►
From YouTube: Distribution Team Demo 2021-08-02 - CRC
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
Welcome
to
the
distribution
demo,
this
was
rescheduled
from
last
thursday,
but
today
I'm
going
to
try
to
be
bringing
up
the
operator
in
crc
the
code
ready
containers
that
red
hat
puts
out
for
bringing
up
a
local
openshift
cluster.
I
will
proceed
with
that.
This
does
take
a
while
to
get
set
up.
The
the
downloads
are
basically
three
gigs
three
and
a
half
gigs
of
container
images
that
get
pulled
down,
and
then
they
get
expanded
into
another
12
gigs
of
space.
A
A
So
I
did
that
today
I
originally
was
planning
on
using
mini
shift
because
it
looked
like
it
worked
reasonably
well,
but
then,
looking
at
things
today
with
mitch,
it
looks
like
that
mini
shift
is
still,
it's
gonna
be
still
running
a
3.0
open
shift
or
three
series
openshift,
even
though
mine
was
reporting
44,
but
that
because
I
had
a
newer
version
of
oc
loaded,
so
it
looks
like
crc
is
about
the
only
way
we'll
be
able
to
get
a
openshift
for
up
and
running.
A
I
am
currently
running
if
I
a
slightly
newer
version
of
the
server.
I
believe
there
is
a
way
with
crc
to
install
specific
versions
of
the
server
I
just
have
not
played
with
it
enough
to
understand
that
yet
so
I
am
running
a
4.8
version
of
openshift
on
a
newer
version
of
kubernetes
2
again,
this
is
all
just
stock
crc
from
the
latest.
A
Versions,
I
could
pull
down
to
run
crc
attacks
if
you're
running
ascf,
it's
actually
pretty
straightforward
to
just
pull
the
crc
distribution
through
asdf
and
get
various
versions.
I
will
tell
you
that
on
os
10,
there
are
some
calls
where
the
newer
versions
of
crc
are
not
being
distributed
in
the
same
way
and
for
os
10,
and
so
it's
not
currently
working
unless
you
manually
install
it
for
os
x.
A
So
after
that
logged
in
I'm
currently
logged
in
as
the
administrator
to
be
able
to
install,
I
think
some
of
the
operators
and
so
forth,
but
I
am
having
an
issue
with
getting
the
cert
manager
installed
going
according
to
the
back
here.
According
to
the
operator
hub
installation-
and
I
I
at
this
moment-
I
can
really
only
as
far
as
I
can
tell-
I
can
only
really
do
it
with
the
operator
hub,
because
I'm
currently
running
this
on
another
box
running
the
open
shift
on
another
box.
A
So
I
don't
have
access
to
the
open
shift
console
at
the
moment
to
be
able
to
just
go
with
the
one
click
install
from
there.
I've
tried
doing
some
port
forwarding
and
things
like
that.
As
soon
as
I
do
the
port
forwarding,
I
get
a
message
on
the
console
url
that
says,
applications
not
available,
and
I
think
the
reason
for
that
is
because
they
are
checking
for
the.
A
The
host
in
the
url,
that's
queried
the
crc
nodes,
do
specify,
and
I
should
probably
put
I'll
put
this
in
the
notes
in
a
minute
here.
I
don't
see
it
at
the
moment
when
crc
gets
installed,
it
runs
with
domain.
I
think
it
was
crc.testing,
and
so
when
I
go
to
the
try
to
go
to
the
console
without
that
in
the
url,
I
think
the
application's
going
is
complaining
that
I
can't
find
the
path
to
that
application.
A
So
that's
another
thing
to
sort
of
fix
and
look
at
how
to
how
to
work
around,
to
use
this
as
a
as
a
full
environment
on
another
machine,
anyways,
the
start
manager.
So
with
certain
energy
you
have
to
load
olm,
which
I
think
I
can
run
this
again,
but
probably
won't
do
what
much
of
anything.
A
So
I
ended
up
installing
it
with
this.
So
basically
curl
the
install
script
and
then
run
it
through
dash
and
it
will
basically
just
go
run
the
to
install
the.
I
think
it
actually
will
complain
that
it
has
already
installed.
Oh
no,
that's
right.
This
one
is
olm
is
installing
a
different
configuration
since
we're
running
openshift.
A
It's
not
a
vanilla,
kubernetes
kernel,
so
our
cluster,
so
it
doesn't
like
it
as
much,
but
when
I
do
try
to
install
the
insert
manager
directly
after
olm
is
installed,
so
it
could
it's
quite
possible.
I
don't
really
have
olm
installed.
A
I
get
the
first
time
it
runs.
It
runs
fine.
I
went
and
deleted
the
subscriptions
the
my
cert
manager
to
subscription
again
and
it
runs
and
installs
the
subscription.
A
A
Except
for,
although
all
the
catalogs
look
pretty
good,
except
for
when
it
looks
like
it's
trying
to
get
the
operator
of
io
catalog,
it
doesn't
like
it,
it
doesn't.
It
seems
to
think
it's
not
there.
I'm
not
sure
where
that's
coming
from
yet,
because
if
I
democrat,
there's
not
not
a
whole
lot
in
the
description
or
skinny
subscription.
A
Full
help,
I
don't
think
I
really
saw
a
url
in
here
which
is
telling
it
where
to
get
that
catalog
from
yeah.
The
only
thing
seem
to
be
is
in
this
spec.
So
almost
seems
like
this
operator.
Io
catalog
is
a
magic
string
which
somehow
gets
pulled
up.
A
A
For
the
open
shift,
oh
well,
so
ideally
it
should
be
picking
it
up
from
there.
So
that's
sort
of
as
far
as
I've
gotten
so
far
this
morning,
getting
things
rebuilt
with
crc
trying
to
get
cert
manager
installed
to
be
able
to
get
us
to
the
point
where
we
can
install
the
operator
and
just
sort
of
chase.
My
tail
right
now
trying
to
figure
out
where,
where
this
operator
hub
catalog
is
coming
from
so
mitch,
I
don't
know
if
you
have
any
ideas
or
seen
stuff
like
this
before.
B
B
Installed
from
a
we
just
have
a
subscription
so
far
right.
A
B
A
C
A
The
operator's
namespace
right
now
and
there's
nothing
nothing
to
point
at
the
moment
and
and
from
like
I
say
I
you
might
have
missed
part
of
what
I
was
going
through.
A
B
Yeah
back
a
few
months
ago,
and
I
was
dealing
with
more
olm
stuff-
I
saw
similar
things.
It's
kind
of
tough
to
debug.
I
know
that,
like
once,
you
have
the
subscription.
That
means
the
sort
manager
operator
itself
will
be
running,
but
it
doesn't
mean
cert
manager
will
be
running.
You
need,
like
an
instance
of
it
right
same
with
gitlab.
So
I
was
just
going
to
see
if
you
did
like
coupsie
tailgate
pods
in
all
name
spaces
and
gripped
for
cert
manager
is
anything
related
to
cert
manager,
running.
A
This
is
this
is
actually
running
on
a
linux
box.
I
am
sitting
right
next
to
me.
There's
a
bit
of
a
problem
with
trying
to
get
crc
running
on
os
x
from
asdf,
which
is
a.
A
That's
that's
just
minor
stuff,
that's
actually.
I've
got
issues
in
with
that
asdf
plugin
saying
these
are
problems
I
actually
created
about
20
days
ago.
It
looks
like.
B
A
No
no
activity
at
this
point,
but
so
I've
been
running
this
on
my
linux
box
here.
B
Okay,
it's
probably
worth
trying.
I
mean
this.
All
m
stuff
is
kind
of
its
own
rabbit
hole.
If
you
want
to
just
try
briefly,
installing,
with
the
that
static,
koopsie
tail
apply
as
long
as
like
the
default
security
context,
constraint
would
get
applied
because
you're
not
specifying
one
in
the
manifests
there,
and
as
long
as
that's
enough
for
cert
manager.
That
should
be
okay,
so
I
would.
B
B
A
I've
not
seen
that
I
go
to
what
they
list
this
documentation
and.
A
A
A
There
it
is
it's
12
gigs
of
memory,
six
six
scores,
I
think,
12
gauge
memory.
So
it's
okay!
It's
not
going
to
be
super
great
for.
B
A
B
A
Are
all
up
and
running
now
this
one
I
was
hoping
to
concentrate
on
a
little
bit
more
this
morning
is
the
far
side
of
it,
but
so
this
is
all
sort
of
a
little
new
to
me
so
mitch
I
might
need
a
little
hand
holding,
but
let's
see
what
we
can
do
here.
Okay,
let's
see
I'm
just
scrolling
through
the
docs
here
since
I
haven't
had
much
time
to
look
at
it.
A
B
A
B
A
A
A
And
I
just
do
like
go
build
here
and
there
you
actually
assume
my
goal
is.
B
A
Well,
yeah,
but
there
it's
okay,
we
got
to
require.
Okay,
do
I
just
need
to
install
crds
and
then
go
get.
B
Yes,
those
make
commands
steps
one
through
five.
It's
probably
one
through
six
right
now,
but
you
can.
You
can
skip
that
install
required
operators,
step
yeah.
A
A
B
B
A
B
A
B
B
A
B
B
A
Because
I
mean
it's
really
just
for
I
think
just
to
configure
the
chart
so
probably
should
be
okay.
B
B
A
A
Now,
okay,
sorry,
although
ideally
I
could
put
it
there
and
then,
when
the
operator
sees
it,
but
yeah.
A
Actually
interesting,
the
oh
sorry,
I
skipped
I
did
skip
step.
I
missed
the
make
deploy
operator
and
say
it
looks
like
you're
installing
the
cr
first
then,
but
yeah
it's
there.
B
B
A
A
B
That
it
relies
on
because,
when
you
send,
when
you
submit
your
gitlab
custom
resource,
that's
you're
submitting
it
to
a
web
hook
that
web
hook
has
certificates
on
it
yeah
so
web
hook,
server
cert
not
found.
B
A
A
A
A
C
A
A
B
Just
delete
any
of
that
old
stuff,
but
I
mean
all
the
steps
are
working
right.
If
cert
manager
was
was
functioning,
then
it
it
created
all
the
other
resources
correctly
and
the
deployment
started
to
come
up.
It's
just
the
pod
can't
start
without
that
secret.
So
it
looks
like
crc
is
functioning
as
we
would
expect
just.
A
I
I
can,
I
can
certainly
crash
crc
and
bring
it
up
again
fresh
and
then
we
can
try
loading
cert
manager
and
get
stuff
closed
again.
I
have
no
problem
with
that.
Okay,
perk.
B
A
C
A
C
A
A
The
problem
I'm
having
is-
and
I
think
I
was
talking
about
this
before
you
get
on
because
I'm
running
on
my
linux
machine-
I
can't
talk
to
the
dashboard
directly,
so
I
have
to
do
a
port
forwarding
over
ssh
and
when
I
do
that
and
try
to
connect
to
it,
I
still
have
it
up
here.
I
get
this.
A
I
get
this
wonderful
message
and
I
think
the
reason
being
is
because
the
console
comes
up
under
this
crc
testing
domain
or
whatever
it
is.
Oh,
I'm
gonna
get
my
full
secret
again.
It
will
because
that's
not
in
because
I'm
going
to
localhost
8080,
because
it's
a
front
forward.
It
doesn't
see
that
domain
in
the
url
and
says
that
application's
not
right
because
it
can't
route
according
to
the
host
name.
So
I
think
that's
what
that's
really
saying.
A
B
Yeah
well,
it's
starting.
I
mean
this
kind
of
makes
me
think
you
know
conceptually.
We
had
this
same
scenario
with
nginx,
where
it
was
an
external
dependency
and
it
had
to
be
installed,
but
because
we
wanted
ssh
support
out
of
the
box,
we
ended
up
shipping.
The
nginx
objects
that
we
forked
as
part
of
our
gitlab
chart.
C
B
A
If
I
remember
correctly,
it's
in
the
it's
a
dependency,
it's
in
the
requirements
camel.
B
B
A
A
A
A
B
Yeah
it
could
have
been
different.
Maybe
it's
every
time
it's
creating.
A
A
network,
while
this
ip
range
is
the
that
ip
range
was
the
range
I
was
getting
for
many
shifts
once
I
get
it.
A
C
A
This
is,
this
is
actually
an
arch,
it's
garuda
linux,
which
is
an
arch
derivative.
C
C
C
So
just
do
a
reverse
net
edit
crc.
C
C
What's
this
forward
forward
type,
is
nat
okay,
so
it's
united,
so
you
what
you
would
need
is
you
need
to
put
routes
between
the
network,
that's
just
spinning
up
when
it
spins
it
up
in
the
other.
The
other
way
you
can
do
this
is.
If
that
box
is
a
has
a
wire
connection,
you
could
use
nmcli
to
create
a
bridge.
C
C
A
Yeah
it,
like
you,
say
the
the
stuff
that
happened
last
well.
I
had
set
up
mini
shift
a
week
or
two
ago
a
couple
weeks
ago,
maybe
to
play
with
it
a
bit
more
and
then,
when
I
went
back
to
work
with
it
on
thursday
morning,
getting
ready
for
the
demo
is
when
everything
just
went
sideways,
and
I
say
it
was
I
I
shouldn't
say
it
changed
ips.
I
I
should
clarify
when
I
tried
to
connect
to
this
mini
shift
cluster
after
using
it
a
bit
ago.
A
A
Many
shift
looks
like
it's
not
going
to
work
out
anyways
because
I
talked
with
mitch.
I
think
I'm
fairly
certain
more.
I
looked
at
it
more.
I
thought
about
it.
It
was
running
a
3.0
openshift,
which
we
can't
use.
B
A
A
Memory,
so
yes,
it's
not
like
super
duper.
I
was
thinking
this
last
day
or
two
and
maybe
I'll
just
go,
buy
something
to
drop
it
into
this
box.
A
A
Okay,
it's
still
got
a
authentication
token
to
allow
me
good
okay,
so
we
are
back
to
scratch.
Where
did
we.
A
Find
that
there
it
is,
try
the
static,
install
of
strip.
B
A
B
A
A
B
B
A
B
B
B
And
you
can
just
watch
that
one
name
space
without
all
that
other
overhead.
A
Right
but
I
thought
I
can't
there
we
go.
A
Well,
I
mean
ideally
normally
what
I
would
be
doing
is
I'd,
be
running
external
dns
so
that
it
would
route
or
create
the
dns
records
up
in
google
or
wherever,
and
then
I
can
just
hit
it
by
a
url
right
now,
if
I
have
to
use
nipio
or
something
to
connect
to
it
enough
to
see.
A
A
B
Yeah,
there's
probably
flags
who
could
pass
it
telling
it
which
versions
to
pick
which
would
be
nice
for
testing,
especially
when
we
get
to
olm
type
stuff.
If
it
has
the
capability
to
run
olm,
which
I
assume
it
would,
then
we
can
kind
of
spoof
what
it
would
be
like
to
show
the
gitlab
operator
in
the
catalog
and
install
it
and
make
sure
that
that
flow
all
works
without
needing
a
big,
expensive,
open
shift.
Cluster
running.
A
B
A
A
A
A
B
A
What
I
saw
earlier
too,
is,
I
think,
was
the
the
kvm
image
was
alligator
nine
about
nine
gigs
of
memory,
the
35
gigs
of
storage,
a
lot
of
that's
used
on
the
front
end
because
it
takes
about
three
and
a
half
gigs
to
download
the
containers
and
when
you
start
start
the
cluster
up
the
first
time.
It
expands
that
into
a.
A
So
it
expands
it
to
about
12
gig,
so
it
takes
a
couple
minutes.
So
it
took
probably
close
to
15
minutes
to
expand
that
for
me.
But
that
looks
like
that's
only
the
first
time.
C
A
Well,
I
may
I
may
decide
to
just
let
it
beat
my
machine
to
hell
for
a
while
and
let
it
start
to
come
up
and
see
where
we
get
to
see.
If
we
see
if
I
can
see
any
errors,
I
can
put
that
in
the
demo
notes.
If
I
see
any
errors
come
up,
but
I
would
say
at
this
point:
yeah,
it
looks
like
things
aren't
good
enough
yeah
coming
up
cool
that
way,
we
don't
have
to
waste
everybody
else's
time.
Just
let
this
run
in
the
background
and
continue
working
on
other
stuff.
A
For
your
for
your
hands,
I
appreciate
your
thoughts
and
help
here.
B
C
B
In
the
crc
docs
they
have
some
information
on
how
to
connect
to
the
instance,
so
we
could
give
that
a
try
outside
of
this
call.
A
But
I
think
I
think,
if
I
were
to
play
a
couple
games
on
my
side
of
faking
out
the
hostname
that
I
could
probably
get
to
the.
A
A
A
Going
to
be,
I
think
it
would
be
doable
it's
just
if
I
was
writing
on
my
local
machine,
then,
yes,
you
could
just
say:
crc
console
will
open
up
in
a
local
web
browser
for
you.
B
Yeah
cool,
so
it
sounds
like
we
could
use
this
for
testing
some
openshift
specific
stuff.
Day-To-Day
testing
could
probably
be
done
in
kubernetes
with
kind
or
whatever
other
local
tool
you
want,
but
at
least
we
have
something
that's
official
from
red
hat.
That
would
help
us
test.
Specifically
these
components,
which
is
good.
A
And-
and
I
earlier
this
morning,
while
I
was
trying
to
just
get
crc
waiting
for
it
to
expand
and
everything,
I
started
modifying
the
asdf
plug-in
to
allow
us
to
run
on
os
x,
but
that's
going
to
take
a
little
bit
of
work,
so
it
might
be
another
day
or
two
before
those
pr's
are
up
to
maintainer.
But
again
the
maintainer
hasn't
responded
in
20
days.
Since
I
recorded
the
issues,
I'm
not
sure
if
he's
gonna
respond
very
quickly
either.