►
From YouTube: CNCF TAG App Delivery
Description
TAG App Delivery Meeting Aug 18th, 2021
Keptn Incubation Proposal
nocalhost Sandbox Proposal
Action Item updates
A
B
Oh
right
yeah
so
johannes
is:
is
it
yeah
he's
going
to
give
us
presentation
for
the
captain
proposal
for
incubation
so
I'll
give
the
voice
to
him.
C
C
And
hopefully,
you
can
see
full
screen
here
I'll
just
make
this
a
little
bit
smaller
the
zoom
controls.
So
if
there
are
any
questions
during
the
presentation,
I
might
not
see
the
chat
but
feel
free
to
just
interrupt,
or
we
can
do
the
q
a
at
the
end.
C
So
thanks
for
for
giving
us
here
the
chance
to
present
captain,
my
name
is
jurgen.
C
I'm
one
of
the
captain
maintainers
and
I
want
to
present
captain
what
it
is
briefly
what
it
is
and
then
mostly
talking
about
how
captain
is
adopted
since
its
sandbox
status,
and
also
how
the
community
is
collaborating,
how
to
captain
contributing
to
captain
what
are
the
community
efforts
around
captain
how
the
project
has
been
like
the
major
achievements
of
captain
since
its
sandbox
status
and
also
a
short
look
on
the
roadmap?
C
What's
coming
next
for
captain,
so
I
think
we
agreed
I
can
take
about
like
15
20
minutes
for
the
presentation,
and
I
try
to
keep
it
short,
especially
the
the
first
part.
C
If
you
already
know
what
captain
is,
but
just
a
very
brief
introduction
why
we
felt
the
urge
to
build
captain
like
what
what
problems
does
it
actually
solve,
and
I
want
to
start
here
with
with
actually
an
adoption
story
already,
because
the
company
ert,
what
the
problem
they
needed
to
solve
is
that
they
have
a
lot
of
automation,
code
in
their
pipelines
and
spanning
this
across
a
lot
of
different
projects.
C
A
lot
of
different
pipelines
actually
did
not
really
scale
well
for
them,
so
actually
they
really
have
the
problem
that
they
ended
up
with
thousands
of
lines
of
automation,
code
and
the
problem
with
all
this
automation
code
is
that
kind
of
each
organization
has
to
solve
this
problem.
They
have
to
build
automation,
code
and
also,
all
these
tool
integrations.
They
have
kind
of
brought
their
automation
pipelines
to
kind
of
to
explode.
C
So
in
captain,
what
we
did
is
we
thought
of
this
kind
of
a
pipeline
is
a
couple
of
different
building
blocks
and
we
remove
the
heart
dependencies
of
what
has
to
be
done
with
who
is
actually
doing
this.
So,
instead
of
having
everything
in
one,
let's
say
pipeline,
we
said:
okay,
let's
build
some
automation
sequences
and
with
this
we
can
separate
them
the
process
from
the
actual
tooling
and
from
the
technical
perspective,
to
allow
this.
C
C
And
so
with
this,
we
established
this,
the
separation
of
the
process,
definition
and
then
the
actual
tool
definition
and
having
a
look
at
a
concrete
example
in
captain,
we
can
see
on
the
this
part
here
in
the
in
the
ammo
that
we
all
love,
we
can
see
a
shipyard
which
is
the
process
definition
of
captain
in
the
uniform
definition
we
can
see.
I
just
put
a
couple
of
examples
here,
like
which.
A
C
C
C
So
with
this
we
we
enabled
actually
this
the
company
irt
to
remove
the
thousands
of
lines
of
pipeline
code
to
just
a
very
small
set
of
automation
codes.
The
other
big
part
of
captain
are
built
in
quality
gate
evaluations.
So
one
part
here
that
I
just
skipped
is
the
evaluation
part.
C
So
integrated
part
of
captain
is
whenever
you
want
to
move
a
new
artifact
from
one
environment,
to
the
other,
like
from
f
to
staging
or
eventually
to
production,
you
can
make
use
of
a
captain
quality
gates,
which
is
in
a
nutshell,
which
is
a
a
yaml
definition
in
terms
of
the
quality
definition
we
are
using
here,
slos
or
service
level
objectives,
and
again
here
we
have
a
strict
separation
of
the
tooling
and
what
the
quality
gate
should
actually
evaluate.
So
the
tooling
in
this
case
a
couple
of
promethouse
queries.
C
They
go
into
a
different
concept
or
a
different
file
than
the
actual
quality
evaluation
and
then
once
triggered
either
in
the
in
the
automation
sequence
or
by
the
api
by
the
cli.
Then
captain
will
reach
out
to
different
data
providers
that
are
connected
to
captain
query
the
data
that
is
defined
in
the
service
level
objective
file.
Do
the
evaluation
can
come
up
with
a
total
score,
and
then
this
can
be
used
for
automated
promotion
to
the
next
stage
or
just
for
investigation
to
decide.
C
Should
this
new
build
actually
make
it
to
production,
or
not
so
with
this
concept,
and
with
this
idea
of
captain,
what
we
achieved
is
for
some
of
our
adopters
to
drastically
reduce
their
automation
code
and
also
kind
of
standardize,
their
automation
code,
because
we
now
have
service
level
objectives.
C
We
have
certain
indicators,
they
have
a
well-defined
format,
how
they
can
define
their
quality
evaluations
and
with
we're
bringing
a
lot
of
different
tools
to
different
tools
in
their
environments.
Together,
we
are
orchestrating
those
tools,
so
we
are
orchestrating
when
to
start
the
tests.
When
the
tests
are
finished,
we
are
orchestrating
the
quality
gate
evaluation
for
exact
the
time
for
the
exactly
for
the
time
frame
for
the
tests
when
they
were
run,
we
can
orchestrate
multiple
different
tests
to
run
at
the
same
time
or
one
after
the
other.
C
So
all
this
orchestration
can
be
done
with
captain
and
it's
not
meant
to
replace
all
the
tooling.
It's
actually
meant
to
orchestrate
the
tooling
that
is
already
available
and
already
present
in
in
the
organizations,
and
we
do
have
already
quite
some
adopters
in
captain
which
we're
really
proud
of
and
for
today
I
just
picked
three
stories
that
I
want
to
talk
about
a
little
bit
more
detail,
not
too
detailed.
I'm
happy
to
answer
more
questions,
but
I
I
want
to
start
with
with
one
company.
C
The
very
interesting
part
here
is
it's:
actually
it's
a
it's
located
here
in
austria,
but
it's
operating
in
in
europe-
and
the
reason
is
also
this
slide
is
part
of
the
slide
is
in
german
because
we
took
this,
we
we
got
permission
from
them
to
use
it,
and
we
directly
took
this.
The
important
part
is
that
at
iphone
software,
they
are
creating
the
software
for
a
mobile
banking
application
and
with
captain
they
are
evaluating
each
release
of
the
software
and
basically
the
captain
qualification
decides.
C
If
this
new
build
will
be
put
into
production.
That
means
it
will
be
adopted
by
nine
local
banks
or
subsidiaries
of
the
bank
and
distributed
to
all
of
the
customers.
If
a
captain
quality
gate
fails.
If
the
captain
quality
is
not
green,
then
the
whole
release
is
blocked
and
they
will
investigate
what
is
the
reason
which
metric
is
not
or
which
slo
is
not
satisfied,
and
they
will
captain
is
the
basically
the
decision
maker
if
the
release
will
make
it
into
production
or
not.
C
Another
adoption
story
I
want
to
share
is
from
vitality
group.
They
are
using
captain
internally
for
their
multi-tenant
environment
and
what
they
need
to
ensure.
Is
that
how
the
different
configurations
they
have
between
the
shared
services
so
some
of
the
services
they
run
their
share
between
multiple
tenants
and
they
needed
to
figure
out
how
this
configuration
is
actually
impacting
their
their
services
and
to
figure
out.
C
We
see
this
here
on
the
right
hand,
side
which
slos
are
actually
failing.
It's
not
always
the
same
slo
that
is
failing,
but
for
them
it
was
very
important
to
identify
which
slos
are
failing
for
different
kinds
of
configuration
for
the
different
for
the
different
services.
So
the
total
score
indicates
if
there
is
any
any
issue,
but
then
in
the
kind
of
detail
screen
here,
they
can
identify
what
was
the
problematic
one
and
they
can.
C
They
can
go
back
and
directly
fix
fix
this
issue
and
for
the
third
story
I
want
to
share
here
with
you,
is
it's
a
little
bit
of
a
different
story?
It's
from
ketoki
a
company
that
offers
kitchen
as
a
service
and
they
needed
to
evaluate
the
resiliency
of
their
micro
service
applications,
and
for
that
they
are
using
a
combination
of
load
tests
and
chaos.
Tests
with
captain
quality
gates
on
top
and
also
they
are.
C
They
are
aiming
for
orchestrating
the
whole
process
with
captain,
but
for
evaluating
the
impact
of
the
chaos
on
the
applications.
They
define
three
different
stages:
no
chaos
stage
that
kind
of
gives
the
baseline
a
light,
chaos
stage
that
needs
to
be
passed.
It
needs
to
be
successfully
evaluated
by
captain
quality
gate
and
a
heavy
chaos
stage
where
they
introduce
a
lot
of
chaos,
they're
ripping
down
their
environment
and
they
expect
the
slos
to
fail
here.
C
That's
kind
of
the
counter
example
or
counter
measurement
that
they
also
validate
their
slos
are
not
too
relaxed,
so
they
the
light
chaos.
Their
application
has
to
survive
this
and
for
heavy
chaos.
They
expect
it
to
fail
and
they
built
all
three
evaluations
with
captain
they're
running
this
already
in
the
nightly
builds
and
last
time
they
they
gave
the
presentation
in
one
of
our
user
groups.
I
think
they
had
around
already
hundreds,
I
think,
around
700
evolutions,
with
captain
on
their
on
their
chaos,
experiments
and
their
low
tests.
C
So
yeah
we
just
talked
about
a
couple
of
our
adopters,
but
it's
not
only
the
adopters
that
we're
proud
of
and
that
we
that
we
saw
really
increasing
since
our
sandbox
status,
but
it's
also
the
the
integration
and
color
rate
collaboration
with
the
whole
cncf
and
the
projects
in
the
scenes
here.
So,
first
of
all,
we
are
heavily
used
in
making
use
of
cloud
events.
That's
the
whole
event
thing.
The
whole
eventing
definition
of
captain
is
based
on
cloud
events.
C
We
have
built
integrations
with
the
litmus
chaos
actually
the
one
that
is
already
used.
Also
at
utopi
we've
done.
This
integration
was
also
presented
at
tubecon.
I
think
around
the
march
time
frame
together
with
cardtape
from
mixmos
chaos
container
of
litmus
chaos.
C
We've
built
great
integrations
with
argo
and
our
rollouts
to
validate
each
stage
of
the
canary
release.
If
the
percentage
should
increase
for
the
canary
release.
Also
using
here
captain
quality
gates,
we
are
integrating
with
artifact
hub.
We
already
have
first
integration
on
the
staging
environment
of
artifact
hub.
C
We
have
already
kept
an
essence
hole
on
the
artifact
hub,
but
we're
working
here
also
to
bring
all
of
our
integrations,
and
I
will
talk
about
the
integration
just
in
a
second
but
to
bring
all
of
our
integrations
from
the
captain
ecosystem
to
bring
this
also
to
the
artifact
hub
to
make
it
more
visible
and
and
also
more
available
to
the
whole
cncf
community.
C
We
are
integrating
and
working
here
also
with
tekton.
We
just
presented
this
yesterday
in
one
of
our
captain
community
meetings,
how
tactum
can
be
used
for
building
or
even
deployment
and
captain
can
be
used
for
the
whole
orchestration
for
a
full
delivery
sequence,
we're
integrating
actively
currently
with
a
cross
plane
for
a
use
case
to
define
temporary
environments
for
tests,
we're
making
use
of
helen
for
deployment.
C
So
it's
a
lot
of
different
cncf
projects
that
are,
in
the
let's
say,
captain
ecosystem
that
we're
either
collaborating
with
or
having
integrations
with
or
directly
used
in
the
captain
project
and
a
lot
of
those
integrations.
They
are
actively
contributed
by
the
community.
That's
a
kind
of
specific
part
of
captain
that
makes
it
so
interesting.
C
Is
that
captain,
as
a
control
plane
allows
others
to
integrate
and
there
is
actually
there
is
no
change
needed
to
the
captain
core,
but
a
captain
integration
or
a
captain
service,
as
we
call
it
can
just
be
added
by
the
community
to
interact
with
captain
and
there
have
been
already
a
lot
of
different
contributions,
starting
from
artillery,
where
the,
where
the
maintainers
contributed
this
locust,
we
worked
actively
with
the
prometheus
community,
also
with
one
of
their
co-founders
to
improve
our
prometheus
integration,
one
chart
and
the
invert
project
crossplane.
C
C
The
progress
that
captain
made
since
it's
a
sandbox
status,
I
want
to
focus
you
on
two
parts
so
once
the
feature
set
of
the
captain
core
project
and
then
on
the
on
the
community
efforts,
so
there
have
been
a
couple
of
actually
have
been
a
lot
of
new
features.
I
just
I'm
just
mentioning
here
a
couple
of
them
and
in
the
interest
of
time
I
think
I'm
already
10
minutes
into
the
presentation.
I
just
maybe
want
to
highlight
two
or
three
of
them.
C
First
of
all,
maybe
the
multi-cluster
support,
which
was
actively
demanded
by
by
the
community,
allows
to
install
the
captain
control
plane
and
the
captain
execution
plane
on
completely
different
clusters.
Different
regions
also
different
cloud
providers.
C
C
Okay
feel
free
to
interrupt.
I
think
there
was
something
going
on
if
not
I'll
just
continue
that
allows
us
to
install
the
execution
plane
and
controlling
with
different
clusters.
C
This
also
allows
to
drastically
reduce
the
resource
consumption
of
captain,
since
we
made
an
explicit
split
between
the
execution,
plane,
control,
plane
and
now
the
default
installation
installation
of
captain
has
been
reduced
by,
I
think
around
50
of
resource
consumption,
and
probably
this
one
is
also
really
interesting.
The
closed
loop
renegation
and
instead
of
having
a
let's,
say,
fire
and
forget
mechanism
for
remediation
actions
in
automation,
sequences
captain
now
treats
this
as
a
closed
loop
remediation.
C
That
means
that
after
one
variation
action
is
executed,
the
character
will
make
sure
to
also
do
an
evaluation.
If
this
remediation
action
actually
did
the
trick
and
resolve
the
specific
issue,
if
not
another
remediation
action
can
be
executed,
and
if
it's,
if
it's
successfully
remediated
the
loop
will
be
closed.
Otherwise
it
can
be
escalated
by
again
utilizing
integrations,
such
as
slag
or
china
to
to
inform
or
modify
the
actual
stakeholders
talking
about
the
community
and
and
the
the
progress
we've
made
here.
C
It's
it's
really
great
to
see
that
most
statistics
have
increased
by
a
factor
of
at
least
2x,
which
is,
I
think,
pretty
impressive,
and
we
talked
about
like
here.
I
think
well,
that's
like
for
around
now,
one
year,
time
frame,
since
we
made
it
to
the
to
the
cncf
sandbox,
so
the
github
stars.
We
could
greatly
improve
this
and
increase
this
feel
free
also
to
star
us.
C
If
you
think
captain
is
an
interesting
project,
twitter
followers,
our
youtube
subscribers,
since
we
put
all
of
our
videos
that
we
do
also
with
other
projects
such
as
the
one
we
did
yesterday
with
with
with
our
friends
from
the
the
cf
and
the
technical
community,
or
integrations
with
litmus
chaos.
We
put
all
of
our
resources
on
a
youtube
channel
and
it's
getting
more
and
more
interest,
also
contributors
to
the
project
increased
by
more
than
doubling,
which
is
which
is
really
great.
C
So
we
see
more
and
more
people
also
actively
contributing
to
the
project,
not
only
in
terms
of
contributions,
but
also
in
terms
of
contributions
to
the
captain
core
and
not
only
in
terms
of
integrations.
So
they
contribute
us
here.
Our
only
body
captain
core
projects
there
are
way
more
that
are
also
contributed
to
the
captain
ecosystem
to
all
those
different
contributions
and
integrations.
But,
as
I
said,
it's
not
necessarily
needed
to
if
for
a
for
a
maintainer
of
integration
to
also
submit
or
to
also
contribute
to
the
captain
board.
C
So
it's
really
great
to
see
also
a
lot
of
our
new
contributors
on
the
capital
core
project,
and
we
also
have
a
lot
of
new
users
in
slack
here.
I'm
mostly
mentioning
the
weekly
active
users
and
there's
a
lot
of
interesting
discussions
going
on
on
on
the
slide,
not
only
in
terms
of
kind
of
a
help
channel,
but
also
in
terms
of
how
to
write
integration
to
captain
how
to
use
captain
and
also
exchange
between
the
captain
community,
around
best
practices
and
and
other
innovations.
C
Yes,
just
seeing
the
chat,
yes,
everything's
public,
also
there.
The
link
to
the
document
is
in
the
in
the
meeting
document,
so
feel
free
to
also
use
this
presentation
or
just
leave
a
comment
in
this
presentation.
Being
part
of
the
cncf
also
enabled
us
to
enable
others.
We
already
have
three
mentees
that
graduated
from
mentorship
projects
that
have
been
governed
and
organized
by
the
cncf,
so
congrats
again
to
our
three
mentees
ankit
raj
in
himach.
They
did
an
amazing
job,
helped
us
to
improve
the
project
in
in
various
ways.
C
It's
always
great
to
see
new
contributors
having
a
fresh
look
at
the
project
and
also
helping
us
improving
the
project
and
we
helped
them,
and
it's
not
only
me
that
mentored
the
mentees,
it
was
the
whole
captain
team,
so
also
thanks
to
the
whole
captain
team
here,
helping
to
enable
others
and
that
really
helped
also
being
part
of
this
of
the
whole
cncf
movement
and
ecosystem
yeah.
C
And
then
there
has
been
probably
you've
heard
about
captain
before,
since
we
had
a
lot
of
different
engagements,
so
we've
been
invited
to
meet
ups
to
talk
about
captain
I
know
alex
is
something
he
has
done.
The
kubernetes
podcast
on
the
kubernetes
podcast
from
google
anna
from
from
gremlin,
talked
about
captain.
C
So
there
have
been
a
lot
of
speaking
activities,
community
activities
around
captain
which
we
are
really
proud
of,
and
if
you
want
to
learn
more,
if
you
want
to
find
out
also
around
resources,
if
you
want
to
probably
talk
about
captain,
I
put
also
here
a
link
to
the
resources
where
you
can
find
all
the
different
all
the
different
topics
before
I'm
closing
here,
I
just
want
to
highlight.
What's
coming
also,
this
is
public.
You
will
find
the
link
to
the
roadmap
here.
C
C
We
are
working
together
with
the
cdf,
the
sister
foundation
of
cncf,
to
define,
let's
say,
industry
standard
for
eventing,
since
captain
has
already
defined
a
set
of
different
of
different
cloud
events
that
define
yeah
all
the
the
different
tasks
and
sequences
in
the
in
the
life
cycle
of
an
application
such
as
delivery
or
operation,
and
we're
working
here
also
with
the
cdf
to
define
this
and
expand
expanded.
C
Also
in
the
ci
part,
so
this
is
actively
going
on
and
we
are
really
looking
forward
to
define
the
symbol
so
having
support
for
this
in
captain
yeah,
there
are
more
parts
here,
I'm
happy
to
talk
about
all
of
them
in
more
detail,
but
in
the
interest
of
time,
I'll,
probably
just
for
now
I'll.
Just
redirect
you
to
this
link
here
we
are
happy
to
hear
your
feedback
on
our
incubation
proposal.
I
provided
here
the
link.
It's
the
full
request,
670.,
it's
also
linked
the
view,
diligence
document.
C
If
you
have
any
questions
any
comments
on
this,
it's
currently
also
a
living
document
feel
free
to
to
add
your
comments.
There
also
directly
attack
me
there,
or
you
will
also
find
me,
the
cncf
slack
or
in
on
the
captain
slack.
I've
provided
all
the
links
here.
C
Hopefully
I
have
not
forgotten
any
link
of
the
captain
project
so
feel
free
to
reach
out
or
have
a
look
at
all
the
resources
we
have,
and
actually
I
want
to
close
here
with
thanking
the
whole
community
and
actually
the
last
words
should
not
be
my
words
but
words
from
the
community.
What
others
are
saying
about
the
project?
I'm
already
opening
up
for
questions?
While
I
leave
this
here,
I
said
I
will
also
share
slides
there
already
in
the
meeting
document.
C
I
will
put
them
in
the
chat
right
after
I
close
the
window
here,
but
yeah
thanks
for
for
giving
us
the
possibility
to
consent
captain
in
this
delivery
tag
and
I'm
happy
to.
If
we
have
time
for
questions,
I'm
happy
to
answer
questions.
B
Thank
you,
jurgen
for
the
presentation
and
all
the
resources.
I
have
a
question
regarding
the
current
state
of
like
the
api
stability
and
also
your
experience
with
like
end
users.
If
there
is
any
report
of
like
upgrades
that
end
users
have
to
have
in
production
or
something.
C
So
for
the,
if
I
understand
for
the
upgrades
we
always
have
when
we
release
new
captain
versions,
we
all
with
each
captain
version,
we
bring
an
upgrade
guide.
So
if
I
remember
correctly,
since
the
last
at
least
five
versions
even
more,
there
was
never
a
breaking
breaking
changes,
but
always
a
single
line
upgrade
it
kind
of
takes
two
minutes
for
a
captain
upgrade
to
bring
one
user
to
the
latest
version
for
the
api.
C
We
have
already
deprecated
api
endpoints.
They
are
still,
you
can
still
use
them
same
goes
for
the
captain
cli.
These
are
also
announced
in
the
release
notes
of
new
captain
versions.
When
something
gets
deprecated,
we
still
keep
it
for
for
some
releases
in
the
source
code
so
that
we
are
not
breaking
something,
but
I
think
same
with
the
kubernetes.
Sometimes
then
there
is
the
the
time
when
deprecated
versions
will
not
be
supported
anymore.
B
B
I
could
test
deployments,
then
the
specific
examples
you
they
were
all
oriented
around,
like
quality
checks
and
quality
gates-
and
you
even
kind
of
mentioned
like
integrating
with
tecton
tecton,
does
the
build
and
deploy
and
captain
does
something
else.
So
I'm
just
kind
of
wondering
are
you
is
your
main
value,
proposition
quality
gates
and
quality
checks,
or
is
this
a
more
general
build
and
deploy
tool.
C
So
I
mentioned
in
the
beginning:
captain
is
not
meant
to
replace
the
tool,
so
if
you,
for
example,
if
you
already
have
tecton
in
place-
and
you
have
all
your-
you
might
have
a
quite
a
long
tactile
pipeline-
that
is
doing
all
the
different
things
then
for
captain
the
job
is
to
orchestrate.
So
you
can
then
split
it
up
into
different
smaller
pipelines
and
reuse
parts
of
the
pipeline.
So
captain
will
orchestrate,
for
example,
text,
but
not
only
tecton,
but,
for
example,
the
the
demo
that
we
did
yesterday
was
a
building
with
tekton.
C
Once
the
build
is
finished,
captain
knows
what
goes
next,
so
there
might
be
now
a
deployment.
Then
the
captain
knows
what
knows
what
goes
next.
It
will
be
the
test
and
deployment
can
be
done,
maybe
again
with
tacton
or
maybe
with
no
champions,
and
the
test
will
be
done
with
a
chameter
or
locust,
and
then
the
evaluation
will
be
done
with
captain
itself.
So
a
lot
of
the
adoption
stories
I
focused
on
the
quality
gates.
Since
this
is
an
integrated
part
of
captain,
we
can
really
see
the
captain
ui
there.
C
We
can
see
the
the
value
that
really
that
the
captain
brings
to
the
quality
part.
The
other
part,
the
orchestration
part,
was
part,
for
example,
at
kitopi.
They
are
working
on
doing
these
integrations
with
captain,
but
we
couldn't
really
see.
The
interesting
part
was
that
they
are
what
they're
using
what
they
want
to
bring
together
and
how
they
evaluate
this.
But
captain
itself
will
not
do
a
build
of
an
image.
It
will
not
do
the
deployment
itself.
C
B
A
And,
and
maybe
to
finish
up
here
now
speaking,
just
part
of
the
captain
team
also
captain
actively
decides
which
release
or
which
version
of
a
service
to
put
into
which
stage
automatically
and
and
when
to
roll
back.
So
it
would,
for
example,
tell
a
github
store
that
would
actively
rewrite
gita
github's
repositories
to
migrate
back
to
all
the
versions
or
move
forward
or
create
release
plans
there,
which
is
usually
something
you
don't
have
in
your
github
tools.
That's
when
github's
tools
and
deployment
tools
usually
kick
in.
D
So
I
have
one
yeah
sorry,
I
have
one
more
comment
like
regarding
the
captain.
Like
I
look
through
the
slides
it,
so
I
think,
like
captain
is
awesome
like
it
provides
like
such
automation
framework,
but
one
thing
since
it's
a
framework,
one
thing
it's
missing
is
that
it
doesn't
like
the
science,
doesn't
talk
about
how
users
can
create
a
new
plugin
for
captain
right,
for
example.
If
I
want
to
create
a
new
task
or
something
like
like,
how
can
I
translate
those
properties?
D
C
Yeah
in
the
slides
I
focused
on
one
specific,
I
only
showed
one
shipyard
and
parts
of
the
shipyard
like
we
saw
the
deployment
test
with
a
couple
of
let's
say:
opinionated
tasks
there,
but
this
can
be
user
defined.
This
was
one
also
of
our
kind
of
major
new.
C
That
we
shipped
since
since
sandbox
that
is
now
possible
to
also
having
user-defined
tasks
and
just
connecting
your
tools
then
to
these
defined
tasks.
So
that's
absolutely
possible.
I
just
mentioned
I
didn't
mention
it.
You're
right.
D
If
you
can
also
jump
in
for
a
second
hi,
I'm
johannes
pm
of
the
captain
project,
I
think,
then
you
also
asked
for
a
plug-in
mechanism
that
allows
me
to
write
my
service
that
acts
upon
a
task
and
therefore
we
also
provide
a
template
that
a
user
can
use
to
customize
and
to
write
it
his
or
her
own
captain
service,
which
then
can
be
plugged
in
into
the
orchestration
of
captain.
D
Sure,
like
like
one
thing
I
want
to
see,
is
that
like
because
I
like,
when
we
my
foot,
it
looks
like
that
the
captain
is
very
similar
to
like
argo,
workflow
and
pekton.
Is
that
that,
like
uses
cloud
events
to
extend
the
the
automation
thing
more
pipelines
on
tasks
into
that,
I
want
to
see
like
how
could
this
be
simpler
than
you
know.
The
container
based
solutions
on
tecton
or
angle,
workflow,
and-
and
this
is
not
clear
right
now-
yeah.
C
Yeah,
I
think
techton
is
a
very
like
general
purpose
pipeline
tools,
so
you
can
do
a
lot
of
things
with
tech
tongue.
They
give
you
all
the
flexibility,
it's
not
very
much
opinionated
in
any
way,
and
with
captain
we
try
to
help
you
establish.
C
With
an
opinionated
way,
helping
you
to
define
automation,
sequences
and
with
this
we
can
then,
with
the
standard
uniform
that
the
captain
has,
we
can
already
kind
of
connect
some
tools
to
the
different
tasks
we
have
like
for
deployment
or
for
testing.
We
have
kind
of
a
default
uniform
that
you
can
already
use
that
will
help
you
there
other.
Like
argo,
workflows,
tecton,
you
would
have
to
build
this,
then
by
yourself.
C
You
can
do
this,
of
course,
but
it's
just
a
more
opinionated
way
in
the
in
a
way
that
really
focusing
on
on
this
automation,
part
and
reducing
the
complexity
of
pipelines.
D
Yeah
I
got
it
like
just
like
you
haven't
seen
like,
because,
like
it's
inside,
to
talk
about
this,
many
interviews
been
done
and
by
intuition
I
I
I
just.
I
thought
that
this
is
not
like
all
done
by
the
chapter
team,
probably
by
the
community
contributors
right
and
and
as
a
contributor.
How
can
I
contribute
a
new
integration
into
captain
and
this
thing
had
better
be
clarified
in
into
the
in
the
dark
as
well.
D
To
how
to
no
no
cre
creating
a
new
integration
like,
for
example,
if
I
like
captain,
doesn't
integrate
with
tectonic
right
now-
and
I
am
from
that-
let's
say
I
built
tecton2
right
and
I
want
to
integrate
into
captain.
How
can
I
do
that
right.
C
C
I
don't
want
to
go
into
detail
here,
but,
depending
on
what
you
want
to
achieve,
there
are
different
ways
for
captain
integrations
they're,
like
easy
ones,
and
and
like
full
implementations
that
you
might
want
to
where
you
really
write
your
own
go
code
and
then
attach
this
and
then
listen
for
for
for
captain
cloud
events
or
others
where
the
whole
listening
for
captain
carl
evans
is.
Everything
is
taken
care
of
by
a
chopped
executor
from
captain,
and
you
basically
only
do
your
your
logic.
So
there
are
different
ways.
C
B
A
E
Okay,
I'm
hi
everyone.
We
are
one
from
samsung
cloud
and
I
will
be
bringing
you
the
local
sandbox
presentation.
E
E
E
E
E
E
No
cohorts
about
a
new
solution
idea
developed
directly
in
container
this
fundamentally
solved
the
problem
of
inconsistence
development
and
runtime
environment.
With
this
approach,
it
is
possible
to
use
existing
development
environment
and
locals,
provide
additional
ide
plug-in
that
require
little
kubernetes
foundation
to
work
smoothly.
E
E
E
E
E
Third,
how
to
stop
the
container
crash
after
replacing
the
main
process
no
holes
will
implement
their
mentions
to
meet.
The
prerequisites.
First
is
synchronize
the
local
source
code
to
container
the
second
is
replace
the
container
imagine
with
development
magic,
and
the
last
is
replace
the
main
process
with
blocking
process
after
entering
development.
E
E
F
Building
cognitive
applications
can
often
be
difficult
setting
up
the
right
environments,
using
suitable
libraries
and
figuring
out
an
easy
way
to
deploy
and
debug.
Your
application
is
time
consuming
to
help
you
solve
these
challenges.
Review
localhost
nokia
host
is
a
cloud
native
development
tool
based
on
ides.
F
You
optimize,
connect,
deploy,
develop
and
debug
process
that
makes
cloud
native
development
easier
and
faster.
No
glow
supports
number
of
ides,
including
rescue
and
jab
rings.
This
way
you
can
leverage
the
tools
you
already
know,
let's
dive
deeper
into
how
locals
optimize,
connect,
deploy,
develop
and
debug
first
connect,
no
curse
helps
you
connect
to
kubernetes
cluster
easily
by
using
qconfig,
and
it
also
supports
multi-clusters
connections.
Localhost
works
with
any
kubernetes
cluster,
whether
it
is
local,
docker,
desktop
or
mini-q,
or
any
cloud-based
clusters.
No
codes
can
connect
by
one
click.
F
It
helps
you
to
get
rid
of
the
pesky
local
environment
configurations.
Let's
see
how
to
deploy
application.
Nucleus
provides
an
easy
way
to
deploy
applications
to
any
namespace
nuclear
supports
to
deploy
applications
through
manufacture,
helm
and
customize
can
quickly
deploy
the
application
to
kubernetes
cluster.
F
Let's
see
how
that
mode
works
once
under
depth
mode
locals
will
replace
the
remote
container
image
with
that
image
use
the
dev
image
to
start
a
new
pop
and
replace
the
existing
port
turn
on
power
forward
in
remote
port.
So
you
can
now
easily
access
remote
application
in
local
start.
Real-Time
file,
synchronization
sync,
the
local
file
to
remove
automatically
automatically
start
terminal
in
a
remote
port.
This
way
you
can
compile
run
remotely
through
the
ide
when
developing
localhost
can
automatically
synchronize
the
code
to
the
container.
F
F
Can
significantly
speed
up
the
female
loops
of
development,
so
you
see,
updates
in
understanding.
You
might
discover
issues
in
your
application.
Knucklehoods
is
here
to
help
by
providing
the
same
debugging
experience
with
using
ide,
even
when
debugging
in
remote
connected
cluster.
If
you
need
to
dive
deeper
nucleus
also
comes
with
a
building
lock
viewer
to
help
you
to
monitor
and
debug
issues
with
your
kubernetes
right
within
your
ide.
Now
you
know
how
neutrals
work
to
improve
your
experience,
creating
cloud
native
applications
within
your
favorite
ide
and
optimizing
the
connect,
deploy,
develop
and
debug
process.
F
E
A
Yeah,
thank
you
for
the
presentation
I
think
you
are
aware
of,
and
we
talked
about
this-
that
obviously
for
a
sandbox
submission.
There
is
a
form
by
the
cncf
in
case
you
have
any
problems
finding
it
please
feel
free
to
reach
out.
Nevertheless,
it's
always
good
to
learn
about
new
projects
and
see
them
shared
here.
So
just
just
that
you're
aware
from
a
process
perspective,
you
need
to
submit
the
form
that
you
see
will
then
review.
D
B
E
No
course
is
just
developed
to
us
it.
It
can
be
a
ci,
a
ci
step.
So
if
you,
if
you
want
to
build
a
docker
image,
you
can
use
some
ci
tools,
such
as
jenkins
and
and
so
on.
E
Thank
you.
I
got
the
demo
here.
E
Let's
I
just
have
two
minutes.
E
A
E
A
Moving
on
to
the
last
two
agenda
items,
thanks
for
the
presentation,
okay,
I
think
we
have
for
today
potato
head
status,
update
on
multi
services.
I
think
thomas
you're
going
to
provide
the
quick
update
there.
Yes,
so
I
actually
put
that
on
whatever
I
put
in
there
yeah,
so
you.
B
B
Yeah
exactly
I
I
just
wanted
to,
let
you
all
know
we're
just
porting
the
stuff
over
getting
into
multi-service,
make
the
docs
a
little
more
consistent,
try
to
add
some
tests
so
that
that's
what's
happening
there.
That's
all
so.
G
So
we
also
changed
a
lot
in
the
whole
ci
pipeline
of
of
captain,
so
we
we
added
some
automatic
tests
after
merge
and
so
on,
so
that
that
we
get
it
a
bit
more
stable
and
replaced
a
lot
of
shell
scripts
against
the
direction.
A
That's
all
for
muscle,
okay,
yeah.
My
question
was:
should
we
reach
out
actively
to
the
projects
that
contributed
to
the
first
version
to
maybe
get
an
update,
because
I
think
it
would
also
be
a
nice
idea
to
have
them
present
on
how
things
work.
I
mean
to
be
fair.
Most
of
them,
I
think,
run
with
the
manifests
or
the
helm
charts.
With
some
exceptions.
B
B
A
Yeah,
I
think
it
is
moving
forward,
so
it
was
obviously
alex
wrote
an
email
to
the
toc
mailing
list.
I
think
a
while
ago
I
just
sent
another
one
requesting
them
to
vote
on
it,
so
I
hope
they
are
well
now
going
to
vote
on
it
rather
soon.
I
think
nothing
really
holds
us
back
from
from
working,
but
I
pushed
it
again
to
the
toc
mailing
list
today
and
I
hope
that
we
get
like
the
official
buy
in
rather
sooner
than
later.
B
G
A
Toc
meeting
alternatives
they
also
reach
out
to
our
toc
liaisons
that
they
maybe
start
the
voting
process
out
there
but
yeah.
I
think
we
are
kind
of
like
victims
to
the
vacation
season
right
now,
where
things
just
got
lost
a
bit,
but
I
assume
this
to
be
moving
forward.
A
Okay,
I
think
thank
you
yeah.
This
is
it
for
today
and
also
to
the
localhost
team
thanks
for
presenting
and
if
you
need
any
help
with
the
forum
for
the
sandbox
submission.
Please
simply
let
us
know
we
can
point
you
in
the
right
direction
there
as
well.