►
From YouTube: CNCF CNF WG Meeting - 2023-01-30
Description
Don't miss out! Join us at our upcoming event: KubeCon + CloudNativeCon Europe in Amsterdam, The Netherlands from 18 - 21 April, 2023. Learn more at https://kubecon.io The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects.
A
A
A
A
A
E
E
C
A
I've
put
the
link
in
the
chat
for
people
to
put
the
name
on
the
list
so
looking
down
the
events
to
to
Wednesday
Cloud
note
security
com
and
we
know
where
I'm
going.
Is
there
any
Telco,
useful,
Telco
events
to
look
out
for
do?
We
know.
E
C
Cfps
are
open,
so
definitely
I
want
folks
to
get
those
in
while
it's
still
open.
A
C
Extended
it
to
the
12th.
C
Yes,
there's
so
many
events
hard
to
keep
up
with
them
it
we
did
hear
that
there
are
going
to
be.
There
will
be
some
Telco
presents
there
from
someone
this
past
week,
but
I'm
not
going
to
make
it
up
to
that
event.
At
this
late
notice,.
A
These
tend
to
be
about
topic,
you
know
the
NFL
projects,
so
this
is
irrelevant
to
Toko
I,
wonder
if
there's
any
that
we
want
to.
D
I
wasn't
checking
before,
but
I
noticed
now
that
these
are
nature.
Topics
which
are
great
like
these
and
even
more,
are
going
to
present
hunting
so
be
nice
to
see
from
Dr
Charles
nephew.
E
A
A
C
Mean
that
I
think's,
a
bigger
thing
for
me,
is
that
it's
so
soon
and
the
deadline
was
the
27th.
So
we'd
have
to
really
have
something
ready
and
ask
you
know,
ask
someone,
can
you
get
it
down
at
this
point
yeah.
C
C
It
would
be
I
mean,
there's
a
a
few
things
that
we
could
look
at,
but
I
think
it's
probably
something
for
a
future
one,
any
type
of
collaboration
which
there's
some
ongoing
talks
that
have
increased
recently
and
then
the
stuff
that
we
were
talking
about
going
up
to
the
end
of
the
year
as
far
as
and
then
as
started
this
year
for
publishing
that
the
dev
best
practices-
and
you
know
if
we
can
get
Maybe-
have
that
as
a
stronger
goal
with
the
timeline
and
then
presenting
something
on
that
which
could
include
some
of
the
stuff
around
energy.
C
The
energy
focused
stuff
environment,
environmental
things
that
we're
seeing
more
of
in
their
post
into
slack
from
the
kubernetes
some
of
the
kubernetes
projects
around
this,
and
we
did
talk
about
best
practices.
There
so
I
think
there's
some
possibilities
for
a
future.
One.
C
E
A
A
We
mentioned
the
comments
Telco
day,
anything
else.
Anyone
wants
to
call
out
or
something
we
should
submit
a
talk
to
Paul.
D
Yeah
good
point,
foreign.
C
I'd
like
to
know
how
Tom
you
and
Victor
think
about
the
open
source
side.
We
talk
about
the
cloud
native
a
lot.
The
open
source
is,
you
know,
underlying
a
lot
of
the
pieces
when
I'm
in
conversations
with
different
orgs
and
trying
to
get
collaboration
a
lot
of
times.
One
of
the
initial
things
is
we,
our
products
are
not
open
source
or
were
having
to
work
with
non-open
Source
Products.
C
A
C
A
So
I
agree,
I,
think
I
think
there's
plenty
there's
plenty
that
could
be
said
about
the
benefits
of
Open
Source
to
telcos
I.
Think
if
you
were
going
to
do
it,
if
someone
was
going
to
do
a
talk
on
that
you'd
have
to
address
the
elephant
in
the
room.
A
We've
seen
seems
to
me
to
be
the
elephant
in
the
room,
which
is
the
the
patent
problem.
That
telcos
are
relying
on
telcos
and
Telco.
Vendors
have
reliable
paint
and
Revenue,
and
that
kind
of
goes
against
the
grain
of
Open
Source
licensing
and
open
source
project
management.
C
That's
true
to
an
extent,
there's
two
paths,
one
that
come
to
mind,
thinking
about
that
one
is
a
little
bit
easier,
I
think
for
to
talk
with
folks
about
not
not
focusing
on
what
you
are
doing
with
your
product
or
how
your
business
works
specifically
and
and
then
more
focusing
on
what
your
business
is
built
on
and
what
your
products
are
built
on
and
I
would.
C
Without
you
know,
this
is
just
me
throwing
something
out
my
own
opinion,
but
I
would
I
would
bet
that
most
products
out
there
are
using
some
open
source,
if
not
a
large
amount
of
Open,
Source
libraries
and
everything
else.
Besides,
if
you're
running
on
kubernetes,
then
especially,
if
you're
doing
like
something
where
it
also
provides
a
platform
type
thing,
but
if
you're
providing
like
operators,
you
may
be
using
some
operator
framework
but
you're
using
specific
pieces
and
that's
without
looking
at
containers
running
on
something
and
dbdk.
C
You
know,
for
instance,
well
lots
of
cnfs
that
are
dealing
with
the
user
planner
using
dpdk,
and
you
just
keep
going
down
that
path.
So
you
talk
about
how
open
source
has
provided
a
lot
of
benefits
and
I.
Don't
know
you
could
talk
about
that
and
I
I
think
it,
for
some
companies
are
more
engaged
with
the
whole
open
source
community.
So
it's
not
just
a
technology
and
there
may
be
feedback.
You
know
and
user
stories
and
stuff
like
that.
C
So
in
my
mind,
that's
kind
of
an
easier
path
because
you're
getting
more
people
to
talk
about
open
source
and
how
they're
involved,
even
if
their
product
is
an
open
source
and
then
the
other
side
would
be
talking
about
different
models.
For
what
does
it
mean
if
you
have
an
open
source
product?
How
how
do
you
deal
with
that
and
there's
a
lot
of
examples
of
of
those
including
like,
if
you
just
looked
at
cncf,
then
a
a
lot
of
the
projects?
I'm
not
going
to
say
all
of
them,
but
a
lot
of
them.
C
There
is
a
commercial
version
of
that
project
where
they
may
call
it
Enterprise
or
whatever
else
they
call.
It
could
be
a
hosted
type
of
service
or
a
version
that
has
more
plug-ins
that
they
have,
but
the
core
would
be
open
source
so
and
that's
just
one
path,
there's
a
lot
of
paths
for
that.
So
I
think
there's
an
opportunity
there
that
similar
to
Cloud
native
a
lot
of
other
domains
have
learned
what
what
does
it
mean
and
how
to
work
within
you
know.
How
do
you
work
within
the
whole?
C
C
I
think
that
telecom
world
is
still
can
benefit
from
that
similar
to
Cloud
native
I'm,
just
not
sure
how
to
approach
it
into
for
Telecom,
because
it's
a
little
bit
different
like
how
do
you
translate
and
show
like
the
benefits
of
the
adoption
foreign.
A
A
Kind
of
within
your
infrastructure
that
is
being
used
to
deliver
products,
but
there's
also
the
use
of
open
resourcing
the
products
that
you
sell
directly
to
consumer
or
business
customers.
A
So
you
know
all
the
degrees
of
separation
between
the
thing
you're
selling
and
the
open
source
project
and
I
think
the
more
degrees
there
are.
Certainly
I
found
the
more
likelihood
you're
going
to
end
up
with
the
patient
discussion
being
had
I,
I.
Think
I.
Think
it's
a
good
idea.
I
think
there's
plenty
of
discussion
to
be
had
plenty
of
talk
material
on
the
topic
of
Open
Source
and
token.
C
So
I'm
interested
in
the
topic
I'm
wondering
how
that
ties
back
to
that.
If
there's
a
we're
talking
about
open
source
Summit
in
a
cfp,
so
is
it
something
that
we
want
to
do
as
CNF
working
group
and
that's
separate
from
you
and
I
Tom
go
hey!
This
is
interesting.
Let's
go.
Do
a
talks
or
Summit
just
completely
separate,
that's
you
know
it
could
be
upstairs
Telecom
does.
C
Long
term
I
think
it
benefits
what
we're
doing
I
mean
we're.
You
know
the
CNF
working
group
is
definitely
engaged
in
open
source
projects.
So
that's
like
there's
something
there
I'm,
just
not
thinking
of
a
more
direct
I'm,
not
making
a
more
direct
connection
right
now
in
my
head.
So
let's
see
February
5th.
D
C
Does
anyone
else
have
any
thoughts
like?
How
would
we
because
you
know
if,
if
we're
going
to
put
time
into
submitting
on
cfp,
if
we
have
an
idea-
and
we
just
want
to
submit
something
fine
but
and
then
we
can
always
decline
later,
but
I
don't
have
something
just
ready,
but
otherwise
we
need
to
decide
on
something
and
we
think
yeah.
This
has
been
official
and
maybe
it
ties
back
to
what
we're
doing.
D
Well,
in
my
case,
my
own
experience
of
having
these
open
source
event
is
in
Austin,
and
one
of
the
things
that
I
noticed
is
mainly
two
things
like
most
of
the
questions
are
very
technical.
They
were
trying
to
explain
some
complexities
of
the
technology,
and
the
other
thing
is
like
I
feel,
like
most
of
the
attendees
were
looking
for
places
to
innovate
like
some
looking
for
some
fresh
ideas
or
tips
like
that,
so
maybe
I
don't
know
if
that
was
my
impression.
D
So
probably
I
don't
know
if
we
can
provide
something
like
we
could
meet
those
criteria
like
I,
don't
know,
probably
explain
a
little
more
like
a
test.
Suite
tool
like
how
it
works
internally
or
like
some
of
the
benefits
are
I.
Don't
know
that
could
be
a
suggestion,
but
I
don't
know
if
that
is
to
be
also
align.
What
you
were
saying
before.
A
B
I
I
just
would
maybe
on
that
same
topic
and
I
know
this
wasn't
necessarily
the
ass,
but
it
feels
like
the
the
there
are
a
couple
different
angles:
I
mean
you
can
take
the
angle
of
sort
of
the
benefits
of
using
open
source
and
the
test,
Suite.
Being
an
example
of
that
right
in
in
with
the
purpose
of
helping
to
drive
Cloud
native
adoption,
you
can
make
me
so
depending
on
which
Forum
right,
which
you
know
I,
was
thinking
you
asked
earlier
and
I
think
there
was
a
comment.
B
Taylor
I
think
you
put
it
on
cloud,
the
dntf.
You
know,
for
example,
if
you
scroll
up
just
a
little
bit,
I
think
that
was
I'm
trying
to
see
this
on
the
on
the
screen
here,
yeah
that
we
had.
You
know,
topic
submission.
You
know
it's
too
it's
too
late.
We
that's
why
I
didn't
make
a
mention
of
it,
but
otherwise
I
would
have
thought
that
would
have
been
a
possible
area.
B
Certainly
since
there
are
it's
early
days,
but
there
are
some
conversations
on
doing
collaboration
around
again
open
source
tool
using
it
across.
You
know
open
source
organizations
that
that
would
be
another
angle
as
well
so
I
guess.
We
were
mainly
focused
on
the
open
source
part
though,
which
was
which
that
one's.
Maybe
we
have
time
for
that,
that
we
won't
have
a
time
for
the
elephant
dnp
F
this
time
around,
but
maybe
one
of
the
future
ones.
C
I'm
I'm,
taking
that
into
consideration
with
what
Victor
had
said
and
I
think
it's
maybe
similar
to
the
higher
level
thing
that
Tom
was
talking
about
and
if,
if
we
can't
find,
if
we
don't
think
there's
going
to
be
a
a
good
audience
for
what
who
are
presenting,
then
we
can
spend
our
time
elsewhere.
That's
it!
We
don't
have
I'd
love
to
visit,
Vancouver
again
right,
I,
think
that
we
should
respect
our
time
and-
and
you
know
know
that
we're
only
going
to
have
so
much
time.
C
C
As
far
as
the
tests
we
open
source
Maybe,
it
does
give
the
opportunity
to
talk
about
collaboration
with
other
orgs
Oliver,
so
that
could
be
like
just
one
part
of
it,
using
certification
and
built
on
open
source
and
using
a
lot
of
Open
Source
projects
took
different
tools,
and
so
there's
potential
for
that
sort
of
thing
to
be
part
of
it.
C
And
I
think
that's
an
okay
fit
for
open
source,
Summit,
potentially
like
pitching
it,
and
why
are
you
going
to
pitch
it
there,
maybe
as
a
a
testing
tool
that
could
be?
You
know,
used
by
other
developers
and
Ops
folks
and
whatever
else,
and
then
just
the
benefit
of
collaboration
groups,
so
I
see
something
there
I'm
I'm,
not
tying
it
back
to
why
we
would
do
it
as
a
CNF.
You
know
working
group,
folks
and
and
how
that
relates
back
to
Telco.
C
So
maybe
let's
table
it.
Unless
y'all
have
some
ideas
like
more
and
if
you
want
to
continue
that's
fine
but
give
the
opportunity
to
move
on
and
if
we
table
it
but
we're
having
some
ideas
in
the
back
of
our
head
either
on
that
what
you
put
forward,
Victor
or
something
else,
I'm
willing
to
set
some
time
for
this
I
mean
I
I,
do
think
that
we
need
to
set
some
specific
goals
and
try
to
push
for
more
involvement
with
the
working
group
and
pishmore.
C
If,
like,
how
do
we
get
people
engaged
in
Cloud,
native
and
open
source
Telecom
and
the
benefits
of
those
things
cicd
and
a
lot
of
these
things
are
driven
by
the
efforts
in
open
source?
So
a
lot
of
it's
just
like
what
are
you
missing
in
methodology
and
thinking
of
problem
solving
anyways?
What
do
y'all
think
table
it
and
come
back.
C
Yeah
well,
I
will
I'm
gonna
just
set
some
time
on
my
calendar
and
if
anyone
wants
thinks
of
something,
then
just
message
me
and
I'll
make
some
time
this
week
and
if
we're
like
nope,
we
have
an
idea,
let's
go
for
it
and
let's
do
that.
C
C
Which
isn't
a
positive
or
A
negative?
It's
just
a
a
thought.
C
A
A
Possibly
focuses
on
total
cost
ownership,
best
selling
the
rules,
development
of
stable
and
reliable
wreck.
A
A
C
They
get
a
lot
of
folks
joining
I've,
seen
that
even
the
in
person,
too
filling
up
a
room,
but
it
seems
like
the
conversations
or.
A
Okay,
don't
worry.
A
A
Okay,
should
we
move
on
to
the
review
in
the
PRS
and
the
issues.
A
E
A
C
C
Yeah,
that
sounds
good,
I'm
sure.
If,
if
we
start
working
on
the
options,
then
we'll
be
able
to
find
some
somewhere
to
talk.
There's
a
lot
of
places.
A
A
You
can
open
a
blank
issue
or
you
can
best
practice
proposal,
and
so,
when
we've
got
these
text
boxes
to
be
able
to
put
in
the
different
sections
of
the
best
practice
proposal
and
do
one
or
all
of
those
or
one
or
more,
it's
not
my
help.
We'll
see.
A
So
reviewing
the
issues-
let's
go
from
the
bottom
up,
do
not
want
to
tell
you
the
privileged
flag.
A
So
it's
included
in
the
standard
certification
as
an
essential
test
currently
assigned
to
Taylor
and
Victor,
with
a
milestone
of
this
quarter.
So
by
the
end
of
March.
C
Oud
cool
I'm
gonna,
send
you
a
calendar,
invite
Vector,
okay,
I've
I've
been
a
little
unavailable
for
at
least
half
of
January,
so
I'm
just
trying
to
catch
up
I
will
I'm
going
to
send
out
a
few,
and
you
just
let
me
know
which
one
works.
A
D
No
well
probably
related
with
the
same
VR.
What
Chinese
shoes
like
the
the
parallel
effort
that
the
current
committee
is
doing?
It's
some
of
the
latest
things
that
they
were
discussing
about
naming
so
I
think
that
they
are
going
to
send
out
survey
based
on
the
number
of
proposals,
so
they
have
narrowed
down
to
achieve
them.
D
D
Yeah:
okay,
this
net
bot,
Network
connectivity
provider,
connectivity
service
and
network
instance.
D
In
fact,
they
tried
to
just
get
all
the
the
idea
in
one
single
Champions.
One
of
the
good
videos
to
to
his
name
could
be
like
in
the
semantics
of
the.
For
example,
if
you
use
Cube
cattle,
like
Cube
cattle,
get
what
networks
it
sounds
like
a.
C
C
I'm
not
as
worried
about
the
name
as
if,
if
it
actually
has
the
functionality
that
we've
been
hearing
and
talking
and
complaining
about
for
years
now,
is
it
Kevin
it's
is
it
covering
some
of
the
main
stuff.
C
E
D
The
last
time
that
I
mentioned
these
two
to
add
he
mentioned
like
what's
doing
the
last
Yukon
and
he
was
saying
like
he
didn't
see
any
point
to
this
effort,
but
yeah
because
I
think
not
this.
This
is
one.
This
is
not
the
person
that
they're
trying
to
do
in
terms
like
offering
multi
multi,
Network,
so
I'm,
given
that
they
have
found
like
another
way
to
mitigate
these
problems,
like
kids,
didn't
see
any
any
particular
advantage
to
keep
you
doing
all
these
things.
A
Interesting,
okay,
let
me
try
to
read
through
this.
Oh,
so
what
are
the
calls
for
this
Wednesday
8
A.M
PST.
A
D
A
A
All
right,
thank
you.
So
that's
relating
to
this
issue,
which
will
remain
on
hold
for
now,
create
a
link
to
non-root
and
non-route
I
just
need
to
get
on
with
that.
A
A
Right
so
this
one
one
process
types:
there
was
a
bit
of
dispute
from
foreign.
C
E
C
Watson
is
going
to
be
talking
with
or
plans
to
talk
with
you,
okay,
I
think
this
week
and
yeah
I
don't
know,
I
could
put
it
in
there,
but
he's
he's
going
to
be
talking
with
him.
Some,
but
they've
been
working
on
more
right
up
about
related
items
that
would
tie
into
this
and
I
think
there's
a
microservice
white
paper
that
is
being
edited
right
now.
C
So
that'll
have
a
lot
more
content
about
this
and
we'd
probably
point
to
that
as
a
reference
and
have
further
discussion.
Ideally,
though,
if
you
know
this
is
going
to
be
something
proposed
and
there's
gurga
actually
even
opened
a
pull
request
to
just
remove
it
from
the
test
Suite
as
a
a
main
essential
test
or
something
I
can't
remember
that's
outside
of
the
working
group,
but
both
ideally
you
know
we
can
get
gargoy
on
here
to
go
in
and
and
provide.
C
C
For
most
cases,
not
all
cases,
so
the
idea
with
any
of
these
like
the
privileged
flag
and
all
these
other
stuff
or
most
of
the
time,
you
should
try
to
do
these
things.
But
when
there's
times
where
you
shouldn't,
then
that's
okay
and
the
way
that
we've
been
writing
up
the
best
practices
communicate
that
including
caveats
that
you're
going
to
run
into.
If
you
don't
do
this
Alternatives
any
anything
like
that.
So
if
there's
a
dissenting
view
that
we
think
is
important
to
highlight,
then
that
can
go
right
into
a
best
practice.
C
If
we
think
that
it's,
you
know
a
best
practice
that
should
be
followed
in
most
cases
and
this
particular
one
is
related
to
how
you
would
be
utilizing,
microservices
normally,
which
would
tie
into
I
think
a
lot
of
the
best
practices
and
Cloud
native
practices
that
you
would
be
following.
C
D
Yeah,
the
the
other
team
Taylor
I,
don't
remember
like
last
last
time
we
were
talking
about
abuse
on
that
also
brought
the
the
topic
about
having
using
what
was
the
name
container
no
process
Handover
in
the
in
the
in
the
key,
because
the
the
container
requires
to
manage
the
finally
improperly
that
probably
the
the
concern
about
like
doing
in
the
right
way,
which
was
related
with
the
topic,
but
not
directly
tied
to
the
one
process
per
for
container.
D
So
one
of
the
topics
that
I
found
was
related
to
the
most
of
the
cases
like
I
mean
you
can
definitely
you
can
manage
multiple
processes
per
container.
One
of
the
major
problems
is
like.
D
If,
if
you
are
not
propagating
those
signals
to
to
The
Container
manager,
you
get
a
problem
so
and
they
were
suggesting
from
articles
were
suggesting
to
use
other
tools
in
the
docker
entry
point
yeah
properly
those
things
I
mean
it
was
related
with
the
the
one
process
per
container,
probably,
but
also
it
was
another
kind
of
best
practice
for
that
hold
up
Docker
grade
the
container
creation.
C
Yeah
well
I
mean
if
this
we
probably
should
just
start
coming
up
with
a
list
of
those
either
as
they
can
go
in
as
issues
or
into
the
discussion
forum.
For
those
ideas,
one
of
the
things
that
were
talked
I
think
we
talked
about
a
little
bit
was
you
can
have
zombie
processes
and
other
stuff,
and
so
you
can
start
looking
into
process
managers
and
there's
software
out
there.
That
does
pretty
good
management.
C
I.
Think
I
may
have
already
mentioned
that
like
email
systems,
so
qml
I
think
was
one
of
the
big
examples
of
separation
of
concerns,
implemented
and
processes
that
are
running
and
actually
isolating
them
under
different
users
and
stuff.
So
if
anything
was
if
any
one
of
them
had
a
problem,
including
like
external
security
issues
and
and
postfix,
which
was
a
a
rewrite
of
that
by
someone
from
IBM,
those
are
two
large
Mill
I.
C
Think
if,
if
you're
into
email
at
all
as
far
as
the
systems,
then
most
people
think
sunmel
and
but
qmail
and
and
postfix
would
be
the
ones
that
they
actually
handle
a
larger
quantity
so
like.
If
we
look
at
growth-
and
they
were
both,
postfix
was
designed
like
qmail,
to
run
on
different
processes
and
those
are
internal
versus
saying
using
in
it.
C
What
is
your
init
process
manager
and
there
are
those
as
well,
but
you
can
actually
build
a
system
that
does
process
handling
and
if
you
look
at
languages
like
erlang,
well,
it's
all
about
having
it's
building
applications
that
have
lots,
and
you
know
tens
of
thousands
of
processes.
It's
designed
for
that
and
and
separation
of
this
using
some
type
of
process
manager,
where
someone
actually
already
knows
what
they're
doing
and
you're
taking
advantage
of.
C
That
is
a
good
path
if
you
haven't
already
been
Building
Systems,
like
postfix
or
software,
the
I
deal
with
a
best
practice
of
one
process.
Type,
not
multiple
processes,
but
one
process.
Type
in
a
container
is,
is
a
method
of
Separation
similar
to
what
postfix
is
doing
with
multiple
concerns
broken
into
processes,
they're,
not
containers,
they're,
just
processes
or,
if
you're
running
something
like
erlang
and
you're
having
a
lot
of
processes.
C
But
each
little
component
may
be
split
off
versus
saying
I'm,
going
to
have
one
container
and
run
running
using
a
process,
Monitor
and
manager
that
can
run
everything,
but
you
end
up
with
lots
of
process
types
which
moves
you
towards
it's
an
anti-pattern
for
micros
following
microservice
model
and
there'd
be
a
lot
of
different
things
that
you
could
start
pointing
out.
But
this
again
is
more
of.
Is
this
the
path
that
works
for
you
like?
Do
you
want
to
take
all
the
benefits,
and
you
have
caveats
or
not
and
they're?
C
You
know
you
could
definitely
have
a
a
container
I,
don't
want
to
say
a
CNF
I'm
going
to
be
more
specific.
You
have
a
single
container
that
has
multiple
process
types
and
you
have
a
very
strong
and
valid
reason
for
that,
but
is
that
the
best
practice
for
everybody?
That's
that's
all
we're
saying
like.
Should
most
people
follow
the
practice
of
trying
to
separate
concerns
and
that's
what
each
process
type
is
there.
You
know
I
would
say
probably
it's
a
different
concern.
It's
handling
something
different.
C
C
Erling
is
hard
for
me
to
think
about
in
kubernetes,
because
it's
in
a
lot
of
ways
and
and
people
do
write.
Erling
and
erling
derivative
applications
running
on
kubernetes,
but
what's
weird
for
me,
is
erlang
in
a
lot
of
ways
is
designed
to
work
like
kubernetes
like
it
is
essentially
kubernetes.
It
has
its
own
system
for
handling,
restarts
and
monitoring,
and
all
of
the
a
lot
of
the
stuff
that
kubernetes
does
a
framework
and.
C
C
It's
also
designed
to
run
tens
of
thousands
of
processes
and
when
you're,
when
you're,
building
an
application,
and
it's
designed
to
split
them
all
up
but
kubernetes
is
designed
to
Warrior.
You
know
you
would
build
lots
of
containers
and
they
work
together.
You
know
not
I'm
running
a
kubernetes
cluster
and
I
have
three
big
applications
that
are
five
gigs
of
pace
with
massive
data
sets.
That's
an
unusual
case
for
kubernetes.
Most
of
the
time.
C
You're
talking
about
thousands,
you
know,
or
people
are
talking
to
thousands
or
tens
of
thousands
of
containers
on
running
on
a
cluster
and
with
erlang
it's.
It
feels
a
little
weirder
to
split
it
out.
I
mainly
brought
it
up
as
an
example
of
that
particular
programming
language
you
would,
if
you're,
if
you're
building
a
container,
it
feels
a
little
weird
to
split
an
erlang
container
out.
C
So
if
you,
if
you
had
an
erlang,
Telcom
application
and
you're
trying
to
run
it
on
kubernetes
I,
can
see
how
splitting
different
process
types
into
different
containers
doesn't
make
as
much
sense,
because
you're,
actually
not
taking
advantage
of
erling
like
the
actual
system,
was
designed
to
have
really
high
speed
connectivity
between
essentially
its
own
little
mini
containers.
What
would
be
running
on
the
erlang
VM,
but
I
would
count
erlang
as
a
a
very
small
subset
of
applications
that
are
targeting
kubernetes.
C
As
far
as
I
know,
I'm
I'm,
not
hearing
about
a
lot
of
telecoms
that
are
going
hey,
we're
trying
we're
having
trouble
I
like
the
language
but
I
think
for
most
languages.
That's
not
the
case,
including
like
Java.
You
know,
C
plus
plus,
you
know.
Besides,
you
know,
and
most
Mustang
language
that
I
can
think
of
right.
Now,
it's
that's
not
an
issue.
A
Just
conscious
of
time,
I've
got
to
jump
to
the
next
call,
I'm
afraid
I
think.
That's
that's
been
quite
interesting.
Useful
discussion
I
know
it's
hard,
but
if
you,
if
we
can
summarize
some
of
that
in
the
in
the
issue,
I
think
it
will
help
whoever
comes
to
write
that.
B
I
know
I,
know
we're
running
over
time,
but
I
just
wanna.
It's
a
comment.
I
think
we
can
address
on
the
next
time
around
on
that
specific
I
think
we're
talking
about
Taylor
you're,
talking
about
it
being
in
some
cases.
You
know
there
may
be
reasons
for
not
doing
this.
You
know
and
you're,
in
other
words,
you're
saying
this
doesn't
necessarily
apply
to
to
everyone.
There
are
caveats:
Etc.
B
The
only
thing
I
think
that
we
need
to
address
somewhere
and
all
that
is
I,
think
people
think
about
the
certifications
and
things
like
that
and
if
it's
an
essential
test-
and
you
don't
do
this,
then
you
lower
your
chances,
perhaps
okay,
being
able
to
certify
so
I
think
that
needs
to
just
be
addressed
some
way,
because
that,
ultimately
we
don't
want
to
turn
people
away
saying.
Well,
if
you,
if
you
have
a
good
reason
for
doing
it,
go
ahead
and
do
that,
but
that
doesn't
you
may
not
get
your.
B
C
So
if,
if
it
turns
out
that
we
have
a
large
set
of
things
that
we're
all
recommending
these
are
essential,
and
then
we
have
the
majority
of
people
not
being
able
to
get
certified,
you
know
we
can
think
about.
What
does
that
mean?
Ideally,
folks
are
getting
helped
and
they're
becoming
more
Cloud
native.
That's
that's
the
point
and
the
the
long
run
I
mean
I
I'm
wanting
people
to
do
these
things,
because
I
literally
think
that
it's
gonna
be
beneficial.
C
E
C
For
talks,
just
Reach
Out
thanks.