►
Description
Last week the product management team went through the list of new features and changed capabilities for OpenShift 4.12 across all the many aspects. This week OpenShift 4.12 was released as generally available (GA), which means it's available for you to use for new cluster deployments!
Today's Ask an OpenShift Admin will focus on the aspects of OpenShift 4.12 that are important to administrators and go into some details that weren't covered by the PM team.
Want to learn more?
Try Red Hat OpenShift: https://red.ht/3uVxkya
Follow us:
Andrew Sullivan https://twitter.com/practicalandrew
Jonathan Rickard https://twitter.com/jrocktx1
OpenShift https://twitter.com/openshift
Subscribe to Red Hat's YouTube channel: https://www.youtube.com/redhat/?sub_confirmation=1
About Ask an OpenShift Admin:
Do you have questions? We want to answer them! Every Wednesday at 11am ET, hosts Andrew Sullivan and Jonathan Rickard sit down to answer your questions about Red Hat OpenShift.
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 on redhat.com https://red.ht/3rYz5ZS
OpenShift on GitHub https://github.com/openshift/
#RedHat #OpenShift #Kubernetes #ContainerPlatform
B
Hello
and
welcome
to
this
week's-
and
this
year's
first
ask
an
openshift
admin
office
hours.
Welcome
to
everyone
in
our
audience.
I
am
Andrew
Sullivan
the
co-host
with
Mr
Johnny
Rickard.
How
are
you
today,
sir
I'm
good
man?
How
are
you
I
just
noticed
our
names
are
not
showing
up
in
here.
Let's,
let's
turn
that
on,
let's
see
here
there
we
go
now
now
I
can
remember
who
you
are
because
it's
been
like
I,
don't
know
a
month
of
Sunday,
since
we
did.
B
Yeah,
you
know
holidays
and
all
those
other
things
they
get
in
the
way,
and
last
week
we
had
a.
We
were
planning
on
doing
a
stream
and
then
turns
out
that
we
were
overlapping
with
the
what's
new
for
openshift
4.12.,
so
serendipitously
that
kind
of
works
out
for
us
right,
because
this
week's
topic
is
exactly
that
so
yeah
yeah
lots
of
exciting
stuff
going
on
and
I'll
go
ahead
and
apologize
now
for
anyone
who
can
hear
all
of
the
noise
that's
happening
in
the
background
here.
B
A
B
B
A
B
I'm
not
going
on
around
here,
but
yeah
we're
two
weeks
in
just
just
over
two
weeks
in
we've
already
got
openshift
4.12
out
the
door.
It's
shipped
GA
yesterday,
so
you
know
we
quick
turnaround.
Sometimes
we
do
those
what's
new
and
it's
another
month
or
five
weeks
before
it
goes
out
the
door
and
sometimes
it's
literally
the
next
week
so
yeah.
A
B
Yeah
well
and
it's
funny,
because
you
know
a
lot
of
times
we'll
get
asked
you
know.
Oh,
when
is
it:
when
is
it
going
to
ship?
When
does
it
go
out
the
door
and
my
response
I'm
sure
I've
said
this
on
the
stream
before
of
it?
It's
it's
always
until
it's
literally
out
the
door
like
in
in
on
the
mirrors
ready
to
download
it
can
always
slip
and
I've
seen
it
happen
where
just
hours
before
it's
supposed
to
ship
they'll
find
an
issue,
and
you
know
they
they
have
to
push
out.
B
You
know
by
I
think
the
minimum
is
I'm
trying
to
remember
it's
it's
a
several
days,
because
it
takes
so
long
for
like
a
full
QE
cycle
to
happen,
and
then,
after
that
it
takes.
You
know
some
time
for
the
build
to
happen
and
then,
after
that,
it
takes
time
to
actually
push
everything
out
to
all
of
the
mirrors.
It's
like
a
12
hour
process
right
to
just
like
everything's
good.
All
of
the
bits
are
built.
B
A
B
A
B
B
So
I
saw
some
folks
were
having
issues
with
that,
where
you
know
I
can't
deploy
or
my
deploy,
fails
well
yeah,
because
it
can't
pull
container
images
so
yeah,
but
anyways
before
I
get
too
distracted
here
by
our
topic
today,
so
yeah.
So
this
is
a
office.
Hours
live
stream
here
on
red
hat
live
streaming,
which
means
that
Johnny
and
I
are
here
for
you
for
our
audience.
B
So
we're
here
to
answer
questions
we're
here
to
talk
about
openshift,
all
of
the
things
that
are
going
on
in
openshift
and
anything
that
happens
to
be
top
of
your
mind.
So
we
try
to
we.
We
almost
always
have
a
topic
that
we'll
talk
about
here
during
the
stream,
but
don't
let
that
limit.
The
questions
that
are
in
you
know
on
your
mind,
anything
that
you're
curious
about
anything
you
want
to
ask
about.
You
know
any
of
that
type
of
stuff
feel
free
to.
B
You
know
type,
a
question
here
in
chat,
so
we
see,
regardless
of
which
platform
you're
watching
us
on
whether
you're
on
YouTube,
whether
you're
on
Twitch
Etc.
We
see
all
of
the
chat
coming
into
our
interface
here,
so
you
can
always
Reach
Out,
always
ask
those
questions
and
then
includes.
If
you
are
not
live,
you
know
lots
of
folks
watch
us
watch
the
recording
or
maybe
you're
listening
to
us
in
the
in
the
car,
or
something
like
that.
Podcast
like
you're,
welcome
to
reach
out
to
us
after
the
fact
as
well.
B
So
if
I
can
find
the
right
button
here,
you
can
reach
out
the
easiest
way
to
reach
both
Johnny
and
I
is
probably
email,
so
you
can
find
me
at
or
andrew.sullivan
redhat.com
super
easy,
first
name
dot.
Last
name
in
that
respect,
I
will
say:
email
comes
in
waves,
so
sometimes
I'm
able
to
get
to
email
quickly,
sometimes
it'll.
Take
me
a
little
bit
of
time
to
turn
through
them
also
on
social
media.
All
the
various
places
at
practical,
Andrew
I
will
admit
that
Twitter
is
not
my
favorite
platform.
B
B
B
A
B
Unless
you
got
some
things,
Johnny,
no
okay,
yeah,
so
really
the
big
one
that
I
wanted
to
bring
up
is.
There
was
an
internal
thread
that
somebody
started
around.
You
know
my
my
customer
is
over
committing
CPU
resources
and
memory
resources
inside
of
their
cluster.
You
know
this
is
bad.
How
do
I?
How
do
I
explain
to
them
that
this
is
bad
and
the
response
that
they
got
both
from
me
as
well
as
from
several
other
folks,
is
well,
it's
not
inherently
bad
to
over
commit
your
resources.
B
B
So
way
down
here
at
the
bottom
of
our
describe,
we
have
this
resource
output
right,
CPU
and
memory,
and
we
can
see
that
roughly
half
52
percent
of
both
CPU
and
memory
on
this
node
is
allocated
to
requests.
Basically,
it's
put
into
a
reserve
status
so
that
nothing
else
can
use
those
resources,
but
we
don't
have
any
limits
so
kubernetes
and
therefore
openshift
won't.
Let
use
have
more
than
100
percent
of
the
resources
tied
up
in
requests
and
right.
B
That's
just
if
a
request
remember
is
tantamount
to
a
a
reservation
and
you
can't
reserve
more
than
100
of
the
resource.
Just
doesn't
work
that
way.
However,
limits
can
be
over
committed.
You
can
see
it
says
it
right
here,
so
I
could
have
limits
in
place.
So
maybe
my
pod
requests
right.
It's.
It
says:
please
reserve
for
me:
500
mL,
CPU,
half
a
CPU
but
put
a
limit
of
64
CPUs
right.
So
that
way
it
can
go
all
the
way
up
to
consuming
some
massive
amount.
That
would
show
up
here
as
being
whatever
percent.
B
B
So
it
does
mean
something
in
that
it
is
an
indication
of
how
much
workload
you
have
and
how
much
contention
there
could
possibly
be.
That
being
said,
it's
not
a
terribly
helpful
metric
in
that
respect,
because
pods
don't
have
to
have
limits,
so
you
could
have,
and
this
host
does
I
mean
you
can
see
here,
there's
a
whole
bunch
of
PODS
41
pods
on
here.
Not
a
single
one
of
them
has
a
request
or
excuse
me
a
limit.
B
B
So
we
want
to
be
able
to
use
as
much
of
the
resources
much
of
the
CPU
and
the
memory
as
we
can
right,
I'm
paying
for
this
100
of
those
resources,
a
hundred
percent
of
those
four
CPUs
and
that
16
gigabytes
of
RAM,
whether
or
not
I
use
one
percent
of
it
or
one
hundred
percent
of
it.
So
as
an
administrator,
if
I'm
artificially
limiting
myself
to
say
50
utilization,
that
means
if
this-
and
this
is
running
an
AWS
I-
have
no
idea
what
the
hourly
cost
is.
B
But
let's
say
it's
two
dollars
an
hour
well
I'm
effectively,
if
I
limit
it
to
50
utilization,
I
have
to
pay
for
two
of
those
to
reach
the
capacity
the
the
resources
of
one
I
hope
that
makes
sense.
I'm.
Looking
at
you
in
my
window
here,
Johnny
make
to
see
if
that
makes
sense,
yeah
it
does
so.
It
means
that
right,
it
gets
more
expensive
to
add
more
resources,
so
I
want
to
be
able
to
utilize
as
much
of
those
as
possible.
So
over
commitment
doesn't
necessarily
mean
anything
bad.
B
A
little
bit
of
contention
is
fine,
especially
with
CPU
right
because
effectively,
you
know,
hey
the
Pod,
it's
it's
requesting
two
CPUs,
but
there's
only
one
and
a
half
CPUs
worth
of
resources
available,
so
it
might
get
throttled
a
little
bit.
It
might
take
a
little
bit
longer
to
run.
This
is
perfectly
normal.
This
is
what
happens
in
literally
every
not
just
Linux
system
every
operating
system
when
it
gets
busy
the
processes
that
are
running,
compete
for
time
in
Linux
in
Rel
and
core
OS.
B
It's
the
CFS,
the
completely
Fair
scheduler
I
think,
is
what
it
stands
for.
It
basically
allocates
that
time
to
those
processes.
So
it's
normal.
It's
expected
it's
okay,
where
it's
bad
is
when
it
starts
being
detrimental
to
the
application,
and
that's
where
we
want
to
be
careful
and
utilize
other
features
of
the
platform
in
order
to
well
make
it
be
as
good
as
possible.
Right
sorry
for
my
poor
English
there,
so
things
like
the
D
scheduler.
So
if
we
look
at
let's
come
over
here
to
docs.openshift.com.
B
But
so
the
the
descheduler
here
has
this
life
cycle
and
utilization
policy.
Where
effectively.
If
it
sees
that
a
pod
has
or
a
node
has
low
node
utilization
or
high
node
utilization,
it
will
begin
to
take
action
to
balance
those
out,
but
CPU
is
considered
a
compressible
resource,
so
it
won't.
There
is
no
equivalent
of
like
the
out
of
memory
killer
for
CPU,
so
we
have
to
have
something
like
eviction
or
ideally,
pod
health
checks.
B
Oh
there
we
go
monitoring
application
elf,
so
with
a
liveness
health
check
configured
on
our
pods
that
health
check
can
do
things
like
you
know,
hey
is
my
application.
Responding,
you
know,
is
the
latency
adequate?
Is
it
getting
enough
resources?
Is
it
doing
all
the
things
that
it
needs
to
do
to
determine
that
the
the
application
is
healthy
and,
if
not,
maybe
it's
being
throttled
and
therefore
application
latency
went
from
five
milliseconds
to
five
seconds.
B
It
basically
can
declare
it
as
unhealthy
and
then
results
in
it
being
restarted
and
potentially
terminated,
which
means
that
it
would
go
through
a
rescheduling
process
and
land
on
a
lower
utilized
node
assuming
a
lower
utilized
node
is
available
so
long
story
short
here
over
commitment
is
good
for
CPU.
We
want
to
be
able
to
utilize
all
of
those
resources
to
the
maximum
potential
that
you
feel
safe
with
again.
B
B
A
B
B
Yeah
Jamie,
that's
a
good
point,
be
careful
of
going
back
to
the
same
defaults
and
you
know
being
able
to
choose
or
decide
what's
good
and
bad
for
you
alert
fatigue
is
real
and
it's
a
real
concern.
So
just
be
careful
of
that.
I
will
say
very
much
to
your
point.
Jamie
of
it
is,
there
is
risk
associated
with
it
right.
The
reason
why
those
emails
are
generated.
The
reason
why
the
default
alert
configs
are
set
at
whatever
value
they're
set
at
that
I,
don't
remember
off.
B
The
top
of
my
head
is
because
it
adds
risk,
and
the
same
is
true
of
virtualization
environments
right
when
I'm,
using
VMware
or
rev
or
whatever
I
happen
to
be
using
when
I'm
over
committing
those
resources
on
that
host.
You
know
with
virtual
machines,
it's
going
to
start
to
generate
errors,
especially
as
the
virtual
machines.
You
know,
utilize
more
and
more
of
those
resources
and
that's
not
necessarily
A
Bad
Thing,
but
we
want
to
adjust
those
for
our
thresholds.
So
I
said
a
moment
ago.
It's
it's
about
risk.
B
B
You
know
that
can
help
to
right
size
as
much
as
it
can.
You
know
right
size
that
and
there's
other
tools
that
do
similar
types
of
things
as
well
as
partners
that
do
those
types
of
things.
Turbonomic
is
one
that
I
was
recently
talking
with
the
other
side
is
The
Human
Side
so
and-
and
this
was
how
Johnny,
when
you
and
I
were
co-workers
as
administrators
a
decade
ago.
This
is
how
we
we
handled
it
right
of.
A
B
You've
requested
you,
you
you're
I,
you've
reserved
all
these
resources,
they're
not
available
to
anyone
else,
and
you
know
you
please,
please
give
us
these
back
like
you,
you
obviously
don't
need
them,
I,
guess,
I!
Guess
another
option
would
be
pay
as
you
go
right.
That's
how
the
hyperscalers
handle
it
it
you
paid
for
it
it's
yours,
but
you
paid
for
it,
whereas
I
think
most
organizations.
Certainly
every
organization
I've
worked
for
as
an
admin.
We
we
never
did
charge
back.
A
Yeah
every
I
think
I
mean
even
even
like
when
I
left
before
I
came
to
Red
Hat
right,
like
we
were
always
talking
about
doing
chargeback,
but
just
never
like
there's,
there's
always
the
we
should
implement
this
because
it'll
help
it
get
money
back
from
some
of
these
bigger
organizations,
but
they
just
never
implement
it.
Yeah.
It's
a
great
idea,
hard
to
hard
to
put
in
practice.
B
Yeah
yeah
because
there's
also
the
budget
stuff
right
and
you
know,
but
it
has
allocated
the
money
for
it.
Not
you
know:
Finance
Finance
has
money
to
do
finance
things,
not
I.T
things.
So
there's
it's
a
very
complicated
problem,
foreign
yeah.
So
let's
jump
back
up.
Let's
talk
about
lucians
and
apologies
for
butchering
anybody's
name
comment,
so
why
why
openshift
and
not
any
other?
You
know
kubernetes
out
there.
B
So
this
is
a
question
that
we
get
I
wouldn't
say
often,
but
we
do
get
its
on
a
regular
basis,
and
my
response
has
always
been
the
same
in
that
openshift
is
kubernetes,
so
anything
you
do
with
any
other
kubernetes
you
can
do
with
openshifts.
The
differentiator
with
openshift
is
that
it
is
an
opinionated
kubernetes.
B
So
what
that
means
is
that
red
hat
out
of
the
box
we
take
kubernetes,
we
take
an
sdn,
we
take
CSI
drivers.
We
take
all
of
these
things
that
you
need
to
have
a
production,
kubernetes
deployment,
and
we
provide
that
we
box
it
all
up.
We
integrate
everything.
We
test
everything
we
verify
everything
is
going
to
work
together
and
then
we
support
all
of
that.
B
But
and
I
think
Docker
used
to
use
this
phrase,
It's
batteries
included
but
replaceable.
So
you
don't
like
openshift,
sdn
or
ovn
kubernetes
that
we
use
that's
fine,
go
you
know,
tigera
and
Calico
works
with
openshift.
You
can
deploy
that
and
you
can
use
that.
There's
nothing
wrong
with
that.
You
want
to
use
a
third-party
CSI
driver.
You
know
NetApp
Dell
EMC,
one
of
the
ones,
that's
not
included
out
of
the
box.
That's
great!
You
can
absolutely
use
those
and
you
don't
break
support
in
doing
that.
A
Yeah
and
specifically
around
the
question
about
being
like
tanzu
versus
openshift
right
I
think
it
comes
down
to
like
what
your
organization's
preferences
for
vendors
are.
You
know
if,
if
you've
already
bought
completely
into
you,
know
Red
Hat
openshift,
then
it
probably
makes
sense
to
stick
with
openshift
versus
tanzu
and
I.
A
Don't
know:
I
I,
don't
have
enough
information
on
tandu
to
be
able
to
say
like
there's
a
here's,
a
you
know
future
for
future
parity
example:
hey
you
get
this
with
openshift
and
you
don't
get
this
with
tanzu
or
whatever,
but
I
think
it
comes
down
to
organizational.
You
know,
taste
appetite
for
for
vendors.
You
know
if
you're
all
in
on
on
VMware
probably
makes
sense
if
you're
all
in
on
red
hat,
probably
makes
sense
to
stick
with
openshifts.
B
I'm
not
gonna
lie
I'm,
you
know,
I'm
wearing
a
red
hat
shirt.
There's
a
little.
You
know:
Red
Hat
openshift
logo
in
the
corner
of
the
stream.
Here
you
know
it's
I
I'm,
a
little
biased
and
that
and
that's
okay
but
yeah.
It's
a
I
I
would
always
encourage
anyone
and
everyone.
If
you're
curious,
go
check
out
learn.openshift.com.
Let
me
reshare
my
screen.
I'll
show
you.
B
I
click,
the
start,
button
notice,
I,
haven't
logged
in
I'm,
not
logged
into
anything
I
hit
launch
and
aside
from
it,
apparently
the
pool
is
exhausted,
so
it'll
take
about
eight
minutes,
but
when
it's
done
I'm
going
to
be
dropped
right
into
an
openshift
cluster.
So
if
you
want
to
experiment
with
it,
you
want
to
test
it.
You
want
to
check
it
out.
You
can
use
these
they'll
walk
you
through
different
aspects.
You
can
get
Hands-On
literally
within
if
the
pool
isn't
exhausted
like
15
seconds.
B
Otherwise
it's
8
minutes
and
15
seconds
so
I
always
encourage
everyone
right.
Don't
don't
listen
to
me,
don't
listen
to
Johnny
like
go
and
experience
it
for
yourself
and
if
you're
going
to
be
at
Summit,
which
I
think
Summit
call
for
papers
closed.
Finally,
but
I
do
know
that
they're
talking
about
having
like
I,
want
to
say
30
or
36
different
Hands-On
Labs
at
Summit
this
year.
B
So
if
you
have
the
opportunity
to
come
to
Summit
you're
going
to
have
a
chance
to
get
to
get
Hands-On
as
well-
and
you
know
be
in
a
proctored
environment
and
go
through
and
see
See
all
the
stuff
that
we
got
going
on
there
I
haven't
seen
a
list
yet
I
know
they.
They
were
asking
for
a
lab
suggestions
as
well.
So.
A
B
I
I
hope
I
think
that
they
finally
listen
to
all
of
us
who
are
complaining
about
it.
It
not
being
technical
enough
not
being
enough
of
a
user
conference
like
it
was.
You
know
five
years
ago,
you
know
I,
think
coveted
and
being
virtual
and
all
that
kind
of
through
through
all
of
that's
through
a
loop
and
made
it
hard
yeah
I'm
I'm,
looking
forward
to
Summit
this
year,
I
I
think
it's
going
to
be
great.
A
A
B
Dwayne,
could
this
be
automated
with
ansible?
Yes,
yes,
of
course,
so
openshift
is
kubernetes,
so
the
ansible
kubernetes
modules
you
can
use.
You
can
use
to
automate
to
do
all
of
the
things
inside
of
there,
as
well
as
ansible
itself.
Ansible
automation
platform
can
be
deployed
onto
openshift,
so
you
can,
you
know,
host
host
it
as
well
as
use
it
to
manage
the
platform
and
then
there's
endless
debates
about
you
know.
B
B
B
B
Yeah
I
know
it's:
it's
like
riding
a
bike
right,
yeah,
it's
funny,
I
was
setting
up
for
the
stream
and
I
was
I,
was
updating
our
our
graphic
that
we
used
with
the
you
know
our
faces
and
the
title,
and
all
that
and
the
last
date
was
like
December
14th
I'm
like
gosh.
Has
it
really
been
that
long
since
we
had
an
ask
an
openshift
admin,
wow.
B
Has
and
yeah
lots
of
stuff
going
on
so
the
way
that
these
usually
work
is
the
product
management
team
will
have
a
live
stream.
That
is
lately
they've,
been
about
90
minutes
in
length
where
they
go
through
kind
of
all
of
the
different
new
features.
New
functions
in
the
release
across
the
whole
gamut,
the
the
breadth
of
openshift.
B
B
So
you
can
go
to
this
oops
live
and
go
here,
and
there
is
a
what's
new
in
openshift
4.12,
so
that
is
the
live
stream
that
they
did
last
week.
You
can
also
go
to
red.ht,
slash
next
and
new,
no
spaces
that'll.
Take
you
to
the
landing
page
here
or
what's
new
in
openshift
4.12,
as
well
as
a
link
to
the
slides
here,
so
that
is
and
we'll
paste
that
link
into
the
chat
here.
B
So
both
of
those
ways
you
can
see
that
recording
you
can
get
access
to
the
slides
and
all
that.
But
what
Johnny
and
do
and
I
do
is,
after
that,
we'll
go
in
and
we'll
identify
the
things
that
we
we
thought
were
interesting
or
maybe
need
some
more
details
or
things
that
are
just.
We
want
to
make
sure
we
call
out
to
you,
know
our
audience
to
you
all
to
administrators.
B
Of
course,
please
feel
free
to
ask
questions
as
we
go
through
or
anything
like
that,
because
effectively,
John
and
I
are
taking
notes
for
future
episodes
as
well.
So
we've
got
a
list
of
six
I.
Think
I
see
six
here,
six
things
that
we're
going
to
effectively
Deep
dive
on
so
the
disconnected
agent
based
installer
ebpf
based
Network
observability,
core
OS
layering
right.
B
All
of
these
things
that
we
want
to
Deep
dive
into
and
we'll
dedicate
a
whole
stream
to,
but
we're
cramming
a
whole
bunch
of
things
that
maybe
don't
need
that
based
off
of
our
judgment,
but
that
doesn't
mean
that
they
might
not
be
interesting
to
you
all
and
you
don't
want
some
more
information
on
them.
So
absolutely
if
you
have
questions,
if
you
want
more
details,
we'll
we'll
tackle
all
that
foreign.
B
Let's
see
just
looking
yeah
logging,
five,
six
we're
working
on
that
one
Our
Hope
nine
we've
been
working
with
those
folks
for
probably
three
months
now:
Johnny
trying
to
get
them
to
come
on
and
I,
think
holidays
and
release
schedules
and
other
stuff
have
gotten
in
our
way.
But
we're
we're
close
we're
so
close.
B
All
right,
so,
let's
see
first
one
on
my
list
here,
I,
you
know
I
I
sat
here
and
for
anybody
who
watched
the
live
stream
I
was
in
chat,
I
was
listening
to
it
the
whole
time
and
all
this
I
honestly
missed
when
they
presented
this
during
the
live
stream.
I
knew
I
knew
it
was
going
to
be
announced,
but
I
missed
it
during
the
live
stream,
and
that
is
that
eus
releases,
so
the
even
numbered
releases
4.12
being
one
of
those
now
have
support
for
24
months.
B
So
previously
all
releases
even
or
odd,
eus
or
not
were
supported
for
18
months,
starting
with
4.12.
Those
even
numbered
releases
are
supported
for
24
months
and
you
still
have
the
accelerated
EOS
to
us
update
path.
So
to
me
this
is
a
big
one.
It
means
particularly
for
eus
releases.
You
know
EOS
to
us.
We
now
have
that
much
more
overlap
between
supported
releases
so
that
you
know
you
all.
You
can
take
your
time
in
testing
and
validating
and
all
those
you
don't
have
I
think
before
it
was
like
three
months.
B
B
Let's
see
now
that
Vector
is
GA.
Will
fips
clusters
be
eligible?
That's
a
good
question.
I,
don't
know.
Let's,
let's
take
a
note
on
that
one
Johnny.
We
can
ask
the
logging
folks
because
I
I,
literally
I
just
saw
an
email
coming
from
them
like,
as
we
were,
hitting
the
go,
live
button
trying
to
schedule
that
lost
what
do
I
need
to
do
to
work
in
Red
Hats
apply
all
of
our
jobs.
All
of
our
stuff
is
available
on
the
Red
Hat
jobs
portal.
B
So,
if
you
happen
to
know
anybody
at
Red
Hat,
you
can
also
ask
them,
or
you
know,
if
they
and
like
it,
has
to
be
somebody
that
you
know
but
like
ask
them.
If
they'll
do
a
referral
for
you,
that's
always
a
great
way
again
the
recruiter
team.
This
I
saw
an
email
earlier
today,
where
it
was
something
like
you
know.
The
the
best
way
to
find
new
red
Hatters
is
for
existing
red
adders
to
to
refer
them,
so
yeah
definitely
reach
out.
B
If
you
know
anybody,
you
know
identify
the
positions
that
you're
interested
in
Johnny.
Thank
you
for
linking
the
jobs
portal
there
and
then
just
submit
against
it.
So
it
is
again
I'm,
biased,
red
Hat's,
a
great
place
to
work
there's
so
it
depending
on
the
role
depending
on
the
position.
It
can
be
competitive.
A
Oh
yeah,
but
yeah,
and
just
so
it
could
take
time
too
right
like
it's.
If
you,
if
you
apply
for
a
job
it,
it
can
take
some
time,
especially
coming
back
after
coming
back
after
the
holidays
into
the
quarter
beginning
the
quarter
of
that
type
of
stuff.
So
don't
don't
lose
faith.
If
you
don't
get
a
response
in
a
day,
you
know
just
give
it
a
little
bit
of
time.
You'll
you'll
be
all
right.
Yeah,
definitely.
B
B
Openshift
sdn,
as
the
name
implies,
has
a
community
of
basically
one
red
hat
right.
We
we
created,
we
maintain.
We
contribute
right,
it's
open
source,
of
course,
but
other
kubernetes.
This
is-
are
not
using
openshift
sdn
as
their
sdn,
OV
and
kubernetes
on
the
other
hands
is
used
by
others,
other
kubernetes
distributions.
So
we
gain
more
Community.
We
gain
the
ability
to
participate
as
well
as
ADD
features.
Let
me
find
my
documentation
link
here
so
we'll
be
in
kubernetes.
B
So
for
context,
for
a
little
bit
of
context,
some
deployments
single,
node,
openshift,
Rosa
anytime,
you
needed
to
do
windows
nodes
which
required
hybrid
networking.
You
already
had
to
use
OV
and
kubernetes
for
those.
If
you
wanted
to
do
like
the
ipsec
encryption,
stuff,
I
think
that
was
in
four
or
eight
or
four
nine
when
they
added
that
where
traffic
traversing
between
nodes
across
the
sdn
is
encrypted
using
ipsec
that
relies
on
ovn
kubernetes
so
effectively.
B
A
B
B
It's
only
candidate
updates
that
are
not
supported,
but
I
understand
most
folks
want
to
wait
for
stable,
but
if
you
upgrade
it
will
not
automatically
migrate
your
network,
you
will
stay
on
ovn
or
excuse
me,
openshift
sdn,
until
you
choose
to
migrate,
that
migration
is
it's,
you
can
migrate
the
network,
but
it
is
I'll,
say
minimally
disruptive.
Basically,
it
has
to
reboot
the
nodes
in
order
to
do
it
they
are
working
on
and
I,
don't
remember
the
precise
release,
but
they
are
working
on
doing
a
a
quote.
B
Unquote
live
migration
which
feels
weird
saying,
because
I
always
use
that
in
context
of
like
rev
or
openshift
fertilization,
but
a
live
migration
from
openshift
sdn
to
OV
and
kubernetes
that
is
non-disruptive,
so
yeah,
so
ovian
kubernetes
default
for
all
new
installs.
It
is
not
automatically
changed
when
you
upgrade
to
4.12
and
the
upgrade
is
a
manual
process.
B
So
you
have
to
reboot
the
nodes
today
because
it
is
it
from
what
I
understand
it
wants
to
change
one
of
the
settings.
Basically,
it
has
to
disconnect
all
of
the
pods
and
then
reconnect
all
the
pods
and
they
don't
either
don't
have
the
testing
or
don't
have
the
capability
today
of
doing
that,
without
basically
stopping
everything
and
the
easiest
way
to
stop
everything
is
to
reboot,
so
I
I,
don't
know
again,
I,
don't
know
if
it's
testing
or
if
it's
adding
a
new
feature
but
it'll
get
there.
B
B
Yeah
so
device
Edge
is
the
officially
productized
name
for
what
we
used
to
call
Mini
shift
in
combination
with
Rel
for
Edge,
and
you
can
see
it
describes
that
here
in
the
in
the
first
paragraph.
So
what
does
this
mean?
So
relfer
Edge
is
Red
Hat
Enterprise
Linux
that
uses
RPM
OS
tree
for
basically
OS
management,
so
very
much
similar
to
core
OS.
You
can
create
Rel
for
Edge
instances
using
the
Rel
image
Builder,
you
can,
you
know
push
those
out,
you
can
push
out,
updates,
ansible
is
and
ansible
automation.
B
Platform
are
usually
the
management
platforms
for
Rel
and
rail
for
Edge,
as
well
as
satellite
I
should
be
sure
to
mention
that
so
the
satellite
people
don't
get
upset
with
me
and
then
minishift
is
I'm
trying
to
think
of
the
best
way
to
describe
this.
It
is
kubernetes
with
the
openshift
API,
so
you
know,
we've
had
many
shift.
We've
talked
about
this
a
couple
of
times
on
the
stream
right.
We
had
the
engineering
folks
on
here.
First,
we
had
Henri
and
Diego
come
on
and
show
us
it
being
used.
B
I
think
we've
got
a
couple.
Other
things
longer
term.
Thank
you,
microshift
for
correcting
me
in
chat,
Johnny
I,
keep
saying
minishift
mini
shift
was
the
thing
to
stand
up
single
nodes,
yeah
Dave
correcting
me.
Words
are
hard
today,
folks,
I'm,
sorry,
so
yeah
microshift
is
kubernetes.
The
the
Red
Hats,
the
the
oh,
the
kubernetes,
underneath
openshift,
along
with
the
openshift
API.
B
B
B
So,
for
example,
there
is
no
like
full
CVO
cluster
version
operator
and
all
of
the
services
and
things
that
cluster
version
operator
deploys.
So
to
get
an
idea
of
that.
Oc
get
Co,
I.
Think
all
of
these
things,
most
of
them
aren't
going
to
show
up
in
a
a
micro
shift
deployment
in
a
red
hat
device.
Edge
deployment,
olm
operator,
lifecycle
manager,
isn't
there
so
there's
no
operator
catalog,
all
that
type
stuff.
It's
designed
to
be
able
to
deploy
containerized
applications
out
to
the
edge
using
the
kubernetes
API,
and
it's
it's
another
option.
B
The
ansible
folks
actually
had
a
use
case
where
they
used
openshifts
and
pipelines
to
build
a
container
image,
and
then
they
use
ansible
to
pull
that
container
image
out
to
a
device
running
at
the
edge
and
run
it
using
podman
so
kind
of
similar
type
of
use
case.
They
were
just
using
ansible
and
podman
instead
of
ACM
and
which
ACM
I
think
microshift
device
Edge
is
on
their
roadmap.
You.
B
B
A
B
B
A
B
B
So
what
I've
done
here
is
I,
just
I
connect
it
to
a
different
cluster
that
I've
got
running.
If
you've
got
good
eyes,
you
can
see
rhe
here.
This
is
one
of
the
Clusters
that
we'll
be
using
for
Tech
Exchange
to
show
off
some
things
and
what
this
cluster
has
is.
It
has
hosted
control
planes,
AKA
hypershift
deployed
inside
of
it.
B
B
B
These
are
all
the
pods
that
make
up
the
control
plane
for
this
cluster,
so
they're
the
control
plane
for
cluster
roads
is
hosted
in
the
Americas
cluster
and
then
the
nodes
for
cluster
roads.
The
the
worker
nodes
are
managed
from
this
cluster
as
well.
So
I
kind
of
do
OC,
gets
machine
I
think
it
picks
over
the
yeah.
It
takes
over
the
defaults
crd
name.
If
you
can
see
that
I
have
four
nodes
that
have
been
deployed
from
that
and
those
are
matched
via
node
pool.
B
B
B
You
don't
need
so
you
just
need
openshift
to
do
this
along
with
the
multi-cluster
engine,
so
just
to
show
that
if
I
come
here
to
operators
and
go
to
operator
Hub
and
search
for
multi
LTI
Oscar,
we
have
this
multi-cluster
engine
for
kubernetes.
So
this
is
what
you
need
to
install
and
from
here
you
can
then
go
in
and
begin
deploying
hosted
control
planes
just
remember
that
their
Tech
preview
or
excuse
me
yeah
Tech
preview,
not
Dev,
for
you
there's
a
bunch
of
resources.
We
had
a
Dell
on.
B
We
talked
about
it.
There's
a
bunch
of
resources
on
cloud.red
hat.com,
we'll
kind
of
I'll
do
my
best
to
summarize
these
and
we
can
share
them
out.
Bench.
Mouse
has
done
several
blogs
on
like
how
to
use
this
with
internal
or
with
local
resources.
So
you
know
you're
running
a
bare
metal
cluster
and
you
want
to
host
control
planes
for
other
virtual
and
bare
metal
clusters
inside
of
there.
Excuse
me
so
how
to
do
that,
how
to
make
all
of
that
happen
really
cool
stuff.
B
I
think
this
has
a
lot
of
potential
for
increasing
density
of
clusters
that
are
deployed
and
optimizing
resources.
I
also
think
it's
really
cool
for
things
like
you
can
host
the
control
plane
in
your
central
Data
Center
and
have
the
worker
nodes
out
at
the
edge
where
again,
resources
might
be
constrained.
So
it's
like
a
almost
an
extreme
version
of
remote
worker
nodes.
A
B
Yeah
yeah
I
we're
going
to
be
talking
about
it
at
Summit.
I
know,
I've
seen
some
of
the
session
proposals.
I
know
it's
something
that
we
want
to
talk
about
there
for
any
red
Hatters
who
are
watching
I.
Think
we
have
several
sessions
about
it
at
Tech,
Exchange,
coming
up
in
two
weeks
three
weeks,
so
keep
an
eye
on
that
there.
You
know,
we've
been
I've
been
experimenting
with
it
breaking
it.
B
B
So,
for
example,
when
you
want
to
like
manage
your
your
cluster
I'm
looking
at
the
wrong
thing
here
so
like
when
I
want
to
configure
authentication
for
this
cluster
I
configure
authentication
on
this
host
cluster
resource,
so
just
like
I
would
go
in
today
and
I
would
say
you
know.
Oh
you
know,
I
create
this
oauth
edit.
B
The
oauth
object
to
add
you
know
whatever
my
authentication
scheme
happens
to
be
with
hosted,
control
planes
I,
basically
add
in
the
yaml
that
stanza
into
the
hosted
cluster
objects,
and
it
pushes
out
that
configuration
so
as
administrators.
It
means
that
we
can
like
scale
really
easily.
All
of
a
sudden.
You
know
before
not
that
it
was
hard
to
do
something
like
that
before,
but
it
was
easier
to
use
something
like
ACM
and
I
use.
Acm
and
I
create
a
policy,
and
if
that
says,
configure
this
oauth
object
inside
of
all
my
clusters.
B
B
Let's
see
moving
down
the
list
console
so
I
just
discovered
this
this
morning,
I
was
I,
saw
it
in
the
slide
deck
that
they
that
they
used
and
I
was
like.
Oh
Dynamic,
plugins.
That's
finally,
going
GA,
you
know:
hey,
that's,
really:
cool
Jafar
who's,
one
of
the
folks
I
work
with
on
solution
patterns.
He
created
a
dynamic
plug-in
to
show
like
the
the
patterns
like
directly
in
the
interface
and
be
able
to
click
on
them
and
use
them
Johnny.
B
So
to
me
this
is
useful
in
a
couple
of
different
ways
as
administrators.
You
know,
if
I'm
you
know
an
administrator
or
the
administrator
team
for
my
open
shifts
and
I'm,
enabling
others
to
access
those
I
can
provide
access
to
resources
that
are
important
to
them.
Hey
here's,
our
internal
Wiki
on
with
documentation,
and
it
shows
up
as
like
a
menu
navigation
item
over
here
instead
of
up
here
where
they
would
previously
be.
B
You
know
hey
if
I'm
maybe,
and
we
used
to
when
I
was
a
VMware
administrator
right.
The
the
NetApp
plugin
into
vcenter
was
great
because
I
could
go
in
and
I
could
access
all
of
my
storage
resources
just
from
vcenter.
Well,
I.
Don't
know
that
any
storage
vendors
have
done
this
yet,
but
in
theory
our
partners
could
go
in
and
they
could
add
in
you
know,
say
under
storage.
B
You
could
have
a
storage
vendor
button
here
that
goes
in
and
you
can
see
you
know,
maybe
what
your
Cloud
native
storage
so
think
and
who
is?
Why
is
the
name
escaping
me?
Johnny
odf,
odf
already
has
this.
If
you
deploy
odf,
they
have
a
button
here
where
you
can
go
in
and
you
can
see
the
status
of
the
odf
cluster.
B
Other
partners
can
now
do
that
type
of
stuff
or
at
a
minimum.
You
can
you
know
you
the
administrator.
If
you're
also
the
storage
admin,
you
can
go
in
and
add
a
thing
here
or
when
you
click
on
it.
It
pops
up
your
storage,
admin
interface
or
something
like
that.
So
I
don't
know.
I
was
strangely
excited
to
see
something.
So
simple
I
say
so
simple
knowing
full.
Well
then,
in
the
background,
there's
a
lot
of
complexity.
There.
A
Yeah
I
think
like
especially
with
as
much
Focus
as
we're
putting
on
the
console
like
you,
want
people
to
be
able
to
have
like
that
one
chain
of
glass
for
all
the
things
right
so
I
think
that
that's
that's
a
cool
feature
that
they
have.
The
way
that
we
do
it
right
now
is,
if
you
look
at
like
the
nine
square
drop
down,
We'll
add
like
a
button
up
there.
A
B
B
So
the
other
thing
that
happened
in
the
console
is
they
added
the
ability
to
customize
the
things
that
you
see
so,
for
example,
and
I
haven't
tried
this
yet
so
we're
going
to
try
this
live
you'll
see
up
here.
I've
got
this
show
getting
started
resources,
so
this
stuff
up
here
we
also
have
I,
can
switch
over
I've
got
the
developer
perspective.
B
So
I
can
select
this
guy
come
up
here
and
go
to
customize
literally.
This
is
the
only
one
that
has
this
by
the
way,
so
I
can
go
to
customize
and
I
can
change
a
bunch
of
different
aspects
of
my
console.
For
example,
I
can
turn
off
the
developer
perspective,
that's
cool
saved,
and
now,
if
I
refresh
this
well
I
say
that
I
don't
know
how
long
it
actually
tries
or
it
takes
to
apply
there.
We
go
I
had
to
do
a
force
refresh
or
a
full
refresh.
A
B
So,
and
there
was
an
email
chain
about
this
internally
two
weeks
ago,
somebody
was
asking
hey
my
customer,
you
know
my
they
don't
they
don't
want
to
have
the
developer
perspective,
they
they
want
to
turn
it
off,
and
the
response
was
just
wait:
it'll
be
there
I,
don't
know
and
I'm
not
gonna.
Try,
because
it
could
break
our
cluster
that
we're
using
here
I
wonder
what
happened
if
I
set
both
of
these
to
disabled.
B
B
So
if
I
go
back
to
my
configuration
here
and
go
back
to
console
and
customize
I
think
what
will
happen
here
is
if
I
select
all
of
these
guys
and
and
move
them
over
and
it's
saved
again
and
I'll
go
ahead
and
do
a
shift
refresh
just
to
be
sure,
I
think
that'll.
Disable
these
things
I'm
not
I'm,
not
positive,
yeah,
see
at
least
one
of
them
disappeared.
B
B
Yeah
yeah,
it's
you
know
the
little
things
I
love
the
slide
that
they
that
they
had
in
the
presentation,
too,
is
like
you
know,
console
user
happiness
or
something
like
that
and
I
think
it
was
Ali
who
was
presenting.
He
had
like
two
dozen
rfes
that
they
had
added
in
and
and
stuff
like
that,
so
yeah
Council,
always
new,
exciting
things
happening
in
here.
B
If
you
didn't
see
the
or,
if
you
aren't
aware
in
4.11,
they
did
remove
the
grafana
interface,
so
you
might
want
to
go
back
and
look
at
that
stream.
If
you
weren't
aware
of
that,
that's
the
other
big
thing
that
I
try
and
bring
up
regularly
like
there
is
no
longer
a
grafana
interface
for
looking
at
the
metrics.
There
is
just
the
the
dashboards
here
in
openshift
console
itself.
B
So
definitely
go
back
and
check
that
out,
including
we
did
a
stream
with
deploying
a
custom.
Grafana
dashboard
as
well
I'll
have
to
dig
that
one
up.
We've
got
it
right
here.
Let's
see
what
time
is
it?
Okay,
we've
got
one
two
three
we're
like
halfway
through
the
list
here
and
15
minutes
to
go
Johnny.
B
B
B
So
if
we
come
over
here
to
installing
and
I
scroll
down
to
vsphere
and
I,
don't
know
if
it's
in
the
preparing
here
we
go
right,
so
you
must
use
Virtual
hardware
version
15
or
later
you
must
use
vsphere,
7.0.2
or
later
version.
6
is
no
longer
supported
foreign,
be
aware
that
openshift
4.13
will
require
vsphere,
7
Dot
7u2.
B
So
I
don't
remember
what
version
that
is
what
seven
dot
whatever
that
is,
but
7.0
U2
is
required
and
this
has
to
do
with
CSI.
So
effectively
we
are
so
CSI
is
and
has
been
the
defaults.
So
as
we
are
removing
the
entry
stuff,
this
is
our
intermediary
version.
Where
you
can
no
longer
deploy
new
entry
stuff
new
entry,
Provisions
PVCs,
you
can
only
do
CSI
PVCs,
which
requires
again
that's
702
or
later
and
version
hardware
version
15
or
later
4.13
will
require
7.0
U2
because
of
CSI
migration.
B
So
keep
that
in
mind,
if
you're,
not
on
7.0
U2
effectively,
it
will
Mark
your
cluster
as
unable
to
upgrade
until
you
get
to
vsphere
7.0
YouTube,
so
that
that's
kind
of
the
big
one
that
that
I
wanted
to
call
out.
You
may
have
seen
in
the
presentation.
I
should
bring
up
the
slides.
Are
you
the
sides?
No
you're,
not
the
slides.
Here
we
go.
Let's
go
to
the
slides,
really
you're,
not
gonna.
Let
me
really
who
added
this
open
link
in
new
tab.
B
A
B
It's
so
funny
to
me
seeing
the
how
the
sausage
is
made
with
the
what's
new,
because,
like
literally
until
we
hit
the
go,
live
button
and
somebody
like
whoever's
running
the
slides,
pushes
the
display
button
they're
making
changes
to
those
slides
and
then
as
soon
as
it's
done,
they
go
back
in
and
they
make
a
whole
bunch
more
changes.
So
it
would
not
surprise
me
if
they
have
added
or
changed
those
slides.
B
So
I
I
need
to
re-upload
these
into
speaker
Deck
with
the
new
version
anyways,
so
Zone
awareness
in
vsphere,
Tech
preview,
so
not
supported.
Yet
oh
slides,
34.,
let's
see,
did
I
just
miss.
It
was
I,
not
yeah.
It's
openshift
fertilization,
though
Our
Hope,
none,
so
Zone
awareness
takes
think
of
the
AWS
region
and
availability
Zone
concept.
So
AWS
has
regions
U.S
east
one
Us,
East,
2,
Us
West,
one
U.S,
West,
two
and
then
inside
of
there
they
have
availability
zones,
U.S
east
1A,
1B
1C.
B
So
what
this
does?
Is
it
effectively
Maps
those
Concepts
to
vsphere,
Data,
Centers
and
DRS
clusters,
so
I
add
a
tag
in
vcenter
to
my
data
center.
That
says
data
center.
One
is
you
know:
region
X
data
center
2
is
region,
Y
data
center
3
is
region,
Z
and
then
DRS
cluster,
a
in
region.
One
is
Zone
x,
one
so
on
and
so
forth.
So
very
much
like
AWS
I
have
this
region
Zone
type
of
layout
and
then
in
vsphere
and
put
it
in
here.
Oh
maybe
it's
in
the
ins,
the
release
notes.
B
B
So
you
can
see
here.
Csi
topology
takes
advantage
of
the
same
thing,
so
this
will
walk
you
through
how
to
set
up
all
of
those
things
how
to
set
up
the
tags
inside
of
vcenter
put
them
onto
those
various
resources,
including
the
data
stores,
and
then
you
can
use
that
to
distribute
availability
across
as
well
as
with
CSI.
B
It's
also
for
you
wouldn't
want
to
have,
for
example,
a
new
PVC
created
in
you
know:
data
center,
one
cluster,
a
when
the
node
is
in
data
center,
3
cluster
M
right
because
most
likely
well,
theoretically,
they're
in
different
rooms,
different
buildings,
different
literally
different
regions,
maybe
inside
of
your
organization.
So
that
would
be
bad,
so
CSI
also
uses
that.
B
So
this
is
a
really
cool
one
again
remember
it's
technology
preview,
so
not
supported,
but
please
do
test
it
out.
If
it's
something
that's
interesting
to
you
and
let
us
know
of
any
errors,
issues
or
problems
that
you
have
the
last
thing
that
I
wanted
to
say
with
regard
to
vsphere
is:
there
is
a
very
nuanced
change
and
thanks
to
Chuck
Douglas
internally
for
pointing
this
out,
if
I
go
here
to
and
we'll
go
with
installing
on
network,
it
is
installing
a
cluster
on
vser
with
network
customizations.
B
So
this
is
an
IPI
install.
Let's
search
for
Ingress
VIP,
remember
with
IPI
in
the
install
config
we
provide
two
IP
addresses
the
API
virtual
IP
address
and
the
Ingress
virtual
IP
address
so
you'll
notice
here
that
this
is
changed.
It's
now:
plural,
API,
VIPs,
Ingress,
VIPs
and
there's
a
handy
little
note
out
here
to
the
side
that
says
that,
as
of
4.12
and
later
API
VIP,
the
singular
instance
Ingress
VIP
API
VIP
is
now
deprecated
and
instead
we
use
a
list
in
the
value
of
API
VIPs.
B
So
why
did
this
happen
because
of
dual
stack
support?
So
now
you
can
provide
a
an
array,
a
list.
Excuse
me
in
the
ammo
of
an
ipv4
address
and
an
IPv6
address
for
those.
So
you
can't
have
multiple
ipv4,
you
know
Ingress
IPS
or
API
IPS,
but
you
can
have
an
ipv4
and
an
IPv6,
so
small
change
I
have
not
tested
what
happens
if
you
say
you
use
an
old
Ingress
or
a
install
config.yaml
that
is
not
pluralized
and
you're.
B
A
B
B
Therefore,
if
it's
not
pluralized,
it
shows
up
as
missing
and
therefore
the
installer
would
immediately
barf
and
say:
oh
I
can't
continue
you're
missing
this
required
field
or
maybe
at
worse
it
would
do
some
sort
of
translation
of
oh
I,
see
you,
provided
this
I'm
going
to
go
ahead
and
translate
this
for
you
to.
You
know
the
ipv4
list,
but
here
let
me
let
me
tell
you
that
this
is
deprecated.
B
You
need
to
fix
your
install
config,
instead
of
going
all
the
way
through
and
deploying
bootstrap
and
and
deploying
control,
plane,
nodes
and
then
all
of
a
sudden
like
it
just
fails
to
finish
bootstrapping
because
it
can't
find
the
and
then
you
got
to
go
and
dig
through
a
bunch
of
logs
and
see
a
bunch
of
errors,
and,
oh
it's
because
it
never
configured
it
I.
I
hope.
B
Because
I
I
was
going
to
test
it.
I
ran
out
of
time
yesterday
and
because
I
had
to
I,
don't
my
home
lab
has
been
decommissioned.
At
this
point
you
know
I
I
I'm
in
a
different
room
right
we're
having
work
done
on
the
house.
I
had
to
turn
everything
off
I
had
to
like
disconnect
a
bunch
of
stuff,
and
so
my
home
lab
is
out
of
commission.
B
A
B
B
So
we
do
this
during
install
time.
We
can
specify
you
know,
hey,
don't
don't
include
these
things
install
so
historically
it
was
things
like
the
I'm
looking
at
the
list
of
things
that
have
been
of
new
options
but
I'm
forgetting
what
the
original
ones
were
and
free.
Thank
you,
so
I'm
going
to
dig
through
and
we're
gonna
still
config
with.
B
B
B
B
I
can
completely
not
deploy
the
console
like
just
have
it
not
there
at
all,
which
is
something
that
surprising
number
of
people
ask
for.
Actually,
you
know
it
makes
sense,
app
teams,
Dev
teams,
they
don't
ever
look
at
the
GUI.
They
do
everything
through
yaml
we
as
administrators
as
we
start
to
gain
more
clusters.
I
feel
like
there's
a
leap
over
to
automation,
tools
like
ACM
and
other.
You
know
other
GUI
based
interfaces
where
we
tend
to
not
use
the
actual
console
as
much
so.
B
Let's
free
up
some
resources,
insights,
you
can
opt
not
to
deploy
the
insights
operator.
So
insights
is
useful.
I
think
we,
we
should
probably
bring
John
Spinks
back
on
to
talk
about
all
the
new
stuff.
That's
happened
in
insights,
but
you
know
if
you're
disconnected
insights
isn't
going
to
do
going
to
do
any
good
right.
It
has
to
reach
up
and
talk
to
the
cloud
service
to
get
information,
so
just
don't
deploy
it
at
all
if
you're
not
using
any
of
the
open
shift
storage
providers.
B
B
Here
we
go
openshift
ships
with
all
of
these
different
volume
plug-ins.
If
you're
not
going
to
use
any
of
these,
then
maybe
you
don't
want
to
install
that
now.
That
being
said,
that
could
be
so,
for
example,
if
you're
using
the
lvm
operator
I'll
get
to
that
one
in
just
a
moment,
lvm
I
don't
think
uses.
Any
of
these.
You
do
need
to
be
a
little
bit
careful
here.
B
A
B
A
B
Yeah
I
see
one
here:
why
isn't
oadp
available
for
okd
I?
Don't
know
the
answer
to
that.
We
we
probably
should
probably
bring
the
oadp
folks
on
to
look
at
that.
Oedp
is
from
my
understanding,
so
take
it
with
a
grain
of
salt
because
Andrew
doesn't
know
what
I
don't
know
about
it,
but
in
my
head,
I
map
it
over
to
Valero
with
some
some
special
sauce,
so
it
effectively
packages
up
an
application
so
as
a
backup
partner,
our
partners
can
come
in
and
say,
o
ADP.
B
Give
me
all
of
the
resources
for
this
application
to
back
up,
and
it
says:
okay,
here's
all
the
yaml
files,
here's
the
container
images,
here's
the
PVCs
right
all
that
stuff,
instead
of
them
having
to
do
that
independently,
but
we
definitely
should
bring
them
on
to
have
a
chat
about
backup
options
like
televm
for
snow.
Looked
in
412.,
good,
good
news,
our
hope.
Not
that
was
the
next
thing
on
my
list,
so
the
lvm
CSI
storage
provider
is
now
generally
available.
A
Yeah
before
in
411
it
was
theirs
Tech
preview,
where
you
could
install
the
lvm
operator.
It's
it
was
awesome.
It's
super
easy
to
use
it.
It
gets
a
little
tricky
if
you
have
any
kind
of
like
file
system
on
there,
so
you
may
have
to
go
and
like
actually
run
like
a
wipe
FS
on
it.
You
know
like
completely
blast
it
out
and
then
rerun
the
lvm
operator,
but
it's
super
easy
to
configure
it's.
It
was.
It
was
very
nice.
B
Yep
yeah,
so
it's
generally
available
with
412
note,
however,
that
if
you
deployed
it
in
411
or
earlier
and
it's
Tech
preview
status,
it
is
not
compatible.
You
have
to
redeploy,
so
you
can't
there
is
no
upgrade
path
there,
unfortunately,
so
just
be
aware
of
that,
if
you're
using
it
today
or
hope,
none
but
yeah
I
thought
that
one
was
pretty.
Cool.
B
So
now
with
lvm,
you
know
operator
being
GA
with
openshift
virtualization
and
all
the
maturation.
That's
happened
there
and
all
those
things
there's
a
lot
of
folks
internally,
who
are
considering
like
maybe
I,
can
like
deploy
what
is
effectively
a
small
footprint
openshift
again
single
node,
openshift,
plus
composable
openshift,
to
remove
some
things.
You
know,
plus
all
these
other
things
now
I've
got
local
volume
management.
B
A
B
B
B
B
B
You
know
leap
or
jump
in
functionality
and
security
when
it
comes
to
those
types
of
things
and
historically,
we've
just
always
assumed
that
well
yeah
the
firewall's.
Always
there
it's
just
transparent
right,
it's
a
kubernetes
openshift
manages
it
for
us.
Well
now
we
can
take
positive
control
and
we
can
manage
those
rules.
B
I
I,
don't
know
the
sharp
edges
yet
so,
for
example,
you
know,
can
I
accidentally
lock
myself
out
of
my
cluster,
like
if
I
create
one
of
these
rules
and
I
globally
apply
like
block
all
Port
443
like
like
or
six
four
four
three,
you
know
would
I
accidentally
lock
myself
out
of
the
cluster
I
I,
don't
know
likewise
could
I
accidentally
block.
You
know,
let's
say
that
I've
got
a
mySQL
database
deployed
and
it's
using
metal
lb.
B
If
I
put
in
something
that
blocks
Port
3306,
the
MySQL
Port
am
I,
gonna
suddenly
block
that
or
is
there
some
integration
there
is
there
something
that
happens
with
services
in
particular
routes
always
run
through
80
and
443.,
so
yeah
I
definitely
some
stuff
to
explore
there.
But
to
me
this
one
is
I'm
keeping
an
eye
on
this,
because
this
is
really
cool.
A
A
No
I
mean
everything
else
was
just
kind
of
like
I
I
had
the
pipelines
as
code
where
that
went
GA
and
if
you're
not
familiar
with
pipeline's
code.
If
you've
used
gitlab
Ci
or
anything
like
that,
you
know
you
could
put
like
a
gitlab
CI
Dash
yaml
file
in
your
in
your
repo
and
then
what
happens
is
gitlab
as
soon
as
anything
happens
in
that
repo
it'll
it'll
pick
up
the
changes
and
start
the
pipeline,
so
the
the
pipelines.
A
This
code
is
essentially
the
same
thing
where
it
allows
you
to
put
a
configuration
in
your
repo
in
your
git
repository,
and
then
it
will
spin
up
the
pipeline
runs
and
the
task
runs
and
all
those
things
that
you
either
would
create
your
own
task
run
for
your
pipeline,
run
for
or
you'd
have
to
go
in,
and
you
know,
go
through
the
the
interfaces
and
create
it
so
just
a
a
new
tool
to
help
automate
even
further.
You
know
development
and
and
openshift,
using
openshift
pipelines
and.
B
Johnny
I'll
I'll,
ask
you
the
question
because
I
don't
I,
don't
remember
if
you
were
on
the
chat
for
the
the
live
stream
last
week,
I,
don't
think
you
were
I
think
you
were
on
PTO.
Somebody
was
asking
there
like
why.
Why
would
they
use
pipelines
as
code
like?
What's
what's
cool
about
this?
Why
do
I
care
about
pipelines
as
code
yeah.
A
So
the
the
reason
why
you'd
want
to
use
pipelines
as
code
is
because,
if
you,
if
you
don't
have
it
and
you've,
created
your
pipelines
and
your
tasks
and
stuff
like
that
in
Tech
time,
then
you
either
have
to
have
a
pipeline
run
or
a
a
task
run
to
run
that
individual
task
or
pipeline
run
to
run
that
overall
pipeline
or
you
need
to
go
into
the
the
user
interface
and
then
fill
out
the
context.
You
know
the
pipeline
context
and
then
that'll
create
the
pipeline
run
which
will
trigger
the
the
pipeline.
A
B
Yeah
so
simplification
rate,
because
yeah
the
pipelines
are
they're,
always
quote
unquote,
code
right,
it's
yaml,
that's
running
and
behind
that
that
defines
the
objects
in
behind
the
scenes.
It's
just
a
question
of
does
Johnny
have
to
go
in
and
Define
the
pipeline
and
point
it
to
the
code
Repository
so
that
it
can
pull
in
the
code
to
build
the
application
and
go
through
the
pipeline
steps
or
does
Johnny
just
point
it
out
the
code
repository
and
it
says
oh,
look,
there's
a
pipeline
in
here.
Let
me
create
the
pipeline
yeah.
A
B
A
Does
all
of
it
for
you
and
it's
if
you
look
at
like
gitlab
Ci
to
me,
is
a
good
example,
because
it's
very
similar
right.
You
have
the
gitlab
CI
file
within
the
repo
and
then
you
that's
essentially
your
pipeline,
so
that
gets
slurped
in
by
gitlab
and
then
it
generates
your
pipeline
and
then
often
away.
It
goes
yeah.
B
Yeah
it's
cool
stuff,
so,
let's
see
so
our
hope9.
Thank
you
for
asking
about
artwing.
He
is
busy.
He
and
I
are
working
together
on
a
number
of
different
things.
Both
Red
Hat
Tech,
Exchange,
related
IBM,
Tech,
related
Bobby,
Carr
related
lots,
lots
of
cool
stuff
going
on
there.
So
for
any
red
Hatters.
You
will
be
seeing
a
lot
of
ortwin's
work
during
Tech
Exchange,
so
I
have
we
have
brought
up
talking
about
service
mesh
again
here
on
the
stream.
B
Johnny
and
I
have
brought
it
up,
so
we're
probably
looking
at
like
late
February,
if
not
March
as
being
the
earliest,
that
we'll
be
able
to
do
that.
That's
a
combination
of
our
schedule
as
well
as
orwin's
schedule.
So
let's
talk
about
what's
happening
over
the
next.
You
know
six
or
eight
weeks
here
on
the
live
stream.
B
So
one
I'm
sitting
here,
looking
at
a
list
of
as
I
said
before
no
less
than
six
things.
Six
future
topics
and
service
mesh
is
a
seventh
that
we'll
be
talking
about
as
we
go
forward
so
doing
deep
dives
into
things
like
again
ebpf
network
observability,
the
security
profile
operator,
topology,
aware
lifecycle
manager
in
ACM
I
did
not
know
about
this.
One
and
I
was
blown
away
seeing
that
functionality
inside
of
the
what's
new
presentation.
That's
really
awesome.
I'm
super
excited
about
having
that
as
a
as
a
topic.
B
A
A
A
B
So
that
is
the
first
February
1st
and
February
8th.
We
will
not
be
streaming
here
on
ask
an
openshift
admin
and
that's
because
we
have
work
commitments
and
just
too
too
many
time
zones
apart
in
a
way
to
make
it
feasible.
Nor
you
know
normally,
when
I'm
going
at
conferences,
I
don't
mind
doing
the
live
stream,
but
it
just
just
won't
work
out
this
time
so
yeah
we
will
miss
you
all.
B
It
feels
slightly
frustrating
that
we've
been
gone
for
a
month,
we're
back
for
two
weeks,
we're
going
for
two
weeks
and
then
we're
we're
we're
back.
We
don't
have
I,
don't
think
either.
One
of
us
has
any
time
away
planned
or
any
other
things
coming
up.
That'll
interrupt
us,
but
yeah
so
keep
an
eye
on
the
schedule.
As
we
get
things
cemented
and
solidified
we'll
be
updating
the
streaming
calendar
we'll
also
be
staging
the
streams.
B
B
B
B
A
B
Hope
nine!
It's
funny!
You
mentioned
that
so
the
the
APAC
folks
August
seminelli
in
particular,
they're
working
on
and
I
think
they
had
their
first
or
maybe
second
instance,
now
of
a
stream
based
in
the
Apec
time
zones.
So
I
I
had
briefly
thought
about
doing
something,
maybe
out
in
that
time
zone
while
we're
out
there.
B
But
if
I'm
being
honest
I,
don't
know
if
time
is
going
to
allow
so
looking
at
our
schedule,
while
we're
out
there
the
days
are
super
packed
like
I'm
presenting
sessions
at
Tech
Exchange
at
seven
o'clock
at
night,
so
I
I,
don't
know-
or
maybe
it's
six
o'clock,
it's
something
like
that.
It's
they're
very
long
days
so
I
don't
know
if
we'll
be
able
to
squeeze
in
something
like
that.
If
we
do
it'll,
probably
be
an
ad
hoc,
but
that
would
be
really
cool.
B
I'd
love
to
include
you
know
that
that
region
in
those
time
zones
more
and
Johnny
and
I
have
talked
about
that
before
we
just
haven't,
been
able
to
figure
anything
out,
yeah
all
right
chill.
It's
good
to
be
back
good
to
see
everybody
Our
Hope,
nine
Dave,
let's
see
I'm
winning
the
list.
Wayne
yeah!
It's
it's
good
to
see
everybody
good
to
hear
from.
You
really
really
appreciate
you
showing
up
and
chatting
with
us
and
all
the
questions,
all
the
comments.
As
always.
It's
it's!
B
What
we're
here
for
it's,
what
we
love
yep,
as
always,
don't
hesitate
to
reach
out.
If
you
have
any
questions
again,
if
you're
not
watching
us
live,
you
can
reach
out
or
if
anything
occurs
to
you
for
any
of
the
folks
who
are
watching
us
live
after
this
stream
ends
feel
free
to
reach
out
andrew.sullivan,
redhat.com
or
johnnyjo
nny
at
Red
hat.com
and
we'll
we'll
get
back
to
you
as
soon
as
we
can
I
got
nothing
else.
Johnny
last
words.