►
From YouTube: Keptn Community Meeting - May 14th, 2020
Description
May 14, 2020
Noteworthy Issues & PRs
Johannes:
Update on keptn.sh/docs: https://github.com/keptn/keptn.github.io/issues/513
Christian K.
Improved platform tests for various K8s versions GKE and Minikube: https://github.com/keptn/keptn/pull/1782 and https://github.com/keptn/keptn/pull/1773
Master Travis-CI Build: https://travis-ci.org/github/keptn/keptn/builds/686974524
WIP: Improve integration test for quality-gates
https://github.com/keptn/keptn/issues/1591
https://github.com/keptn/keptn/pull/1786
Christian H.
Go-Utils https://github.com/keptn/go-utils/pull/157
Florian:
Upgrader 0.6.1 - 0.6.2: https://github.com/keptn/keptn/issues/1660
Link to docs
Manage open approvals in configuration service: https://github.com/keptn/keptn/pull/1776
Ermin:
Bridge Better time visualization Better time visualization in Bridge for events that are faster than 1 minute · Issue #1584 · keptn/keptn
Menu for Environment and Services screen https://github.com/keptn/keptn/issues/1698
Jürgen:
Cluster internal load generator: https://github.com/keptn/examples/pull/94
Working in progress:
https://github.com/keptn/keptn/projects/29
Backlog Refinement / Next Sprint:
see https://github.com/issues?q=is%3Aopen+is%3Aissue+user%3Akeptn+user%3Akeptn-contrib+label%3A%22next-sprint%22
Planning Board
https://github.com/keptn/keptn/projects/29
Attention: new dates for upcoming community meetings
Wednesday, May 20
Wednesday, May 27
Wednesday, June 10
Please subscribe to the community calendar to stay up to date
Learn more: https://keptn.sh
Get started with tutorials: https://tutorials.keptn.sh
B
A
To
discuss-
and
we
will
also
take
a
look
at
the
backlog
and
we'll
do
some
planning
but
before
I
hand
over
to
the
captain
development
team.
What
I
will
want
to
announce
is
that
please
be
aware
that
for
next
week
and
in
two
weeks
we
will
move
the
captain's
community
meaning
to
the
Wednesday.
So
it
will
be
the
same
time.
But
when
we
on
the
Wednesday
is
only
in
20th
and
may
27th
will
be
the
next
community
meetings
and
then
again
we
have
to
move
to
the
Wednesday
in
the
middle
of
children.
A
A
Winston's
and
we
still
have
our
regular
weekly
cadence,
but
we
just
have
it
on
a
different
date.
If
you
want
to
get
all
these
updates
make
sure
to
subscribe
to
our
community
calendar.
We
have
the
link
down
here
in
our
document
and
also
we
will
announce
this
on
in
our
captain's
slack
workspace
and
also
on
Twitter,
so
make
sure
to
also
follow
us.
There.
B
A
Links
and
the
very
top
of
this
document,
so
if
you
enter
this
document,
you
will
find
all
the
links
and
talk
and
with
BIST
I
guess
we
can
already
jump
to
our
first
TRS
here
I'll
hand
over
to
you.
Hannes
I
also
see
a
couple
of
folks
from
the
captain.
Community
are
joining
and
promoting
them,
so
they
can
also
share.
If
you
have
something
to
share
and
yeah,
let
me
just
pause
line
or
stop
my
screen
sharing
and
please
Johannes
go
ahead.
Alright,.
C
Yep,
you
should
now
see
my
screen.
Well,
I
have
one
update
to
share
it's
about
the
captain
documentation.
Actually,
this
was
a
joint
work
between
you
and
myself,
and
we
did
restructuring
of
the
captain.
Dogs
page
and
I
just
want
to
show
you
what
changed
or
what
will
change
with
the
ordered
seven
release
when
it
comes
to
the
documentation,
maybe
just
an
additional
note
on
that.
The
main,
the
main
reason
why
we
did.
C
C
C
B
C
See
here
then
I
call
it
topic
areas.
In
other
words,
you
find
sections
that
have
have
a
specific
topic
or
target,
and
here
you
see,
for
example,
a
section
for
operating
captain
manage
captain,
then
a
section
for
the
use
case
of
continuous
delivery,
one
for
quality
case
automated
operations.
Then
here
one
for
custom
integrations.
This
will
contain
documentation
how
you
can
write
your
your
captain
service
and
and
keep
low
it
with
captain
or
not
and
use
it
with
captain
and
then
he'll
further
down.
You
get
information.
How
to
set
up
monitoring
then
still.
C
This
is
still
available.
The
reference
section
where
you
get
infos
about
the
captain,
API
bridge
and
CLI,
and
the
last
item
in
this
menu
is
now
a
troubleshooting
section
and
when
I
click
on
one
of
those
categories,
let's
take
the
operate
captain
and
then
you
still
get
all
the
information
that
was
available
in
the
previous
captain
documentation.
C
But
now
it's
not
aligned
to
a
tutorial.
It's
just
how
to
how
to
install,
for
example,
this
Xilai,
how
to
upgrade
how
to
uninstall
and
also
some
additional
information
like
the
community
support
matrix,
is
also
part
of
of
this
operate.
Captain
section
and
I
take
another
example,
and
then
I
will
stop
because
you
can
scroll
or
you
can
explore
it
by
your
own,
but,
for
example,
let's
take
continuous
delivery
here
you
get
also
informations
how
to
set
up
a
multi-stage
delivery.
C
All
right,
yeah,
as
I
said,
please
feel
free
to
click
through
the
new
documentation.
Whenever
you
find
a
broken
link
fix
it
by
your
own
or
just
shoot
me
an
a
select
mesh,
the
message
then
I
can
take
care
of
fixing
the
problem
link
or
something
else
all
right
yeah.
This
was
the
update
from
my
side.
It's
about
the
captain.
Talks
page
and
now
I
will
hand
over
to
Christian
thanks.
A
A
We
have
a
very
simple
use
case,
quality
gates
test
and
before
this
pull
request,
you
only
tested
on
the
on
a
very
old
mini
cube
version,
which
was
I,
think
version
1.2
and,
with
this
test,
I
basically
extended
our
test
matrix
on
Travis
CI,
to
cover
multiple
versions
of
mini
cube,
which
essentially
mean
that
we
cover
communities
versions.
One
point
fifteen
to
one
point:
18.2
with
our
tests
again
only
for
the
use
case,
quality
gates
in
that
scenario,
so
not
for
the
full
installation.
A
The
files
that
are
changed
is
essentially
only
Travis
dot,
llamo
and
the
way
it's
written
now
is
using
llamo,
templating
syntax.
So
I
can
basically
just
say
down
here.
Give
me
the
mini,
cube,
standalone
test
from
above
and
override
the
mini
cube
version,
and
that's
what
I
did
basically
for
the
for
the
other
tests.
A
So
if
we
need
to
test
a
new
mini
cube
version,
it's
as
simple
as
just
adding
three
new
lines
and
if
you
count
the
new
line
for
wine,
the
second
one
is
about
Google
Cuba
need
Ascension,
where
we
also
have
an
automated
test
and
Travis
CI.
This
one
got
extended
a
little
bit
further.
Now,
basically,
we
have
the
option
to
reuse,
East,
yo
or
to
reuse
the
ingress
gateway
in
that
sense,
that
gke
already
delivers
East,
you
version,
I,
think
1.2
or
something
so
a
rather
old
version,
but
it's
already
pre-installed
and
Google
manages
it.
A
We
basically
now
test
google
kubernetes
with
and
without
SEO
in
version,
one
point,
fourteen
for
Cuban
leaders
and
1.15
for
Cuban
eaters.
So
that's
four
test
cases
now
and
within
that
pull
request.
I
also
prepared
the
tests
for
Google
Cuban
eaters
version.
One
point:
sixteen
and
1.17
with
the
same
or
very
similar
templating
syntax
that
you've
seen
just
before.
So
this
is
just
to
be
yeah
to
be
enabled
when
we
need
it.
A
Unfortunately,
as
of
right
now,
one
point:
sixteen
is
not
directly
available
on
google
kubernetes,
but
from
what
I
read
it
should
become
become
available
or
like
in
the
next
couple
of
weeks
and
1.17
is
scheduled
for
later
this
year.
But
one
point
sixteen
is
something
we
definitely
want
and
need
to
test
with
the
full
installation.
A
A
A
We
have
a
cron
that
actually
failed,
but
it
failed
down
here
with
mini
queue,
but
this
is
a
cron
that
runs
once
a
day
just
to
make
sure
that
we
did
not
break
anything
in
the
full
installation,
but
obviously
the
tests
for
the
fuel
installation
they
take
like
a
little
bit
longer
than
the
tests
with
mini
cube.
Therefore,
we
don't
run
them
every
time,
also
with
google
communities.
There
is
some
cost
involved
and
we
also
don't
want
to
let
these
costs
explode
at
the
moment.
A
Alright,
if
you
have
any
questions
or
concerns,
shoot
me
a
slack
message
afterwards
or
comment
on
the
pull
request.
I
would
also
like
to
show
one
that
still
work-in-progress
now
so
I
started
improving
the
integration
test
for
quality
gates,
I'm
not
quite
finished
there.
So
it's
still
working
progress
just
want
to
show
a
couple
of
things
here,
so
you
can
see
my
approach,
so
what
I
do
basically
is
I,
have
a
slide-out
llaman
files
for
dynaTrace
that
I
test
that
would
be
default.
A
Metrics
and
I
also
extended
them
with
non-default
metrics
I
have
a
different
entity
select
or
on
the
very
right
side
of
the
sli,
and
essentially
I
do
have
two
SLO
files
that
I
will
test
with
that.
Actually,
if
a
third
one
as
well,
the
first
one
has
two
metrics
that
have
criteria
and
a
third
one.
That's
an
info
metric
and
the
second
file
basically
adds
to
more
info
metrics.
That
would
be
the
number
of
calls
to
other
services
at
the
time
spent
in
those
calls.
A
So,
for
instance,
if
my
service
calls
the
database
what
if
it
calls
another
micro
service,
then
that
would
be
something
that
I
want
to
see
and
in
the
third
file.
I
also
say:
I
want
to
add
a
warning
at
the
past
kathiria
for
the
number
of
calls
to
other
services
and
with
that
that
basically
can
write
an
integration
test.
A
I
will
spare
you
the
details
of
the
test,
because
it
basically
is
a
lot
of
bash
scripting,
at
least
right
now,
maybe
in
the
future
we
will
switch
to
writing
the
tests
in
go
you'll,
see,
but
essentially,
I
can
write
tests
that
make
sure
that
these
criteria
work
that
make
sure
that
adding
new
as
allies
and
us
allows
work
and
so
far
let
me
see.
Is
there
a
job
running?
Yet
it's
one
running.
A
That's
that's
in
my
future
branch
that
I'm
working
on
and
this
one
actually
has
just
passed
12
minutes
ago
and
essentially
I
do
check
the
response
of
the
SLI
provider
and
the
response
of
lighthouse
service
in
this
case,
I
check
that
this
is
a
neighbor
evaluation
done
event
and
basically
I
also
check
the
log
output
of
some
services
to
see
what's
going
on
actually
and
to
see
if
everything
is
working
as
expected
and
yeah.
So
this
is
work
in
progress,
but
it's
about
to
be
finished.
I
think
it
will
be
done
tomorrow.
A
A
C
A
D
D
Come
on
Lord,
as
you
can
see
mine,
my
computer
is
a
little
bit
slow
today,
but
for
the
Scylla
enhancements,
I
already
implemented,
most
of
the
yeah
required
see
the
I
commands,
get
projects,
project,
services,
service
cards
and
so
on,
and
what
is
missing
right
now
is
the
output
of
the
rotator
here
for
the
output
young.
This
I
need
to
implement,
but
therefore
I
need
to
adjust
the
go
utils
one
more
time
to
actually
get
this
stuff
so
picture
out
it
to
you're
honest.
He
helped
me
a
lot
and
one
day
so
yep,
that's
it.
D
D
E
Yes,
okay,
perfect
today,
I
want
to
show
you
two
things,
so
the
first
one
is
an
operator
for
a
pretty
American
installation
from
0.6
two
point:
two
zero
six
two,
and
for
that
upgrade
we
have
used
a
concept
that
has
proven
to
be
quite
useful
in
the
last
release,
which
is
to
employ
a
human
at
this
job
that
would
take
care
of
operating
all
the
resources.
I
can't
make
it
a
little
bit
bigger.
Please
first.
E
Alright,
thank
you.
So
in
essence,
I
won't
go
over
their
code
in
too
much
detail,
but
what
it
does
is
first,
it
will
check
if
all
the
manifests
are
required
to
upgrade
the
installation
and
are
reachable
by
it
up.
Then
it
will
check
if
currently,
you
have
version
zero,
six
one
installed
and
and
that's
the
case
it
will
proceed
with
upgrading
all
the
components
so
applying
the
manifests.
E
E
Also
worth
mentioning
is
the
procedure
of
the
plotting
or
a
plan.
This
operator
is
exactly
the
same
for
the
full
installation
as
well
as
product
on
its
own.
The
installation
so
based
on,
what's
installed
currently
in
your
cluster
upgraded
shop,
will
detect
of
installation
required
actions
automatically.
A
A
E
Part
of
the
manual
approvals
operator
working
on
right
now,
because
yeah,
what's
that
issue
or
pull
request,
is
about
was
an
extension
of
the
configuration
service,
because
now
we
have
the
requirement
of
maintaining
a
list
of
open
approvals
or
a
certain
service
between
the
configuration
service
and
this
we
have
extended
response
sent
by
that
service
for
the
get
projects,
get
services
and
get
stages,
endpoint,
etc.
So,
in
addition
to
the
previous
property,
like
the
creation
dates
and
less-developed
types,
we
also
now
have
a
list
of
open
approvals
which
looks.
E
D
E
E
E
Configuration
service
there
is
now
a
new
endpoint
where
can
create
an
approval.
This
is
actually
needed
by
the
gatekeeper
service
by
this
endpoint.
As
soon
as
is
perceived
and
an
improver
trigger
event,
then
of
course,
there's
a
get
method
for
the
proven
endpoint,
which
will
return
list
of
open
approvals
for
certain
service
and
then
in
order
to
close
an
approval.
B
E
Also
have
any
delete
endpoint,
which
will
remove
a
certain
approval
for
a
service
and
what
is
now
missing
for
the
complete
manual
approval
service
is
the
possibility
for
the
CLI
to
first
retrieve
all
ultimate,
prove
us
and
tend
to
send
an
approval
finished
event.
So
that's
what
I'm
currently
working
on
and
it
should
be
finished
tomorrow
or
next
week.
F
F
The
first
issue
was
that
with
time
with
civilization,
as
that
unit
that
we
have
been
showing
was
in
minutes
fixed,
so
when
in
event
the
duration
of
an
event
was
lower
than
one
minute,
it
showed
zero
minutes
instead
of
detailed
information
about
how
many
seconds
it
or
even
if
it
was
longer
than
60
minutes,
then
it
was
60
minutes.
Then
it
was
showing
just
the
minutes
and
not
not
more
information
about
that.
F
F
So
now
we
introduced
two
views
and
then
we
will
enhance
them
with
more
features
and
yeah
I
will
show
you
that
as
a
live
demo
as
we
can
see
when
we
enter
the
bridge
now
when
I
select,
the
project
I
enter
the
services
scream,
since
this
screen
has
the
most
information
currently
so,
as
we
are
used
to,
we
can
expand
the
service
and
yeah
select
the
root
event.
Where
we
can
see
all
the
traces
here,
I
will
show
you
the
improvement
of
the
time
with
civilization.
F
As
you
can
see,
if
the
duration
of
an
event
is
yeah
several
minutes,
then
it
shows
minutes
and
seconds,
and
if
it's
only
just
it,
if
it's
less
than
a
minute,
then
it
shows
only
the
seconds,
for
example,
or
if
it
would
take
longer
than
one
hour.
It
will
also
show
hours,
minutes
and
seconds
yeah
and
on
the
environment
screen.
We
have
the
stages,
so
here
we
have
currently
only
an
overview
of
this
stages
and
this
services
that
they're
displayed
that
are
deployed
in
that
stage.
A
If
I
can
interrupt,
I
just
thought
that
deep
link
functionality
does
not
work
with
the
environmental
services
screen.
I
think
we
need
to
add
that,
because
if
you
click
on
services,
then
I
guess
the
deep
link
should
always
reflect
that
you're
going
to
services
and
then
go
to
environment.
This
should
also
reflect
that
your
in
environment
etc.
Yeah.
F
F
F
B
C
F
F
F
That's
here
in
the
duration
of
the
events,
for
example,
especially
for
the
evaluation
events,
it's
very
interesting
to
have
the
duration
of
that
event,
and
it
was
showing
here,
for
example,
now
it's
showing
8
5,
58
minutes
and
8
seconds.
It
was
showing
this
unit
only
minutes
yeah.
It
was
showing
only
only
58
minutes
or
in
case
of
seconds.
It
was
showing
zero
minutes.
Yeah.
C
B
A
A
F
A
B
A
Want
to
walk
through
the
tutorials
of
captain.
There
is
no
need
to
start
the
load
generation
in
a
separate
terminal
or
to
start
it
on
the
local
machine,
but
instead
you
can
start
it
directly
within
the
cluster.
Nicholae
I
should
go
to
the
conversation
first,
so
it's
basically.
What
we
wanted
to
achieve
is
that
you
just
execute
what
you
just
apply
when
it
is
manifest,
and
you
will
start
the
load
generation
inside
the
cluster
to
all
the
three
stages
that
are
typically
on
board
it
or
create
it.
A
When
you
do
that,
tutorials
with
captain
inlets
shopping,
cart
service.
So
in
this
K
is
it's
very
simple:
no
generation
script.
It's
a
shell
script
so,
but
there
is
no
need
for
you
to
execute
the
shell
script.
Basically,
you
just
start
the
load
generation
and
we
have
a
couple
of
yeah
memo
files
that
can
be
applied
for
different
scenarios.
So
you
have
the
cards
note
generation
base
which
will
just
start
it
with
like
a
base
traffic.
That's
going
on!
There's
a
little
bit
of
sleep
time
between
the
items
that
are
added
to
the
shopping
cart.
A
Then
there
is
a
faster
version
where
you
can
imagine
the
sleep
time
is
just
a
little
bit
slower
and
then
there
is
another
version
which
is
which
adds
some
faulty
items
that
will
cause
the
card
service
to
need
more
CPU
to
process
these
items,
and
it
just
has
different
items
here.
So,
of
course,
you
can
also
change
this.
A
So
nothing
more,
not
fancy,
but
we
just
hope
it
may
makes
it
easier
for
you
when
you
go
through
the
captain,
tutorials
that
there
is
no
need
to
run
any
low
generation
script
for
the
use
cases,
and
also
we
suggest
now
to
already,
after
the
first
onboarding
of
the
service
to
start
the
load
generation.
So
the
monitoring
tools,
in
the
background
good
baseline,
when
you
move
on
to
the
quality
evaluation
and
also
to
self-healing
parts
of
the
captain
tutorials,
so
that
the
monitoring
tools
already
got
their
stable
baseline
and
the
anomaly
detection
is
working
correctly.
A
If
not,
we
just
take
a
look
here
at
our
agenda.
Next
part
will
be
take
a
look
at
the
working
progress
items
and
I
think
I'm,
also
painting
over
here
to
Christian
and
Johannes,
so
I'm
stopping
my
screen
sharing.
We
have
a
couple
of
more
questions,
but
I
think
they
fit
more
when
it
comes
to
the
backlog
refinement.
So
we
just
keep
the
questions
and
we
will
ask
them
yeah,
also
we're
running
a
little
bit
short
on
time.
A
Alright,
so
one
of
the
next
steps
we
want
to
go
through
with
you
is
to
take
a
look
at
our
working
items
and
as
a
first
thing,
I'd
like
to
discuss
the
things
that
are
in
progress
or
assigned
and
committed
so
things
that
are
not
done
yet.
So
basically,
the
issue
here
is
taken
by
Kristen
taekman.
He
already
presented
of
his
issue
he's
still
working
on
it
same
is
for
my
issue
here:
improved
the
integration
test,
I've
presented
it
I
think
I
will
be
done
tomorrow.
A
F
A
F
A
F
B
A
E
A
C
Mine
last
week
I
forgot
to
talk
with
andrea
andreas
about
this
issue
and
I
still
need
his
feedback
on
that.
I
left
a
comment
there
and
yeah.
It's
it's
still
open
and
need
to
be
discussed,
but
this
is
the
first
issue.
First
Tasker
will
approach
on
Monday
morning
to
get
this
one
done
and
also
to
get
one
step
further
when
it
comes
to
the
self-healing
use
case.
C
A
A
A
B
F
C
F
G
F
C
A
B
A
Picture
is
already
on
it.
I
will
remove
myself.
So
basically,
Patrick
has
created
us
a
little
issue
that
basically
goes
down
to.
We
should
add
a
better
documentation
for
the
upgrader
that
Florian
has
shown
us
today,
because
when
you
execute
the
upgrader,
it
tells
you
to
delete
the
existing
upgraded
job,
but
that
statement
might
not
work
on
all
humanities
platforms.
We
should
add
the
namespace,
in
this
case
the
default
namespace
to
that
statement.
So
it's
essentially
just
a
documentation
issue
and
I'll.
Just
actually
I'll
just
do
that.
Take
that
if
that's
okay
for
everyone.
C
Okay,
tan,
the
next
one,
is
an
issue
that
dry
storm
takes
a
look
on
kubernetes
and
which
Cooney's
versions
we
are
supporting
without
at
seven,
because
last
week
we
got
the
request
funds
from
captain
user,
and
this
person
asked
us
what
community's
version
are
we
going
to
support
in
the
future
and
for
answering
this
question
I
also
provided
some
documentation
there.
Let
me
just
click
on.
Oh.
C
Well,
on
the
captain
talks
page,
you
have
a,
but
it's
a
closed,
PR
and
I.
Think
we
don't
get
to
preview
there
or
do
we
know
we
don't
well,
let's
bring
it
to
the
point
with
ordered
seven,
we
wanna
support
kubernetes
1.14
to
1.18,
and
this
need
to
be
verified.
Currently
we
are
supporting
1.15
as
the
highest
cuneus
version
we
support
and
in
the
future
we
are
going
up
to
1.18.
C
Christian
is
already
doing
great
work
when
it
comes
to
testing
captain
on
various
communities
versions,
as
he
showed
us
today
with
with
his
automation
in
regard
to
mini
cube
and
but
to
have
a
clear
answer
and
that's
working
on
the
different
cloud
providers
as
well.
We
need
to
take
this
task
and
we
need
to
invest
some
time
in
verifying
the
support.
C
Here
is
and
ok
well,
the
definition
of
tan
is
that
when
we
are
sure
that
Katniss
working
on
those
versions,
then
we
need
to
remove
the
diversion
check
which,
which
is
implemented
in
the
CLI
and
in
case
there
are
some
limitations
or
some
that
occurred.
When
running
captain
on
any
Kunis
version,
then
we
also
need
to
list
the
de
limitations
in
the
release.
Note
all
right
who
is
up
with
that
one
meaning
this
one
is:
is
yep
testing
out
captain
on
various
communities
versions.
A
A
Johannes
has
written
a
nice
test
matrix
for
us
and
some
test
scenarios
that
we
can
test
with
like
manual
approvals,
automatic
approval
and
so
on.
It's
a
very
long
text.
I'm,
not
gonna,
go
through
all
of
that.
So
basically
it's
writing
an
integration
test
for
men
who
approval,
as
the
title
says,
anyone
up
for
that.
A
A
C
Well,
at
beginning
or
the
day,
yeah
at
the
beginning
way,
we
wanted
to
take
this
task
to
provide
an
upgrade
mechanism
for
the
approval
strategy,
but
since
we
are
missing
the
upgrader
for
ordered
seven,
it's
also
all
about
creating
the
first
version
of
the
upgrader
for
ordered
six.
Two
two
two
ordered
seven
and
it's
also
kind
of
researched
tasks,
because
actually
I
don't
know
right
now.
What
other
is
it
is
something
needed
for
upgrading
the
approval
strategy
or
for
upgrading
captains
or
resisted
that
implement
their
approval
strategy.
A
A
C
A
So
I
guess
these
two
go
hand
by
hand.
That's
the
one
that
Florian
was
talking
before
about
exposing
the
event
endpoint
of
MongoDB
datastore,
why
our
public
REST
API
so
like
in
one
of
the
previous
meetings,
we've
shown
that
we've
changed
the
REST
API
gateway
a
little
bit,
and
that
should
be
a
rather
simple
task.
I
guess
we
just
need
to
deprecate
the
existing
endpoint
and
and
forward
like
this
value.
I
of
the
MongoDB
data
store.
C
C
A
F
C
A
Obviously,
if
anyone
is
watching
this
afterwards
on
YouTube
or
other
channels,
just
let
us
know
wise
like
if
you
want
to
do
any
of
the
remaining
issues
or
there's
always
the
good
first
issue
list
that
we
have
I.
Think
it's
linked
in
the
captain
community
meeting
document
on
top
somewhere
I
think
it's
linked
some
work
below
other
than
that
I
think
we're
done
with
the
planning
and
I
think
we
can
hand
it
back
to.
A
Thanks
I
just
share
one
last
time,
but
it's
actually
only
because
I
again
want
to
highlight
said
for
next
week
we
already
meet
on
Wednesday
and
the
big
afterwards
are
also
on
Wednesday.
So
if
you
have
already
subscribed
to
the
Google
Calendar,
it
should
be
already
up
to
date,
so
we
updated
the
calendar
already
might
just
be
a
synchronization
issue
that
it
does
reflect
to
get
to
your
calendars,
but
just
be
aware,
we
start
next
next
week
already
and
then
we
have
a
couple
of
new
issues.
A
So
thank
you
all
for
participating
here,
also
in
the
community
meeting
in
the
chat.
So
we
got
a
couple
of
feature
requests
there
and
there
will
be
like
you
trigger
discussion,
also
in
maybe
for
the
last
two
or
three
minutes.
I
also
want
to
try
to
kind
of
give
a
short
summary
of
the
discussion
and
it
was
around.
Is
the
image
tag
required
for
all
the
captain
cloud
events
also,
if
captain
is
not
doing
the
deployments
raw
I
think
this
was
the
actual
question
yeah.
G
I
can
just
elaborate
a
little
bit
on
that
if
you
want
oh
yeah,
sure
thankful,
sure
yeah
I
mean
I'm
just
thinking,
you
know
the
future
of
Captain,
where
it
can
deploy
things
presuming.
This
is
the
division
that
captain
can
deploy
things
besides
to
just
kubernetes
and
it
can
do
things
besides
just
a
deployment
of
an
image.
So
one
I
wanted
to
kind
of
validate
that.
G
But
then,
if
that's
the
case,
should
we
more
when
we
have
these
cloud
events
that
all
currently
specify
or
require
that
you
specify
an
image
name,
two,
maybe
more
generalize
that
so
that
it's
more
like
you're
you're,
you
know
not
locking
yourself
into
requiring
image
in
all
over
the
place.
I
hope
that
makes
sense.
I.
A
G
F
G
It's
so
it's
not
like
something
you'd
probably
solve
immediately,
but
just
right,
if
it's
more
an
artifact
or
but
also
not
just
aren't,
it
could
be
a
configuration
change
how
you
abstract
that
or
title
that
I
don't
know,
but
it's
just
more
I'm
thinking.
You
know,
as
you
think,
about
the
cloud
events,
maybe
consider
that
I
don't
have
a
solution.
Just
an
observation.
A
Yeah
thanks
I
I
agree
with
that
observation.
I
don't
have
a
solution
either.
Maybe
Johannes
has
one
I,
don't
know,
but
we'll
actually
consider
a
lot
of
different
things
for
the
future,
not
just
in
a
cloud
events
also
for
what
captain
is
doing
and
we'll
need
to
see
if
we
can
get
rid
of
this
image
tag
combination
somehow
in
the
end,
I
guess
whatever
you
want
to
have
as
a
version
could
be
a
string
or
a
chaise,
an
object
or
whatever,
and
it
doesn't
necessarily
need
to
be
specified
by
captain.
A
G
C
C
There
are
situations
where
you
maybe
don't
deploy
an
image,
or
you
have
no
reference
to
an
image
at
all.
Yeah
thanks
for
bringing
it
up,
I
will
or
I
will
think
about
that
when
we
go
further
and
discussing
the
architecture
and
how
we
are
doing
or
how
we
are
specifying
the
next
generation
of
cloud
events.
A
Okay,
thank
you
for
all
the
discussions
since
we're
already
at
the
end
of
the
schedule
meeting
time,
I
would
like
to
end
the
meeting
here.
We
will
put
the
meeting
the
recording
online
on
YouTube.
If
there
are
any
objections,
please
let
me
know
otherwise
we
will
put
it
on
there,
so
you
can
also
take
a
look.
There.
We'll
also
put
another
link
to
our
meeting
document
there,
so
you
can
also
have
a
look
there
and
yeah
I
just
want
to
let
you
know
again.