►
From YouTube: Kubernetes sig-aws 20171215
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Good
morning,
everyone
it
is
Friday
December
15th.
This
is
the
Korean
basic
AWS
bi-weekly
meeting
the
meeting
is
recorded
and
will
be
posted
on
YouTube
I
promise
and
we
do
have
an
agenda.
So
if
it's
pretty
light
with
the
December
workload,
but
if
anyone
wants
anything,
please
do
so
right
now.
The
only
thing
we
have
an
agenda
is
something
I.
Don't
know
that
b19
release
is
very
likely
to
land
today,
if
not
this
weekend
or
this
weekend-
and
there
are,
there-
are
a
bunch
of
features
for
AWS.
There
are
volume,
improvements
and
vme
support.
A
B
D
I
have
a
follow
up.
Pr
I
posted
in
cig
docks
right
now,
I
think
there's
until
1.9
is
out,
there's
like
a
docks
freeze
on
release,
1
9
and
the
so
I
can't
edit
I
don't
want
to
make
a
PR
to
release
1
9
I'll
have
to
make
it
to
master
once
that
gets
merged,
but
I
have
a
patch
already
to
make
and
I'll
make
a
PR
later
today
with
fixes
for
that
and
I.
D
B
D
A
It
is
possible
by
the
way,
to
run
NEC
I
built
locally.
They
do
get
deleted
every
month
or
so.
But
if
you
have
a,
if
you
have
a
even
a
PR
will
have
a
publicly
available
one
and
you
can.
You
can
look
at
the
cops
test
to
see
the
URLs
that
it
needs.
There's
basically
a
tree
that
mirrors
the
release
tree
and
you
can
use
those
artifacts.
You
can
pass
the
base
URL
the
cops,
which
is
how
the
testing
works
or,
if
you're,
using
something
different.
A
B
And
I
think
that
would
be
a
nice
blog
post
as
well,
that
you
know
I
want
to
set
up
a
random
kubernetes
version
cluster
on
AWS
and
co-ops
allows
me
to
do
that
and
how
do
I
do
that?
Because
the
more
we
make
it
visible
that
oh
I
want
to
test
this
particular
PR,
and
here
is
how
I
did
it
so
maybe
I'll
you're
giving
me
ideas
of
my
new
blog
post.
Essentially
that's.
A
B
A
The
easiest
way
is,
if
you
go
to
a
PR
and
click
through,
you
should
go
to
see
the
logs
and
see
the
build
log,
and
then
you
get
the
full
build
log
and
in
there
it
will
have
a
you
search
for.
If
you
search
for
HTTP
colon,
slash,
slash,
that's
pretty
the
easiest
way
to
find
it
because
it
will
be.
Oh
actually,
I'm,
not
even
sure
if
they
should
be
asking
us.
Let's
Ruby
issue,
yes,
but
then
black
yeah
well.
A
A
The
hunt
on
your
blog
post
instead
I,
don't
know
thing
was
to
claim
to
create
and
tag
I'm
reasonably
sure.
I
didn't
do
that.
It
is
a
fairly
simple
PR.
If
someone
wants
to
do
that,
the
idea
being
that
create
and
tag
are
two
operations.
So
there
is
a
window
where
you
create
an
EBS
volume,
but
it's
not
tagged,
so
it
could
in
theory,
leak.
Yes,
in
March
issue
announced
an
atomic
operations,
new
EBS,
create
and
tag,
and
we
should
just
switch
to
a
previously.
A
A
Ok,
let
me
know
if
you
wanna
claim
it
about
that.
Alright,
let's
see
we
have
some
closing
tender
already,
which
is
great
well,
the
to
finish
up
the
1:9
thing.
There
are
some
notes
if
you're,
like
a
device,
has
more
features
than
most
so,
do
you
have
a
quick
look
at
the
release,
notes
and
yeah?
I
NLB
is
great
in
alpha.
I
wouldn't
recommend
it
in
production,
yet
I
think
that's
fair
right
like
that
and
or
like
yeah,
but
like
try
it
out
in
your
dev
cluster
that
we
know
we
found
introducing
with
two
ports.
A
Anyone
I
know
I
would
suggest
that's
a
really
great
introduction,
so
there
we
go
and
yeah.
Anyone
wants
to
look
at
the
release.
Notes
do
fire
a
comment
but
I
think
they're
they're
starting
to
get
lock
them
down
for
for
final
release
so,
but
we
can
always
do
them,
one
on
one
note
or
try
to
get
enough
to
it.
Otherwise,
I'll
move
on
to
the
next
thing.
A
E
Me,
let
me
say
a
word
about
1:9.
If
you
don't
mind
Justin,
so
1:9
is
upgrading
CNI,
the
actual
interface,
which
makes
me
needless
to
say,
very
nervous,
so
big
changes
around
networking
ipv6
is
an
alpha.
So
little
caution
warning
fingers
crossed
definitely
don't
run
want
a
no
in
production.
Let's
wait
till
our
one
night
on
productions
right,
let's
wait
till
193
or
192,
so
everybody
gets
the
networking
hiccups
worked
out
is
I
guarantee
you
there's
a
bunch
of
CNI
providers
that
are
just
gonna,
be
on
fire
afterwards.
B
E
Dns
yeah
Justin.
Do
you
know
if
they're
an
alpha
I
think
they're
an
alpha
four
one:
nine,
that's
what
I
could
I'm
sorry,
Cordy
Ness
Oh.
Somebody
else
chimed
in
so
for
DNS
is
an
alpha
four
one,
nine
they're,
dropping
and
supporter
into
cops
right
now
so
but
yeah.
If
we
can
start
testing
that
on
AWS
I'd
be
wonderful
as
well,
because
I
know
it
hasn't
been
burning
in
AWS
watch.
Well,.
B
I
mean
one
thing
we
have
done:
is
the
nice
part
of
the
AWS
workshop
that
we
are
built
for
kubernetes?
We
have
like
a
chapter
which
talks
about
how
to
use
code
BNs
anyway,
and
that
chapter
has
been
tested
by
a
few
folks,
I
know
because
it
has
had
some
updates
and
all
so
that's
one
part
we
have
verified
and
that's
what
was
wondering
when
there's
a
cube
dns,
the
core
DNS
change
happens.
E
Cni
one
font
or
the
previous
version
works
in
Gua
by
the
way,
with
one
nine,
because
that's
all
we've
been
testing
against
Justin
in
cops,
so
we
haven't
even
tested
the
new
CNI
provider.
Four
one:
nine
on
AWS,
that's
gonna,
get
fixed
all
right!
Well,
merge
my
PR
for
final
assets
and
I'll
fix
it
right
after
that,
anyways.
A
B
It
would
be
nice
to
get
to
this
level,
but
I'm
not
really
getting
a
clear
number
that
what
is
it
that
the
limit
should
be
increased
to,
because
once
I
get
the
number
that
okay
110
whatever
that
concrete
number
is
nice
to
is
good,
but
is
that
the
number
that
you
want-
or
you
want
a
bigger
than
that
number
so
that
I
want
to
go
back
and
ring
the
bell
once
and
then
say?
Okay,
you
know
what
this
is.
The
number.
A
So
this
is
about
concurrent
testing
on
ete
infrastructure
and
what
happens
on
the
Google
side?
Is
they
have
something
called
boss
cos?
I
was
trying
to
remember
that
name
and
what
boss
cos
does.
A
That
does
not
happen
on
AWS
currently
and
so
I
think
the
answer
is
we
don't
have
a
limit
technically
right
now,
but
I
would
say:
let's
pick
a
sensible
number
and
then
let's
implement
boss
cos.
So
if
the
number
is
gonna,
if
100
sounds
great
and
then
like
I
feel
like
the
other
issues,
if
we
have
a
hundred,
if
we
have
20
clusters
running
in
one
account,
we're
gonna
start
hitting
other
rate
limit
issues.
You
know
the
natural,
a
diverse
API
rate
limits,
because
we
don't
we're
not
light.
A
The
et
tests
are
not
light
on
those
kubernetes
api.
They
you
know
they
will
try
to
create
volumes
and
do
things
like
that.
So
I
would
say
let's
if
a
hundred
is
doable
this
to
100
and
then
we
can
implement
boss
cos
or
whatever
the
equivalent
would
be
so
that
we
limit
the
amount
of
concurrency.
That
happens
because
you
know
100
100
instances
per
let's
per
region.
A
E
A
E
A
A
A
A
There'll
be
more
than
we
want,
I
think
I
think
that
yes,
but
so
that
would
be
great
I
think
you
know
we
could
probably
we're
not
really
gonna
want
to
run
more
than
10
tests
in
a
region,
no
matter
what,
because
we'll
hit
other
limits
which
will
make
the
tests
flaky
and
we
have
to.
We
have
to
deal
with
those
things,
but
we
have
to
deal
with
those
issues,
but
we
don't
want
the
tests
to
be.
We
want
to
test
them
as
a
genuine
test.
We
want
to
be
randomly
if
causing
PRS
to
flake,
so.
B
A
Just
we
do
is
we
have
one
region
where
we
one
region
or
one
accounts,
where
we
have
a
limit
that
will
enable
us
to
do
whatever
test
you
want
I
think
we
can
have
a
limit
of.
We
have
a
limit
on
one
instance,
type
right
that
we
could,
let's
figure
out
what
instant
site
that
is
in
that
one
region
or
one
account.
We
will
have
a
limit
that
is
much
higher.
Okay,.
B
Okay,
fair
deals,
I
guess,
I'm
still
hearing
that
we
have
an
approximate
idea
of
what
would
we
like
it
to
be,
but
we
don't
have
the
concrete
details.
So
I
guess
my
point
is
I'd
sent
a
mail
to
send
this
morning.
I
would
like
to
know
those
exact
details
so
that
once
I
have
the
details.
I
just
want
to
send
one
email,
but
I
want
14
regions
all
pumped
up
200.
B
A
B
F
E
A
I
mean
that's
like
whatever
we
don't
have
to
be
on
the
same
schedule.
I
think
I've
run
if
there
is
a,
if
there's
an
instance,
type
or
class
in
a
region
where
we
were
it's
better
for
you
know,
because
it
is
an
expensive
test
to
run
not
anywhere
brain.
This.
E
Is
complex
enough
I
would
almost
recommend
setting
up
like
a
call
for
whoever
wants
to
be
involved
with
it
because,
like
we
have
to
babysit
this
sucker
the
first
couple
times
it
runs
and
I
want.
You
know,
I,
don't
want
it.
The
worst
thing
we
can
do
is
slam
a
region
hard
on
you
guys.
You
know
impact
production
with
their
stupid
testing.
E
B
No
I
understand
that's
none
of
my
concern
at
this
point
of
time,
because
I
don't
think
I
have
the
clear
requirements
on
what
exactly
the
configuration
we
are
looking
for.
So
babysitting
is
one
part
of
it,
but
before
we
even
babysit
before
I,
even
go
back
to
the
AWS
staff
that
okay
guys,
this
is
what
I
increase
the
limit
to
be.
We
need
to
understand
what
the
frequency
is.
B
B
Okay,
what
I'll
do
is,
maybe
I'll
ask
you
know
so,
Chris
you
and
Justin.
You
are
on
that
thread
as
well
and
by
the
way.
What
I
want
to
do
is
I
want
to
take
this
moment
to
introduce
somebody
on
the
call
Bryce
Carmen
Bryce
works
for
Amazon,
and
he
is
the
eks
leading
in
here
here
he's
the
lead
architect
for
he
casts.
So
what
I'll
do
is
I'll
Bryce
in
the
discussion
as
well.
A
G
B
Oh
yeah
I
think
I'm
not
related
to
AWS
SiC
per
se,
but
I
wanted
to
show
this
new,
so
docker
as
well
as
a
dr.
captain.
I
got
limited
access
to
how
you
can
set
up
kubernetes
using
docker
for
Mac,
so
that
was
a
brand
new
tool
that
was
introduced
literally
yesterday.
Well,
two
days
back
actually
and
I've
been
playing
with
this.
B
B
H
My
name
is
Henry
harlot
I'm,
a
you
know,
software
architect
and
I've
been
working
with
getting
spring
boot
rest
services
running
in
docker
and
kubernetes
on
my
Mac
I've
got
some
spring
boot.
Open
source
project
work
that
I've
published
on
github
and,
at
this
point
I
think
I've
got
through
the
growing
pains
of
getting
darker
running
locally,
and
you
know
my
spring
boot
service
running
in
kubernetes,
locally
I've
also
used
data
stacks
driver
to
connect
to
Cassandra
both
locally
or
on
the
Amazon
Cloud
and
I
was
disappointed
with
data
stacks
image
missing.
H
So
I've
got
a
single
node
Cassandra
running
with
bit
Nami
on
AWS
ec2,
so
I'm
happy
to
share
whatever
work
that
I've
done
since
I've
done
the
compatibility
testing.
What
I
don't
have
visibility
into
is
how
successful
it's
gerber
datings
on
AWS
but
I
think
what
I've
learned
the
kubernetes
pods
are
not
configurable,
which
I
think
according
to
Google
was
they
were
calling
it
templating.
H
So
the
concept
of
configurable
pods
was
not
offered
so
I
added
configurability
to
a
rest
service
by
adding
REST
API
to
configure
the
pods
themselves,
mainly
to
configure
remote
IP
address
for
the
AWS
ec2
instance
and
login
and
password
for
the
Cassandra
node,
not
my
preferred
method
of
configuring
pods,
but
that
was
the
only
fallback
option.
So
that's
what
I
have
at
this
point.
I.
A
A
A
H
A
H
B
I
think
this
is
very
interesting.
You
know
I
mean
we
would
love
to
hear
more
of
these
and
hear
what
people's
concerns
are
on
running
kubernetes
on
EWS,
so
anything
is
valuable
than
that
sense.
Alright,
so
I
was
jumping
to
the
point
where
I
was
ready
to
share
my
screen.
I'm
gonna
go
here
and
share
my
screen
here,
and
so
here
I
am
like.
If
I
look
at
darker
here,
essentially
I
can
go
to
now
preferences
and
in
preferences.
B
I
see
a
brand
new
tab
here
which
says
kubernetes
now,
I
can
I
mean
I've
already
enabled
it,
but
essentially,
if
I
enable
it
I
can
say,
show
system
containers
as
well
now
I
have
docker
is
running
and
kubernetes
is
running,
so
both
of
them
are
essentially
running
on
my
local
machine,
and
this
is
not
mini
queue
at
all.
This
is
not
using
mini
queue.
This
is
using
docker
format,
the
docker
desktop.
So
essentially
now,
if
I
go
here
and
I
say
cue
card
o
to
convey
get
context
if
I
can
type
correctly.
B
It
shows
me
to
context
here.
There
is
one
context
for
mini
queue
and
there
is
one
context
docker
for
desktop
and
if
I
say
to
cuddle
get
nodes,
it
shows
me
the
nodes.
You
know
if
I
say
pod,
it
will
show
me
the
pods
now
I
can
deploy
my
applications
over
here
and
then,
if
I
say
docker
word
run
here.
It
says
the
orchestrator
is
kubernetes.
B
So
one
of
the
challenges
that
we
have
had
in
our
workshop
is
even
getting
mini,
cube
to
install
because
it
requires
VirtualBox
and
people
have
an
older
version
of
VirtualBox
and
then
they're
not
able
to
install
it,
but
they
typically
do
have
a
darker
desktop
installed
on
it.
So
once
this
does
become
a
ga,
the
way
I
was
thinking
is
the
local
development
environment
could
use
docker
desktop,
and
then
we
will
say,
oh
by
the
way
you
have
multiple
contexts
set
up
over
here
switch
your
context
from
your
desktop
to
a
cop
space
cluster.
B
Pushing
out
a
blog
literally
in
a
day
or
so,
which
talks
about
how
do
you
run
communities
locally
on
your
machine?
One
is
using
mini
tube,
the
other
one
is
using
dock
of
a
desktop.
Those
are
the
sort
of
credible
options
and
then
I
think
this
is
something
good
that
people
should
be
aware
of
and
start
looking
at
it.
B
The
other
value
that
it
also
provides
is
if
you
are
born
into
the
dock
or
ecosystem,
and
you
are
building
your
applications
using,
say,
docker
compose,
then
you
can
take
a
docker
compose
application
and
because
your
Orchestrator
is
now
set
to
kubernetes,
you
can
deploy
your
docker
compose
applications
directly
to
docker
and
it
will
use
the
kubernetes
scheduler.
Essentially
hey.
F
B
A
B
Don't
know
because
all
I've
done
is
I've
enabled
kubernetes
I
was
able
to
deploy
a
pod
or
so,
and
my
next
venture
I
literally
got
this
working
yesterday,
but
my
next
venture
would
be
to
actually
start
deploying
some
real
applications
to
it
and
I'll
take
a
look
at
what
type
:
load
balancer.
Does
you
know
how
does
service
discovery
work?
Those
all
those
functionalities
really
look
cool.
So
that's
what
I'm
going
to
be
hitting
on
today
and
maybe
early
next
week,
yeah.
E
A
I
don't
know
if
this
year,
but
Seth
pointed
me
towards
something
which
cloud
flare
did
where
they
have
an
ingress
controller,
which
does
a
warp
connection,
so
that
might
make
ingress
actually
work
on
your
laptop,
which
would
be
pretty
cool
sort
of.
Like
I.
Think
someone
described
this
n
grok
for
for
ingress.
B
B
I
should
have
some
sort
of
a
working
experience
like,
of
course,
when
I
go
to
AWS
it'll
be
a
ELB
or
NLB,
whatever
works
in
a
more
scalable
way,
but
at
least
on
a
local
dev
experience
by
default
it
should
fire
up
an
engine
X
or
what
am
I
ingress
controller
I.
Don't
care
really,
but
give
me
an
English
English
controller,
get
my
app
up
and
running
so
that
I.
Can
these
get
a
similar
behavior?
B
C
B
And
that's
my
intent
as
well,
because
now
the
discussions
for
the
cloud
provider
working
group
has
started
already,
but
maybe
they
have
been
going
on.
We
just
getting
into
it.
Yeah
private
slack
channel
for
cube
you
for
captain's,
I,
think
and
then
I'm
gonna
give
this
feedback.
That
I
would
like
docker
guys
to
join
the
working
group
or
cloud
provider
and
have
an
experience
by
which
we
can
actually
have
a
working
service.
So
the
type
:
load
balancer
should
not
explode.
Yeah.
F
E
A
D
B
B
A
A
D
A
F
Kind
I
mean
I'm,
really
I'm,
really
interested
in
that
as
well.
Actually,
it
might
be
interesting
to
have
like
a
quick.
You
know
sort
of
house
you
session
if
somebody
would
be
willing
to
host
that,
because
I
was
looking
at
some
stuff
as
Chris
love
about
at
CD
and
I.
You
know
just
didn't
even
know
how
to
get
started
so
I
might
be
real
interested
if
I
answered
that
Michael
would
be
good.
So.
E
A
I'm
happy
to
share
my
experiences.
I'm,
not
sure
I
am
well
we'll
just
get
one
of
the
guys
from
cig
test
as
well.
Yes,
they
can,
they
can
like
I
can
give
like
the
practical.
This
is
what
you
probably
need
to
know
and
then
maybe
suggest
important
like
everything
that
I
was
wrong
about,
but,
like
I
got
away
with
before
a
reasonable
it
is
it
just
works
but
yeah.
It's
there's,
there's
a
bunch
of
things
to
sort
of
understand,
but
I
have
a
superficial
understanding
and
I
can
make
it
work.
A
B
E
A
A
Mars
has
also
been
like,
because
I
think
there's
another
renegade
group
like
this
meeting
somewhere
in
secret,
so
reach
out
to
him
he's
sort
of
organizing
that
I
don't
know
what
their
plans
are.
I
think
previously
was
sort
of
unclear
who
really
was
responsible
for
how
to
operate
it
and
that's
gonna
change,
I
think.
Hopefully
they
actually
I
saw
they're
gonna
win
out,
look
at.
They
are
now
deprecating
at
CD
2
in
one
nine,
so
the
clock
will
start
ticking
and
that
will
mean
that
everyone
will
have
to
sort
of
take
this
more
seriously.
A
The
in
terms
of
the
configurations,
although
a
lot
of
people
run
a
separate
at
CD,
cluster
I
think
the
only
configuration
that's
actually
tested
in
ete
right
now
is
with
ed
CD
on
localhost,
so
the
sort
of
tight
coupling
of
API
server
and
Ed
CD,
and
so
you
will
see
different
behaviour
in
terms
of
latency.
For
example,
if
you
spit
them
out,
but
you
have
nicer,
you
know
full
behaviour
like
if,
if
it
CDs
on
isolated
cluster,
then
if
your
API
server
consumes
all
your
memory,
you're
not
gonna
lose
that
CD
yeah.
B
A
E
So
every
time
I've
asked
for
data
and
that's
that's
usually
what
I
base
my
guidance
on
I
really
haven't
gotten
solid,
actionable
data,
so
I
think
through
the
testing
that
we
do
in
Bryce.
I
know
you
guys
are
doing
internal
tests
in
it.
If
we
can
get
kind
of
take
the
ETH,
a
testing
that
we're
doing
and
the
testing
that
you
guys
are
doing
and
combine
them
I
think
we
can
come
up
with
a
couple
recommendations
and
more
like
scenarios.
G
E
Performance
impact
to
the
different
controllers:
what
happens
when
you
go
from
three
to
not
the
you
know
nine.
What
happens
you
know?
Basically,
so
we've
got
single,
headed
H,
a
super
H,
a
alright,
so
and
and
what
a
ruin
referenced
was.
We
still
have
some
open
questions
about
how
the
exact
implementation
of
crap,
the
H,
a
naming
or
the
the
name
way
they
get
quorum
escapes
me
right
now,
but
we
have
some
questions
about.
E
A
Yes,
yeah
the
I
hope
we
can
find
a
venue
to
discuss.
This
is
a
much
broader
discussion
and
we
probably
want
to
have
in
I
mean
it's
obviously
very
relevance.
I've
run
in
this
channel,
but
it
also
is
relevant
to
you
know
everyone
that
is
operating
in
kubernetes
cluster.
So
so
you
trust
a
life
cycle
and
or
sig
caster,
ops
and
ie
I'm
I
don't
want
to
yeah
I,
don't
know
I,
it's
a
huge
topic
and
I.
A
They're,
not
they're,
not
even
gonna,
be
a
miss
sake,
so
it'll
be
good
to
figure
it
out,
but
I
think
yeah
I,
don't
know
I
guess
the
next
I'll
find
out
whether
secrets
still
X
I'll,
put
it
on
the
agenda
and
try
to
update
the
agenda
with
what
I
find
out,
but
I
guess
to
find
out
whether
it's
official,
the
successor
lifecycle,
now
owns
that
CD
and
whether
we're
gonna
have
a
working
group
or
what
the
plan
is.
I.
A
B
A
Yeah
I
think
there's
a
there's,
a
PR
that
I
think
basically
looks
good
I.
Think
there
was
some
discussion
open
discussion
about
whether
it
were
there
was
remaining
issues,
but
my
my
idea
is
that
we
should
probably
merge
it
and
those
issues
merged
with
a
suitable
caveat
and
then
people
experiment
and
find
those
issues
and
be
able
to
use
the
plug-in
in
a
you
know,
integrated
way.
Yeah
I.
E
B
E
We
all
get
that
it's
alpha
my
question,
the
question
that's
on
the
PR
is:
do
we
have
it
hidden
behind
a
cops
feature
flag
which
forces
the
user
to
set
in
an
environment
variable
to
even
be
able
to
use
it,
or
do
we
just
release
it?
As
is
and
say,
hey,
you
know,
be
warned
it
ain't
production.
You
know
it's
great,
it's
awesome,
but
please
don't
run
in
production.
Yet
yeah.
G
B
A
President
is,
we
are
not
as
good
on
network
providers
in
terms
of
getting
them
as
we
are
in
other
places,
I
would
say
we
would
just
enable
it
and
say
you're,
not
gonna
accidentally
enable
the
provider.
You
know
without
typing.
You
know
network
whatever
that
magic
word
is
and
I
think
it's
kind
of
AWS,
Ian
I
see
and
I
something-something.
So
it's
not
it's
not
gonna
happen
by
mistake.
A
So
I
think
that's
that's
been
fine
to
date.
Whether
we
want
to
haven't.
We
want
to
be
structure
about
that
in
cops.
We
can
do
that,
but
I
I
don't
think
we
need
to
gate
this
particular
PR.
On
on
that
we
know
it's
alpha.
We
can
document
as
such
and
I
yeah
I
did
put
the
link
in
there.
It
is.
It
is
a
nice,
it's
a
cool
thing
because
I
know
Greg,
you
don't
cops
yet,
but
but
the
PR
is
there
and
I'm
sure
it
links
to
the
actual
plugin.
A
Yes,
that
would
be
awesome
yeah
exactly
like
you
know,
if
you
want
it,
if
you
want
to
try
the
the
latest
yeah
I'm,
really
the
gen
2
of
Cox's
are
we're
kubernetes
that
we
go
alright,
and
then
we
have
some
links
or
a
link
to
the
spring
boot
stuff,
which
quantum
infusion
crews,
who
is
a
renaming?
What
is
I
missed,
you're,
gonna,
I'm,
sorry.
A
A
A
H
Alright,
so
some
I
can
introduce
myself.
So
to
my
background,
I've
been
building
microservices
I
I
was
a
platform.
You
know
platform
architect
at
Verizon,
so
much
of
what
I've
done
has
been
battle
tested
on
Amazon
in
a
straight
easy
to
environment
with
spring
boot,
and
so
that
was
that
was
the
problem
in
2015.
H
B
H
H
It
supports
kubernetes
running
locally
and
it's
configurable,
so
you
can
point
to
a
database
back-end,
whether
it's
on
Amazon
or
running
local.
It
basically
runs
Cassandra
and
and
it
integrates
with
with
data
Stax's
AWS
as
well.
But
basically
you
just
point
to
the
Cassandra
backend
and
you
know
the
service
layer.
Rest
layers
is
a
separate
concern
from
the
database
itself.
A
Very
cool
stuff,
thank
you,
and
it
seems
like
that
is
gonna,
be
like
an
increasingly
big
topic.
I
know
like
Brendon
burns
did
a
keynote
about.
You
know
putting
in
annotations
in
the
code
which
sort
of
feels
like
you
know.
The
next
level
is:
how
do
we
actually
want
applications
on
communities
in
the
real
world,
and
this
is
great
to
see.
H
Exactly
so
so,
this
could
be
used
as
a
as
a
test
case.
You
know,
I've
also,
you
know
vetted
much
of
my
work
with
the
pebble
guys
that
write
the
spring
boot
stuff,
so
I
made
some
substantial
improvements
on
that.
There's
also
some
crypto
layers,
which
I've
not
fully
integrated
into
a
rest
service,
but
I've
been
working
on
end-to-end
encryption
as
well,
which
is
a
separate
project
not
related
to
this
discussion,
so
that
that's
kind
of
that's
kind
of
what
I've
offered
here
is
a
general
purpose
use
case
that
could
be
used
with
docker
kubernetes.
H
A
Thank
you
and
I
think
that
was
the
last
thing
any
other
drop-ins
they
want
to
fire
in.
Otherwise
everyone
gets
9
minutes
back
and
happy
holidays
as
you
choose
like
them,
B
holidays.
Are
we
meeting
in
two
weeks,
I
said
yes,
so
I
think
oh,
oh
they'll
be
December
30th,
December
29th
right.
A
Don't
we
we'll
have
a
meeting,
but
we
won't.
We
won't
decide
anything
too
important
like
we
won't
have
any
binding
votes
unless
I
decide
to,
but
we
will
have
a
meeting.
Framin
wants
to
show
up,
but
I
can't
imagine
that
here
and
if,
if
I
hope,
everyone
was
wearing
a
silly
hat.