►
From YouTube: OKD Community Development Meeting - 11-29-2022
Description
The OKD Working Group's purpose is to discuss, give guidance to, and enable collaboration on current development efforts for OKD, Kubernetes, and related CNCF projects. The OKD Working Group includes the discussion of shared community goals for OKD 4 and beyond. Additionally, the Working Group produces supporting materials and best practices for end-users and provides guidance and coordination for CNCF projects working within the SIG's scope.
https://okd.io
A
B
Development
meeting
for
November
29th
2022
and
take
a
quick
look
at
the
agenda,
see
if
there's
anything
that
has
been
missed
or
that
you'd
like
to
add
or
move
around
for
any
reason.
Let's
take
30
seconds
and
look
at
that
and-
and
let
me
know
if
there's
anything
that
you'd
like
to
change.
C
Okay,
so
I
haven't
heard
anything
about
the
Sno
I
know
you're
involved
in
that
Luigi
I
didn't
whether
you've
got
an
update,
yeah.
D
Yeah
absolutely
I
met
this
morning
with
Benjamin
Branch.
He
calls
himself
Dwayne
for
so
I
was
a
bit
confused
because
all
these
emails
is
Benjamin
brunch
and
then
he
says
this
is
Dwayne,
you
know,
and
so
anyway,
so
we
did,
we
did.
We
did
meet
this
morning,
so
we
have
a
plan
forward
on
Friday
we're
going
to
be
setting
up
the
initial
environment.
D
You
know
the
going
through
a
document.
I've
created
work
through
it
and
then
once
we've
got
the
whole
sort
of
setup,
a
worked
out
and
an
installation
of
okay
be
working
up
and
running
in
single
node
using
the
upis
methodology.
D
We
would
then
create
a
video,
so
I
mean
I've
already
started
on
the
introduction
of
the
video,
and
you
know
I've
created
a
set
of
slides,
so
it's
just
getting
Dwayne
up
to
speed.
You
know
he
says
he's
there's
some
things
he's
not
sure
about.
So
we're
going
to
go
through
that
and
so
yeah
we're
making
we're
making
progress
there.
So
that's
that's
great
he's
he's
fairly
available.
D
B
B
We
can
see
you,
but
we
can't
hear
you,
but
we
we
see
that
you're
walking.
So
okay
I
put
a
little
note
there
that
I
have
started
working
on
the
the
tecton
operator
install
so
just
to
reiterate:
I'm
going
to
start
with
just
the
basics
and
then,
after
our
first
version
of
the
blog
post,
I'll
move
on
to
doing
the
open
shift,
modifications
and
including
those
there
is
a
separate
repo
that
has
those
openshift
editions,
I,
think
they're
called
or
something
like
that.
Yeah.
B
C
B
It
should
be,
this
I
think
it's
the
same
source
and
then
they
they
flip
I'll,
also
look
at
the
repo
again
to
see
what
it
is,
but
you
literally
use
a
flag
to
have
it
installed.
The
open
shift
extras,
basically
and
interestingly
enough,
when
I
asked
about
I
reached
out
through
Red
Hat
person
who
reached
out
to
like
one
of
the
product,
managers
and
I
said
hey,
you
know
it
would
be
nice
to
have
you
know
the
the
pipelines
operator
available.
B
I
I
need
to
verify
that
actually,
with
my
hands
before
I
commit
to
that
being
the
case,
though,
right:
okay,
anything
else
about
technical
documentation.
C
Just
to
say
that
we've
had
quite
a
lot
of
inputs
into
the
the
tutorials
on
the.
Let
me
put
it
there
and
paste.
The
discussion
thread
is
just
put
it
in
the
chat
on
all
the
tutorials
there
is.
There
is
a
lot.
There
are
a
lot
of
links
put
in
that
discussion
forum,
so
yeah
I
mean
I,
I
guess.
C
The
question
is:
where
do
we
go
from
here
in
terms
of
because
I
I
notice
again,
there's
been
some
more
chat
this
week
where
people
are
looking
at
four
nine
and
back
level
versions,
because
they're
following
tutorials
and
instead
of
jumping
to
the
latest,
there's
also
been
some
questions
about?
Well,
why
aren't
we
supporting
back
level
versions
the
assumption
that
we're
following
the
ocp
model
and
applying
fixes
to
back
level
versions,
so
I
I
think
there
is,
although
these
are
really
really
useful
and
that's
how
I
got
started
by
finding
an
old
one?
C
C
I
mean
there's
a
lot
there,
I
mean
if
you
look
through
and
we've
got
the
age
of
them
as
well.
So
there's
quite
a
lot
from
one
or
two
years
ago,
there's
quite
a
few
from
three
four
years
ago,
so
yeah
I
think
we've
we've
got
enough
to
make
a
good
start.
There.
E
C
I
mean
one
thing:
I
was
going
to
ask
Jamie
is:
do
you
have
the
results
of
this
survey?
Do
we
know
what
platforms
everyone's
using
because
I
think
that
would
give
us
a
clue
as
to
where
we
should
be
focusing
our
effort,
because
with
sort
of
openstack
over
their
metal
VMware,
there's
a
and
then
we
get
into
the
cloud
providers?
There's
a
lot
of
different
potential,
different
options
and
I.
Don't
think
we've
got
the
bandwidth
to
do
them
all
Justice,
so
it'll
be
useful
to
think
to
actually
know
do
from
the
survey.
B
Yeah
we
so
Mike
and
I
actually
talked
about
this
when
we
presented
at
kubecon
and
I,
haven't
massaged
all
the
data
yet.
But
if
we
I
don't
have
the
most
recent
download
of
the
data
right
now,
but
as
of
the
one
I
pulled
in
September,
it
was
a
surprising
amount
of
410
and
I.
B
Don't
know
if
this
is
like
self-selection
that
the
people
who
bothered
to
do
it,
but
a
lot
of
410,
as
of
some
of
these
responses,
are
July,
August,
September
and
like
two
four
sixes
and-
and
this
was
out
of
like
40,
some
odd
responses
so
but.
C
B
Well,
it's
a
select
releases
can
right,
so
select
releases
have
edges,
that'll
work
with
it,
and
if
you
look
at
the
the
output
of
the
pipelines,
you
can
see
that
where
they
are
and
which
ones
you
can
transition
to
the
next
minor
version
up.
A
B
B
I
think
was
from
four
eight
to
four
nine
or
something
like
that
where
we
struggled
for
a
while
for
folks
to
be
able
to
go
from,
or
maybe
it
was
14,
it
was
four
eight
to
four
nine
or
four
nine
to
four
ten,
where
folks
really
struggled
and
we
didn't
have
that
release
to
be
able
to
go
up
to
the
next
minor
version.
B
Vmware
are
yeah,
let
me
let's
see
here,
how
did
we
do
it
because.
C
F
B
Yeah,
let
me
let
me
let
me
put
this
in
I,
put
it
in
a
Google
spreadsheet
and
shared
it
out
to
the
folks
like
a
while
ago,
but
it's
not
the
most
up-to-date.
So
let
me
pull
the
data
again
and
put
it
into
a
spreadsheet
yeah,
because
I'm
not
seeing
off
the
top
of
my
head
anything
that
I
can
tease
out
just
while
we're
talking.
We,
you
know
what
we
didn't
actually
ask
that.
B
B
Actually
hold
on
no
actually
I'm
I'm
wrong.
We
did
yeah,
we
did
it's.
It's
been
so
long
now,
yeah,
interestingly
enough,
the
majority
are
bare
metal,
so
it's
a
bare
metal,
then
vsphere,
then
AWS
and
then
a
smattering
of
like
open,
stack
and
overt
and
stuff
like
that.
So
bare
metal
has
the
most
then
vsphere
than
AWS
in
that
order.
B
B
Yeah
I
did
a
okd
install
and
just
a
base
install
ends
up
being
like
it
was
like
358,
or
something
like
that.
If
you
didn't
like
reserve
your
instances
now,
if
you
reserve
them,
obviously
you
could
get
that
price
down.
Maybe
you
know,
but
it's
it's
pretty
pricey
US
Dollars,
so
yeah
I
mean
let's,
let's
look
at
that
and
then
like.
B
So
what
do
we
really
want
to
tease
out
I
mean
if
we
look
at
what
we
have
here
bare
metal,
which
I
think
we
kind
of
already
knew
that
right,
like
bare
metal,
is
the
thing
that
people
really
want
the
help
with,
and
it
was
the
one
that
about
two
years
ago,
when
I
first
got
involved,
there
was
some
hesitance
to
do
a
lot
of
documentation
for
bare
metal
because
it
was
so
unique
and
each
install
was
unique
and
whatever,
but
I
think
we've
sort
of
moved
away
from
that
philosophy.
B
G
Hey
this
is
this:
is
back
Charming
in
so
I
I
just
wanted
to
mention
that,
just
because
it's
like
bare
metal
doesn't
necessarily
mean
you're.
Actually,
installing
on
on
Hardware
right
I
mean
you
can.
This
is
actually
what
what
we
are,
what
we're
doing
at
CERN,
so
we
are
actually
just
installing
on
our
own
VMS
and
then
kind
of
deploying
our
own
cloud
controller
on
top,
so
it
kind
of
like
open,
openshift
things.
G
It's
a
bare
metal,
but
it's
it's
actually
not
and
I
think
if
people
might
be
doing
this
in
their
home
Labs
or
something
like
this,
where
maybe
also
just
like
using
some
nodes
and
or
you
have
like
a
VMware
or
whatever,
then
it's
also
going
to
look
like
a
bare
metal
or
you're.
Gonna
select
bare
metal
in
the
survey,
but
well
it
kind
of
is.
But
you
know
it's
it's
a
little
bit
misleading.
G
So
maybe,
if
we
want
to
focus
on
that
aspect,
then
like
pulling
out
some
interesting
bits
like
okay,
if
you
didn't
Deploy
on
on
one
of
the
quote-unquote
supported
platforms,
how
can
you
still
automatically
provision
new
machines
and
put
them
into
your
cluster,
or
how
can
you
still
automatically
provision
storage-
something
like
this,
like
kind
of
as
a
extension
tutorials
like
how
to
extend
your
own
okd,
because
of
course,
if
you,
if
you
just
do
the
regular
deployment
that
you
just
described
on
AWS,
for
example,
you're
gonna
get
everything
out
of
the
box,
but
then
yeah.
B
Yeah
I
think
that's
a
great
idea
if
we
can
tease
out
a
little
bit
the
specifics
of
those
and
then
write
stuff
up
or
direct
people
to
the
specifics
and
whatnot
yeah.
C
H
No
worries
it's
a
good
discussion.
H
I
think
we
have
to
be
a
little
careful
here
in
terms
of
like
the
language
we're
using
when
we
talk
about
bare
metal
versus
kind
of
like
what
I
would
consider
just
a
a
UPI
install
like
yes,
UPI
install
is
kind
of
like
a
bespoke
installation
right,
you're
installing
it
on
whatever
custom
machines
you've
created,
but
there
is
also
what
we
call
bare
metal
in
terms
of
a
platform
type
and
bare
metal
in
terms
of
a
platform
type
does
allow
some
interaction
with
like
some
of
the
clustered
life
cycle
kind
of
stuff,
where
you're
you're,
defining
pools
of
bare
metal
machines
that
could
be
possible
for
growing
the
cluster
and
shrinking
the
cluster,
and
so
I
think.
H
My
personal
feeling
is
that,
when
we're
talking
to
our
users,
we
should
probably
be
careful
to
kind
of
first
talk
about
the
distinction
between.
Are
you
just
trying
to
do
a
bespoke
installation
on
the
hardware
you
have,
and
so
you'd
like
to
have
openshift
running,
or
are
you
trying
to
do
something
that
looks
more
like
what
we
consider?
H
You
know,
like
kind
of
a
cloud-based
installation
where
you
want
to
be
able
to
use
openshift
to
manage
the
underlying
infrastructure,
because
then
that
you
know
bare
metal
could
be
included
in
that
as
well.
The
official
bare
metal
platform
and
so
I
think
like
if
we're
just
telling
users
okay,
you
have
a
handful
of
machines.
Here's
how
you
install
it
on
the
machines,
you
control,
that's
one
really
well-known
pathway
and
probably
there's
not
a
lot
of
commonality
there.
H
But
if
we
start
to
talk
about
bare
metal
as
the
infrastructure
platform,
I
think
that's
where
it
gets
a
lot
more
complicated
because
configuring
a
bare
metal
like
data
center,
to
look
like
an
AWS
installation.
It
is
a
more
complicated
task
and
so
I
think
you
know
defining
documentation
for
just
doing
a
normal
UPI
install,
that's
good
and
then
also
talking
about
this
other
distinction
where
it's
like.
Okay,
if
you
have
infrastructure
like
vsphere
or
something
like
that,
then
we
should
start
to
look
at
okay.
H
What's
the
most
popular
like
infrastructure
deployment,
you
know
like
like
I
commented
earlier,
like
you
know,
vsphere
is
very
popular
when
you
look
at
the
Telemetry
around
openstack
or
openshift,
you
know
so
it's
like
okay.
Well,
that
might
be
a
place
to
focus
on
if
we,
if
we
can
determine
that
our
community
is
really
into
that
as
well.
Maybe
that's
a
good
place
to
start,
but
I
think
it's
a
it's
a
distinction.
We
should
probably
make
you
know
for
the
users
so.
C
B
I
think
Dwayne
had
a
good
point,
which
is
really
a
glossary,
would
be
helpful
for
this
because
we
there's
this
conversation
about
what
is
bare
metal
keeps
coming
up
right
and
our
users
might
not
know,
and
we
ourselves
may
not
really
know
what
we
mean
with
this.
So
having
a
little
something
about
bare
metal
and-
and
it's
different
interpretations
might
be
kind
of
helpful
somewhere
and
are
you
volunteering
Dwayne
is
that
is
that
what
I'm,
seeing
okay.
C
C
E
Volunteer
for
the
glossary,
because
you
learn
a
lot
getting
to
being
able
to
distinguish
the
terms.
That's
that's
great
learning.
A
F
C
Want
to
sort
of
get
so
you
can
actually
go
and
I
don't
know
whether
you've
actually
done
anything
on
OK
data
IO.
But
if
you
want
me
to
go
through
how
that
site's
managed
and
how
you
add
content
up
to
do
that,
we
can
just
ping
me
on
on
slack
or
something
and
set
some
time
up.
Okay.
Yes,
we
can
do
that.
F
So
one
piece
that
I
needed
to
understand,
or
rather
loop
back
to
your
documentation
issues,
is
that
when
I
go
back
through
and
look
at
stuff,
one
case
where
I
was
having
problems
with
node
ports
trying
to
troubleshoot
that
all
it
said.
F
Oh,
you
have
to
have
a
red
hat
account
and
you
have
to
have
full
requests
for
it
and
the
like
and
that
I'm
wondering
why
that
Keeps
On
Slipping
through
or
if
we
end
up
copying
stuff
from
the
red
hat
release
to
the
next
okd
release
and
it
never
we're
continuing
to
have
to
go
back
and
update
those
things.
B
Yeah
I
think
that
this
is
a
it's
a
little
bit
of
of
several
things,
so
one
of
it
is
copying
stuff
over
directly
from
the
ocp
documentation.
The
other
is
there
are
some
resources
that
are
not
accessible.
If
someone
used
the
the
dummy
pull
Secret
and
the
dummy
pull
secret
was
suggested,
and
then
the
idea
was
sort
of
backtracked
and
they
came
up
with
an
actual
I
think
like
a
new
pull
secret,
that
actually
sort
of
works
for
ookd
as
opposed
to
just
putting
something
random
in
there.
B
So
I
think
that's
part
of
it
and
then
the
other
part
of
it
is
that
a
third
component
of
that
I
think.
Is
that
folks,
we
don't
really
have
we
don't
really
have
the
the
okd
documentation
at
the
level
of
things
like
node
port
and
some
of
the
networking
and
other
stuff,
like
we've,
really
always
just
sort
of
copied
over
from
the
ocp
documentation
and
I,
don't
know
that
we
can
get
into
that
level,
because
that's
that
changes
a
lot
right
or
could
change
a
lot
and
we'd
have
to
maintain
those
docs.
B
F
Now
the
issue
that
I
was
having
is
that
I
had
an
ocp
install
and
an
okd
install
of
the
same
genre,
four
nine
and
on
ocp
the
Node
ports
work
on
okd.
The
Node
ports
did
not
work
and
through
the
VIP
Ingress
VIP.
F
B
F
I
finally
got
around
to
not
putting
in
the
doc
request
on
that,
but
I
never
saw
anything
feedback
on
that,
but
it's
still
there's
random
spots
that
I
see
matter
of
fact,
you're
talking
about
the
the
fake
pull
request,
pull
Secret,
that's
still
in
the
docks
yeah.
F
So
there's
lots
and
lots
of
places
that
things
are
still
out
of
sync
with
for
people
that
are
doing
this,
all
the
time
versus
people
that
are
just
coming
to
it.
A
Well,
but
I
think
the
fake
full
secret
actually
is
accurate.
It's
just
that
before
it
was
a
random.
H
D
A
Has
to
be,
you
know.
B
C
Know
yeah
so
so,
just
to
help
Larry
I
put
a
link
in
the
chat.
There
is
a
sort
of
page
on
the
okd,
which
is
how
we
try
and
get
the
the
official
okay
docs
updated.
So
it
is
pulling
in
a
request.
We
need
to
update
this
page
because
it's
still
got
Diane,
but
we
need
to
copy
Michael
Burke.
F
Github
page
specifically
or
the
GitHub
site
for
the
docs.
F
B
F
Just
yeah
side
note:
it's
interesting
that
once
I
had
my
all
sorts
of
issues
around
my
stuff,
four,
nine,
not
upgrading
after
a
fresh
install
because
all
all
sorts
of
pieces
were
saying:
oh,
these
have
been
modified.
This
is
what
I
didn't
modify
anything
the
other
piece
that
is
really
strange
is
that
I've
torn
things
down
and
then
now
I
can't
get
it
to
reinstall.
It
just
blows
up
at
bootstrap
and
I'm
now
struggling
to
figure
that
out.
B
A
B
Yeah,
that's
that's.
I
I
wrote
a
script
that
cleans
it
but
yeah.
If
you
just
delete
the
whole
directory
that
you
were
doing
your
install
from
because
there's
hidden
files
that
contain
certificates,
a
log
yeah,
exactly
they're,
hidden
dot
files
as
checkpoint.
So
so.
B
F
B
That's
a
good
point
that
install
State
openshift
underscore
install
underscore
State
file.
That's
the
one!
That's
really
the
the
the
kicker
in
all
of
this
Okay
Okay
cool,
so
yeah
so
put
some
in
the
channel.
If
you
want
us
to
to
Hash
that
out
more
happy
to
help
you
out
on
that,
and-
and
we
should
put
a
note
in
for
documentation
on
that-
one.
F
D
B
B
C
B
Okay
and
that's
another
one
I
would
add,
is
someone
asked
earlier
today
about
the
Street
about
selecting
or
just
adding
one
of
the
other
update
streams,
the
fast
stream?
And
so
we
had
to
explain
no
there's
only
one
per
minor
version,
one
stream,
and
that
would
be
a
good
FAQ
to
add
as
well
I
think
now
Jack
go
ahead.
G
It's
totally
fine
that
okay,
he
doesn't
like
I
I,
really
understand
that
it's
not
really
feasible,
not
easily
feasible
with
all
of
the
testing
that
is
required,
but
it
just
should
just
be
pointed
out
very
clearly
that
when
people
start
using
it
that
they
don't
expect
to
get
similar
updates
to
what
they
would
expect
or
what
they
could
get
when
they're
using,
let's
say
I,
don't
know,
send
us
8
or
send
us
streamline
or
whatever,
because
there
you
could
still
get
patch
updates
updates,
which
you
don't
get
on
okd
and
it's
absolutely
not
clear
unless
you
want
to
kind
of
parse
the
release,
the
releases
from
the
from
the
GitHub
that
that's
the
case.
B
Yeah,
that's
that's
a
great
point
and
if,
if
Christian
was
here,
I
would
ask
if
they
feel
that,
with
the
new
infrastructure
we
might
be
able
to
start
providing
patch
updates
to
previous
versions.
That
might
be
a
question
for
next
week's
meeting,
with
all
of
the
new
infrastructure
and
and
the
ease
with
which
things
are
getting
built.
Now,
maybe
there
will
be
an
opportunity
for
that,
but
I
don't
know.
We'd
have
to
ask
yeah.
G
I
mean
I
mean
infrastructure
and
testing
is
one
thing,
but
then
you
also
kind
of
need
to
support
it
right.
So
if
someone
comes
along
with
like
a
4
8
cluster
and
something
is
broken
and
all
of
us
so
to
speak-
are
already
using
for
10
or
for
11
or
412
with
I,
don't
know
S
course
or
whatever.
G
Then
no
one's
really
going
to
be
able
to
help
them
except
say:
well,
have
a
look
at
the
docs
or
try
this
or
try
that
so
I
think,
it's
perfectly
fine
to
say:
hey
we're
just
going
to
support
one
One
release
at
a
time.
I
think
it's
a
really
reasonable
strategy,
especially
because
the
people
who
are
choosing
okd
I
think
they
are
somewhat
willing
to
to
to
be
on
there
to
be
on
their
own
anyway
or
to
customize
in
any
way.
B
It
might
be
helpful,
then,
to
have
a
very
clear
upgrade
document
like
we've
got,
there's
some
stuff
on
the
that
was
on
the
front
page
of
the
repo
and
I.
Don't
know
if
it's
still
there,
but
I,
don't
think
we
turned
it
into
a
dock.
That's
basically
like
this
is
how
you
pull
down
the
latest
tools.
You
know
for
a
release
for
the
next
release,
and
this
is
how
you
upgrade,
or
this
is
how
you
change.
You
know
your
stream
Etc,
so
it'd
be
nice
to
actually
have
like.
B
Okay,
we're
on
411
now
to
get
to
411.
Here
are
the
steps
that
you
can
follow
to
to
get
there,
and
here's
here
are
the
versions
that
you
need
to
to
be
able
to
do
that,
to
sort
of
tease
that
out,
we've
sort
of
relied
on
people
going
to
the
the
testing
pages
and
the
and
the
the
CI
pages
to
sort
of
see
where
those
edges
are
and
stuff
like
that,
but
it'd
be
nice.
B
If
we
provided
that
or
highlighted
it
for
folks
I
mean
we
have
to
keep
it
updated,
I'd
be
happy
to
volunteer
to
do
that.
But
I
think
that
would
help
our
public
interactions.
A
lot
is
for
people
to
see
like
okay
you're
limited
in
this
way,
but
this
is
how
you
get
to
the
next
version.
In
these
simple
steps.
Yeah.
C
We
are
very
good,
yeah
I
was
going
to
say
yeah.
We
do
need
to
go
and
change
the
documentation
because
I'm
just
looking
now
and
of
getting
clusters
on
okd
Dot
docs.okd.io
and
it's
got
candidate,
Channel
fast,
Channel,
stable
channel.
So
right
this
is
the
ocp
stuff.
So
you
do
need
to
go
in
and
get
this
updated,
because
this
is
the
red
hat,
ocp
yeah.
It
doesn't
apply
to
a
PD,
so
yeah.
We
need
to
get
that
removed
all.
B
Right
we'll
get
a
ticket
into
Michael,
he
did
show
up,
but
then
for
folks
that
didn't
catch
it
two
weeks
ago
he
did
show
up
at
the
end
of
the
of
the
meeting,
but
it
was
was
it
or
was
it
last
week
he
did
show
up,
but
it
was
like
an
hour
late.
He
showed
up
at
one
instead
of
noon.
G
Yeah
I
just
wanted
to
to
like
kind
of
bring
this
back
to
a
topic
we
had
before
about
the
Telemetry
or
the
survey.
Do
we
have
any
data
on
on
who
is
actually
using
the
official
channel
that
is
provided
by
okd
versus
who
is
kind
of
rolling
their
own
in
terms
of
releases
or
cluster
versions?
G
Because
because
we
do
like
we,
we
never
used
any
release,
Channel
we're
just
building
our
own
okd
versions
and
then
just
upgrading
from
point
to
point,
basically,
which
might
be
totally
weird
but
I,
don't
know,
maybe
also
a
lot
of
other
people
are
doing
it
that
don't
don't
really
have
any
insight
there.
B
One
thing
that
I
do
know
and-
and
you
know
Amico
you
might
be
able
to
speak
to
this
as
well-
is
that
the
Telemetry
data
Vadim
had
access
to
it
and
could
give
us
sort
of
generalized
well.
This
is
what
I've
seen,
but
we
can't
make
it
public
or
we
don't
have
access
to
it
as
a
group
as
of
right
now,
I
don't
know
amigo.
Do
you
got
any
insight
into
that
as
well?.
H
Yeah
I
mean
this
is
kind
of
why
I
was
being
a
little
delicate
about
how
I
was
talking
about
the
Telemetry
before
I.
Think
I.
Think
vadim's
approach
is
probably
right,
like
I'm
I'm,
probably
not
allowed,
or
you
know,
Luigi
and
I
aren't
allowed
to
talk
about
the
actual
numbers
in
the
Telemetry,
but
maybe
we
could
talk
about
Trends
or
something
so
I
and.
H
Part
about
this
too,
especially
to
Jack's
question.
You
know
we're
still
working
towards
the
point
where
we're
gonna
give
more
like
I've
talked
about
this
install
flexibility
effort,
but
in
the
past,
like
we're
working
towards
providing
more
opportunities
for
users
to
be
able
to
customize
how
openshift
is
deployed,
especially
with
respect
to
the
you
know,
to
these
like
infrastructure,
specific
components
right
now,
though,
we
don't
have
any
sort
of
telemetry.
That
shows
us
when
people
have
done
those
kind
of
customizations
or
whatnot.
So
we
really
don't.
H
You
know,
there's
really
no
way
to
detect
right
now
to
say
like,
oh,
you
know,
someone
is
running
a
cluster
and
they've
added
some
sort
of
customization
that
will
persist
through
upgrades
I.
Think
we're
starting
to
see
some
of
that
with
like
the
olm
and
people
who
install
things
through
through
the
catalog,
and
we
can
see
those
catalog
upgrades,
but
the
type
of
customizations
that
I
know
you're
doing
at
CERN
and
probably
other
people
are
curious
about
too.
H
G
G
This
type
of
data
like
hey,
are
you
using
the
the
okd
stable
Channel,
or
do
you
do
something
else,
just
as
a
general
sentiment
like
I
totally
understand
that,
like
we
don't
want
to
have
all
of
the
glory
details
of
the
of
the
Telemetry,
especially
because
I'm
not
sure
on
how
many
okd
installations
that
actually
works,
because
it
doesn't
on
hours
and
I've,
never
spent
any
time
figuring
out
why
I
thought
it
was
just
because
we're
not
using
ocp?
To
be
honest,.
B
Jack
can
I
handle
that
to
you
as
a
task.
Can
you
write
up
a
question
that
that
asks
simply
what
it
is
that
you're
getting
at
and
then
we'll
throw
it
into
the
to
the
form
for
the
survey.
B
B
It's
a
little
bit.
Frustrating
I,
I
realized
last
week
might
have
not
been
a
good
week
for
U.S
folks,
but
osbo
also
has
some
non-us
folks,
so
I
was
kind
of
hoping
for
a
response.
So
let
me
email
them
once
again
this
week
and
then,
if
we
don't
hear
anything
in
two
weeks
from
now
at
our
next
meeting,
maybe
we
got
to
think
about
Alternatives
because
it
was
nice
for
them
to
to
initially
say
that
they
were
going
to
help
us
out.
B
B
Will
yeah,
okay,
well
I
think
the
goal
was
to
to
shoot
for
after
December,
because
a
lot
of
folks
are
going
to
be
out
starting
next
week
because
of
the
Red
Hat
holiday?
What
is
you
guys
get
like
a
whole
month
or
something
like
that?
Isn't
it
or
like
at
least
a
couple
weeks
of
like.
B
Okay,
so
yes,
and
we
will
invite
I,
think
for
for
January
or
do
folks
think
it's
okay
to
to
invite
in
December.
Some
folks
are
really
busy
in
December,
but
if,
if
folks
want
to
do
that,
invite-
and
have
it
be
maybe
the
next
meeting,
because
we
only
have
two
in
December
right-
we've
got
the
13th.
Actually
we
well.
Actually
we
should
answer
this
now.
Do
we
want
to
have
a
meeting
on
the
27th,
because
if
not,
then
we
only
have
one
meeting
next
month.
B
H
A
B
Yeah
all
right:
well,
let's
shoot
then
for
we'll
do
the
invite
for
the?
What
would
that
be?
That
would
be
so
we're
skipping.
We
would
skip
the
27th,
so
it'd
be
the
10th.
Let's
invite
her
to
the
meeting
on
January
10th.
Does
that
sound
good
or
just
in
January,
one
of
the
January
ones,
starting
on
January
10th?
That's,
probably
good:
okay,
oh
okay,
social
media
I've
been
tweeting
about
releases
and
some
of
the
videos
I'll
tweet
more
of
the
video
stuff.
D
I
was
chatting
to
just
before
this
meeting.
I
was
just
catching
up
and
syncing
up
with
Diane,
and
she
said
we
should
bring
it
up
because
it
sounds
very
interesting,
maybe
set
up
in
a
candle
I
think
it.
The
the
more
important
one
she
mentioned
is
getting
the
the
email
Alias
so
that
we
could
have.
D
You
know
some
some
form
of
email
to
0.24
for
control
of
Mastodon
and
maybe
even
Twitter.
So
yeah
from
my
my
my
perspective,
I
think
it's
a
good
idea.
I,
don't
know
what
the
rest
of
the
guys
think.
B
B
Know
no
one
really
investigated
if
red
hat
would
be
cool
with
that
or
not
like
that's
the
other
part
of
this
is
like
okay,
we
create
a
whole
new
domain
and
email
World
aside
from
okd.
Is
that
cool
with
them
or
not?
You
know
I
mean
it
is
a
community
project,
but
it's
also
a
community
project
supported
by
Red
Hat,
pretty
heavily
so
I.
H
C
H
You
know
because,
like
we
can
keep
like
I'm
sure,
Luigi
and
I
can
keep
bugging
people
in
the
hospital
to
see.
If
they'll,
like
answer
the
question
here,
but
that's
you
know,
it
doesn't
really
give
us
the
power
to
do
what
you
know
doesn't
give
the
community
the
power
to
kind
of
like
choose
their
own
destiny
right.
B
You
know
elsewhere
within
Red
Hat
to
make
sure
that
this
isn't
is
going
to
conflict
with
such
and
such
and
then
they
never
got
back
on
that
part
of
it,
so
they
actually
offered
which
kind
of
made
this
complicated,
because
if
they
just
said
no,
we
don't
know
or
whatever,
but
they
offered
and
said
they
would
do
it
and
then
they
wanted
to
check
something
and
then
they
just
never
got
back
so
yeah
I.
Think,
let's
at
the
next
meeting,
we'll
come
up
with
a
plan
with
how
to
move
forward.
B
H
Or
you
can
so
the
thing
to
know
about
Mastodon?
Is
that
like
it's?
You
know
it's
decentralized,
so,
like
there's
a
lot
of
different
servers,
so
the
first
thing
we'll
have
to
do
is
kind
of
find
a
server
that
that
we,
you
know
the
community
is
happy
with,
and
there
are
some
there
are
some
like
open
source
based
servers
that
are
probably
more
appropriate
than
others.
You
know,
and
there
are
some
large
instances
to
like
you
know
that,
might
that
might
be
appropriate
so
like
once
then
yeah?
H
You
just
use
an
email
to
create
your
account
once
your
account
is
created,
though,
like
you
could
you
could
change
the
email
later
or
whatever
that
wouldn't
necessarily
be
important
to
us,
but,
like
you
know
so,
like
the
first
thing
would
be
to
choose
is
like?
Are
we
going
to
be
like
you
know,
at
okd
project
or
whatever
it
is
on?
H
You
know,
use
the
same,
handle
as
Twitter,
but
just
move
us,
you
know,
put
it
on
a
mastodon
server
and
then
set
up
some
replication
between
the
two
Twitter
and
Mastodon
side,
and
then
on
the
Mastodon
side.
You
have
to
start
getting
into
like
okay
that
then
you
know
connecting
to
other
servers
and
whatnot
connecting
people's
feeds.
H
That'll
be
a
little
slower
process
than
it
is
on
Twitter,
because
the
discoverability
isn't
quite
the
same,
but
you
know
we
can
navigate
through
that
process
and
I
think
the
big
deal
is
like
just
making
sure
that
we
choose
a
server,
that's
pretty
large
and
doesn't
have
like
a
crazy
amount
of
block
lists
or
whatever
associated
with
it
like
that'll,
be
kind
of
the
biggest
thing
for
us
to
do.
I.
Think,
okay,.
H
H
C
B
So
yeah,
oh,
go
ahead!
Sorry.
C
So
yeah
I,
I,
I
I
think
that's
the
biggest
one
and
then
also
maybe
we
do
need
to
provide
some
guidance.
So
community
members
that
want
to
sign
up,
maybe
suggest
I
mean
I.
I
went
down.
Do
I,
look,
do
I,
use
a
a
regional
base
one
or
do
I
use
a
technology-based
one.
What
are
the
pros
and
cons
and
I
never
really
got
to
the
bottom
of
it
yeah
so
and.
B
This
was
kind
of
my
concern
as
I
know
and
Brian
in
the
past
you've
sort
of
voiced
a
concern
about
us
spreading
ourselves
thin
across
different.
You
know
different
Communications
channels
and
stuff
like
that.
A
B
Concern
is
okay
if
we
I
getting
that
synchronization
in
place
so
that
what
we
put
on
Twitter,
what
we
put
on
Facebook,
what
we
put
on
whatever
they're
all
the
same
or
as
close
to
the
same
as
possible.
So
we're
not!
You
know,
Reinventing
the
wheel,
every
single
time,
that's
one
aspect
and
then
the
other
aspect
is
people
being
able
to
get
to
us
there
and
do
we
support
people
getting
to
us
there
or
do
we
just
say
well.
C
Just
said
that
that's
the
problem,
if
we
have
a
presence,
we
do
have
to
service
that
presence
and
I
think
we're
a
small
community.
We
we're
trying
to
push
everything
to
that
one
GitHub
discussion
group
and
have
all
the
conversation
there,
so
the
more
channels
that
we
have
the
more
frustration
we're
likely
to
give,
and
we
saw
that
when
we
were
actively
using
the
the
the
the
the
Google
sort
of
group.
C
We
had
two
slack
channels.
We
had
the
git
Channel,
you
got
people
wanting
a
response
and
just
broadcasting
everywhere
and
we
had
the
same
person
sort
of
putting
six
messages
on
six
different
channels
because
they
wanted
a
response
and
they
weren't
sure
where
they
got
response
from
so
they
just
spammed
everything
and
that
then
caused
so
I'm
upset
within
the
community
and
so
I
I.
Think
we,
yes,
it's
nice
to
sort
of
say
people
can
follow
us
where
they
are,
but
I
think
we.
C
A
G
Yeah
I
I
would
agree
like
looking
at
the
so
actually
it's
the
first
time
that
I'm
looking
at
the
Twitter
presence
now
of
4kd,
to
be
honest
and
and
I
mean
looking
at
the
things
that
are
being
announced
there
they're
kind
of
interesting.
But
it
also
seems
like
a
relatively
big
commitment
to
to
keep
them
up
to
keep
that
up
and
yeah
like
Brian
was
saying
not
not
sure
if
we,
if
we
actually
can
or
or
want
to
do
that
in
the
long
run.
G
But
then
the
question
is:
where
else
do
we
publish
news
that
are
not
release
updates
but
because
for
releases
I'm
just
subscribing
to
the
GitHub
RSS
feed?
But
then
what
I
find
really
interesting
is
like
okay,
like
when
our
OPD
working
group
meetings
in
person
or
like
there
was
some
stuff
about
last
year
about
first
dim
where
there
were
some
presentations,
so
in
general,
when
like
presentations
are
happening.
So
where
do
we
publish
those
if
not
on
a
social
media
presence.
G
I'm.
Sorry,
if
my,
if
my
if
my
note
here
does,
does
not
answer
any
questions
by
the
way.
B
No
I
think
it
poses
more
questions,
which
is
really
what
this
does
is.
It
poses
questions.
If
we
start
talking
about
this,
you
know
I've
been
posting
to
Twitter,
but
you
know
it's
not
like
we're
consistently
like.
We
have
a
lot
of
stuff
coming
out
that
we
want
that.
That
gives
us
a
Twitter
we're
not
building
that
social
grid
connectivity.
Yes,.
G
H
And
solid
yeah
people
have
already
done
this
too.
There
are
other
organizations
you
know
they're,
especially
with
what's
gone
on
recently
with
Twitter.
There
are
a
lot
of
organizations
that
have
started
to
set
up
alternative
presences
in
the
What's
called
the
fediverse,
which
is
how
they
refer
to
the
the
network
of
Mastodon
servers,
and
you
know
that
are
linked
together
and
there
people
have
already
set
up
like
replicators,
they
just
replicate
stuff
from
their
Twitter
and
just
when
they
tweet
something
it
goes
out
onto
Mastodon
as
well,
so
I'm,
not
I.
H
Don't
necessarily
think
that
you
know
Mastodon
will
become
this
like
this
two-way
communication
pathway,
where
we're
having
to
like
do
a
lot
of
stuff
with
it,
but
it
might
be
another
place
just
for
us
to
broadcast
information,
and
then
probably
you
know
and
I
just
looked
quickly
to
see.
Matthew
Miller
who's
like
the
Fedora
lead.
He
happens
to
have
a
a
mastodon
account
on
the
same
instance
that
I'm
on,
which
is
a
a
fast-based
instance.
H
So
that
might
be
something
where
it's
like.
Well,
you
know
you're
talking
to
a
community
of
like
40
or
50
000
people
who
are
all
there
because
they're
interested
in
like
open
source
software.
So
it's
like
if
we
choose
a
server
that
probably
gets
us
to
most
of
the
people
who
might
be
curious
to
read
about
this
stuff.
You
know
that
could
go
a
long
way
towards
doing
what
we
want
to
do.
Anyways
so,
but.
B
B
If
you
could
that'd
be
fantastic,
okay,
we
are
at
time
the
only
other
thing
that
I
wanted
to
mention.
Real
quick
is
Zoom.
Feedback
are
folks,
cool,
okay,
that
sounds
good
and
then
the
other
thing
is
was
just
announced
in
Boston.
B
Do
we
want
to
reach
out
to
our
liaison
person
to
make
sure
that
we
get
some
space
at
Summit
thumbs
up?
Yes,
okay,
all
right
cool!
We
will
do
that.
That'll,
be
part
of
the
the
invite
on
this
is
to
talk.
I
didn't
see
when
submissions
are
due
by,
but
I'll
look
at
that
real,
quick.
Let
me
get
off
this
call
any
last
minute
things
before
we're
done
with
the
meeting.
E
Yes,
I
have
one
last
thing:
I
guess
like
similar
to
the
Fedora
Foundation
I.
Guess
they
you
know,
do
a
lot
of
projects.
Is
there
a
process
or
a
way?
I
could
maybe
write
a
proposal
for
a
project
with
lkd
back
in
Raleigh
North
Carolina
I
was
part
of
a
granite
award
for
1.1
million
dollars
from
Carolina
Central
to
focus
on
dealing
with
the
pipeline
for
devops
steps
that
costs
and
ml
Ops,
and
it's
just
three
students
but
I'm
thinking.
E
If,
if
we
could
use
okd
as
a
learning
platform,
so
they
can
learn
is
what
I'm
trying
to
do
so.
You
know
you
know.
Red
Hat
was
you
know
written
in
there
with
mass
open
Cloud,
but
you
know
I
I
would
love
to
do
it
with
okd,
so
I
don't
know
how
to
do
that.
So
that's
what
I'm
asking?
Is
there
a
procedure
because
I
you
know
the
grant
may
start
next.
Next,
yes,
next
semester,
but
again,
I,
don't
know
myself
with
me
being
in
Qatar.
E
I
have
to
figure
that
out
again
but
yeah
operate
first.
It
sounds
like
a
good
link
sounds
like
a
great
link
to
help
a
new.
You
know
some
high
school
students
undergraduates,
you
know
try
to
get
into
this.
B
B
Meeting
and
you'll
notice,
the
videos
are
going
to
start
going
up
much
quicker
now,
they'll
actually
be
up
in
like
a
day
or
two.
So
we'll
get
this
up.
Pretty
quick.
All
right,
I'll
talk
to
you
again
soon,
thanks.