►
From YouTube: 2021-11-04 Crossplane Community Meeting
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
All
right,
hello
and
welcome
everyone
to
the
cross,
plane
biweekly
community
meeting.
It
is
what
is
it
november
now
november,
4th
on
a
thursday
at
our
regularly
scheduled
time.
It
looks
like
we've
got
some
a
few
folks
here
and
some
other
folks
filtering
in,
but
we're
gonna
go
ahead
and
kick
it
off
just
to
make
sure
to
be
respectful
of
everyone's
time.
It
does
look
like
there
may
be
some
new
folks
on
the
call.
A
A
few
things
I
want
to
mention,
please
feel
free
to
drop
your
name
in
the
google
doc,
which
is
in
the
zoom
chat
right
now.
Everyone
should
have
comment
or
access
and
we'll
make
sure
to
approve
those.
Also,
if
you
have
something
to
add
to
one
of
the
sections,
please
feel
free
to
do
that
as
well.
A
We
do
have
a
few
things
already
on
the
docket
for
today
that
we'll
make
sure
to
cover,
but
before
we
do
any
of
that,
I
do
want
to
give
folks
who
may
be
new
to
the
call
and
who
feel
comfortable
the
opportunity
to
introduce
themselves
and
tell
us
who
you
are.
What
you're
interested
in
you're
also
welcome
to
to
just
listen
in
as
well,
if
you
prefer.
A
All
right
well
we'll
go
ahead
and
kick
it
off.
Then.
If
you
do
join
or
if
you're
here
right
now
and
decide,
you
want
to
speak
up
later,
feel
free
to
drop
a
message
in
the
chat
or
just
speak
up.
That's
also
welcome,
so
I'm
gonna
go
ahead
and
get
us
started
here.
A
All
right,
so
biggest
news
of
the
past
couple
weeks
has
been
some
patch
releases
that
went
out
as
well
as
a
new
minor
release.
V
1.5,
let's
start
off
by
looking
at
v
1.5
and
going
through
some
of
the
updates
there.
You
know
this
was
a
pretty
pretty
tame
release
in
terms
of
there
wasn't
too
many
updates,
but
there
was
a
few
bug
fixes
that
have
been
patched
back
as
well
as
some
new
documentation
and
and
various
other
fixes.
A
So
you
can
take
a
look
at
the
release
notes
here
to
look
into
that.
One
of
the
things
I
definitely
want
to
call
out
is
a
a
big
thing
that
was
in
this
release.
Is
we
had
a
lot
of
new
contributors
getting
their
first
contributions
to
core
crossplane
in
here?
Some
folks
already
worked
on
providers
and
that
sort
of
thing,
but
it
was
definitely
really
exciting
to
see
a
lot
of
new
names
on
here.
A
A
So
let's
hop
over
there
and
see
that
if
you
go
down
to
concepts,
composite
resources,
you'll
see
some
really
nice
diagrams
some
updated
information
and
just
an
overall,
cleaner
explanation
of
how
composition
works.
So
it's
definitely
really
exciting
to
see
that
and
last
one
here
it
looks
like
movavic,
who
I
don't
think
was
able
to
make
it
today,
but
this
pr,
I'm
actually
blanking
on
this
specific
one.
But
it
looks
like
do
not
couple
life
cycle
of
crds
and
xrds
with
revisions
and
yep.
A
Revisions
are
deleted,
so
that
basically
means
that
if
you
clean
up
an
old
provider
vision
or
have
it
automatically
garbage
collected,
then
your
xrds
and
and
crds
excuse
me
will
stay
around,
but
your
your
revision
will
be
gone
and
when
you
actually
delete
the
provider
configuration
the
parent
of
those
revisions,
those
crds
and
xrds
will
get
cleaned
up.
A
So
that
was
great
to
see
from
movavik
like
I
said,
there
were
also
a
number
of
bug
fixes
in
this
release,
and
many
of
them
were
back
ported,
potentially
almost
all
of
them.
A
So
we
can
take
a
look
at
a
couple
of
these,
a
a
big
fix
here
that
was
really
important
to
get
back
for
some
folks
who
were
using
the
functionality
where
you
can
actually
install
packages
when
you're
installing
crossplane,
which,
if
we
hop
over
to
the
docs
again,
we
should
be
able
to
see
in
our
api
documentation
or
actually
potentially
in
our
reference
install.
A
If
you
go
down
and
look
at
packages
potentially
provider
packages
and
configuration
packages,
you
can
essentially
provide
a
list.
I'm
gonna
install
those,
but
the
way
that
those
were
being
used
and
being
installed
meant
that
every
time
that
you
upgraded
crossplan
and
changed
a
version
of
the
revision,
you
were
getting
a
a
brand
new
package
install
which
was
then
causing
that
to
fight
so
that's
fixed.
So
anyone
who
is
is
using
that
functionality
should
have
that
cleaned
up
now
and
that
is
fully
backwards
compatible.
A
So
you
can
still
see
versions
in
the
name
of
packages,
but
you
will
upgrade
successfully
to
the
new
version,
just
without
changing
the
name,
so
that
was
great
to
get
in,
and
so
next
up
is
going
to
be
version.
1.6,
which
looks
like
it
says
it's
going
to
be
sometime
around
the
new
year.
I
believe
we
should
have
a
date
already
in
place
for
that
yep,
so
the
target
date
is
december.
A
21St,
that's
obviously
pretty
close
to
at
least
some
american
holidays
and
some
other
holidays
around
the
world,
and
so
we'll
see
if
that
gets
changed.
But
that
is
the
at
least
tentative
plan
right
now,
if
not,
then
it'll
be
on
a
date
close
to
that
all
right.
I
don't
think
that
many
folks
have
gotten
opportunities
to
kind
of
brainstorm
what
they
want
included
in
v
1.6,
but
I'll
pull
up
these
boards,
and
we
can
take
a
look
anyway.
A
So
it
does
look
like
we
have
a
few
things
in
progress,
we'll
talk
about
some
scale
stuff
coming
up
specifically
related
to
installing
crds,
but
I
also
know
that
nick
has
done
some
work
already.
A
That
will
be
in
v.
1.6
I'll
jump
over
here
see
if
I
can
find
it
support
exponential
back
off.
So
nick
has
reconfigured
our
reconcilers
a
bit
to
use
exponential
back
off,
and
this
is
pretty
much
ready
to
go
I'll,
approve
it
as
soon
as
the
the
dependency
on
crosslink
runtime
is
updated.
So
you
should
see
some
better
performance
in
both
the
composition
and
package
controllers.
A
I'm
not
super
familiar
with
the
other
things
in
progress
here.
Did
anyone
have
something
they
wanted
to
comment
on?
That's
currently
on
the
roadmap
board.
B
Maybe
I
can.
I
can
briefly
talk
about
like
the
design,
the
second
design,
which
is
support
for
pluggable
secret
store.
B
It
was
like
a
design
which
is
in
progress
for
for
some
time
right
now,
but
recently
we
had
like
updated
it's
based
on
all
the
feedback
we
get
so
right
now,
it's
I
believe
it
is
close
to
get
merged.
There
are,
I
think,
two
controversial
points
that
we
need
to
agree
on,
so
I
would
appreciate
if
anyone
can
like
have
a
look
and
support
either
of
the
solutions
or
designs
there
or
like
just
review
the
document
and
add
comments.
I
really
appreciate
that.
A
Awesome
thanks
hassan,
and
thanks
also
for
your
patience
and
going
through
this.
I
know
it's
been
a
long
review
process
and
some
other
things
have
come
up,
but
definitely
you
know
community
members
as
well.
I
know
some
folks
have
already
commented
on
this
and
it's
been
great
to
see
folks
who
are
potentially
cross-plain
users
or
just
folks
with
thoughts
on
the
matter
or
contributors
weighing
in
here,
and
I
know
a
lot
of
users
are
looking
for
this
so
definitely
feel
free
to
do
that.
A
All
right,
let's
jump
back
over,
I
believe
there
are
a
few
other
things
that
are
on
the
roadmap
here.
One
that
I
wanted
to
bring
up.
That
is
still
top
of
mind
for
me.
A
A
This
is
less
related
to
scaling
in
terms
of
performance
and
more
related
to
isolation
and
multi-tenancy,
and
so
definitely,
if
folks
have
thoughts
on
this,
take
a
look.
I
think
this
is
something
I've
heard
folks
ask
for,
but
there
hasn't
been
a
ton
of
comments
on
the
on
the
issue.
A
A
This
provider,
runtime
interface,
could
potentially
allow
you
to
run
your
controllers
externally.
So,
as
you
know,
a
function
somewhere
or
just
you
know,
have
a
central
control,
plane
and
controllers
executing
elsewhere,
which
has
some
implications
in
terms
of
permission
boundaries.
A
One
thing
that
is
probably
more
relevant
to
folks-
and
you
know
today's
usage
of
crossplane
is
this-
is
currently
what's
holding
back
promoting
the
controller
config,
which
is
a
highly
used
resource
by
a
lot
of
folks
who
want
to
do
things
like
configure
their
service
account
for
their
controller
and
and
so,
if
you
have
thoughts
on
that
or
want
to
weigh
in
definitely
feel
free
to
take
a
look.
A
I
know
there's
a
few
things
related
to
custom
compositions
and
that
sort
of
thing,
but
I
believe,
moveoffic
or
nick,
is
probably
the
best
to
talk
about
that,
and
neither
one
of
them
are
here
so
yeah.
C
Yeah,
maybe,
although
there
are
some
also
related
items,
as
I
see
in
the
roadmap,
then,
as
you
know,
we
are
also
progressing.
We
are
also
making
some
progress.
You
know
in
scalability
of
crossplane,
maybe
it's
worth
mentioning
that
you
know
with
the
new
terraform-based
providers.
C
C
You
know
forcing
kubernetes
boundaries,
especially
in
gk
clusters,
and
you
know
we
are
in
touch
with
the
upstream
humanities
community
to
have
our
you
know,
use
cases
covered
and
you
know
we
are
also
making
progress
in
the
scalability
dimension
of
cross
plane,
very
suspects.
A
Awesome
thanks
for
that
update.
Maybe
we
should
just
roll
into
that
topic
in
general
there.
I
know
a
few
of
us
were
at
the
sig
api
machinery
meeting
this
week,
but
I'll
pair
did
you
have
anything
else
that
you
want
to
go
into
more
depth
that
you
know
specifically
related
to
this
issue
or
do
you
feel
like
it's
pretty
self-explanatory
at
this
point?
What's
going
on.
C
Yeah
I
mean
there
is,
as
you
mentioned,
we
have
attended
this
api
missionary
meeting
this
wednesday
and
there
is
an
open
pr
related
to
the
open
api
schema
publishing,
which
is
the
current
big
blocker.
C
A
Awesome
thanks
for
that
update,
yeah
folks
want
to
follow
along
with
this
pr.
This
is
the
one
specifically
related
for
the
api
server
and,
as
I've
ever
said
there
is,
there
was
pretty
strong
agreement
that,
as
long
as
the
cherry
picks
are
not
too
cumbersome,
then
this
should
be
something
that
is
included
in
patch
releases.
Coming
up
definitely
wanted
to
give
a
shout
out
to
all
of
the
folks
who
worked
on
this.
A
Specifically
all
pair
move
off
and
negs,
and
I
believe
hassan
as
well.
Definitely
some
really
in-depth
investigation
here,
and
it's
also
nice.
You
know
we
as
a
community
should
always
be
striving
to
contribute
back
up
upstream
to
kubernetes.
So
it's
really
cool
that
we
get
to
have
some
impact
there
as
well.
A
All
right,
I
want
to
go
more
generally
into
the
terra
jet
update,
and
I
know
we
have
a
few
folks
who
are
working
on
that,
maybe
hassan
or
I'll
pair,
if
you
all
wanted
to
give
an
update
based
on
some
of
these
things
in
the
dock
here.
B
Okay,
thank
you
so
yeah,
basically
like
in
the
last
sprint,
we
focused
on
like
making
some
abstractions
and
stabilizing
stabilizing
the
telogen
interface,
so
that
new
telejet
based
providers
could
be
contributed
from
like
anyone
who
who
wants
to
like
generate
so
right
now.
B
We
only
have
like
two
two
terraform-based
providers,
one
tf
azure
and
on
tf
aws,
and
we
basically
like
use
that
as
an
opportunity
to
to
like
try
different
cloud
providers
or
different
providers
and
then
like
find
the
comments
and
differences
and
try
to
cover
that,
so
that
we
can
have
a
common
api
so
for
others
as
well.
So
that
was
the
work
that
that
we
like
that
we
did
in
the
last
sprint.
So
it
should.
B
It
also
like
enables
us
to
document
the
steps
to
generate
a
telejet
based
provider.
So,
like
we
have
a
like
guide
coming
soon,
there
is
a
draft
pr
open
and
with
that
guide,
we
expect
anyone
to
like
just
generate
a
crossplane
provider
from
an
existing
telephone
provider.
B
And
finally,
I
would
like
yeah.
There
are
also
some
like
bug
fixes
and
improvements,
like
you
can
add,
custom
connection
keys
to
any
resource
and
also
some
some
improvements
to
speed
up
the
provisioning
process.
B
And
finally,
I
would,
I
would
add,
like
provider
tfgcp
is
also
coming
soon.
There
is
already
a
repo
initialized,
but
not
like
not
much
resources
landed
there
yet,
but,
like
I,
I
believe
in
the
latest
in
the
following
week,
we
expect
to
have
it
and
like,
as
you
have
discussed,
there
is
some
api
servers,
availability
performance
investigations.
B
I
I
like
mentioned
this
under
terrorism,
because
this
unrevealed
actually
that's
problems,
but
you
have
you
already
mentioned
that
I
don't
know.
If
I'd
do
you
have
anything
else
to
add.
C
Yeah,
as
you
mentioned,
we
first
hit
this
these
issues.
You
know
when
doing
the
development
and
testing
of
terraform-based
providers,
and
there
are
multiple
issues
opened
regarding
regarding
this.
Two
of
those
issues
are
opened
in
the
terror
jet
project.
C
So
for
folks
who
are
interested
in
the
origins
of
these
problems,
how
we
supported
them
and
how
we
analyze
them,
I
will
put
the
links
there
yeah.
A
Thanks
for
that
update
folks,
lots
of
exciting
stuff
coming
there
and
definitely
a
lot
of
hard
work,
so
I
definitely
appreciate
y'all's
efforts
and
I'm
excited
to
see
a
big
release
coming
up
as
well.
A
Let's
see
looks
like
we
had
a
provider
aws
release
and
this
fixed
a
problem
in
eks
cluster,
and
I
believe
this
was
potentially
fixed
by
christopher,
who
is
a
new
maintainer
of
provider
aws,
and
we
want
to
take
a
moment
to
congratulate
christopher
on
that
and
welcome
you
as
a
maintainer
and
also
thank
you
for
this
awesome
fix
and
if
you
want
to
tell
us
a
little
bit
about
what
went
into
this
and
what
was
going
wrong.
You're
definitely
welcome
to.
D
Yeah
thank
thanks
for
this.
Yes,
I
I
saw
that
we
bumped
the
sdk
version
and
we
had
the
problem
that
our
two
headers
are
missing
for
present
urls
and
we
added
again
the
two
headers
I
think
in
this
match
request.
We
have
a
yeah,
a
very
detailed
description,
what
we
are
looking
for
and
what
we
have
fixed.
D
Yes,
awesome
and
now
it's
working
again
for
that
we
can
use
the
cubeconfig
for
provider,
hand
for
provided
kubernetes,
yes
yeah
here,
you
see
it
in
detail
that
a
little
bit
up,
and
then
you
see
that
we
are
missing
two
headers,
so
the
x
amazon
expired
time
and
also
the
x
amazon
signed.
D
Headers
was
without
yes,
they
both
are
missing
and
I
think
at
the
end,
it's
a
little
bit
a
problem
in
the
sdk,
because
amazon
told
us
in
the
few
issues
that
the
x
amazon
empire
expire
time
is
set
automatically
in
the
latest
updates
of
the
sdk.
To
I
don't
know
15
minutes,
but
for
eks
they
lost,
they
lost
it
in
the
latest,
sdk
versions
and
for
the
headers
there's
a
problem.
That's
not
added
so
yeah
and
I
fixed
it
and
now
it's
working
again.
A
Awesome
awesome
thanks
for
this
extensive
write-up
explaining
what
went
wrong
as
well
as
the
fix.
Definitely
a
nice
model
for
folks
to
follow
in
the
community
in
terms
of
really
explaining
clearly
what's
broken,
and
then
what
the
fix
looks
like.
So
definitely
really
appreciate
that,
and
congratulations
again
on
becoming
a
maintainer.
A
For
sure
all
right,
so
we
we
mentioned
christopher,
I'm
not
sure.
I
think
we
also,
let's
just
look
at
our
list
of
maintainers
here.
I
believe
carl
might
have
been
mentioned
on
the
last
community
meeting,
but
carl
is
also
now
a
provider
aws
maintainer,
and
so
definitely
congratulations
to
carl
and
I
believe,
also
on
the
last
community
call,
but
I
want
to
make
sure
we
we
did
mention
this.
We
also
have
gabrielle
is
a
new
maintainer
on
provider,
gcp
so
really
exciting
to
see
new
folks.
A
All
of
these
folks
have
already
contributed
a
ton
to
the
community.
So
we're
super
grateful
for
your
contributions.
Crossland
users
are
super
grateful
and
these
these
positions
are
well
deserved.
So
very
glad
to
have
you
all.
A
All
right
provider,
gcp
updates,
bug,
fix,
can't
delete
bucket
policy
member
that
pertains
to
non-existing
bucket,
and
it
looks
like
this
has
been
fixed.
A
Some
some
good
work
by
folks
I
got
reviewed,
sounds
good.
Is
anyone
on
the
call
familiar
with
what
happened
here
and
yeah.
B
B
Yeah,
actually,
if
you
click
on
the
issue
that
that
this
pr
fixes
it
has
more
details
like
basically
when
you
like,
first
delete
the
bucket
and
then
try
to
delete
the
bucket
policy
policy
member
which
relies
on
that
bucket
object.
Previously,
it
was
just
getting
stuck,
but
with
this
fix
like
it
is
possible
to
like
identify
or
differentiate,
the
errors
is
actually
caused
by
a
non-existing
pocket,
and
in
this
case
it
is
safe
to
delete
the
bucket
policy
member.
So
basically,
this
is
fixing
that
that
problem.
B
A
Awesome
thanks
for
that
description,
hassan
and
thank
you
forehead.
Excuse
me.
If
I
I
mispronounced
that,
but
thank
you
for
your
contribution.
Definitely
much
appreciated.
Do
we
have
a
or
have
we
already
made
a
patch
release
that
includes
this
hassan.
A
C
Yeah,
so
a
new
contributor
cross
playing
contributor,
sergey
aldson,
is
working
on
this
and
he
is
with
us
today.
Maybe
sargan.
Can
you
explain
what
you
are
working
on
yeah
sure?
Actually,
the
mysql
server
configuration
management
server.
I
am
working
on
this
managed
resource
and
actually
the
before
the
postgre
sql
server
configuration
bench
source
was
implemented
by
a
parent.
I
am
working
on
this,
the
similar
topic
and
for
now
this
is
the
draft
pierre.
A
C
And
we
have,
I
think,
spotted
some
problems
also
related
to
the
servers
mysql
server
right,
sargan,
oh
yeah.
Actually
we
realized
that
in
the
evening
this
problem-
actually
it's
about
when,
for
example,
the
configurations
are
the
term
determined
by
the
actually
the
server
side.
C
So
you
cannot
create
a
new
configuration
and
you
can
only
update
the
this
configuration
so
actually,
when
you
try
to
update
an
invalid
configuration
the
server
side,
it
does
not
give
you
an
error
in
the
update
function,
and
so
you
should
before
get
the
configuration
and
if
it
is
not
existing
of
the
server
configuration
site,
you
should
report
them
to
before.
So
we
realized.
We
realized
this
behavior
on
the
server
side,
mysql
server
set
yeah,
and
I
think
we
should
open
an
issue
yeah
to
fix
this.
Yes,
yes,
right.
A
Awesome
thanks
for
that
update,
sounds
good
and
also
welcome.
As
a
new
member
of
the
crossplane
org,
I
saw
that
go
through
yesterday.
So
as
a
reminder
for
folks
who
are
contributing
to
crossplane
and
part
of
the
ecosystem,
you
can
open
up
an
issue
when
you
have
sponsorship
from
two
existing
community
members.
The
whole
process
is
defined
here
in
the
org
repo.
A
So
definitely
take
a
look
at
that
if
you're
contributing
and
you
can
become
a
member
all
right,
it
looks
like
next
up.
We
have
our
section
on
content,
so
there
are
a
few
different
things
here.
A
First,
one
here
jared
and
I
were
on
a
podcast
at
cubecon,
it's
one
of
the
changelog
podcasts,
where
we
generally
kind
of
talked
about
crossplaying,
and
what
the
community's
up
to
and
jared
really
focused
in
on
reaching
incubation
status,
which
is
definitely
a
cool
conversation,
so
feel
free
to
give
that
a
listen
from
our
good
friend
gerhard,
who
is
a
big
cross
playing
advocate?
A
It's
also
this
talk
that
I
believe
was
from
kubecon
here,
and
it
is
talking
about
crossplane
as
well,
so
feel
free
to
give
a
look
at
that.
I
think
it
was
part
of
what
was
mentioned.
A
There
looks
like
victor,
you
might
have
had
a
presentation
at
api
world.
B
Yeah
I
had
last
week
in
api
world
and
also
there
was
devops
experience,
but
recordings
are
not
yet
ready.
So
unless
you
attended
it
live
you'll
have
to
wait.
A
Sounds
good
we'll
make
sure
to
get
that
updated
when
it's
ready
all
right.
I
did
not
add
these
links
here,
but
this
is
a
link
to
my
blog
post.
A
So
you're
welcome
to
take
a
look
at
that,
but
it's
mostly
just
me
talking
about
how
I
think
about
crossline
packages,
so
you
can
give
that
a
read
if
you
like
other
one
once
again,
looks
like
it's
been
added
here,
I'll,
be
speaking
at
the
risk
5
summit
on
december
6th
and
I'll
be
talking
about
open
hardware
for
the
open
cloud.
A
So
basically
talking
about
how
both
the
cloud,
ecosystem
and
processor
ecosystem
have
evolved
over
time
and
how
we're
moving
towards
more
heterogeneous
solutions
and
how
we
need
common
interfaces
to
access
those.
So
if
you're
interested
in
that
kind
of
thing
feel
free
to
take
a
look
at
that,
I
it's.
A
I
know
it's
virtual
friendly,
so
I'm
not
exactly
sure
what
the
pricing
is
on
that,
but
I'm
sure
stuff
will
be
available
on
youtube
afterwards,
all
right,
so
we
are
going
to
move
into
community
interest
items
here,
which
it
looks
like
some
folks
have
added
to
the
list
here.
So
I'm
going
to
go
ahead
and
click
on
them.
I
don't
think
we
have
folks
names
here,
but
if
you
added
it
to
the
list,
this
one
looks
like
excluded.
A
Resources
are
being
included
in
crds,
please
feel
free
to
speak
up,
looks
like
aaron's
on
this
one
here.
D
Yeah,
I
think
aaron
opened
this
issue
and
at
the
end
we
are
blocked
at
the
moment
in
provider
aws
for
all
in
for
all
new
resources
in
ec2,
for
example,
so
for
transit
gateway,
transit
gateway,
attachment
and
routes,
and
hopefully
aaron
can
get
us
a
little
bit
deep
detail.
What
has
happened
under
the
hood,
because
it
looks
very
strange
since
we
merged
the
sdk
upgrades
in
the
version
20
0
20.0.
E
In
most
cases,
the
newer
version
of
the
api
is
the
generated
one
and
it
takes
on
the
storage
version,
so
we're
fine,
but
in
in
some
cases
the
newer
version
or
the
higher
api
is
the
manually
generated
spec,
and
so
it
is
completely
lacking
the
fields
that
a
generated
version
would
would
put
into
the
crd
and
it's
taking
the
storage
version,
meaning
that
the
crd
becomes
invalid.
You
can't
actually
store
the
fields
that
you
need
in
order
to
generate
the
resource.
I
hope
that
was
clarifying.
A
Makes
sense
to
me:
yes,
anyone
have
any
questions
on
that.
D
I
think
movav
added
a
comment
before
the
community
meeting,
so
we
can
read
and
erase
that
we
can
open
a
merge
request
or
pr
in
the
code
generator
he
dropped
the
link.
We
can
have
a
deeper
look
at.
A
All
right,
s3,
late
initialization,
let's
see
who's
on
this
one.
It
looks
like.
A
Cool
well
at
least
it's
reopened,
now.
Apologies
for
not
getting
that
reopened
earlier,
but
definitely
looks
like
something
we
need
to
tackle
here.
If
anyone
is
interested
in
working
on,
this
definitely
feel
free
to
comment
and
I'm
sure
mubafik
or
chris
or
someone
will
come
around
and
make
sure
this
gets
triggered.
Yeah.
F
Thank
you,
that'd
be
great
also.
I
think
we
saw
an
issue
with
the
payment
configuration
but
and
also
on
migration.
We
were
just
trying
to
isolate
issues
but
I'll,
we'll
open
up
a
separate
issue
if
they're
not
related
to
latent,
if
it's
not
related
to
late
initialization,.
A
Cool
sounds
good.
Thank
you
all
right,
let's
see,
what's
up
next
support
patching
of
common
resources,
so
patching
from
common
data
source,
it
looks
like
did
someone
add
this
specifically
to
the
agenda
and
wanted
to
talk
about
it.
I.
F
Did
I
saw
that
nick
had
linked
that
to
another
to
another
proposal,
so
I
was
just
kind
of
curious
as
how
the
conversation
was
going
along.
A
Cool,
I
am
not
familiar
with
this
and
it
looks
like
mubafik
and
nick
are
not
on
the
call,
but
I
believe
there
was
another
contributor,
potentially
yep,
looking
at
a
resource
management
policy
here.
So,
to
be
honest,
oh
it
looks
like
hassan
commented
here.
Hassan,
do
you
wanna
weigh
in.
B
Yep
yeah
yeah.
Basically
this
is
a
design
pr
from
a
community
member
and
it
is
actually
like
proposing
adding
some
some
extra
features
to
provider,
kubernetes
and
nick
actually
commented
and
added
that
these
like
matches
with
the
some
of
the
like
requested
features
in
in
crossplane
but
like
at
the
end
of
the
day.
B
We
we
ended
up
like
to
to
have
them
in
provider
kubernetes
first,
so
that
we
can
use
it
for
a
while
and
see
how
it's
how
it
goes
and
then
we
might
consider
to
like
move
it
to
crossplane
runtime
so
that
all
other
providers
could
leverage.
So
this
is
basically
mentioned,
as
you
can
see
mentioned
here,
but
looks
like
this.
This
will
first
land
in
provider
kubernetes
and
I
couldn't
say
that
all
of
them
could
simply
apply
to
to
crossplane
other
crosstalk
providers.
B
So
yeah,
let's
see
how
it
goes.
A
Cool
thanks
for
that
update
and
also
wanted
to
thank
morningspace
here
for
writing
up
a
design
doc.
That
is
really
awesome
and
definitely
helps
us
make
decisions
faster.
A
Certainly,
if
you
are
have
questions
about
whether
something
is
useful,
you
can
open
an
issue
right
to
make
sure
that
it's
something
we
should
explore
or
not
or
if
there's
any
consensus
on
it,
but
in
general
writing
one
pages
and
design
docs
go
a
long
way
for
helping
us
to
be
able
to
make
big
improvements
like
this.
So
super
cool,
also
great
to
see
that
we
are
kind
of
using
this
provider
as
kind
of
like
an
incubation
place
for
features
in
core
cross
spine.
A
All
right,
composition,
revision,
versioning,.
F
Yes,
so
there
was
a
proposal
to
bring
composition
revision
to
beta
and
there
was
a
breadcrumb.
This
is
the
issue
that
I've
created
and
I
believe
it
was
nick,
nick
or
movabic
that
suggested
a
an
impul
that
would
solve
yep.
I
think
it
was
nick
that
I
think
that
would
solve
our
problem
with
labeling
to
see
if
we
thought
about
that
being
able
to
do
to
use
that
and
implement
it
as
a
way
to
version
and
having
this
done
prior
to
beta
or
as
part
of
beta.
F
We
really
really
love
composition
revision,
but
we
also
manage
lots
of
lots
of
kubernetes
clusters,
so
we
need
to
try
to
keep
a
way
to
keep
the
versioning
consistent
across
clusters
for
packages
for
compositions.
A
Absolutely
absolutely
cool
yeah.
It
sounds
like
from
what
I'm
seeing
in
this
proposal
here
it.
This
makes
a
lot
of
sense
to
me.
I
know
there
is
urgency-
and
this
is
just
kind
of
like
a
general
comment
on
philosophy
of
api
versioning
is
it's
kind
of
like
an
up
or
out
policy
right?
We
don't
want
things
to
stay
at
a
non-uh,
stable
version
which
would
be
alpha
or
potentially
beta
for
for
an
extended
period
of
time.
A
So
I
would
imagine
I
know
nick
is
very
keen
to
get
this
to
v1
beta
1.
So
I
would
imagine
if
this
is
the
direction
that
is
stated
and,
like
I
said
this
looks
pretty
reasonable
to
me,
then
I
I
would
imagine
this
would
happen
fairly
fairly
soon.
Potentially,
in
that
release,
perfect
cool
sounds
great
all
right.
Well,
that
is
the
end
of
the
things
that
were
added
to
the
agenda.
So
the
way
we
usually
do,
this
right
is,
if
you
don't
have
anything
else.
A
This
is
the
formal
conclusion
of
the
community
meeting,
but
if
folks
have
other
topics
they
want
to
bring
up,
then
you're
also
welcome
to
do
that.
But
if
you
need
to
drop
off,
thank
you
for
joining
this
week
and
it's
definitely
a
privilege
to
have
this
many
folks
in
the
community
meeting.
If
you
haven't
added
your
name
to
the
roster
up
here,
definitely
make
sure
to
do
that.
A
Yeah,
let's
see
I'm
going
to
copy
it
into
the
document
here,
just
so
everyone
so
cross
playing.
I
haven't
seen
any
resource
garbage
collector.
Do
you
want
to
give
a
little
more
detail
on
what
you'd
see?
The
purpose
of
the
garbage
collector
in
this
case
is.
G
Yeah
yeah
sure,
that's
that's
actually
the
case
I'm
facing
day
to
day
my
my
job.
So
basically,
when
people
when
people
create
resources
and
cloud
resources,
they
often
forgot
to
shut
them
down,
especially
it's
it's
the
case
and
companies
who
work
on
developing
some
infrastructure
tools.
So
people
just
bring
up
new
clusters,
then
then
then
test.
Something
then
forget
to
shut
them
down.
Clusters
resources
so
basically,
and
they
stay
around-
and
this
is
a
stay
around
and
that's
like
a
company
is
paying
for
that
yeah.
G
So
actually,
I
think
that
that
it
might
be
interesting
to
have
a
feature
which
is
like
maybe
policy,
based
that
we
have
policy,
for
example,
that
by
default
for
a
group
of
developers,
resources
for
example,
clusters
they
live
for
two
days,
then
we
can
extend
the
lease
or
not
so
yeah.
It's
just
general
idea,
and
I
was
I
was.
I
just
started
exploring
crossplane
and,
as
I
mentioned,
I
haven't
seen
any
any
garbage
collector.
Maybe
I
haven't
found
it
yet
and
it's
something
I
would
really
interested
in
working
on.
A
Awesome
yeah.
I
know
I
think
that
is
a
a
great
use
of
the
fact
that
crossplane
is
standardized
on
the
kubernetes
api.
I
don't
think
that
this
is
something
that
we
would
build
into
cross
playing
proper,
but
I
think
it's
something
that
definitely
makes
sense
as
a
part
of
the
community
ecosystem
in
terms
of
just
like
general
policy-based
cleanup.
A
You
know
the
simplest
thing
that
I
could
see
as
a
solution
here
is
literally
just
like
a
cron
job
right
that
just
ran
and
looked
at
resources
and
looked
at
their
creation
time
or
something
like
that
and
delete
them.
Based
on
that,
I
wouldn't
be
surprised
if
there
wasn't
already
a
project
somewhere
in
the
kubernetes
ecosystem.
A
That
did
this
and
probably
allows
you
to
configure
any
type
of
resource,
and,
if
that
you
know
if
it
runs
for
deployments
or
things
like
that,
it
could
also
work
for
kubernetes
objects,
so
that
might
be
one
route
to
look
into
also,
if
this
is
something
that
you're
interested
in
working
on,
you
know
in
the
cross
plane
community
or
something
like
that.
I'm
sure
we
could.
You
know
help
get
that
set
up
and
and
give
you
anything.
You
need
to
get
started
on
that.
G
Yeah
yeah
so
yeah
I
was
thinking
so
the
the
cro.
The
current
job
is
not
that
easy,
because
I
have
already
started
working
this
project,
but
it
is
a
separate
project
and
I
I
thought
it
can
be
as
easy
as
cruncher,
but
apparently
it's
not-
and
so
I
was
thinking
maybe
about
some
pluggable
feature-
maybe
not
that
maybe
not
a
part
of
a
crossplan
engine
but
but
but
maybe
plug-in
that.
So
I
think
it's
this.
This
topic
was
a
separate
discussion.
G
Maybe
I
just
I'll
just
keep
exploring
keep
thinking.
Maybe
I
can
bring
it
up
next
time.
A
Awesome
yeah.
That
sounds
great.
I
definitely
think
that
would
be
useful
for
a
lot
of
folks.
I
could
even
see
this
almost
being
a
provider
type
of
you
know
package
that
you
could
install.
That
would
do
this,
for
you.
G
Yeah
yeah,
that's
that's.
I
think
it
fits
the
provider,
the
yeah
it's
gonna
work
as
provided,
but,
as
I
said,
I
don't
have
much
details
yet.
So
I
just
wanted
to
check
if
you
guys
thinking
about
something
like
that
or
working
on.
B
G
G
Actually,
that's
that's
my
that
was
my
second
thought
because
in
a
project
I'm
working
on,
I
hit
the
so
I
started
with
pretty
simple
policies
like
yaml
defined
policies,
but
I
hit
the
ceiling
pretty
fast
because
many
conditions
we
have
this
group,
this
growth
and
this
condition
for
this
group
and
something
else
so
I
think
oppa
will
be
great
to
use,
and
I
know
you
have
you
guys
talked
about
integration
with
opa,
but
I
don't
know
have
you
done
something
about
it
or
not?
How
you?
G
What
what
progress
you
you
have
with
this
with
open
integration?
So
that's
something
again.
I
need
to
check
so
yeah
some
policy
based
and
that
I
was
also
thinking
about.
If
we
have
something
like
lease,
so
the
developers
or
users
can
use
resources
and
they
have
lease
period
and
if
lee's
expired
and
the
developers
don't
don't
want
to
extend
the
lease,
the
cost
plan
will
clean
it
up.
G
Just
because
the
lease
expired
doesn't
nobody
wants
to
use
those
resources?
Yeah,
yes,
and
I
actually.
B
B
No,
I
was
about
to
say,
if
you
open
an
issue
or
something
like
that,
and
you
see
me
that
the
issue
I
can,
I
can
probably
bring
people
from
from
kyberno
and
the
gate
and
the
api
gate.
Sorry
and
the
opi
opa
gateway
into
the
conversation
and
see
whether
that
would
make
sense
some
kind
of
collaboration
with
them.
G
Yeah,
that's,
I
actually
have
worked
with
students,
so
I
can
talk
to
him.
Also.
G
Okay,
yeah
sounds
good
yeah,
so
if
you
I
don't
have
anything
like
that.
So
I
I'm
gonna
keep
exploring.
F
Me
personally,
we
nuke
our
clusters
are
like
for
any
for
developers
every
week
and
we
just
started
scaling
down
every
night,
our
clusters,
so
that,
yes,
that
you
don't
do
a
lot
of
resource
consumption.
G
Yes,
that's
that's
what
called
custodian,
for
example
does,
but
it
doesn't
always
work
because
sometimes
to
it,
it
takes
time
for
developers
to
bring
up
and
and
configure
cluster
right,
and
then
they
don't
want
to
do
it
every
night
and
some
resources,
even
they.
If
they
shut
down,
they
still
eat
money.
G
Yeah
and
many
companies,
so
I
talked
to
to
two
guys
from
from
startups
and
so-
and
I
think
many
guys
from
startups,
especially
who,
like
b2c
series
when
they
don't
have
enough
resources
to
spend
on
watching
resources
they
they
hit
this
issue
on
a
regular
basis,
because
many,
like
bigger
companies
like
bigger
companies,
they
have
the
internal
tools
to
watch
the
resources,
but
but
smaller
companies
they
don't,
and
I
think
it's
it's
it's
good
to
have
something
like
butcher
like
this.
A
Absolutely
well,
thanks
for
bringing
that
up
and
and
victor
made
a
great
point
if
you
want
to
open
an
issue,
I'm
sure
we
can
get
some
folks
commenting
on
that.
So
thanks
for
bringing
that
up
cool
does.
Does
anyone
else
have
any
topics
they
want
to
cover
before
we
wrap
it
up.
A
All
right,
well,
as
always
slack
channel,
is
open
for
any
questions
or
comments
or
thoughts,
so
definitely
feel
free
to
join
us
there.
If
you're
not
already-
and
thank
you
for
taking
time
out
of
your
week
to
be
here,
we
definitely
appreciate
folks
participating
and
bringing
their
thoughts
and
ideas.
So
have
a
good
rest.
Your
week
and
a
a
good
weekend.