
your
cloud
config
as
well.
If
you
have
something
added
there,
that
is
only
specific
for
whatever
stem
cell
version,
you
have,
you
just
need
to
add
ubuntu
bionic
there
as
well.
Okay,.
E
For
nfs,
maybe
that's
that's
a
very
good
question.
I
mean
for
nfs,
but
maybe
that's
the
very
specific
point.
You
need
to
spend
extra
time
when
testing
on
your
sandbox
environment,
making
sure
it's
working
perfectly
yeah.
That's
a
question
that
that
might
be
a
question.
I
don't
have
the
answer.
You
can
check.
G
I
appreciate
the
comments
I
do.
It's
just
been
something
that
our
security
team
has
been
kind
of
poking
me
on
and
I'm
like.
I,
I
don't
know
the
answers
to
right
now
and
I
see
the
direction
that
you
guys
are
going
and
we
like
it,
but
we're
not
sure
that
the
time
frame
is
going
to
fit
with
our
schedule.
But
I
do
appreciate
you
take
giving
me
a
couple
minutes
of
some
some
time.
Thank
you
very
much.
C
A
Right,
I'm
slightly
off
topic
but
fun
off
topic.
So
no
problem
justin
you
can.
I
mean
this.
Is
the
community
call
so
you're
more
than
welcome
to
find
some
time
here,
but,
like
most
people
said,
slack
is
probably
the
best
place
to
get
answers.
But
yes,
if
folks
did
not
have
much
else,
I
think
we
can
give
us
all
20
minutes
back
and
see
you
on
the
cap
call
for
march.
Thank
you
very
much.
Bye,
bye,
bye,
bye,.