►
From YouTube: App Runtime Deployments Working Group [Apr 21, 2022]
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Working
group
meeting-
okay,
sorry,
I
had
to
shift
this
meeting
because
of
a
vacation
day.
I
also
had
a
quite
a
struggle
with
daylight
saving
time
adjustments.
A
So
I
think
this
time
now
is
correct
and
I
will
adjust
the
meeting
series
to
this
time
that
we
have
now
so
currently,
it's
one
hour
too
late,
because
I
messed
around
with
the
calendar
good,
then
there's
not
too
much
on
the
agenda
for
today,
as
one
announcement
as
you
know,
we
have
this
story
for
setting
up
our
own
concourse.
A
We
had
some
trouble
at
sap
with
yeah
other
priorities
and
now
yeah
this
story
here.
Oh
let's
do
here
is
a
moment
here.
Putting
this
really
now
in
progress
and
yeah,
we
might
may
have
to
do
some
refactoring
first
before
we
can
reuse
this
project,
but
then,
hopefully
we
soon
we
have
our
own
concourse
nice,
good
yeah,
and
that's
basically
all
from
my
side.
Then
I
would
hand
over
to
dave.
A
So
we've
seen
that
after
quite
some
efforts,
you've
managed
to
create
a
bionic
release
of
cf
deployment,
and
we
would
just
like
to
tell
us
what
obstacles
you
you
hit
and
how
you
overcame
them,
and
so
that
we
get
an
idea.
A
What's
up
for
the
next
stem
cell
upgrades.
B
Sure
so
we
have
a
fairly
automated
system
in
general.
Let
me
find
that
as
my
conquest
there,
it
is
so
if
I
go
to
update
releases-
and
I
can
share
my
screen-
so
this
is
the
the
pipeline
in
general-
that
does
all
of
the
release
management.
I
think
I've
shown
this
in
the
last
meeting,
but
I
wanted
to
focus
on
this
first
group
of
of
jobs
here,
because
this
is
how
the
stem
cell
bumps
are
managed.
B
B
Oh,
I
need
to
log
in
apparently
I
don't
know
why
that's
not
public,
I
don't
think,
there's
anything
sensitive
in
here.
Basically,
it
detects
the
the
news,
the
difference
between
the
new
stem
cell,
that's
coming
in,
and
the
stem
cell.
That's
currently
on
main
of
cf
deployment
and
figures
out.
Is
this
a
major
or
a
minor
bump
and
it
just
outputs
a
file
which
I
I
remember
right
that
goes
to
gcs.
Currently,
that
says
it's
either
a
major
or
it's
a
minor
that
then
triggers
one
of
two
jobs.
B
Actually,
does
it
do
that?
No,
maybe
this
one
doesn't
do
that
this
one
just
says:
okay,
I
need
to
redeploy
all
of
the
releases
with
the
new
stem
cell
export
them
to
compile
them
updates.
All
of
the
the
ops
files
that
are
necessary
puts
all
the
releases
in
gcs,
and
this
then
pushes
to
develop
and
that
flows
through
the
the
regular
pipeline
has
all
the
validation
that
that
has,
and
then
we
we
kind
of
released
manually.
B
If
it's
a
minor
stem
cell,
then
it
says
okay.
Well,
I
don't
need
to
recompile
everything.
I
don't
have
to
stop
everything
else
from
happening,
so
it
just
grabs
one
environment
and
then
that
will
just
do
a
a
validation
deployment.
Just
like
a
release
bump
run
smoke
test
make
sure
everything
is
still
fine.
B
It
doesn't
need
to
recompile
the
the
releases,
so
it
will
commit
directly
to
maine
and
actually
create
a
release
and
then
so
that
everything
else
continues
to
work.
It
then
makes
the
same
commit
it
merges
that
into
develop
that
flows
through
the
rest
of
the
pipeline,
but
the
release
is
out,
so
this
is
kind
of
like
a
short
circuit
way
of
getting
a
minor
stem
cell
bump
out
quickly
without
it
having
to
a
flow
through
the
pipeline
and
b.
Wait
for
us
to
then
kind
of
release.
B
So
we
should
now
that
we
we've
got
we're
on
bionic.
This
job
will
start
triggering
unless
they
create
version
two
or
higher.
If
we
get
a
minor
bump,
this
job
will
trigger
and
we'll
we'll
get
automated
releases
again,
which
is
nice
as
far
as
the
the
challenges
that
we
had
really.
The
biggest
challenge,
I
think,
was.
B
That
I
was
probably
a
little
too
eager
in
doing
the
bump
on
develop
and
there
were
components
like
mfs
volume,
services
and
smb
volume
services
that
weren't
ready
that
I
didn't
realize
because
we
had
a.
B
We
had
a
pipeline
that
was
testing
bionic
with
an
ops
file,
but
it
wasn't
testing
those
components,
and
so
I
didn't
realize
that
they
were
broken
and
so
because
the
change
made
it
on
to
develop.
B
We
were
then
stuck
in
a
situation
where
we
were
blocked
on
that,
but
it
was
very
hard
to
back
out
because
once
it
made
it
to
develop
all
of
the
new
release,
bumps
that
came
in
after
it
were
being
compiled
for
that
stem
cell.
So
it
wasn't
just
a
case
of
chase
the
stem
cell
back.
All
of
that
stuff
would
have
had
to
be
wound
back
and
then
replayed
with
the
old
stem
cell,
and
so
we
kind
of
got
into
a
place
where
it's
just
like
okay.
B
Well,
we've
just
got
to
wait
until
these
things
are
ready.
Fortunately,
that
didn't
take
too
long,
but
I
think
that
was
the
biggest
learning
for
me
was
okay.
B
We
need
to
have
a
better
idea
of
what
needs
to
be
validated
before
we
do
the
next
part,
the
other
the
bump,
to
jammy,
because
I
think
that
will
that
will
help
us
to
have
a
much
smoother
transition,
and
then
I
think
the
other
thing
that
came
through
were
was
the
dynamic
asg
change
and
that
caused
ci
to
go
red
for
a
while,
because
there
were
some
issues
with
cats
that
they
had
to
help
us
with,
and
so,
even
once
we
started
getting
some
of
the
fixes.
B
We
couldn't
get
those
to
flow
through
the
main
pipeline,
because
it
was,
it
was
red
for
other
reasons,
and
so
there
were,
there
were
just
a
bunch
of
things
happening
at
the
same
time,
so
I
think
coordination
of
oh
okay,
we
know.
Okay,
nothing
else
is
happening.
We
can
do
the
jammy
bump
in
isolation,
we'll
we'll
be
we'll
make
things
a
lot
smoother.
I
think.
A
B
I'll,
let
you
process
all
of
that
and
see
if
you're,
ready.
A
B
B
Get
in
touch
with
that
team
and
get
them
to
kind
of
release
of
that
they
also
manage
smb,
and
that's
when
somebody
else
realized.
Oh
smb's
also
got
a
similar
problem
and
it
was
just
because
of
utilities
that
are
are
packaged
into
those
releases
that
needed
to
be
updated,
and
so
once
they
realized
that,
and
I
could
impress
on
them
the
urgency
of
getting
this
done.
It
happened
fairly
quickly.
At
that
point,.
A
B
I
did
add
smb
volume
services
to
the
part
of
the
pipeline
that
tests
disaster
recovery,
but
I'm
not
sure
that
that's
giving
us
full
coverage
of
either
nfs
or
smb
volume
services,
because
someone
was
reporting
an
issue
with
actually
creating
a
volume
and
mounting
it
into
an
application
container
that
I
didn't
see
when
running
that.
B
You
know
keeping
an
eye
on
those
releases
and
making
sure
that
they're
ready
and
talking
to
that
team
and
saying
hey
we're
planning
this.
Can
you
validate
that
you're
ready
to
go
kind
of
thing.
B
A
D
Yeah,
maybe
we
can
add
the
nfs
and
smb
components.
Also,
then,
for
the
gemmy
validation
pipelines
before
everything
goes
into
cf
deployment
directly.
D
B
To
do
a
like
a
final
validation,
yes,
we
can
definitely
do
that.
We
just
got
to
figure
out,
because
I
don't
think
there
might
be
something
in
cats
that
does
basic
validation
of
them.
I
seem
to
remember
that
the
volume
services
team
added
that
so
it
might
just
be
a
case
of
finding
a
cat's
job
that
has
the
appropriate
include
flex
turned
on.
B
B
D
On
this
journey
of,
let's
say,
setting
up
an
own
conquest
and
then
moving
over
the.
D
Step
by
step
and
yeah
at
one
point
in
time,
it
would
be
nice
if
everybody
from
the
working
group
could
cut
a
release
actually,
which
requires
probably
some
yeah,
some
some
training,
but
first
of
all,
we
need
all
the
excess
etc.
Is
there
some
blocking
things
that
you
could
foresee
when
we
start
with
this
new
concourse,
for
instance,
can
we
set
it
up
and
still
use
the
old
yaz
accounts
for
the
landscape
so
that
we
can
do
this
step
by
step.
B
I
think
so,
yes,
we've
just
got
to
transfer
the
the
credentials
into
the
appropriate
secret
manager,
but
I
don't
see
any
reason
why
not
they
are
currently
in
separate
gcp
accounts,
so
there's
nothing
that
relies
on
everything
being
in
the
same
account
on
in
the
same
project.
So
I
don't
think
that
that
will
be
a
challenge.
B
B
B
D
Yeah
yeah,
first
of
all,
we
need
to
get
the
conquest
set
up
and
running.
That's.
D
A
D
A
B
Yeah,
I
did
have
one
question
related
to
ci,
so
I've
been
tracking
the.
B
The
recommendation
that
is
a
pr
only
workflow,
which
I
think
is
great
for
contributions,
but
I
didn't
know
how
that
applied
to
automation
and
the
reason
I
say
that
is
right.
Now
we
have
a
lot
of
automation,
especially
the
release
bumps
and
stem
cell
bumps
that
commit
directly
to
develop
as
a
bot
user,
and
I
did
are
we
supposed
to
because
if
we
have
to
change
those
two
pr's
then
does
that
mean
we
have
two
bots,
the
one
that
creates
the
pr
and
then
another
bot
that
immediately
approves
it.
D
That
sounds
that
sounds
strange
right.
I
mean
the
pr
makes
sense
if
somebody
reviews
that
if
there's
a
bot
creating
the
pr
and
directly
merging
it
behind,
then
I
don't
know
what's
what's
the
value,
that's
at
least
my
point
of.
D
I'm
aware
that
they
are
recently
got
many
bot
users
added
to
all
the
working
groups
and
projects.
So
I
guess
this
is
common
practice
everywhere,
for
dependency
updates
depend
upon
okay,
they
create
aprs
right.
Yes,.
B
D
B
D
D
Day,
the
important
thing
is
that
we
get
cloud
foundry
releases
out
of
the
door
regularly
and
that
we
have
the
whole
working
group
being
able
to
to
cut
those
releases
so
that
you
have
to
do
all
the
work
and
everything
and
all
the
time.
And
we.
C
B
A
A
A
A
B
Okay,
the
only
other
thing
I
had
on
my
list
of
things
to
talk
about.
If
we're,
if
we're
done
with
that
is,
I
wondered
whether
we
needed
to
have
more
frequent
meetings
in
the
short
term
just
to
help
get
things
up
and
running.
B
A
A
B
A
B
A
Yeah,
okay
yeah,
but
it
just
certainly
makes
sense.
Then,
in
the
beginning,
good
yeah,
I
guess
the
real
fun
starts
when
we
move
the
first
pipelines
and
try
to
build
something
useful
with
them.
Then
this
is
when
the
real
action
starts.
This
should
hopefully
soon
be
the
case.
B
Yeah
and
if,
if
you
have
kind
of
somewhat
advanced
notice
of
when
you're
going
to
work
on
that,
I
can
be
available,
you
know
early
every
morning
like
from
you
know,
8
8
30,
to
help
and
and
have
more
synchronous
time.
You
know
we
can
either
jump
into
zoom
or
just
chat
on
slack,
but
at
least
you're
not
doing
something
asking
a
question
waiting.
24
hours
coming
back
the
next
day,
yeah.
D
Maybe
one
question:
do
we
have
everything
in
place
that
we
need
for
setting
up
the
congos?
I
mean
this
is
still
something
where
we,
if
it
may
need
open
costa
for.
A
So
we
can
ask
ruben
for
the
technical
stuff
of
this
github
com
cloud.
Foundry
bosch
community
stem
cell
project.
I
guess
I
don't
guess
he
has
implemented
most
of
it
and
can
certainly
help
when
we
start
refactoring
it,
because
it
contains
a
lot
of
hard-coded
bosch.ci
stuff
in
it
yeah.
We
have
to
agree
on
a
domain
name
for
this
one.
A
So
we
have
what
is
it
so
bosch
ci?
This
is
the
one
for
for
stem
cells
up
and
running,
and
we
need
then
the
next.
The
new
one
could
be
something
like
I
don't
know:
release
integration,
ci
cloud,
foundry
org.
D
A
A
B
A
Okay,
so
work
more
with
issues
could
be
a
solution.
They
have
checked
all
the
other
working
groups
what
they
are
doing,
but
there
seems
to
be
no
good
common
sense
what
to
use.
But,
okay,
we
don't
have
to
solve
this
now
so
far,
it's
still
overseeable
what
we
are
doing
good
so
for
the
domain
name,
something
like
deployments:
dot,
ci.
D
To
yeah
working
group
deployment,
so
g
minus
deployment
that
you
know
this
is
not
just
for
the
project.
A
Let's
go
for
this
okay,
good
and,
of
course,
as
we
have
christened
the
call
chris
you
mentioned
that
you
can
provide
us
an
gke
account
yeah.
I
can
get
that
set
up.
E
E
A
E
A
Yeah
I
mean
for
developing,
we
would
have
something
yeah.
We
also
have
gcp
accounts,
but
if
we
get
the
official
one,
this
would
also
be
fine.
Where
would
we
store
the
credentials
or
other
credentials.
A
They're
also
deploying
a
cretub.
Yes,
I
guess
it's
for
the
let's
encrypt
integration.
Ruben
has
the
details
here:
okay,.
A
Maybe
also
for
the
concourse
secrets,
database
secrets
and
the
like:
yes,
so
it's
quite
sophisticated,
but
it
should
have
everything
we
need,
most
importantly,
the
let's
encrypt
integration,
so
that
we
have
a
valid
certificate
and
this
github
teams
integration
so
that
you
can
easily
log
on
with
your
github
account
like.
I
do
now.
D
D
D
Interfaces
for
in
the
copy
team,
because
here
you
need
some
credentials
to
upload
bosch
releases
into
some
buckets,
for
instance
a
new
ruby
version,
or
something
like
that,
and
here
I
think
philip
and
florian
commit
us
as
they
received
some
credentials
to
access
a
certain
password
manager
that
obviously
the
copy
team
had
there
in
place.
I
don't
know
if
there's
a
general
solution,
or
I
mean
if
every
team
has
a
cut-up,
then
yeah,
maybe
that.
A
D
I
guess
this
is
not
the
best
option.
You
can't
read
the
credentials
in
github
sequence,.
A
B
It
sounds
like
they've,
maybe
shared
access
to
to
lastpass,
which
is
what
we
were
using
for
credential
sharing
within
release
integration,
so
I'll
have
to
reach
out
to
them
and
find
out,
if
that's
the
case,
because
if
so
then
we
can
definitely
do
the
same
thing.
We
have
a
shared
folder
that
has
all
of
the
the
secrets
necessary.
B
D
C
C
A
E
B
If
you
need
help
research
and
slack,
if
I'm
online,
which
will
be
the
end
of
end
of
day
for
you
beginning
of
day
for
me,
then
I'll
I'll
help
in
any
way
again.