►
From YouTube: Sync on Kubernetes Deployment for Pages Post-Rearch
Description
Supporting deployment of Pages with Kubernetes
B
That's
great
thanks
for
the
intro
and
Craster's
for
a
little
bit
more
background.
We,
the
maturity
level
of
the
cloud
native
install
today,
is
viable
and
we
have
a
goal
to
get
it
eventually
to
lovable.
But
as
a
first
step
to
complete
by
middle
of
this
year
is
our
hope
and
one
of
the
things
that
we
identified
to
get
it
to
complete
is
that
we
have
pages
support
in
which
I
it
seems
like
we're
blocked
on
because
of
the
NFS
storage
aspect
of
pages.
And
so
we
know
that
there's
an
epoch
that
you're
working
on.
B
C
So
at
least
I
know,
like
the
major
problem,
like
you
said,
is
the
need
for
a
Sherrod,
shaft,
disc
storage
via
anus
in
a
safe
and
removing
the
need
for
this
was
one
of
the
major
drivers
for
the
work
we
are
doing
right
now.
So
pages
is
using
local
or
any
safe
storage
for
two
things.
One
is
for
to
build
the
configuration
for
all
the
domains,
its
serving,
which
also
was
quite
slow
because
it
was
traversing
the
whole
directory
structure
for
the
pages.
C
You
know
on
github.com
that
takes
a
while,
and
then
it
serves
those
pages
files
from
from
the
same
any
set
story.
So
the
work
we're
completing
now
is
to
move
the
config
from
like
from
the
disk
to
like
talking
to
get
lap
rails
up
via
API
and
so
that
the
API
returns,
the
the
config
for
a
domain
that's
requested,
and
once
we
have
this,
our
goal
is
to
start
working
on
object
storage.
C
So
so
the
pages
can
be
somewhere
in
like
a
stray
or
some
other
object,
storage,
and
once
this
is
done,
I
think
that
will
be
enough
to
eliminate
the
issue
with
the
NSA,
local
storage
and
I'm,
not
sure
at
least
I
don't
know.
Are
there
any
other
technical
limitation
in
the
context
of
kubernetes
deployment
for
pages,
but
I
think
that's
does
the
main
one
and
we
are
so
working
through
with
a
word
from
the
serverless
team
cause
they
identified
a
use
case.
That
pages
was
supporting
most
of
it
for
them.
C
So
with
moving
to
the
API
config,
we'll
be
able
to
actually
have
like
different
types
of
things
that
page
itself,
so
it
can
be
a
local
file.
It
can
be
a
file
from
object,
storage
and
in
their
case
it
can
be
just
a
proxy
to
a
server
list
function,
I'm
just
saying
just
to
like
for
more
context
to
for
you
to
know
who
are
working
with
so
yeah.
C
We
are
like
close
to
I'm
just
waiting
on
like
one
or
two
most
requests
to
be
merged
so
that
we
can
get
this
in
production
on
comm
and
test
it
incrementally
before
we
release
it
for,
like
everyone
on
pages,
just
a
config
change,
and
then
we
will
start
work
on
the
object
storage
which
it's
like.
We
haven't
looked
too
much
into
details
in
this
one.
So
I
don't
know
what
issues
we
may
or
may
not
face
so
I
guess
it
will
take
us
at
least
a
couple
of
releases
to
get
it
ready.
D
No,
the
plan
is
pretty
much.
What
I
would
have
expected
is
separate
that
configuration
file,
because
when
they
all
no
matter
how
many
pages
you
have,
you
still
have
to
have
the
same
configuration.
This
is
one
of
the
big
aspects.
The
aspect
of
how
page
is
actually
deployed
to
be
served.
That's
gonna
be
their
tricky
part
because
they're,
basically
artifact
tarballs,
so
you.
D
It
generates
a
tgz
file
that
is
then
pulled
after
the
fact
and
deployed
onto
the
shared
file
system
and
then
added
to
the
configuration
for
pages
and
then
pages.
This
goes
up.
This
is
my
domain.
This
is
where
my
father
is
I'm.
Gonna
go
read
the
files
out,
so
there
is
still
some
items
that
will
get
more
complex
as
we
progress
through
this
and
that's
going
to
be
the
larger
challenge
that
they
face.
C
Well,
ideally,
there
will
be
nothing
on
a
disk
and
now
I.
Remember
the
other
thing.
That's
causing
issues
I'm,
not
sure
for
kubernetes,
but
for
containerized
deployments,
which
applies
to
kubernetes.
Is
that
because
of
the
local
access,
the
the
cause
of
the
access
to
local
disk
pages?
Not
is
currently
doing
the
the
jails
route
so
that
it's
creating
a
jail
for
the
processor,
if
in
case
an
one
managed
to
like
do
something
they
can
escape
out
of
it,
and
that
was
causing
problems
in
containers
with
resolving
DNS
and
stuff.
C
So,
like
a
side
effect
of
removing
the
need
from
local
storage,
we
are
hoping
to
get
rid
of
all
that
jailed
sourcing
which,
which
is
complicating
things
and
causing
problems
at
the
same
time.
So
yeah,
that's
another
thing
that
should
be
gone
ideally,
so
my
only
worries
are
at
the
moment
and
to
think
I
don't
know
well,
not
only,
but
one
of
the
worries
I
kept
about
the
object.
Storage
support
is
like
how
fast
and
efficient
it
will
be
in
practice
when
we
start
using
it.
But
we
can't
say
this
until
we
have
something.
D
B
C
Well,
if
you
have
the,
if
you
are
willing-
and
you
have
the
capacity
we
can
use
some
help-
I
guess
when
we
start
working
from
the
object,
storage
and
cause
that
is
like
will
it's
kind
of
a
self-contained
piece
of
work?
That's
not
very
well!
Its
page
is
specific,
but
not
that
much
so
so
we
can
use
any
help
with
that
likes
in
America.
We
now
work
with
the
serverless
team
on
the
config
changes,
so
I'll
be
happy
to
what
we
do.
If
you
get
the
time
and
the
will
to
do
this
so.
C
C
Sorry
I'm
just
like
I'm
just
saying
that
it
we
may
use
help
for
you.
May
we
may
use
help
for
the
object
storage
support,
because
the
config
work
is
like
almost
done
it.
It
doesn't
make
much
sense
to
get
someone
new
involved
at
that
point
cause
where
it
basically
the
if
they
entered
it.
It
will
be
easier
if
someone
starts
working
on
something
after
it.
A
So
Larissa
I
think
what
would
be
best
just
for
I
can
set
up
a
followup
with
human
I
heard
12.9
planning
halfway
through
12.8,
and
if
something
is
closer
to
getting
work
queued
up,
we
can
see
if
there's
work
that
makes
sense
to
schedule
on
your
team's
release
for
12.9.
If
it
doesn't,
if
any,
if
that's
nothing
makes
12.9,
and
then
we
will
do
the
same
exercise
for
middle
of
12.94,
12.10
yeah.
B
That
sounds
good.
If
we
have
some
issues
created,
then
we
can
check
our
capacity.
We
do
have
a
fair
bit
coming
up
before
the
release
of
the
next
major
release
in
May
13
point.
Oh
so
I'm
not
sure
how
much
capacity
we'll
have,
but
at
least
if
we
have
issues
created,
we
can
look
at
them
and
see
if
there's
something
that
makes
sense
for
distribution
to
help
with.
C
C
C
D
C
D
We
will
and
they'll
spawn
off
of
issue
37,
but
really
because
of
the
change
to
pull
the
configuration
instead
of
a
file,
but
instead
pull
it
through
the
API
interactions.
A
lot
of
that
won't
actually
be
necessary.
It'll
be
more
along
the
lines
of
actually
enabling
the
configuration
on
the
rails
outside
and.
B
D
C
Don't
know
I,
don't
know
much
about
how
the
charts
will
works,
but
that's
something
we
might
miss
because
like,
for
example,
the
pages
in
the
API
they
talk
and
they
use
shared
secret
to
generate
tokens
and
I've
made
the
changes
to
only
push
that
generate
those
secrets
in
like
push
them
for
both
pages
in
the
app
but
I
totally
neglected
the
huge
art
or
whatever
so
I'm,
not
sure
how
easy
to
burn
using
compass
or
not.
So
there
might
be
some
some
like
support
work
needed
in
regarding
there.
D
D
D
C
C
D
A
Things
with
some
follow-up
that
we'll
need
to
do
KRAS
we'll
have
to
create
the
issues
to
break
down
the
object,
storage
work
and
then
I
can
groom
them
with
Larissa.
So
we
can
understand
what
makes
sense
for
her
team
to
help
partner
with
us
on
and
get
those
scheduled
to
the
next
following
milestones
and
then
Larissa
I'll
just
set
something
up
below
of
12.8.
That's
a
check
point
and
then
we
can
cancel
it.
If
we
don't
have
any
issues
for
that.
Okay.