►
From YouTube: TAG General Meeting - 2023-01-18
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).
E
E
Have
some
very
interesting
topics
on
the
agenda
today,
working
group
proposal
and
presentation
of
score.
C
That's
me
I'm
filling
in
for
Souza
she's,
not
feeling
well.
So
oh.
E
And
then
they'll
also
eventually
back
from
vacation,
so
I
know
that
Thomas
won't
join
today,
as
he's
traveling
for
vacations,
not
sure
about
Alex
on
the
tech
side.
Otherwise,
given
that
we
have
most
likely
a
pretty
packed
agenda,
giving
people
time
to
properly
presents,
it
seems
like
not,
many
people
are
joining
the
reading
today
directly,
but
rather
would
then
watch
the
recording
so
for
both
of
you,
Andrew
Paolo,
also
check
in
later
on
our
discussions
in
the
app
delivery
select
Channel,
where
people
might
chime
in
on
those
topics.
E
A
E
Up
if
you're
watching
this
later
on
in
the
recording
and
you're
interested
try
to
reach
both
of
today's
presenters
in
the
Tag
app
delivery
Slack,
since
he
has
like
a
workspace
all
right
Josh,
should
we
just
kick
it
off
yeah
we
just
follow
through
the
edge
and
I
think
both
topics
yeah
try
to
take,
ideally
about
like
20-ish
minutes
and
yeah.
C
All
right,
let
me
share
my
screen
hold
on
just
one
second,.
E
C
C
There
we
go
okay,
let
me
try
now.
All
right
is
that
is
that
better,
you
see
my
screen
foreign.
C
All
right,
thanks
for
coming
everyone
I,
just
wanted
to
kind
of
go
through
the
slides,
Souza
couldn't
be
here
today,
so
I'm
filling
in
for
her
I'm
substitute
teacher.
So
try
to
be
nice,
but
definitely
I
can
answer
questions
and
she,
of
course
the
answer
questions
we're
both
on
the
cncf
slack.
So
you
can
find
us
over
there,
but
we
can
definitely
answer
any
questions.
E
C
On
my
on
my
screen,
but
it's
not
on
everybody's
mind
all
right
so
score
one
sector
to
to
move
them
all,
so
it
was
launched
as
an
open
source
project
in
November
of
2022.
Not
that
long
ago
the
pre-launch
title
was
paused.
C
So
if
you've
heard
anything
around
something
called
pause,
that's
similar!
That's
not
a
mistake!
There.
They
were,
they
were
you
know,
pause
was
just
the
working
title.
Scores
of
what
we
released
as
score
is
primarily
been
a
part
of
a
Humana
Tech
project,
but
it
is
broadening
out
now
quite
a
bit
through
forks
and
other
people
using
it,
I
am
not
a
oh
I.
Am
a
customer
engineer.
C
She
updated
these
slides,
like
she
update
engineer
at
humanitek,
filling
in
for
Sousa,
who
is
our
product
manager
of
support?
C
So
what
exactly?
Do
we
mean
by
store?
What
is
for
it's
an
open
source
project
that
is
specifically
designed
to
allow
developers
to
enumerate
dependencies
and
resources
in
a
very
you
know,
a
generalized
and
platform
agnostic
way
to
that
makes
it
much
easier
for
them
to
you
know,
sort
of
State
exactly
what
they
need,
and
then
it
allows
platform
teams,
Ops
teams,
you
know
whatever
you
want
to
call
them
to
interpret
that
on
their
own.
C
So
you
know
the
the
score
developer
says:
I
need
an
RDS
database,
you
know,
and
then
it's
kind
of
up
to
the
the
the
the
Ops
Team
the
the
details
of
that
now,
the
developer
can
you
know,
get
into
the
weeds
if
they
want,
but
they're
not
required
to
it.
Eliminates
the
you
know,
the
the
need
for
for
them
to
get
into
the
weeds
there.
C
So
again,
like
I
was
saying
the
the
configuration
okay
consistencies
having
to
know
all
of
the
Gory
details
of
how
things
started,
staging
versus
how
things
run
in
Dev,
or
all
of
that
can
be
abstracted
away
from
from
a
developer,
and-
and
you
know,
then
this
also
allows
you
know
when
it's
no
longer
in
the
purview
of
developers.
C
It
also
allows
the
platform
team
to
you,
know,
establish
you,
know,
configurations
and
golden
pass,
and
things
of
that
nature
and
standardization
that
you
know
that's
going
to
make
them
happy
as
well.
B
And
is
this
mainly
Drew?
Is
it
meant
mainly
for
development
time,
or
would
this
even
apply
like
to
production
like
development
time
I
used?
You
know
an
embedded
SQL
light,
but
in
production
I
want
RDS
or
something
yeah.
C
That's
the
whole
idea
that
you
know
you
say
you
know
you,
you
say:
I
need
an
RDS
database
and
you
know
you
generate
a
Docker
compose
file
and
it's
going
to
run
that
RDS
database
in
a
container
for
you
know
kind
of
on
your
laptop.
But
then
you
know
when
you,
when
you
push
it
up
to
Dev.
You
know
it
runs
on
a
VA.
C
You
know
it
runs
on
a
VM,
that's
been
up
for
10
years
in
Dev
and
then
in
staging
it
runs
on
a
cluster
and
then
in
production.
It
runs
on
a
bigger
cluster,
with
an
F3
bucket
to
back
it
all
up
to
that
kind
of
that
kind
of
information
gets
abstracted
away.
They
don't
have
to
worry
about
that.
They
just
say:
hey
I
need
an
RDS
payment,
so
that
is
exactly
the
idea.
Yeah.
C
And
the
idea
here
as
well
is
when
you
kind
of
abstract
that
away
you
take
that
off
of
the
developer's
plate.
You
know
if
the
developer
doesn't
want
to
know
the
the
nitty-gritty
of
connecting
a
workload
in
kubernetes
to
a
VM.
You
know,
then
they
don't
have
to
they.
You
know
that
that
can
be
handled
by
the
platform
orchestration
that
can
be
handled
by
the
platform
team,
and
then
you
know
beyond
that
things
running
in
production
they
get
more
complicated.
C
Even
so,
you
know
the
the
the
the
the
amount
of
knowledge
required
to
work
in
that
is
even
higher
and
then
and
there's
no
standardized
approach.
There's
no
there's
no
Surefire
way
to
say
to
the
the
platform
team.
This
is
what
I
need.
You
know.
This
is
the
connection
string
or
you
know
here
is
a.
There
is
a
parameterized
version
of
the
connection
string
that
I
might
need.
There's
no
standardized
way
to
pass
that
information
over
to
the
platform
team.
C
So
so
this
is
kind
of
the
the
problem
that
we're
solving
here
is
standardized
way
of
communicating
your
dependencies,
both
inside
and
outside
of
kubernetes,
and
then
what
the
the
benefit
reap
from
the
standardization
is.
You
know
your
your
developers
aren't
getting
into
the
nitty-gritty
of
you,
know:
load,
balancing
and
and
and
persistent
volume
claims
and
kubernetes,
and
that's.
C
So
the
infrastructure
Centric
development-
this
allows
the
you
know.
Well,
it
doesn't
allow
but
they're
they're
that
it
causes
your
platform
team
to
be
overly
involved
in.
E
C
Yes,
absolutely
because
again,
what
happens
is
all
right
now,
I,
oh
I,
need
to
know.
You
know.
I
I
need
to
know
what
my
programming
language
and
then
on
top
of
that
I
need
to
understand
how
it
runs
in
on-prem
in
openshift,
but
also
how
we're
going
to
be
moving
it
over
to
AKs.
C
So
now,
I've
got
to
learn
AKs
and
the
other
thing
that
happens
is
you
start
your
development
starts
being
designed
for
whatever
Tech
fact
you
happen
to
be
working
with
so
then
your
your
development
becomes
more
and
more
specialized
itself,
such
that
when
you
do
want
to
maybe
make
a
shift
from
openshift
to
to
Azure
or
something
along
those
lines.
C
It's
going
to
be
a
lot
more
work
because
you
you're
you've,
already
you've
involved
your
developers
so
deeply
in
your
infrastructure
side
that
you
know
they've
written
code
perfectly
designed
to
your
infrastructure
that
day,
not
necessarily
future
looking
or
even
more
agnostic
to
allow
you
know
a
nimbler
footprint
and
a
wider
footprint,
so
yeah
we're
we're
still
in
the
problem
statement,
but
then
score
comes
in
giving
you
this
developer-centric
platform,
agnostic,
spec
and
eliminates
the
configuration
and
consistency
between
the
two.
C
The
way
it
does.
That
is
through
this
thing
called
it's
of
course,
it's
the
yaml
file
you're
a
surprise
to
no
one,
that's
going
to
capture
a
workloads,
runtime
requirements.
Now
one
thing
to
keep
in
mind
is
score
is
like
I
said
earlier:
it's
still
not
just
to
kubernetes,
but
to
your
platform
as
a
whole,
so
the
score
can
allow
you
to
declare
kubernetes
based
resources
and
dependencies.
C
So
you
know
if
it
needs
a
persistent
volume
claim
or
it
needs
an
indirect
configured
in
a
certain
kind
of
way
that
can
be
specified
in
your
score,
but
you
can
also
specify
things
like
S3
bucket
database.
Vector,
no
databases
other
things
that
don't
necessarily
reside
within
kubernetes
but
are
integral
to
its
running.
C
The
nice
thing
about
scoring
is
a
very
tightly
scoped
spec,
so
you
know
using
score.
You
can
only
get
so
far
into
the
weeds
just
because
it's
designed
to
be
platform
agnostic,
so
you
know
you're
not
going
to
worry
too
much
about
translating
that
to
compose
or
Hell
or
Nomad
or
kubernetes
or
or
whatever
it
is
you're
running
them.
So
so
that's
the
idea
of
the
benefit
from
the
the
developer
side
and
let's
take
a
look
at
what
a
score
file
would
look
like.
C
So
this
is
an
example
of
a
a
tip.
This
is
an
example
of
a
test
score
file.
This
is
a
score
file
we
use
in
in
our
training
and
our
tutorials.
But
you
know
you'll
see
here
it
looks
suspiciously
like
a
kubernetes
deployment
file.
C
It
has
a
lot
of
typical
kubernetes
style
information
within
there,
but
it
also
has
again
these
resources
and
database
declarations
that
are
going
to
be
external
to
kubernetes,
and
we
have
you
know
if
you
go
to
score.dev,
that
is
the
that
is
the
the
web
page.
There's
an
entire.
The
spec
is
completely
documented
there
with
lots
of
tutorials
and
and
other
ways
to
use
it.
So,
but
this
is
what
it
looks
like
the
idea
is.
C
You
would
keep
a
copy
of
your
score
yaml
in
the
same
place:
You're
Gonna,
Keep,
Your,
Docker
file,
so
the
gift
that
contains
your
Docker
file
without
also
contain
your
story
ammo,
and
it
would
accompany
that
Docker
file
as
it
gets
compiled,
and
then
it
would
accompany
that
image.
You
know,
as
it
goes
from
Dev,
to
staging,
to
production.
C
All
right
so
I
I
think
we've
pretty
much
that
that
is
the
the
two
parts
of
it
that
we've
covered
so
far.
It
provides
the
platform
of
specification
it
eliminates
inconsistencies
by
by
allowing
your
platform
team
to
set
up
things
like
golden
paths
and
those
types
of
specifications
now
score
in
action
and
it's
the
explore
implementation
DLI
and
there
are
multiple
one.
Currently
we
have
score
composed.
We
have
score
held.
We
have
sport
humanotech
as
well,
and
then
what
happens?
C
Is
you
use
this
CLI
file
to
generate
a
you
know
a
Docker
compose
file
or
to
generate
a
a
Helm
chart
or
to
generate
a
a
Delta
for
your
specific
platform
orchestrator.
So
so
this
is
kind
of
the
a
graphic
representation
of
the
the
previous
slide
here.
So
you
have
your
score
specification.
You
pass
it
through
the
CLI
tools
or
compost
or
help.
You
know
score
my
private
platform,
orchestrator
we're
coming.
C
You
know,
there's
going
to
be
score
for
things
like
Nomads
or
for
things
like
DC,
yes,
coming
as
well
down
the
pike,
and
then
you
know
coming
out
of
that.
Is
your
Docker
compose
your
help
or-
and
you
know,
if
your
own
private
platform,
orchestration
setup
that
you
know
the
yaml,
that's
going
to
then
be
fed
into
your
system.
That
will
then
interpret
all
right,
RDS
database
connection
string
in
staging
this.
Is
you
know,
setting
up
this
database
in
the
the
staging
cluster
and
you
know
providing
the
connection
string
to
the
workbook?
C
So
so
this
is
kind
of
the
idea
of
score
in
action,
I'm
trying
to
be
good
on
time
here
so
and
then
yeah.
So
what
what
happens
here
is
developers
declare
their
workloads
separate
from
any
platform
environment.
The
target
environment
is
responsible
for
making
sure
everything
is
running
and
set
up,
so
the
the
the
it's
the
platform
team
that
is
going
to
be
responsible
for
interpreting
the
output
of
the
score
CLI
and
then
running
that
so
in
this
type
of
environment,
the
the
the
you
know
the
all
of
the
worry
of
having
to
learn.
C
You
know
deep
tied
into
you,
know,
cnis
and
and
mesh
networks
and
and
istio,
or
something
like
that.
You
know
you
take
that
out
of
their
hands
and
off
of
their
off
of
their
plate
to
allow
them
to
focus
more
on
just
the
code.
B
C
No,
the
the
underlying
CLI
tool
would
be
determined
by
your
platform.
So
you
know
if
you're
using
Docker
compose
there's
one
called
score
human
attack.
Where
you
know
Humana
Tech
is
a
platform
orchestrator
there's
going
to
be
one
coming,
I
believe
for
ECS.
B
Okay,
how
does
score
like,
when
I
score
run
against
this
gamble?
How
does
it
know
which
you
know
whether
to
use
compose
or
Helm
or
Humana
Tech.
C
It
yeah
these
Peppers
CLI,
so
you
know
in
the
case
of
you,
wanted.
C
E
D
Okay,
hi
hello,
Andrew.
Thank
you
for
our
presentation,
I
Rise
to
my
brand
my
hand
because
I
I'm
doubt
about
when
you
said
configuration,
but
your
answer.
The
next
slide
said:
what's
your
focus
focus
on
the
environment
when
you
say
developers,
workload,
Centric
development
or
Development,
Center
developer-centric
configuration
in
my
in
my
mind,
I
I
talked
to
him
not
only
environment
but
also
configuration
for
the
application.
So,
but
your
sentence
now
explained
that
this,
for
the
configurations
configuration
here
is,
is
focus
on
the
platform
on
the
environment.
Right,
yes,.
C
Yes,
okay,
you're,
very
great
yeah.
Sorry
I
missed
your
question.
I
think
I've
got
the
thing
that
will
show
hands
now,
hopefully
was
that
there
was
another
question
as
well:
I'm,
sorry
about
the
this
particular
more.
E
E
So
so
my
question
I
think
it's
a
bit
more
on
the
maybe
5
000
feet
level
wire
was
held
not
good
enough.
Why
didn't
you
choose
to
go
so
something
like
you
Baylor
or
like
service
catalogs.
C
That's
a
very
good
question,
so
the
reason
why
we
love
Helm-
this
is
not
yeah.
We
love
Helm,
Humana
Tech
is
is
very
helm-like
in
the
way
that
it
deploys
to
kubernetes,
but
Helm
is
scoped
to
kubernetes
and
we
wanted
to
make
something
scoped
to
the
entire
platform.
So
so
that's
kind
of
the
idea
there.
C
As
far
as
cubella
goes,
that
is
the
instead
of
that
is
workload
resource.
So
she
volatiles
your
workload
specifically
to
the
resource.
Score
is
more
about
tying
your
resources
to
the
environment,
so
you
know
in
Cuba
you
specify
all
right.
I
have
this
I
have
this
workload.
This
is
how
it
runs
in
Dev.
This
is
how
it
runs
in
staging
it's.
This
is
how
it
runs
in
production
and
that
is
tied
to
the
workload
itself,
a
human
attack
or
I'm.
Sorry
in
score.
C
This
is
what
happens
when
you,
you
send
a
that's
a
substitute
teacher
in
store
you,
your
your
your
expensifications
are,
are
are
tied
and
interpreted
by
the
environment,
not
the
specific
workloads.
So
you
don't
have
to
you
can
just
say:
hey
I've
got
this
workload
and
then
the
platform
team,
the
platform
orchestration
team,
can
dynamically
decide
how
that
workload
is
going
to
run.
You
don't
have
the
the
the
the
the
way
score
does
it
is
it
moves
it
one
step
out
from
the
actual
workload.
E
C
And
human
attack
and
then
coming
out
I
believe
next
is
going
to
be
ECS
and
then
possibly
Nomad
I'm,
not
sure
there
are
several
people
that
are
working
on
using
score
for
their
people.
I
mean
organizations.
There
are
several
organizations
that
are
working
to
use
score
on
their
private
in-house,
a
platform
orchestration
so
I,
know
of
that
as
well,
but
that
those
are
the
current
ones.
B
Does
it
I
think
this
is
similar
to
Eloise's
question?
Is
it
that
you,
your
organization,
needs
to
write
the
score
CLI,
that's
appropriate
for
your
environment
in
order
to
get
these
benefits.
B
Like
how
will
how
does
it
know
that,
in
my
organization,
when
I'm
testing
I
want
to
use
RDS
and
in
prod
I
want
to
use
Azure
postgres
like
there's
not
like
somebody
has
to
write
that
logic?
Right?
Yes,.
C
A
C
All
of
the
above
of
those-
and
you
know
the
the
thing
with
platform
orchestrators
is
everybody
already
has
one,
but
most
of
ours
are
pretty
messy,
and
so
that's
why
this
this
whole
field
is
kind
of
opening
up
now,.
B
C
I
mean
I,
wouldn't
recommend
doing
that.
I
would
recommend.
You
know
something
more.
You
know
something
more
along
the
lines
of
you
know,
humanitect
or
or
some
other
off-the-shelf
orchestrator,
but
you
know
absolutely
I.
You
know
that
at
its
core
we're
all
already
doing.
B
Yeah,
will
you
have
any
sort
of
like
live,
I
mean
I
guess
so
you
have
score
composed
score
Helm.
Those
seem
pretty
generic.
So
like
do
you
have
anything
which
is
like
deploy?
You
know,
database,
let's
say
across
any
given
cloud
or
use
sqlite
locally.
Is
there?
You
know
that
kind
of
thing
certain
example
of
that
or.
C
Yeah,
absolutely
we
do
that
all
the
time
with,
with
our
with
Humana
Tech
yeah,
absolutely.
B
C
Whole
I
could
give
you
a
whole
demo
on
on
on
how
that
would
look.
Absolutely
yeah.
That's
not
a
problem
and
you
we
have.
There
are
score
launches.
You
know,
there's
four
tutorials
in
the
deck
here
as
well.
Sorry
I
didn't
get
to.
E
Them
I
mean
some
some
feedback
here
and
I
understand
that
everything
on
this
planet
is
kubernetes,
focused
and
obviously
talker
compose
is
nice
as
well,
but
especially
as
we
look
at
projects
we,
we
tie
them
back
to
like
the
core
principles
of
cloud
native
development.
The
operator
pattern,
even
if
it's
an
Implement
as
an
operator
but
I,
think
one
of
the
major
changes.
E
If
you
look
especially
in
I,
am
now
calling
it
a
deployment
use
for
the
sake
of
Simplicity
here
and
I
know
that
this
is
not
just
what
it
is,
but
does
it
also
have
some
ideas
around
reconciliation
when
environment
is
not
in
shape,
item
component
deployment
of
configurations
or
what's
even
the
capability
for
running
upgrades
on
on
the
existing
environments,
or
is
this
left
entirely
to
the
CLI
how
to
handle
this
because
the
way
I
understand
it
I
have
this
like
very
generic
file
and
to
do
Josh's
point
also
in
all
the
logic
it
is
like
kind
of
in
the
CLI
needs
to
be
re-implemented
in
in
all
of
those
individual
clients,
I
think
about
a
bit
of
long
questions.
E
C
It's
implemented
in
the
end
of
a
Market
Street,
so,
yes,
it
absolutely
does
support
that
and
it
is
implemented
in
the
platform.
Orchestrators.
E
B
E
C
That
I
don't
know:
you'd
have
I'm
sorry
that
I'm
gonna
beg
substitute
teacher
on
that
one
I,
don't
know
what
you
know:
yeah
that
one
I
don't
know
sorry
but
I'll
get
with
Susan
that
yes
and
ask
her
yeah.
E
It
is
Cuba
like
we
have
cross
playing
always
so
many
projects
in
this
space
together
with
skit-ups
tooling,
and
it
would
be
really
good
to
understand
how
you
see
this
fitting
into
the
entire
landscape
and
how
people
should
use
it,
not
just
whether
you're
going
to
submitted
as
a
Sandbox
project
right
so
I
think
just
for
people
like
building
their
stack
and
charge
here
is
actually
driving
a
lot
of
the
initiatives
around
like
platform,
engineering,
I.
Guess
people
build
out
their
platforms.
E
It
would
be
good
to
understand,
like
which
piece
of
the
platform
puzzle
does
it
really
fit
in,
and
how
does
it
work
together
with
with
others?
That's
what
I
was
I'm
still
honestly
struggling
a
bit
with
this
presentation,
I
kind
of
I
think
I
get
the
idea
where
you
want
to
be,
but
I
think
that
these
are
also
questions
that
might
come
up
so
I'm
already
using
Argo.
How
does
this
kind
of
help
me
I
can't
run
and
see
line?
Articles
is
happening
before
these.
Are
we
just
kicking
in
later?
E
How
do
you
work
with
my
say,
for
example,
cross
plane
resources
if
I
want
to
use
this,
or
is
this
just
to
be
totally
agnostic
and
tweeting
things
as
any
type
of
kubernetes
resources,
I
think
a
bit
more
of
an
understanding
in
this
front,
I
mean
I,
know
you'd,
be
a
people
jumping
in
here,
but
that
would
help
me
and
that's
what
I
would
consider
helpful
here
honestly.
So.
C
In
the
case
of
cross-plane
again,
I
I
have
seen
that
we
do
need
that.
You
know
your
your
platform.
Orchestrator
would
just
spit
out
a
pro
across
the
ammo,
based
on
the
information
it
got
from
the
score.
B
C
Because
I
mean
that's
how
we
do
it
now,
and
you
know,
for
the
case
of
in
the
case
of
get
up
your
platform,
orchestrator
will
spit
out
the
the
the
magic
yammals
and
put
them
in
the
magic,
get
Ops
directory
and
get
up
we'll.
Take
it
from
there
and.
C
B
A
lot
of
like
Argo
CD
has
like
a
renderer
for
customize
and
Helm,
and
it
doesn't
have
compose,
but
the
like.
It
comes
out,
I'm
kind
of
coming
to
the
conclusion
that
score
is
pretty
much
a
renderer
where
you're
taking
a
bunch
of
parameters
and
rendering
it
into
a
Docker
compose
template
a
kubernetes
template.
You
know
exactly
yes,.
C
Yeah,
just
forgive
me:
I
am
not
on
the
open
source
side
of
humanity,
so
Souza
is
she's.
You
know
she
lives
and
brief
open
story.
So
yeah
forgive
me
for,
for
you
know
not
being
more
happy
for
sale
kind
of
guy.
Oh.
C
E
Absolutely
I
could
run
out
of
time.
Yeah
no
I
mean
it's
good,
that
it's
good.
We
have
the
discussion,
but
I
just
want
to
ensure
that
we
also.
D
Okay,
so
first
nice
to
meet
you
everyone.
My
first
time
here
I
know
the
the
group
a
lot
of
time
and
follow
the
group
I
love
that
a
lot
of
time,
but
I
decided
to
start
to
come
to
reports
and
be
be
more
active
here
this
year.
So
thank
you
so
much
for
this
this
opportunity,
so
I'm
sharing
my
screen
and
I.
Don't
know
what
the
white
spouses.
D
Let
me
get
the.
D
I
I
shared
in
the
slack
and
also
stop
again
like
maybe
oh,
let
me
try
I
I
will
do
I
will
try
downloads
for
my
machine
and
use
it.
D
In
order
to,
let
me
try,
so
the
the
idea
is:
try
to
try
to
develop
a
user
group
work
group
or
interest
special
interest
group
to
start
and
share
again
and
I.
Don't
know
it's
supposedly.
D
E
C
D
If
you
can
okay,
the
point
is:
there
is
a
a
lot
of
time.
We
have
the
club,
Cloud
native
definition,
Cloud
native
territory,
round
around
around
many
things
and
one
of
most
important
is
Microsoft
Concepts
architecture.
D
So
we
can
see
that,
in
the
definition
we
have,
microservice
is
a
key
point
for
architecture
for
definition
of
cloud
native,
and
you
you,
if
you
go
ahead,
you
can
see
also
to
the
since
2016
Cloud
Navy.
Mentioning
Microsoft
is
a
key
button
for
cloud
Naval,
architectural
definition.
D
So
we
had
events
when
softer
circles.
Another
one
that
our
chair,
CSF,
chair
member,
said
I
forget
the
name
of
sorry
but
I
had
to
put
here
but
I
will
put
I
will
update
here
with
the
name:
the
guy
talk
about
club,
Cloud
native
territory,
Cloud
Nate
pattern
and
using
microservices
a
key
pattern
for
columnism.
D
So
if
you
can
go
go
ahead,
please
in
the
oh
here,
Alex
Richardson
Alex
said
presented
all
the
times
about
the
the
Microsoft's
oriented
act
that
were
being
key
for
for
Club
native,
and
we
can
say
that
the
three
pillars
of
our
principles
of
Cloud
native
architecture
is
are
sorry.
Microsoft.
D
Sorry,
could
you
go
ahead
number
one
and
more
one,
please
the
key
design
principles
for
Microsoft,
so
we
can
say
based
on
documentation,
based
on
the
definition
based
on
presentation
everywhere
we
can
say
that
we
have
the
two
pillar:
Microsoft
oriented,
orchestrator,
Dynamic,
characteristic
and
containerization.
From
the
perspective
of
cloud
cncf,
you
can
see
in
all
tags
work
groups,
you
can
see
the
infrastructure
or
platform.
D
Pillars,
containerizing
and
orchestration
being
well
leaded
by
all
the
the
security,
also
with
many
groups
working
on
this,
but
specific
about
Microsoft
from
the
developers
perspective.
The
patterns
the
how
how
improved
create
a
architecturally
that
get
all
the
the
possibilities
of
the
infrastructure
that
we
are
offering
cncf
all
projects
and
since
have
all
starts
and
best
practices.
D
We
don't
have
a
specific
group
studying
the
first
leg,
in
my
opinion,
the
first
leg,
the
actuctor
of
the
application
to
get
all
this
work
well
over
the
the
infrastructure,
the
platforms
that
we
are
creating.
So
my
my
proposal
and
you
can
go
go
ahead.
Please
I
put
some
slice
to
create
the
the
the
the
concept.
This
is
all
them
come
from
since
seven
slides.
D
D
75
bar
one
market
go
ahead.
Go
ahead.
Please
I'll
be
running
because
the
slides
itself
is
so.
The
point
is
God
native
betrayal
say
we
start
with
a
container
Rising.
So
the
point
is
no
from
the
perspective
of
application.
D
If
you
have
a
a
monolithic
application,
you
can
deploy
it
over
containerized
infrastructure
or
orchestrator
or
inside
using
observability,
but
at
the
end
you
cannot
get
capture
all
the
advantages
of
you
use
cloud
native
when
you
have
application
a
monolithic
application,
so
the
first
step
for
First
Step
could
be
move
this
application
or
build
a
new
one
or
build
a
new
functionality
of
this,
the
the
Legacy
application
in
Microsoft's
approach
using
a
Microsoft's
approach.
D
So,
in
my
opinion
the
first
item
should
be
build
or
construct
microservice
and
in
seconds.
How
you,
of
course,
Constructor
is
the
the
first
phase,
a
build
test,
make
everything
works
well
and
now
the
sequences
is
containerizing
orchestrating
observability
observability
Etc.
So
the
first
point
is
how
you
we
make
a
good
microservice,
how
we
build
a
good
Microsoft
so
using
the
best
products
Etc.
So
this
the
idea
of
the
mission
that
I
plan,
if
the
next
slide
I
put
this.
D
The
idea
of
a
mission
for
the
work
group
is
collaborative
with
all
areas
or
groups
and
others
organizations
to
build
the
best
practice
of
Book
of
Knowledge.
D
You
know
for
Microsoft
Psychiatry
and
also
a
great
with
this
relationship
with
other
other
groups,
a
better
way
to
build
microserves,
better
way
to
to
or
create
standards
that
make
sense
for
microservice.
For
example,
you
talk
about
in
the
later
presentation.
We
talk
about
configuration
when
it's
look
to
the
application.
You
have
one
pattern,
usually
well
not
by
by
developers
from
12
factors.
There
is
externalized
configuration
in
this
pattern.
We
can
use,
for
example,
in
in
in
Java.
D
You
have
a
platform
that
help
us
to
externalize,
externalize
configuration
and
but
for
example,
I
you
can.
You
cannot
use
this
for
for
for
a
gold
link
application
at
the
same
time,
we
are
in
the
later
application
will
talk
about
how
make
a
standard
way
to
with
the
pla
that
platform
score,
how
you
can
externalize
and
control,
govern
the
infrastructure
or
platform
configuration,
but
from
the
perspective
of
application,
own
application
configuration.
D
You
don't
have
this,
and
it
could
be,
for
example,
an
item
to
to
start
to
stand
on
our
side
and
also
works
with
other
other
foundations.
For
example,.
D
Eclipse
Foundation,
that
has
a
project
that
one
super
is
configuration
so
why?
Why
not
work
together
and
create
a
kind
of
standard
for
configuration
that
could
be
used
by
all
kind
of
language
or
platforms
or
Frameworks?
Using
such
kind
of
pattern
like
like,
we
have
a
similar,
we
have
for
cloud
events.
Cloud
events
is
a
pattern,
as
is
a
standard
that
you
can
use
it
to
communicate
in
events
in
Cloud.
D
Why
not
has
something
for
Microsoft's
from
the
perspective
of
a
good
Cloud
native
architecture,
and
that
you
use
all
all
the
the
platform
that
you
have
also,
we
can
see
something
similar
when
you
have
open
API
and
when
you
have
a
a
synchronous
API.
D
So
it's
it's
a
kind
of
development
that
you
can
put
in
a
good
or
best
practice
for
for
for
microservices
and
can
maybe
we'll
help
us
to
deploy,
build
and
deploy
Microsoft
and
get
the
the
best
execution
runtime
from
the
cloud
or
cloud
and
Cloud
native
with
cook
sales
cabernetes
Etc,
and
we
can
work
with
this
standards
and
best
breaks.
Also
for
the
point
of
view
with
a
relationship
observability.
D
We
have
many
things
from
observability
that
is
on
the
top
of
the
application,
but
not
either
the
application,
and
there
is
a
practice,
architectural
practice
instrumentation
that
we
can
work
with,
create
and
standard
or
improve
or
increase
the
Prometheus,
for
example,
and
to
create
applications
items
for
oversabability,
not
only
the
execution
from
the
perspective
of
environment
but
from
the
business
or
the
large.
We
can
work
on
this
and
improve
the
capability
of
observability,
give
a
deep
dive
in
their
application,
not
only
in
the
infrastructure.
Also
I
think
that
we
have.
D
In
my
opinion,
you
have
a
lot
of
stuff
to
do
in
the
correlated
with
other
other
groups,
and-
and
the
question
is,
is
we
don't
have
this?
Yes,
we
have
many
many
work
being
developed
for
folks
from
outside
cncf.
D
We
have,
for
example,
the
open
Group.
The
open
group
doping
group
is
the
the
group
that
developer
took
off
Enterprise
architectural
framework.
They
tried
they
in
the
past
2016
I
I'm,
not
sure,
but
they
developed
a
separated
best
best
practice
Book
of
Knowledge.
D
Okay,
it's
good,
and
this
group
creates
also
I'll,
try
to
create
something
for
microservice,
but
this
development
stopped
in
2021
and
I
think
that,
in
my
opinion,
cncef
could
be
the
best
house
or
the
best
home
to
this
kind
of
develop,
because
Microsoft
is
something
very
important
for
cloud
native
architecture
and
for
the
definition
of
Club
native,
and
we
have
so
much
people
talk
about
Microsoft's
governance,
supply
chain
and
Microsoft's
in
the
context
boxes
in
the
context
of
supply
chain
Etc,
and
in
my
opinion,
you
can
grab
all
this
work
and
create
something
that
integrated
here
in
CNF
I
gave
I
put
some
examples
that
that
I
I
said
here
in
this
slides-
and
this
is
my
offering
my
idea,
my
offering
for
the
group
to
application
delivery
group-
try
to
think
stretch
analyze
this
this
proposal
and,
if
possible,
we
we
build
this
group
and
work
in
this
Concepts
together.
D
D
So
is
this
my
my
suggestion
for
the
group
and
I
mean
I
want
to
help
if
possible,.
E
Yeah
I
mean
this.
This
was
kind
of
always
in
scope
from
the
very
beginning
when
we
started
app
delivery.
Honestly,
we
never
exactly
knew
what
like
what
very
specifically
to
to
work
on,
because
it's
just
such
a
big
topic.
So,
as
you
pointed
out,
I
think
the
big
challenge
is
that
it's
very
ambitious.
E
What
you're
trying
to
achieve
here,
which
is
good,
don't
get
me
wrong,
but
my
question
would
be
like
what
is
the
first
thing
so
like
when
we
have
been
successful
in
the
past
and
I've
seen
things
working
out
is
we
did
around
operators
like
with
Jen
here
who
worked
on
the
operator
white
paper
with
Josh?
E
He
was
done,
work
on
or
he's
not
doing,
work
also
on
the
on
a
regarding
platform,
engineering
and
I
think
that
success
of
an
activity
like
this
will
very
much
depend
on
getting
a
very
clear
focus
and
addressing
a
very
clear
need
in
the
market
while
it
can
be
actually
more
there
and
you
need
to
get
people
involved,
It's,
usually
the
biggest
challenge
in
a
working
group,
and
this
has
been
also
a
challenge
in
the
past.
E
So
we
know
that
people
usually
tend
to
get
involved
in
tasks
favor
to
reasons
because
they're
honestly
paid
to
do
it,
I
mean
to
be
fair.
Many
people
are
working
to
get
ideas
that
there
are.
Companies
are
working
on
and
elaborating
them
across
the
community,
which
is
a
fair
point,
and
especially,
and
us
you
get
for
solving
a
very
specific
problem
that
they're
dealing
with
right
now
anyway.
E
D
Yeah
but
I
think
in
my
opinion,
the
first
goal
is
producer
of
what
what
paper
that
defines
that
defines
I'll
try
to
Define
what
the
community
you
understand.
The
cloud
native
communities
understand
what
is
the
best,
the
best
products
or
the
Microsoft's
architecture.
What
what
is
this
oh
could
be.
The
definition
from
March
Fallen
Faller
could
be,
but
in
my
my
opinion
is
not
only.
D
This
might
Fallen
build
a
definition
many
years
ago,
Mata
for
work
come
from
there
until
now
and
from
the
perspective
of
the
first
thing
is
create
a
white
paper
that,
as
a
base
Central
Point
to
to
Define
in
the
community
opinion,
what
is
a
good,
a
good
microservice
and
after
that,
if
you
have
the
base
the
the
base
of
this
knowledge,
we
can
try
to
Deep.
D
We
can
say
deep
dive
in
specific
points.
For
example,
what
I
say
observability
how
we
improve
observability
for
Microsoft's,
using
the
context
of
this
white
paper
of
the
content
of
this
architecture
or
how
we
can
Define
the
better
way
to
externalize.
D
D
Okay,
perfect:
how
can
I
define
this
as
a
standard
that
someone
can
change,
for
example,
Transformer,
a
provider
of
database
for
another
provider
of
that
store
and
Define
this
in
a
way,
that's
this
interface
make
this
easily
to
all
implementation
works
on
this.
There
is
a
point
as.
B
You
said
they
come
to
mind
for
me
like
there's,
maybe
a
little
bit
unique
with
microservices
but
a
lot
of
the
patterns.
Wouldn't
they
apply
to
pretty
much
any
kind
of
cloud
native
application
like?
Is
it
worth?
You
know
rehashing
the
entire
landscape
for
for
microservices,
or
you
know.
Maybe
we
can
just
generalize
that,
as
part
of
you
know,
tag
observability
has
you
know
a
lot
of
thoughts
about
how
to
observe.
D
Other
architectures
I
I,
oh
Josh,
I
I,
don't
know
if
I
understand
it
exactly
what
you
asked
me,
but
yeah.
B
I'm
just
asking
like:
should
we
think
about
it?
What,
in
what
way
would
microservices
be
so
different
from
the
other
Arc
I
mean
we're,
definitely
think
microservice
is
one
of
the
top
ones
we
think
about
in
all
of
the
domains.
What
would
make
it
like?
What
would
be
the
difference
between
what
we're
doing
now?
Let's
say
in
in
what
in
this
like?
Can
we
just
generalize
microservices
into
Cloud
native
applications,
and,
and
you
know
we
do
application
development
and
delivery.
So
is
there
something
specific
about
microservices.
D
Oh,
oh
I.
In
my,
in
my
opinion,
we
today
we
we
already
have
a
a
good,
a
a
good
starting
point
for
definition
of
Microsoft
and
our
general
in
the
general.
D
We
have
this
also
Define
it.
My
point
is:
is
that
we
don't
have
work.
D
We
have
this
work,
make
for
difference,
people
from
different
groups-
and
we
don't
have
a
at
a
Target-
is
not
a
good,
but
it's
a
target,
a
Thai
relationship
between
what
we
Define
in
the
design
aspect
with
what
we
are
doing
in
the
runtime
aspect.
It's
it's.
There
is
not
a
group
working
on
this
next
step.
We
have
definition
from
the
design
part.
D
We
have
Frameworks
each
one
work
with
this
aspect
of
the
design,
but
no
one
is
working
on
the
that
three
pillars
together.
D
So,
for
example,
what
folks
from
the
eclipse
Foundation
thought
about
observability
in
microservice,
creating
Frameworks,
creating
specifications
what
they
are
doing
related
with?
What
we
are
we
are
doing.
Cnsf
sensor
are.
B
D
That
node.js
is
is
another
Foundation
yeah
that
works
to
developer
node.js.
But,
for
example,
no
one
say
how
Bear
Run
node.js
in
the
over
over
containerize
it
approximation
yeah,
but
I
I
try
to
simplify.
But
the
point
is
the
I
think
that
we
have
this
point
Fair
generalized,
a
way
to
see
in
the
markets
by
very
good
folks,
from
the
design
point
of
view
from
the
implementation
of
application.
D
We
have
groups
different
groups
who
work
on
that
on
that,
based
on
design
and
some
of
them
working
personally
with
us
with
CSF,
for
example,
I
know
that
there
is
a
a
lot
of
functions
on
on
Jakarta,
looking
for
what
we
are
doing
in
CSF,
but
looking
for,
but
not
together,
not
working
together
or
not
with
a
kind
of
a
specification
best
practice
that
we
can
Leverage
The
the
boof
size,
the
the.
E
Yeah
so
because
we're
running
out
of
time
and
I'd
like
to
get.
E
Are
you
very
enthusiastic?
That's
great,
so
you're
like
really
deep
into
this
topic.
I
think
there
is
something
there.
We
have
never
figured
out
what
it's
going
to
be
in
the
past.
I
see
two
things
that
you
talk
about.
A
lot
of
organizations
The
Next,
Step
I
would
do,
is
get
those
people
involved
and
committed
to
dedicating
time,
because
this
will
eventually
be
your
biggest
challenge.
E
Then
I
would
pick
a
small
enough
problem
like
even
like
defining
what
the
cloud
native
service
looks
like
like
almost
really
happy
with
the
fact
replication
or
whatever
it
is,
and
don't
want
to
tell
you
what
to
do
but
and
I.
Think
creating
a
working
group
like
the
organizational
unit
is
the
least
of
your
problems,
so
the
organizational
unit
can
be
created.
Honestly,
we
have
decided
for
some
of
the
new
activities.
We
don't
even
create
a
dedicated
working
group
anymore.
E
That
might
be
honestly,
and
that
would
be
my
recommendation
as
a
next
step
like
if
you
get
all
those
people
together
and
agree
on
a
small
enough
deliverable
that
you
can
realistically
well
right,
given
the
timing
realistically
have
ready,
maybe
for
a
birth
of
a
feather
for
kubecon
India
to
get
people
together
and
then
get
momentum
there
then
the
working
group
is
the
again
is,
is
a
formal
act
that
you
might
need
or
might
not
need,
depending
on
where
I
want
to
take
this,
but
I
think
that
that
would
be
really
what
what
I
would
focus
on.
E
If
you
want
to
have
a
dedicated
meeting
on
this
topic,
we
can
help
you,
obviously
with
calendar
and
scheduling,
but
I
think
that's
that.
That's
that's
really
what
what
I
would
focus
on,
because
having
the
working
group
won't
get
the
time
and
dedication
from
from
the
people.
This
is
actually
only
needed
at
a
certain
point
where
you
need
the
organizational
support
which,
to
some
extent,
can
be
provided
as
well
like
you
can
have
a
regular
meeting
on
working
on
microservice
related
topics
already,
but
I
don't
know.
E
D
Ed
thanks
a
lot
I'm
pronouncing
correct,
Alloys,
Alloys,
yeah,
perfect,
yeah,
okay,
thanks
so
much
yeah,
I
think
so.
Yes,
I
I
put
the
word
group
name,
but
just
because
I
I
understand
that
is
a
group
first
stud,
but
not
need
to
be
exactly
formal
worker,
I,
I
I.
Think
in
my
opinion,
we
can
work
in
a
small
piece
of
of
a
small
piece
of
gold.
Sorry
swap
yourself
deliverables
very
conscious,
very
simple
in
the
first
time
and
I
will
try
to
get
all
others
interesting.
D
E
And
feel
free
to
share
and
gain
interest.
I
think
the
reason
because
he
had
like
really
12
Factor
up
on
here,
I
think
what
was
great
about
the
12th
Factor
app
back
in
the
days
when
it
was
done,
because
it
was
giving
simple
architectural
guidance
to
not
get
things
terribly
wrong,
getting
just
a
collection
about
Lessons
Learned
and
what
most
people
got
wrong
without
trying
to
necessarily
Implement
a
standard,
because
standards
for
architectures
are
usually
very
very
hard,
if
not
impossible,
to
implement
you
might
get
to
design
patterns.
E
I
think
that's
also
why
the
book
is
is
called
patterns,
but
things
you
really
should
not
get
wrong,
and
you
have
some
examples
like
what's
the
basic
observability
requirements,
you
should
have
like
how
you
should
expose
how
a
service
is
doing.
There's
like
lots
of
things
like
even
service
versioning.
What?
How
do
you
need
to
build
a
microservice
so
that
it
can
be
deployed
and
upgraded
independently,
like
talking
to
a
lot
of
people
and
gaining?
E
This
is
what
we
did
wrong
first
and
this
it
helps
and
I
think
this
will
get
the
auto
traction,
because
practical
advice
is
usually
what
people
are
looking
for.
That's
I
I
think
generally
we're
just
using
more
people
getting
involved.
If
you
solve
a
problem
or
discuss
a
problem
that
they're
dealing
with
today,
then
you
will
definitely
get
a
lot
more
momentum
quicker
and
that's
again,
I
think
I'm,
repeating
myself.
I
think
that
your
vision
is
clear.
E
Like
your
end
goal,
I
think
it's
really
about
that
first
step
to
make
it
small
enough
to
to
Really
build
up
that
momentum
again.
E
Overall,
I
think
it's
totally
live
with
what
this
working
group
is
doing.
Let
us
know
how
we
can
help
what
we
can
do.
We
can
definitely
use
our
channels
to
share
something
even
getting
this
into
a
smaller
blog
post.
That's
also
how
Josh
started
a
lot
of
the
platform
from
related
work.
It
usually
helps
if
you
can
like
put
it
in
like
half
a
page
to
one
page
where
you
want
to
go
and
ideally
have
already
people
from
those
other
organizations
contributing.
E
It
has
been
a
challenge.
We
never
really.
We
knew
that
we
had
to
do
something
in
that
space
But.
We
were
never
entirely
sure
what
it's
going
to
be.
So
it's
great
that
you're
taking
the
initiative
and
let's
try
to
follow
up.
Maybe
the
next
meeting
or
the
meeting
after
visit,
probably
more
concrete
proposal.
Maybe
it
takes
you
a
month
or
so
at
least
to
get
to
talk
to
people
I.
Think
I
would
do
that
you're
at
your
own
pace,
but
that
would
be
my
purpose
on
how
to
continue.
If
this
sounds
good
for
you.
A
I
just
posted,
like
a
sample
of
like
the
charter,
like
maybe
as
a
guidance
on
like
how
to
start
putting
it
into
like
writing
and
then
getting
people
like
how
some
some
groups
started
like
by
drafting
a
charter.
A
And
then
maybe
you
can
come
help
you
construct
a
narrative
and
yeah
gather
interests
like
we
started
like
this
is
on
the
GitHub,
but
we
start
like
with
we
share
Doc
and
then
and
then
send
it
to
people,
and
then
people
start
like
having
interest
and
commenting
and
stuff
and
then
in
the
future
becomes
like
a
charter
and
then
proposed.
But
again
some
groups
will
start
working
with
it.
Having
meetings
even
without
any
any
official
naming
or
something.
E
E
E
That
also
helps
with
a
lot
of
the
discussions
when
they
tend
to
get
broader
again,
so
you
have
to
be
very
clear:
are
you
proposing
how
to
write
a
Java
microservice
or
not,
or
is
it
something
that's
handled
by
framework?
Are
you
telling
people
how
like
that
stacked
in
architecture?
Is
it
something
that
they
need
to
do
when
they're
from
like.
A
E
Do
you
want
to
do,
and
what
do
you
don't
you
want
to
do
is
that
they
don't
I
think
is
even
more
important
than
than
they
do.
D
Okay
at
first
Jennifer.
Thank
you
so
much
for
sharing
the
simple
chapter,
I
I
started
to
write,
write
down
draft
I
shared
with
this
shares
here,
the
of
course
it
it
is
a
very
simple
draft
that
I
started
and
I
will
I
will
work
on
the
on
the
suggestions.
Thank
you.
D
Thank
you
so
much
and
what
at
in
the
end,
what
what
I
want
is,
try
to
develop
the
the
concept
and
develop
the
the
idea
behind
of
the
has
the
portion
of
development,
the
portion
of
the
the
engineering,
not
the
portion
of
the
execution
and
they
put
out
together
in
the
sequence,
so
thank
you
so
much
I
will
work
on
the
suggestions
and
I
think
for
your
time.
E
Bye,
yeah
I
think
we've
done
for
today
we're
running
a
couple
minutes
over
great
conversations
thanks
for
sharing
and
yeah.
Let's
let
us
know
when
it
is
a
good
time
to
be
have
a
more
follow-up
discussion
here.