►
Description
The latest Red Hat OpenShift release is at hand, so what does that mean for upgrades? In today's episode, we take an inside look at what's coming in OpenShift 4.9 and how these new changes and features can help you be successful in the hybrid cloud. Among some of the latest and exciting features for admins in this release, we'll explore MetalLB support, single node OpenShift with UPI, certified Helm Charts, Kubernetes 1.22 changes, and much more. So join us live and bring your questions!
Learn More:
https://www.youtube.com/watch?v=Q1j_mt_XLqE
About Ask an OpenShift Admin:
Do you have questions? We want to answer them! Every Wednesday at 11am ET, hosts Andrew Sullivan and Chris Short sit down to answer your questions about Red Hat OpenShift.
Follow us:
Andrew Sullivan https://twitter.com/practicalandrew
Chris Short https://twitter.com/chrisshort
Subscribe to Red Hat's YouTube channel: https://www.youtube.com/redhat/?sub_confirmation=1
About OpenShift:
Red Hat OpenShift is an open source container application platform based on the Kubernetes container orchestrator for enterprise application development and deployment.
Openshift.com https://red.ht/33Fh6uV
OpenShift on github https://github.com/openshift/
OpenShift on redhat.com https://red.ht/2o4ccDk
#RedHat #OpenShift #ContainerPlatform
A
C
Evening,
wherever
you
happen
to
may
be
just
in
case
that
got
cut
off
hello
everyone
good
morning,
good
afternoon
good
evening,
so
first
of
all
welcome
to
today's
ask
an
open
shift.
Administrator
live
stream,
so
you'll
probably
notice
something
pretty
substantially
different
today,
and
that
is
that
well,
I'm
here
by
myself,
I
don't
have
my
my
long
time,
co-host,
slash
helper,
slash,
question,
asker,
slash,
keeping
andrew
honest
co-host
with
me
chris
shorts.
C
So
if
you
happen
to
miss
that
announcement,
so
last
week
chris
and
stu
miniman
did
a
stream
here
on
red
hat,
live
streaming.
That
talked
about
basically
the
future
and
more
specifically
chris's
future,
as
well
as
the
past,
and
if
you
aren't
aware
chris
has
decided
to
pursue
an
opportunity
outside
of
red
hat.
So
congratulations
to
him.
Thank
you
so
much
chris
for
all
of
the
help.
All
of
the
work,
all
of
the
things
that
we
have
been
able
to
achieve
together,
you
will
be
severely
missed,
but
the
good
news
is.
C
I
am
not
alone,
so
I
it
doesn't
mean
that
I
am
completely
by
myself.
Instead,
we
have
a
a
producer
who
has
been
helping
us
behind
the
scenes
for
quite
some
time,
and
she
has
been
kind
enough
to
kind
of
step
up
and
really
take
the
reigns
here.
So
you
see
sometimes
those
chat
messages
that
come
from,
like
the
red
hat
official
accounts
and
or
the
red
hat
openshift
account
and
stuff
like
that.
That
is
stephanie
stephanie
is
the
producer.
That's
working
in
the
background
there
so
say
hi
to
her
say.
Thank
you
to
her.
C
She
is
the
one
who
kind
of
keeps
me
organized
in
the
background
and
will
help
me
to
hopefully
keep
on
top
of
things.
So
I
will
also
point
out
that,
yes,
this
is
there's
a
lot
of
things
going
on
on
my
monitor
right
now.
So
my
apologies.
If
I
happen
to
miss
your
questions
or
something
like
that,
if
I
I
don't
see
them
after
a
few
minutes,
please
don't
hesitate
to
you
know
call
me
out
and
be
like
hey.
C
You
know,
andrew
you
missed
this,
you
know,
can
you
can
you
please
address
my
question
type
of
thing
and
I'll
I'll
be
sure
to
get
to
that.
So,
yes,
again,
congratulations
chris!
Thank
you,
everyone
for
acknowledging
that
I
don't
know
whether
or
not
he's
watching,
but
I
am
very
happy
to
poke
him
in
the
background.
You
can
also
find
him
he's
still
on
social
media
and
everything
at
the
same
place
that
he
has
always
been
so
by
all
means,
say
hi
to
him
if
you
so
choose.
C
C
So
anything,
that's
top
of
mind.
Anything
that
you'd
like
to
ask
about
by
all
means:
please
bring
it
up
and
we'll
be
happy
to
address
that,
regardless
of
whatever
it
is,
the
topic
of
the
day
happens
to
be
so.
You'll
also
notice
that
with
every
one
of
these
streams-
and
these
happen
weekly,
we
have
a
theme
or
a
topic
that
we
try
and
focus
on
and
today
that
is
openshift
4.9.
C
It's
a
two-hour,
long
presentation,
that's
given
by
the
entire
product
management
team
for
openshift
effectively
they
go
through
all
of
the
things
that
have
happened
in
4.9.
They
give
you
a
quick
little
blurb.
You
know
that
that
one
slide
highlight
of
what's
changed
and
then
we
try
and
cram
as
much
of
that
as
we
possibly
can
in
you
know,
understanding
that
engineering
has
put
you
know
three
plus
months
of
effort
into
a
release,
so
there's
a
lot
of
things
that
change
and
we
try
and
pull
out
those
highlights.
C
C
C
Yes,
I
I
hope
so
as
well
that
that's
actually
one
of
the
things
that
I
wanted
to
talk
about
so
I'll
talk
about
that
in
just
a
few
minutes
when
we
get
to
our
top
of
mind,
topics
on
the
on
the
stream
so
housekeeping
wise,
again,
sorry
to
see
chris
go,
you
know
see
thanks
for
all
the
fish.
Welcome
stephanie,
thank
you
so
much
for
all
your
help.
I
see
you
in
the
chat
already
so
by
all
means,
please
don't
hesitate
to
interact
with
her
and
she
will.
C
C
I
probably
won't
have
a
stream
next
week.
So
if
you
weren't
aware
red
hat
summit
this
year
was
pretty
different
than
in
you
know
normal
years,
so
we
had
the
first
summit.
If
you
will
back
in
april,
it
was
primarily
led
by
or
it
primarily
was
leadership
folks
talking
about.
You
know,
strategy
and
big
picture
things.
C
Then
in
june
we
had
a
second
summit
right
or
summit
part
two,
if
you
will,
where
we
had
some
more
technical
sessions
and
stuff
like
that,
it
was
hosted
very
much
like
summit
was
last
year
right
where
you
joined
all
of
those
mostly
pre-recorded.
Some
pseudo
live
some
fully
live
sessions
happening
on.
I
don't
even
remember
the
platform
now,
but
it's
all
at
summit.redhat.com
and
then
what
we
are
now
in
the
midst
of
is
summit.
C
C
C
So
one
of
those
happens
to
be
here
in
raleigh,
where
I'm
located
and
next
wednesday
is
when
it's
happening,
and
I
am
a
presenter
for
an
ask
the
expert
session.
So
if
you
happen
to
be
in
the
raleigh
area-
and
you
want
to
attend
that
by
all
means
come
out
come
see
us
it's
an
all
day
thing.
I
think
I
have
to
be
there
for
my
first
booth
duty
session
wow.
I
haven't
had
booth
duty
in
like
two
years,
so
that'll
be
interesting,
but
you
know
my
first
boot.
C
C
So,
let's
see
just
catching
up
on
chat
here,
stephen
congrats,
on
the
upgrade
to
4.9.
Yes,
I've
had
a
few
clusters
running.
One
of
the
things
I
want
to
look
at
today
is
some
of
the
ux
changes
which
I
am
still
discovering.
It's
kind
of
funny
that,
despite
the
fact,
our
team,
the
tech
marketing
team,
has
a
monthly
meeting
with
the
ux
team.
I'm
still
surprised
by
all
the
things
that
they
managed
to
cram
in
there,
so
I'm
excited
I'm
looking
forward
to
exploring
some
more
about
4.9
as
we
go.
C
I
also
see
you
asking
if
you
had
to
choose
between
youtube
or
twitch,
really
we
have
no
preference.
So
if
you
didn't
know
we
stream
the
we
stream
to
twitch,
we
stream
to
two
separate
youtube
channels.
One
is
the
openshift
channel.
One
is
the
red
hat
channel,
which
one
of
those
you
choose
is
completely
up
to
you.
It
makes
no
difference
to
us
and
the
technology
we
use
in
the
background
restream.
C
It
also
re-broadcasts
all
of
your
chat
messages
across
all
of
those
so
kind
of
wherever
you're
at
whichever
one
you
choose.
That's
the
right
one
to
use.
You
can
also
look
at.
We
also
have
a
discord
channel
that
you
can
use.
All
of
the
chat
is
accumulated
into
discord.
You
can
chat
through
there
as
well,
and
some
of
us
are
are
I
at
least
I
I
infrequently
hang
out
there.
You
know
langdon
when
he
was
still
here
doing
the
level
up
hour.
C
He
used
to
spend
a
lot
of
time
in
there,
and
also
you
know
if,
if
you
don't
mind,
be
sure
to
subscribe
to
all
the
various
channels,
I
know
sometimes
it
can
get
a
little
bit
noisy.
I
don't
know
if
it's
possible
to
limit
those
youtube
notifications
or
anything
like
that,
but
it
is
nice
to
to
get
a
reminder.
Well,
nice,
for
me
anyways,
that's,
which
is
the
better
better
alternative.
C
C
So
second
thing
so
reminder:
next
week
there
will
be
no
stream,
almost
certainly
no
stream
unless
something
happens,
keep
an
eye
on
the
streaming
calendar
for
any
updates
there.
If
it
happens
to
get
cancelled,
we'll
basically
remove
it
from
the
streaming
calendar.
So
you
can
keep
an
eye
out
on
that,
as
that
one
single
place
next
on.
My
list
is
what
I
think
is
a
pretty
big
change
here,
and
this
is
something
that
I'm
going
to
spend
a
couple
of
minutes
talking
about.
C
So
this
blog
post
is
comes
from
our
product
management
team
and
it
is
effectively
describing
that
we
are
changing
the
way
that
eus
or
extended
update
support
works
for
openshift,
so
historically
open
shifts
right
up
until
openshift's.
Only
eus
release
thus
far
has
been
4.6.
I
won't
say
only
only
in
the
4.x
line.
3.11,
of
course,
is
the
us
as
well
so
with
4.6.
C
C
C
C
So
if
you
read
through
this
blog
post
and
the
blog
post
is
so
mike
barrett,
the
author
here
does
a
pretty
good
job
of
explaining
a
lot
of
different
things
inside
of
here,
including
kind
of
why
we
did
this,
how
we
came
up
with
the
specific
time
frame.
Some
of
the
requests
that
we're
hearing
from
you
from
our
customers.
C
C
C
C
So
you
know
that's
two,
two
of
those
periods
that
they
have
to
go
through
as
opposed
to
with
this
new
process,
this
kind
of
skip
level
or
skip
update
for
the
compute
nodes
right.
We
reduce
that
down
to
just
one
of
those
updates.
So
that's
really
going
to
be.
The
big
benefit
of
an
eus
release
is
having
that
validated.
If
you
will
skip
update
process
associated
with
it,
so
I
definitely
encourage
you.
You
know
please,
if
you
have
a
moment,
go
through
read
the
blog
post.
C
C
So
again,
please
don't
hesitate
to
reach
out.
If
there's
any
questions
there,
I
think
we're
working
on
an
faq
in
the
background,
so
for
any
red
hatters
who
happen
to
be
watching.
Please
don't
hesitate
to
reach
out
as
well.
You
know
we'll
make
sure
to
include
all
of
that
information
out.
I
don't
know
if
we'll
update
the
blog
post
with
some
of
those
faq
questions
or
if
we'll
create
a
separate
document,
or
something
like
that.
But
just
be
aware,
you
know
please
any
questions.
You
have
don't
hesitate
to
ask
our
hope.
C
C
C
C
C
So
whether
or
not
that
linear
or
sequential
rather
4.6
4.7
4.8
has
to
happen,
and
then
you
can
do
the
4.8
to
4.10.
I
don't
know
the
answer
to
that.
I'm
working
on
an
answer
for
that.
Hopefully,
if
we
have
or
if
we
know
that
before
I
create
the
blog
post,
which
will
go
out
tomorrow.
Well,
I
I
send
it
tomorrow
afternoon
we
published
friday
morning.
C
So
if
I
have
an
answer
for
that,
I'll
include
it
in
the
blog
post,
so
keep
an
eye
out
on
what
is
it
cloud.redhat.com
blog
now
I
have
to
think
about
it
because
I
just
changed
from
openshift.com
blog,
I'm
guessing.
There
might
be
unique
issues
not
yet
envisioned
our
hope,
nine.
I
am
sure
you
are
right.
There's
always
some
unexpected
things.
You
know
our
qa
qe
folks
are
really
good.
C
They
test
a
lot
of
scenarios,
but
there
are
you
know
many
many
many
many
customers,
many
clusters,
many
more
clusters
than
customers,
so
we
always
find
those
kind
of
edge
cases
other
things
that
crop
up
as
they
go
on
all
right.
So
next
thing
on
my
list
here
and
don't
hesitate
to
ask
those
questions.
If
you
have
any
here
in
the
chat,
I'll,
try
and
keep
up
with
those
yeah,
I
hope
nine.
C
C
C
What
you'll
see
is,
if
you
do
like
an
oc
version.
It'll
say
you
know
oc
version
4.8.,
something
or
4.8.0
dash
blah
blah
blah
blah,
even
if
you're
running
a
4.8.12
or
a
4.8.14
cluster.
So
it
looks
like
the
version
of
the
cli
tools
that
you
download
from
inside
the
cluster
here
is
a
mismatch
with
the
version
of
the
actual
cluster.
C
However,
the
only
it's
a
mismatch
in
name
only
so
the
way
that
engineering
explained
this
to
me
is
effectively
all
of
the
builds
happen
inside
basically
they're
decoupled
from
the
version,
so
they
go
through
and
they
get
a
version
string.
That's
something
like
you
know:
4.8.0
dash
date,
timestamp
of
when
the
build
is
happening
and
it's
not
until
after
it
goes
through
the
process
and
becomes
you
know
a
fully
approved,
build
and
that
it's
assigned
a
build
number
and
because
these
tools
are
extracted
from
one
of
the
images
that's
built.
C
C
So,
yes,
they
look
different,
but
in
reality
they
are
an
exact,
precise
match
for
this
particular
version,
and
you
can
kind
of
ignore
that
version
mismatch
that
you
happen
to
see.
If
you
are
doing
that,
and
actually
I
wonder
if
I
were
to
use
the
command
line
terminal
if
it
would
show
the
same
thing,
because
I
think
it
pulls
the
tools
from
the
same
place.
So
let's
check
that
out
real
quick
in
case
you're
not
familiar.
C
So
I
click
this
little
thing
up
here
in
the
corner
and
it's
opening
a
command
line
terminal
for
my
cluster,
that's
actually
an
operator.
So
if
you
go
to
the
web
terminal
operator,
let's
minimize
that
and
if
I
search
for
terminal,
you
can
install
this
web
terminal
operator
and
it
gives
you
this
link
up
here
after
it
installs
just
refresh
the
console
page,
and
then
you
have
this
really
convenient
well
as
soon
as
I
get
it
to
come
up,
so
you
see
version
yeah.
So
this
one.
C
Actually,
this
one
doesn't
look
like
it's
been
updated.
So
I
wonder
if
it's
pulling
the
tools
image
instead
of
pulling
the
one
that's
inside
of
the
cluster,
so
that's
interesting.
I
had
noticed
that
so
our
hope
9
to
your
point
earlier
of
things
that
are
still
falling
out
from
the
update
process,
it
looks
like
this
pod
simply
hasn't
been
updated.
C
So
if
you
are
currently
on
a
4.8
cluster,
please
know
that
or
excuse
me
a
4.7
cluster.
Please
know
that
updates
have
been
blocked
again.
So
do
I
have
this
open?
Oh,
I
do
so.
I
think
I've
talked
about
this
before,
but
I
just
want
to
reiterate
if
you're
not
subscribed
to
errata,
so
I'm
just
gonna.
I
just
click
this
button
here.
This
is
simplyaxis.comaccess.redhat.com.
C
Errata,
so
if
you're
logged
in
you
have
this
notifications
preferences
button,
you
can
click
on
this
and
it'll
come
in
here.
Eventually
anyways,
so
it'll
ask
you,
you
know
what
what
errata
notifications
you
want
to
get.
You
can
see.
I
have
mine
and
they
are
disabled
because
it
sends
you
errata
notifications
for
all
products
associated
with
your
account.
C
So
as
an
employee,
we
have
basically
every
product
associated
with
our
account,
which
means
that
we
get
a
lot
of
errata.
So
I
just
disabled
mine,
but
you
I
would
suggest,
enabling
those
notifications
and
then
using
like
client-side
filtering.
You
know
gmail
or
outlook
or
whatever
you're,
using
to
filter
those
down
and
what
you
want
to
look
for
is
these
with
a
synopsis
where
it
says
notification
of
blocked
edges.
C
So
in
this
instance
we
have
a
bug
that
I'm
now
forgetting
what
is
actually
the
content
of
the
bug
we
can
come
in
here
and
we'll
open
the
bz
here
anyways
there
is
now
a
bug.
Oh
egress
ips
are
not
being
scheduled,
so
they
decided
to
make
this
a
blocker
for
those
updates.
I
believe
the
fix
has
already
been
created
and
is
in
the
current
candidate
for
4.7
4.8.
So
hopefully
these
will
be
unblocked
soon,
but
nevertheless,
at
this
moment
the
updates
from
4.7
to
4.8
are
blocked
again,
at
least
temporarily.
C
C
So
if
you
have
any
questions,
please
again
be
sure
to
highlight
those
in
chat,
I'm
trying
to
keep
up
and
make
sure
that
we
are
answering
all
of
those
as
they
come
up
again.
If
I
happen
to
miss
your
question,
you
know,
after
a
couple
minutes,
don't
be
afraid
to
raise
it
again,
but
today
we
want
to
talk
about
4.9
and
some
of
the
new
features,
new
things
that
have
come
out
in
4.9
that
are
really
specific
or
really
interesting
to
us
as
administrators.
C
C
C
That's
a
hard
one,
and
it's
not
it.
If
you
look
at
my
command
command
line
prompts.
Let
me
see
if
I
can
share
this.
Let's
see,
let's
stop
that
one,
and
then
we
want
to
share
this
one.
So
my
command
line
prompt
here
you
see,
I
use
colors
and
stuff
like
that,
I'll,
actually
utilize
different
colors
and
even
change
my
terminal
backgrounds
to
represent
which
color
that
or
which
cluster,
rather
that
I'm
connected
to,
because
it
can
be
confusing,
and
of
course
you
can
also
use,
though,
see
who
am
I
no?
C
I
can
see
I'm
not
actually
connected
to.
I
didn't
export
my
kube
config
here
but
anyways
that
that
is
how
I
help
keep
organized
that's
an
old
habit
from
a
long
time
ago,
literally
way
back
in
the
esx
esx,
not
esxi,
like
3.x
days
when
we
used
to
have
to
ssh
into
them
for
pretty
regular
things,
so
that
may
not
be
the
best
way.
That's
just
the
way
that
I've
identified
to
use
it.
C
So
if
we
switch
back
to
my
console
here,
you
know
another
way
is
maybe
using
this
web
terminal
operator
the
downside
to
the
web
terminal
operator.
Is
it
drops
you
into
a
pod
right?
If
I,
if
I
look
inside
of
here,
you
know,
I'm,
you
can
see
I'm
in
a
fairly
limited
environment
and
I
don't
have
access
to
all
of
the
scripts.
All
of
the
things
that
I
might
be
used
to
right.
C
I
don't
even
know
I
do
have
git,
but
you
know
if
I
have
a
bunch
of
ammo
files
that
I
you
know
want
to
work
with
on
the
command
line,
I
have
to
figure
out
how
to
get
those
in
there
or
or
do
other
things.
So
it's
helpful
for
some
things.
It's
not
helpful
as
a
general
administration
tool.
So
I
I'm
sorry
to
say
that
I
I
don't
have
a
really
good
answer
there.
If
anybody
else
has
any
suggestions,
I
I
would
very
much
welcome
them.
C
Yeah
multiple
cube,
configs
and
then
yeah
do
the
export.
That's
what
I
tend
to
do
in
each
one
of
those
shell
sessions.
I
don't
have
it
configured
anymore,
but
I
used
to
have
like
a
bash
alias
for
each
one
of
my
clusters.
That
would
do
that
kube
config
exports
and
set
that
terminal
color.
You
know
kind
of
a
bunch
of
different
ways
to
tackle
that
all
going
back
to
our
hook.
Nine's
suggestion
of
discipline.
C
Okay,
so
I'm
I'm
gonna,
try
and
keep
an
eye
on
chat
here.
Y'all
keep
keep
those
suggestions
flowing
in
so
I
I
would
very
much
be
interested
in
how
you
all
are
tackling
that
john
moscow
cube
dash
ps1,
that's
interesting.
I
like
I
like
powershell
so
anytime,
we
use
powershell
that
could
be
that
could
be
entertaining
so
openshift
4.9
released.
Ga.
I
think
the
bits
actually
landed
on
the
mirrors
monday
afternoon,
so
it's
been
available
for
a
little
while
now
you
know
48
hours,
basically
our
hope,
9.
C
I
think
it
was
you
who
pointed
out
that
on
the
releases
page,
you
can
see
all
of
that
information.
That's
out
here,
you'll
see
the
fast
and
the
stable
releases.
Now.
One
thing
to
note
is
that
if
we
were
to
look
at
the
update
process,
let
me
post
this
link
just
in
case
that's
the
releases
page,
which
tells
you
all
of
the
currently
available
versions
and
even
versions
that
aren't
available
anymore
and
what
the
active
release
numbers
are.
C
C
This
is
normal.
This
is
expected
so
with
the
last,
and
I
think
we
talked
about
this
a
couple
of
weeks
ago,
so
with
the
last
few
updates,
we're
averaging
somewhere
around
50
to
55
days
for
a
new
release,
so
in
this
case
4.9
to
be
available
in
the
stable
upgrade
channel,
and
that
was
true
of
4.7
to
4.8.
It
was
true
of
4.6
to
4.7
right.
It
just
takes
time
for
the
engineering
team,
the
folks
who
are
looking
at
all
of
the
telemetry,
all
of
the
information
that's
coming
in
to
declare.
C
Yes,
we
want
to
enable
stable
updates,
for
I
think
it's
something
like
98,
97
or
98
of
clusters
that
only
use
the
stable
channel.
However,
if
you're
using
the
fast
channel.
So
if
I
switch
over
here
to
fast
4.8
and
select
my
same
4.8.14,
you'll
notice
that
I
can
make
the
leap
to
4.9
so
updates
in
stable
are
blocked
at
the
moment.
Updates
in
fast
are
available.
C
So
if
you
have
a
test
cluster,
if
you
have
you
know
a
non-production
cluster,
if
you
really
want
to
be
brazen
right,
if
you
have
a
production
cluster
that
you
want
to,
you
know
use
on
the
fast
channel.
Absolutely
you
can
do
that
if
something
goes
wrong,
pick
up
the
phone
call
us
we'll
help
you
with
that,
and
it
was.
It
was
a
couple
weeks
ago
with
rob
zumske.
C
C
So
if
you
have
a
chance,
if
you
have
an
available
cluster,
that
that
you
don't
mind
putting
on
fast
or
you're
willing
to
test
with
fast,
that
would
be
really
appreciated,
stable
4.8
will
never
have
4.9.
That's
not
true
for
it'll,
eventually
have
it
after
again
that
that
roughly
two-month
time
period
candidate
is
not
a
supported
release.
That
is
correct.
I
hope
not
so
candidate
releases
are
kind
of
bleeding
edge.
If
you
will
so
they
can
have
issues
come
up
and
we
we
don't
support
those.
C
You
can.
I,
I
think,
because
I
I
worked
with
one
customer,
what
they
had
done
was
they
accidentally
upgraded
to
a
candidate
release,
and
it
just
so
happened
that
that
candidate
release,
because
so,
basically
once
a
release
is
cut
so
say:
let
me
switch
over
here
to
where
is
my
releases
page?
So
if
we
look
in
the
candidate
channel
down
here
so
4.7.34,
so
you'll
notice,
it's
not
in
fast
or
stable.
C
C
It's
not
going
to
change.
So
if
you
accidentally
end
up
in
candidates-
and
you
know,
then
what
you
didn't
mean
to-
and
I
worked
with
a
customer
who
did
this
essentially,
you
would
wait
until
the
next
release
happens.
That
does
end
up
in
fast.
So
let's
say
that
4.7.35
ends
up
in
fast
right.
It
goes
candidate.
It
ends
up
at
fast,
so
you
would
be
able
to
use
the
candidate
channel
to
update
to
4.7.35.
C
Then
change
your
channel
to
either
fast
or
stable,
whichever
one
maps
up
with
that
and
then
go
from
there.
So
it
is
possible
to
kind
of
re.
You
know
steer
yourself
back
off
of
that
path
if
you
need
to
homeland
has
been
on
fast
for
a
while
yes
same
here,
although
I
will
say
stephen
that
I
I
don't
often
I
don't
often
run
real
workload
in
my
home
lab.
I
probably
should
do
that
more.
C
Is
practically
checking
bugzilla
to
see
what
is
addressed
in
a
particular
fast
release
worth
it
beforehand?
Yes,
I
would,
I
would
say,
definitely
do
that
you
can
also
check
the
errata
or
the
release
notes.
I
don't,
I
think
the
release
notes
get
updated
when
it
goes
into
the
fast
channel,
because
it's
supported
at
that
point,
but
yeah,
sometimes
the
you
know
bugzilla
can
be
a
little
overwhelming
and
not
always
clear,
particularly
if
you're
you
know
we
have
as
an
employee,
you
can
see
the
hidden
comments
which
sometimes
provides
better
information
about
what's
happening.
C
If
you
can't
see
those,
it
can
be
a
little
confusing
to
go
from.
I
have
an
issue
and
then
so
you
know
comment
one.
I
have
an
issue
comment.
37
issue
resolved
and
everything
in
between
is
all
hidden
because
it
contains
customer
data
or
something
like
that,
so
bz
isn't
always
or
bugzilla
isn't
always
the
best
place,
but
it
is
an
option
I
always
suggest
errata,
even
though
it
can
be
a
little.
C
It
can
also
be
a
little
noisy
errata
is
the
next
best
fastest
way
to
find
out
and
then,
as
our
hope,
9
pointed
out,
the
release
notes,
the
release
notes
will
be
the
the
easy
to
read
the
the
best
way
to
digest
method
and
I'll,
even
throw
our
account
teams
under
the
bus.
Talk
to
your
red
hat
account
team
right.
They
should
be
keeping
up
with
that
type
of
stuff,
so
you
can
always
send
them
an
email
and
be
like
hey.
C
C
So
yes,
4.9
is
available
now,
updates
and
stable
are
not
available.
Yet
updates
are
available
in
fast,
fast
is
fully
supported.
Definitely
if
you
have
the
time
you
know
we'd
love
to
see
you
adopt
those
or
if
you
have
a
cluster,
that
you
can
use
we'd
love
to
see
you
use
fast.
That
way,
we
can
get
that
telemetry
and
really
make
good
decisions
about
what's
happening.
C
C
So
you've
heard
me
talk
about
in
a
couple
of
episodes
now
or
a
couple
of
streams.
Now
I've
been
using
single
node
openshift
clusters,
just
as
kind
of
testing
right
validating
that
you
can
do
it
and
single
node
openshift
has
been
available
since
the
4.9
time
frame.
Here's
I
was
already
on
the
window.
I
was
looking
for
so
they've
been
available
since
the
4.8
time
frame,
maybe
even
4.7
through
the
assisted
installer.
C
So
if
I
come
here
to
my
clusters
view
and
I
go
to
create
cluster
and
I
go
to
the
data
center,
you
see
we
have
this
assisted
installer,
so
assisted
installer
itself
is
technology
preview.
We
did
a
we've
done
a
couple
of
streams
on
it
now,
with
reese
oxenham
I'll,
dig
up
the
links
and
include
those
in
the
blog
post,
but
the
goal
here
is
effectively
to
provide
a
really
easy,
really
simple
way
to
deploy
your
cluster
right.
You
don't
have
to
worry
about.
You
know
creating
and
install
config.yaml.
C
You
don't
have
to
go
through
this
process
of
you
know,
generating
ignition
files
and
all
that
other
stuff.
All
of
that
is
kind
of
removed,
and
instead
I
create
this.
You
know
use
this
interface
to
define
everything
that
I
need
for
my
cluster
in
this
case,
I'm
going
to
call
it
single,
node,
openshift
snow.
I
want
to
choose
this
snow
and
down
here
I
can
select
4.9.0.
C
C
It
will
give
me
an
iso
right
if
I
click
the
button
down
here.
I'm
surprised
it
didn't
gripe
at
me
about
not
providing
an
ssh
key.
You
do
want
to
do
that.
So
it'll
generate
this
iso.
I
can
download
that
once
I
do
that,
I
boot
my
node
to
that
iso
and
then
it
will
appear
in
the
list
here.
I
can
choose
to
make
configuration
edits.
So
do
things
like
set
the
host
name?
C
If
I
need
to,
if
I'm
deploying
a
regular
cluster,
not
a
single
node
openshift
cluster
I
can
set
which
nodes
are
going
to
be
control,
plane,
nodes
which
nodes
are
going
to
be.
You
know,
standard,
compute,
nodes
so
on
and
so
forth,
and
then
click
the
go
button
and
it
will
deploy
that
cluster
for
me
so
with
single
node
openshift,
one
of
the
things
that's
important
here
is
assisted.
Installer
will
enforce
this
bare
minimum.
C
It
will
not
deploy
unless
you
have
at
least
eight
cores
and
32
gigabytes
of
ram.
So
if
you
have
something
less
than
that
so
say
you
know
16
gigabytes
of
memory
and
four
cpus.
You
know
six
cpus
24
gigabytes
of
memory.
Whatever
that
happens
to
be,
you
can
probably
deploy
single
node
openshift.
You
won't
have
a
lot
of
room
left
over
for
applications
or
anything,
but
you
can
probably
deploy.
I
haven't
tested
it
with
the
4
and
16.
C
I
wouldn't
go
any
lower
than
that,
but
it
is
an
option
if
you
use
the
command
line
method.
So
what
is
the
command
line
method?
You
may
be
thinking
to
yourself
and
it's
pretty
straightforward.
So
let
me
switch
over
to
my
terminal
here
and
you
might
have
noticed
that
I
was
already
in
this
sno
directory
single
node
openshift.
C
C
So
this
install
config
looks
to
be
pretty
standard
right.
So
up
at
the
top,
I've
got
my
standards.
You
know
base
domain
definition
down
below
I'm
defining
my
network
information,
so
it's
I'm
using
ovn
kubernetes
with
a
standard
machine
network.
This
is
just
my
whatever
the
subnet
is
for
the
mach,
the
network.
I
am
deploying
this
node
to
for
my
compute
stanza
you'll
notice
that
I
have
replicas
set
to
zero
and
my
control
plane
stands
at
replicas
set
to
one.
C
I
must
have
the
platform
set
to
none.
We
don't
currently
support
deploying
to
you
know
with
a
cloud
provider,
integration
or
anything
like
that.
I
think
it
will
technically
work,
but
again
it's
not
supported,
so
I
wouldn't
recommend
doing
that,
certainly
not
in
a
production
instance
and
then
importantly,
we
have
this
bootstrap
in
place
stanza.
C
C
Typewrite,
I
am
using
version
4.9,
so
I
just
pulled
this
this
morning.
I
think
so
I
can
do
that.
Install
or
excuse
me
that
explain
and
if
I
look
up
here,
I
have
this
now:
have
this
bootstrap
in
place
option
right,
so
it
kind
of
walks
through
all
of
the
things
that
I
need
to
here.
So
I
can
go
to
the
explain,
install
config.bootstrap
in
place,
and
it
tells
me
oh,
this
is
the
installation
disk
that
you're
going
to
use
for
coreos
installer.
C
So
with
that
out
of
the
way,
let's
see
what
this
actually
looks
like
inside
of
my
cluster
directory
here
I
have
an
install
config.yml.
This
one
does
have
a
pull
secret,
so
we're
not
going
to
look
at
that
one
and
I
want
to
do
an
open
shift,
install
create
and
let
me
pull
up
my
cheat
sheet.
Real
quick,
open
shift,
install
create
single
node
ignition
config.
C
C
If
I
look
inside
of
there,
it's
got
the
cube
admin
password
and
coupe
config,
just
like
any
other
install,
but
we
only
have
this
one
ignition
file
and
they
very
in
not
so
subtly
hint
where
they
expect
you
to
use
this
with,
which
is
the
live
iso
so
great.
I
now
have
this
bootstrap.
If
I
were
to
look
at
it.
B
C
C
So
the
next
thing
that
we
need
to
do
is
figure
out
how
we're
going
to
get
this
ignition
file
to
our
node.
Now,
historically
or
typically,
we
would
do
something
like
attach
it
as
a
vm
option.
If
I'm
doing
you
know
a
vsphere,
upi
ipi
deployment
or
if
I'm
doing
a
non-integrated
deployments,
I
would
host
it
on
a
web
server,
and
you
can
still
do
that.
You
can
put
it
on
a
web
server
if
you
want,
but
it's
going
to
be
substantially
easier.
C
C
Happening
yeah
dmi,
three
yeah.
What
do
we
do
on
this
openshift
stream?
We
well
we
install
openshift.
I
wish
I
could
say
that
I
remember
the
last
week
that
went
by
that.
I
didn't
install
openshift
and
I
honestly
cannot.
I
don't
know
if
that's
good
or
bad
so
once
this
pulls
down
we're
going
to
use
the
core
os
install
or
excuse
me
chros
installer
command.
C
C
C
So
all
I've
done
is
pull
down
this
coreos
installer
binary,
and
that
is
what
I
have
or
that
that
is
what
I'm
going
to
be
using
here
in
just
a
moment.
So
let's
reshare
our
terminal
window,
because
that
download
should
be
done
and
it
is
so
now
I
have
in
my
directory.
I
have
my
ignition
file
and
I
have
my
live
iso.
C
C
So
we're
going
to
use
this
ignition
file
and
we're
going
to
add
it
to
this
iso,
so
it
hit
hit
enter
there
and
it
comes
back
almost
immediately
with
no
output.
So
we
can
do
an
echo
on
dollar
sign
question
mark
to
verify
that
the
output
of
the
last
command
was
successful.
Right,
return,
code,
zero
and
now
I
have
my
my
file
right-
we
can
see
that
it
was
updated.
Right,
went
from
october
18th
to
october
20th,
so
I've
got
an
iso
great.
What
do
I
do
with
it
now?
C
D
C
C
C
I
can
use
nm2e
nmcli
to
edit
the
network
config,
however
complex
it
may
be,
and
then
provide
the
location
of
that
ignition
file
from
a
web
server
and
have
it
install
coreos
that
way,
both
of
those
will
work.
Fine,
if
you
need
to
I
happen
to
be
using
a
dhcp
reservation
for
mine,
so
I'm
just
going
to
let
it
boot
but
effectively.
What
it's
going
to
do
here
is
go
through
and
install
coreos
right.
C
It's
first
thing:
it'll
do
here
is
install
coreos
it'll
then
do
the
bootstrap
process
it'll
then
go
through
and
do
everything
else
that
it
normally
would
for
an
open
shift.
Install
you
can
see.
It's
already
pulled
that
one
thing
that
I
didn't
talk
about
here
is
the
prerequisites
so
prerequisites
you
just
need
to
have
dns
entries,
so
you
need
to
have
a
dns
entry
for
the
node
name
itself.
So
you
can
see.
I
have
this
sno-node
zero.
C
C
C
So
while
that's
doing
its
thing,
I'm
going
to
switch
over
to
my
ssh
window
here,
because
I
can
do
an
open
shift.
Install
wait
for
bootstrap
completes
and
you
can
see
it's
already
gotten
past
the
you
know
it's
it's
in
the
process
of
bootstrapping
right
now.
So,
while
that's
up
I'll
talk
about
a
few
other
things
that
I
wanted
to
highlight
here,
moving
windows
around.
C
So
if
you
happen
to
be
using
rel
seven
worker
nodes
today,
so
this
is
not
uncommon,
although
it
is
not
always
something
that
we
suggest,
because
rel7
worker
nodes
can't
be
managed
through
machine
config,
the
same
way
that
coreos
nodes
can.
But
if
you
are
using
rel7
worker
nodes,
those
are
now
deprecated.
C
If
I
look
for
the
deprecated
and
remove
features,
you'll
notice
down
here,
bring
your
own
rel7
commute.
Compute
machines
is
and
has
been
deprecated.
So
the
expectation
is
that
now
that
rel
eight
worker
nodes
are
supported,
we
should
all
be
migrating
to
rel,
eight
and
then
eventually
support
for
those
rel.
Seven
worker
nodes
will
be
removed.
C
C
C
So
with
openshift
4.8,
we
shipped
a
tech
preview
of
the
vsphere
csi
driver
from
red
hat.
It
continues
to
be
in
tech
preview
with
4.9
right
now.
What
I'm
being
told
is
that
it's
expected
to
go
ga
and
4.10,
but
will
be
optional
in
4.10,
so
csi
integration
with
vsphere
requires
virtual
hardware,
which
is
an
oxymoron
to
be
at
least
version
15
or
later.
C
So,
if
you
are
on
version
13,
which
essentially,
if
you've
done
like
an
ipi
install
and
you
haven't,
modified
the
template
that
it
uses
or
anything
like
that,
so
if
you
haven't
updated
that
now's
the
time
to
begin
doing
that,
I
don't
know
if
we
have
it
documented
anywhere.
I
know
I'm
working
on
some
things
that
we're
going
to
get
published
on
how
to
do
that.
C
The
short
version
is
essentially
you
shut
down
the
node,
so
do
a
cordon
and
drain
power
off
the
nodes,
you
right,
click
it
and
then
say,
update
virtual
hardware
version
and
you
can
go
to
anything
version
15
or
later
and
then
turn
it
back
on
when
it
comes
back
into
the
cluster
uncoordinate
and
move
on
to
the
next
one.
So
it's
a
bit
of
a
hassle.
You
know
it's,
it's
not
right
now.
The
way
that
I
just
described
is
manual.
C
I
don't
know
if
we
will
be
automating
that
you
can
update
the
template
that
you
use
for
your
machine
sets
so
I'll
play
update
that
template
virtual
machine
to
at
least
version
15.
So
that
way,
when
you
create
those
machine
sets
scale
up
scale
down
right,
all
of
that
will
be
taken
care
of.
Actually,
I
guess
that
is
an
automated
way
of
doing
it.
If
you
update
the
template,
you
can
scale
up.
The
machine
set.
C
C
Other
things,
oh,
this
is
one
that
a
couple
of
people
have
asked
about.
Can
you
customize
the
routes
that
are
used
for
oauth
for
the
console,
etc?
So
if
we
look
in
my
cluster
over
here
and
I
go
down
to
networking
and
routes-
and
I
want
to
show
all
projects,
you
can
see,
for
example,
that
I
have
this
oauth
openshift
route
and
it's
over
here,
it's
oauth,
apps,
dot,
cluster
name.
We've
also
got
the
console
route
right,
console
dash,
open
shift,
dash
console,
so
you
can
customize
these
now.
C
See
if
we
posted
that
one
yet
or
not,
I'm
going
to
post
that
one
so
over
here
in
the
docs,
you
can
see
that
we
edit
the
ingress
config
for
the
cluster,
and
then
we
set
this
component
route.
You
can
do
this,
for
you
can
see
here.
This
is
the
console.
You
can
do
this
for
the
download
routes.
You
can
also
do
this
for
the
oauth
route,
and
that
includes
things
like
providing
custom
secrets
or
excuse
me
certificates
right
so
serving
start
to
keep
here.
C
So
you
can
customize
those
now
in
a
fully
supported
way.
You
can
change
all
of
those
to
make
it
match.
Whatever
your
scheme
happens
to
be
somebody
on
the
stream
a
couple
of
weeks
ago
is
asking
about.
You
know
how
do
I
have
two
clusters
that
use
the
same?
You
know
subdomain
name
well,
this
is
how
you
do
it.
This
is
the
fix
to
that.
So
fortunately,
that
one
was
was
already
well
on
the
way.
So
that's
an
easy
answer
now
that
4.9
happens
to
be
released,
yeah
steven
reeves
finally,.
C
So,
let's
take
a
step
back
so
scd
is
the
data
store
that
is
used
for
everything
on
the
control
plane
right.
It
maintains
the
state
of
the
open
shift,
kubernetes
really
configuration
so
all
of
the
deployments,
pods
services
routes
right.
All
of
that
stuff
exists
inside
of
there,
so
etcd
can
grow
in
size.
You
know,
if
you
think
of
it
as
being
a
database
right,
it
grows
in
size
and
over
time
it
can
get
defragmented
and
defragmented
here
means
that
the
data
is
spread
out
and
or
fragmented.
C
Rather,
it
means
that
it's
spread
out
and
there's
gaps
inside
of
it,
so
it
can
consume
a
lot
of
extra
space,
even
though
there
may
be.
You
know
free
space
interspersed
throughout
as
well
as
it
can
increase
things
like
you
know,
latency
as
it's
having
to
find
that
so
historically,
with
openshift
4
scd
defragmentation
happened
when
the
control
plane
nodes
were
rebooted.
C
The
expectation
was
that
you
know
hey.
We
release
a
new
version
roughly
every
two
weeks,
so
most
folks
will
be
rebooting.
Their
control
plane
nodes
every
two
to
four
to
six
weeks.
It
probably
won't
be
an
issue
right.
You
know,
it'll
get
defragmented
when
the
node
comes
back
up
and
then
you
know
everything
will
be
great,
so
it
turns
out,
as
you
might
expect,
as
we've
seen
more
folks
using
larger
and
larger
cluster
sizes,
the
the
database
etcd
can
get
fragmented,
faster
and
in
between
those
times
so
with
4.9.
C
What
they've
done
is
basically
added
functionality
to
the
xcd
cluster
operator
to
periodically
defragment
lcd
based
off
of
several
criteria.
So
they
talked
about
this.
There
was
a
dedicated
slide
for
this
in
the
what's
new
session,
I'll
link
to
that,
because
it
has
some
additional
details
on
like
what
those
criteria
what
those
thresholds
are.
C
C
C
C
C
So
all
of
these
apis
have
been
deprecated
and
ultimately
removed.
So
kubernetes
open
shift
no
longer
supports
these
v1
beta
1
apis,
but
that
doesn't
mean
that
you
might
not
have
something
that
is
trying
to
use
those
so
with.
If
you
have
openshift
4.8
today,
it
will
block
updates
to
4.9.
If
you
have
anything
using
those,
so
I
would
recommend
we
did
a
dedicated
stream
to
that.
Let
me
dig
up
that
link
real
quick,
because
I
just
sent
it
out
to
somebody
this
morning.
C
This
blog
post
talks
about
the
stream
that
we
had
dedicated
specifically
to
these,
so
rob
zumski
talks
about
and
demos
exactly
how
to
go
and
look
and
see
which
of
those
apis
are
being
used
by
what
and
what
they're
being
used
by
so
that
you
can
go
and
find
those
things
you
know.
Maybe
it's
a
application
developer
and
you
know,
shame
them
into
updating
their
code
to
use
the
correct
versions.
C
C
So
long
before,
openshift
4.9,
you
know,
4.8
to
4.9
updates,
go
stable,
they'll
be
available.
Everything
will
be
good
there,
we're
working
with
our
partners,
I've
seen
a
number
of
emails
between
our
partner
management
teams
and
the
partners
for
the
ones
that
I
interact
with
saying:
hey
did
you
know
your
operator
is
still
using
these
apis
because
you
should
probably
fix
this.
C
So
all
of
those
are
getting
updated.
Just
keep
an
eye
on
that.
Again,
openshift
itself
won't
let
you
update,
if
you're,
using
one
of
those
apis
that
that
won't
work
with
4.9
kubernetes
1.22.
So
hopefully
that
answers
your
question.
If
not,
please
be
sure
or
please
feel
free
to
clarify
yeah
jp
date.
I
I
think
you
were
one
of
the
people
who
asked
me
for
that
link
this
morning.
C
So
a
lot
of
changes,
stephen
reeves
already
on
the
v1
yep
yeah,
so
all
of
the
ones
that
were
removed,
they
they've
basically
moved
from
a
v1
beta
to
basically
v1.
So
it's
just
a
generally
available
api
and
many
of
them
have
been
depleted
for
a
lot
of
versions
as
well.
I
think
I
made
the
joke
with
rob.
Zumski
of
like
back
in
the
early
kubernetes
like
1.345
days
when
there
was
pet
sets
right.
Pet
sets
were
the
original
name
for
what
we
now
call
stateful
sets.
C
You
know
getting
rid
of
all
that
craft,
so
yeah.
Let's
let
me
check
this
guy
real
quick.
No,
my
my
cluster's
still
bootstrapping,
so
I
won't
there's
nothing
really
exciting
to
show
there,
but
just
know
that
it
does
work
exactly
as
I
expected.
I
tested
it
yesterday.
So
I
I
verified
it.
I
think
we've
got
a
video
for
it.
If
not,
I
think
we're
working
on
a
video
for
that.
C
If
that's
up
before
I
post
the
blog
post
this
week,
I'll
be
sure
to
link
to
that,
but
other
than
that
you
know
we're
we're
just
a
couple
minutes
past
our
normal
ending
time.
So
we're
actually
seven
minutes.
Past
wow
time
gets
away
from
me,
so
we're
a
little
running
a
little
bit
over
I'm
happy
to
to
answer
any
questions
that
you
happen
to
have
come
in.
That's
all
the
stuff
that
I
had
to
talk
about
today.
Actually
it's
not!
C
I
do
have
a
couple
of
other
things,
but
I'm
not
going
to
spend
the
next.
I'm
not
going
to
go
too
much
over
to
talk
about
those
in
particular
metal
lb.
So
we'll
talk
about
metal
lb
with
openshift
4.9
in
an
upcoming
episode,
probably
two
weeks
from
now
again,
there
will
be
no
stream
next
week
but
be
sure
to
keep
an
eye
on
all
of
the
other
streams
that
we
have
going
on
here.
So
rasheed.
When
is
3.11
going
to
be
end
of
life,
I
want
to
say
it
is
next
year.
C
And
we
click
on
this
guy
and
up
at
the
top,
so
the
default
search
results
you'll
get
at
least
with
duckduckgo
is
for
openshift4,
so
at
the
top
here
it
has
this
link
for
non-current
openshift
versions.
So
this
will
take
us
back
to
v3
and
if
I
scroll
down
here
june,
2022
is
when
3.11
will
reach
end
of
maintenance
and
no
longer
be
supported,
and
then
it
is
in
this.
Actually,
that's
not
true.
It's
not
no
longer
supported
it.
No
longer
receives
patches
or
anything
like
that.
C
So
the
end
of
extended
life
phase,
which
is
explained
right
here,
ends
in
2024,
so
extended
life
phase.
Andrew's
interpretation-
please
read
this
for
yourself
interpret
it
for
yourself.
Ask
your
account
team
for
clarification.
If
you
needed
andrew's
interpretation
is
essentially
you
can
download
the
bits
you
can
call
us
if
it's
something
we
already
know
about
something.
We
know
how
to
already
know
how
to
fix
we'll
help
you
with
that.
If
it's
something
new
right,
if
there's
a
new
bug,
that's
discovered
there.
You
know
you
can
see
up
here
during
this
phase.
C
C
Okay,
so
thank
you
everyone.
I
really
appreciate
you
joining.
I.
I
know
that
this
one
has
been
at
least
for
me.
It's
been
a
little
more
stressful
because
chris
hasn't
been
here
to
kind
of
keep
me
on
track,
but
it's
also
been
a
little
less
stressful
because
I've
had
stephanie
in
the
background
helping
me.
So
thank
you
very
much.
Everybody
really
do
appreciate
you
tuning
in
today.
C
If
you
have
any
questions
that
we
didn't
get
addressed
that
come
up
as
you're
watching
this
you
know
after
the
live
stream,
please
don't
hesitate
to
reach
out.
You
can
reach
me
at
practicalandrew
on
twitter
or
andrew.sullivan
redhat.com
via
email.
You
can
contact
me
anytime.
The
worst
thing
that's
going
to
happen
is
just
like
here
on
the
stream.
I
might
say
you
know
I
don't
know
the
answer
to
that,
but
let's
find
the
right
people
and
and
get
you
an
answer
for
that.