►
From YouTube: App Runtime Deployments Working Group [May 12, 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
Welcome
to
our
bi-weekly
meeting
now
so
there's
not
too
much
on
the
agenda
for
now.
So
first
is
the
new
conquerors.
So
the
story,
let
me
share
my
screen
now
it's
reinstalled
now
I
can't
share
the
screen
yeah,
fantastic,
so
the
story
from
our
project.
A
A
So
here
in
the
chat,
this
is
the
storyboard
chris
has
provided
us
a
gcp
account
thanks
for
that
that
yeah
worked
like
a
charm,
and
now
we
yeah
have
a
new
account
or
project
where
we
can
set
up
everything.
So
far,
things
run
quite
smoothly,
but
well
we
may
we
are
already
hitting
you
know,
hitting
the
first
problems,
and
maybe
we
have
to
ask
reuben
for
assistance,
but
nevertheless
it's
also
a
good
practice
for
refreshing
kubernetes
know-how.
A
So
it's
countered
yeah.
So
this
story
is
now
really
in
progress.
B
A
Yeah,
okay:
this
is
the
best
suggestion.
We've
come
up
with:
okay
release
dash
integration.ci.cloudfoundry.org.
B
A
B
Bit
but
yeah,
that's
fine,
whichever
you
you
feel
is,
is
best
while
she's
watching,
because
it's
like
it's
a
project,
it's
a
team.
It's
it's
everything.
Yeah.
A
B
D
I
think
we
are
setting
up
the
second
one.
Second,
one
was
the
first
one,
but
there
were
no
working
groups
back
then.
C
C
A
Make
a
small
survey
and
then,
if,
if
nothing
better
comes
out,
we
keep
this
this
release
integration,
but
I'm
fine
with
anything.
A
C
A
Update
story:
actually,
the
plan
was
that
we
so
from
sap
are
working
on
this
bionic
release
to
get
some
practice
in
creating
cf
deployment
releases.
Now
already,
if
you've
already
done.
B
I
mean
there's
a
there
was
a
bit
more
with
this
one
than
there
would
be
with
a
normal
release
just
because
it
involved
pipeline
work
so
carson,
and
I
took
care
of
that,
but
it
would
be
good
to
start
pairing
with
with
one
of
your
two
to
do
releases.
I
know
carson
mentioned
it.
B
I
think
on
the
last
one,
the
challenge
is
going
to
be
the
time
zone,
overlap,
figuring
out
how
we
we
can
coordinate
that
we
may
have
to
to
say:
okay,
we
will
actually
schedule
the
timers
okay,
we're
going
to
kind
of
release
here
and-
and
maybe
you
know,
we
have
to
figure
out
okay.
What
features
need
to
be
there
which
c,
which
component
bumps?
B
Are
we
waiting
for
if
anything
or
we're
just
going
to
take
whatever's
on
release,
candidates
and
say:
okay,
we're
cutting
all
these
from
that
make
sure
that's
ready
to
release,
but
I
think
that
would
be
better
to
actually
shed
you
all
the
time
and
say:
okay,
yep,
say
friday
morning
or
well,
that's
hard,
because
that's
the
end
of
week
for
you,
so
I'm
not
sure
friday
evening
is
great.
We
may
have
to
circle
around
and
say
like
tuesday
mornings
or
something,
but
it
all
depends
on
on
your
availability.
A
B
A
Okay,
good
so
yeah,
if
you,
if
you
can
send
or
proposal
in
our
slack
channel
when
on
on
which
day
you
would
want
to
have
to
someone
to
pair,
and
we
can
try
to
arrange
this
yeah.
B
Okay
sounds
good
and
both
carson
and
I
can
cut
releases,
so
it
could
be
either
one
of
us
pairing
with
you,
so
yeah
I'll
start
a
thread
in
let's
start
a
conversation
in
slack
and
we'll
we'll
figure
out
a
date
and
go
from
there.
B
Speaking
of
stem
cells,
I
had
a
conversation
with
the
infrastructure
working
group
about
the
germany
stem
cell
work
that
they've
just
started
on
and
coordinating
with
them,
because
they
they
were
asking
when
cf
deployment
would
switch
to
jammy,
and
I
said
that
our
policy
has
been
not
to
switch
as
a
default
stem
cell
until
it
has
a
one
dot
x
release,
because
we
want
that
guarantee
of
stability.
B
They
said
that
they
were
hoping
to
actually
get
validation
in
cf
deployment
before
1.x,
so
that
they
could
actually,
they
were
still
free
to
make
breaking
changes
if
needed,
and
so
what
I
proposed
was
similar
to
what
we
had
for
bionic,
which
is
having
an
ops
file
ncf
deployment
that
would
allow
them
to
deploy,
with
whatever
jammy
stem
cell
they'd
uploaded
to
the
director
instead
of
the
default.
B
B
B
So
wait
to
hear
what
happens
with
that.
B
That's
my
understanding,
yes,
is
that
they
like
they
said
they
were
waiting
on
something
for
cred
herb
and
I
think,
there's
also
a
question
of
the
compatibility
with
the
root
fs
that
they're
not
sure
if
they
need
cf
linux,
fs4
or
not.
D
A
cloud
cloud
controller
won't
work
as
well,
because
we
have
not
yet
updated
ruby
3.1.
This
is
pending
and
will
take
also
some
time.
Yeah
3.0
release
just
yeah
broke,
and
we
have
to
start
again.
B
B
I
may
suggest
to
them
that,
if
it's
more
in
flux,
they
could
maintain
and
keep
the
ops
file
themselves
for
now
and
it's
not
in
cf
deployment
until
it
has
a
certain
level
of
stability
and
maturity
that
it's
not
going
to
be
them
submitting
a
pr
every
week
to
change
the
instance
groups
that
it
applies
to
and
that
kind
of
thing,
but
we'll
see
what
they
say.
B
I'm
not
sure
yet
that's
one
of
the
questions.
I
need
to
ask
them
because
with
bionic
I
had
a
pipeline
here,
but
I
could
apply
it
to
all
the
instance
groups
and
it
was
just
a
case
of
validating
that
everything
worked
which
by
and
large
it
did
so
I
I
don't
know
whether
they
in
the
short
term
but
planning
on
having
a
pipeline
here
to
test
it,
while
it's
it's
fairly
unstable
or
what
they're
thinking
there
I'll
I'll
start
up
her
because
they
they
reached
out
to
me
through
vmware
channels.
A
B
Generally
speaking,
we've
never
done
maintenance
branches,
yeah,
we've
just
rolled
forward
to
whatever
the
the
new
versions
are,
and
if
someone
needed
to
to
revert
back
to
something
older,
they
can
craft
their
own
ops
files.
For
that,
that's
the
beauty
of
of
cf
deployment.
Anyone
could
craft
an
ops
file
to
to
make
any
change.
They
need
to.
A
B
B
B
A
B
A
Without
a
proper
requirement,
we
don't
need
to
create
additional
work:
yeah,
okay,
yeah,
good,
okay,
fine!
Then
I
would
move
this
to
done
for
now.
Yeah,
okay,.
A
One
small
thing
popped
up
this
morning:
someone
wanted
to
know
why
there
is
no
cf
22
branch
in
the
cf
acceptance
tests.
So
I
yeah,
when
do
you
usually
create
those
branches?
Does.
B
B
So
the
way
this
works,
if
you
go
back
to
the
the
main
pipeline
view.
Yes,
every
time
a
new
cf
deployment
is
released,
it
triggers
all
the
way
on
the
left
there
and
it
will
actually
redeploy
a
long-lived
environment
with
the
new
version
and
then
every
time
cats
runs,
on
the
right
hand,
side
that
the
two
boxes
in
in
parallel
there
and
it
then
triggers
bless
cats.
B
One
of
the
things
that
that
does
is
it
creates
a
branch
on
cf
deployment,
with
the
version
that
that
environment
is
deployed
with
okay,
the
last
successful
deployment,
the
last
successful
run
of
cats
was
against
cf
deployment
19..
And
so,
if
you
look
in
here,
you
should
see
that
the
branch
that
it
created
well,
this
one
this
one
doesn't
get
to
run
the
the
van
cats.
So
we
need
to
there's
a
problem
with
what
looks
like
an
internal
certificate.
B
Getting
an
error
about
the
certificate
that
cloud
control
the
internal
cloud
control
is
set
that
is
being
specified,
so
I'm
not
sure
whether
that
just
means
that
the
app
needs
to
be
updated
to
ignore
the
validity
of
the
search
or
whether
there's
something
we
can
do.
It
seems
hard
to
provide
the
ca
used
to
create
cloud
controllers,
internal
search
to
the
application,
because
that
would
mean
I
would
have
to
go
into
the
diego
trust
store.
B
I
think-
and
so
I
just
I
need
to
sit
down
and
actually
look
at
what
the
best
way
of
solving
this
is.
This
is
a
new
test
that
was
added
very
recently
by
the
the
runtime
team,
yeah.
A
B
B
B
Yes,
the
test
in
the
test
they
chose
to
try
and
make
a
request
to
a
10
dot
ip
address
which
was
denied.
Then
they
added
the
dynamic
asg
that
it
worked.
Then
they
removed
it
again
and
confirmed
that
it
didn't
work,
and
I
think
it
was
just
the
fact
that
they
chose
to
use
the
internal
address.
They're
like
okay,
well,
try
and
talk
to
power
control,
whether
there's
a
better
alternative.
I'm
not
sure.
C
B
A
Yeah,
okay,
do
we
have
an
issue.
B
Already
for
this,
or
I
don't
believe
there
is
one
no,
I
can,
I
can
create
one.
B
So
I
know
we
have
like:
we
have
the
cf
deployment
project,
but
then
we
also
have
the
working
group
project
and
there's
also
I
I
would
imagine
I
think
I
created
a
cats
project
when
I
was
before
the
working
group
was
was
formed.
I
think
there's
a
a
repo
specific
project
for
cats
as
well.
If
you
go
over
to
the
projects
tab,
I
think
yes.
A
A
D
I
would
maybe
talk
to
to
whom
about
that.
They
tried
a
lot
of
things
out
and
if
I
remember
correctly,
they
came
up
with
one
project
working
group
and.
C
D
Have
an
automation
where
all
the
issues
and
prs
of
all
the
different
projects
gets
an
aggregated,
because
that
that
makes
it
easier
to
get
an
overview,
and
otherwise
you
would
need
to
look
through
20.
Okay.
Our
working
group
here
is
pretty
small,
but
the
infrastructure
working
group
or
the
interfaces
working
group.
They
have
50
plus
projects
and
then
it
doesn't.
D
B
D
A
Yeah,
I
think
using
issues
to
to
track
tasks
and
stories
is
a
good
idea.
I
already
tried
this
automation
from
ruben
with
the
better
projects
and
I
think
I
did
not
manage
to
aggregate
all
the
single
projects
we
have
into
one
I
can
try
again
and-
and
we,
as
many
already
mentioned,
the
disadvantage
is
that
you
cannot
create
notes
here.
So
you
have
to
use
then
issues
and
requests
to
track
every
piece
of
work.
Yeah.
A
Okay,
it
would
be
a
good
approach
and
if
the
other
working
groups
do
the
same,
of
course
we
would
follow
yeah,
okay,
but
that's
yeah
other
topic.
Okay,
so
thank
you.
If
you
create
an
issue
for
this
blocked
cats,
then
then
we
can
respond
in
the
channel
and
then
this
is
done.
A
A
Good
fine,
then
we
have
only
one
last
topic
on
the
agenda
bound
times
on
cf
update.
A
For
the
latest
sea
of
deployment
releases,
we
notice
that
there
may
be
down
times
in
safe
push.
So
if
you
update
to
what
was
it
18,
where
the
the
log
cache
is
split
from
doppler,
there
may
be
short
down
times,
because
the
start's
endpoint
is
not
reachable.
A
A
They
also
noticed
that
stat
server
is
temporarily
unavailable.
While
this
update
to
see
if
deployment
18
is
running.
So
I
I
don't
know
if
you,
if
you
you're,
not
testing
a
cf
push
availability.
A
A
B
So
if
you
open
up
the
task
for
bosch
deploy,
if
you
scroll
down
it's
all
near
the
bottom
boss
deploy
cf
develop,
then,
if
you
scroll
past
all
of
the
upload
releases
output,
you'll
see,
we
have
a
tool
here,
this
yeah
it's
called
up
timer
and
this
basically
tests
app
availability.
So
we
have
a.
We
push
an
app
before
the
upgrade
starts
and
it
gets
polled
consistently
during
the
upgrade
to
make
sure
that
the
app
is
available.
B
For
how
many
failures
we
can
see,
and
so
if
you
go
all
the
way
to
the
bottom
after
the
upgrade
is.
B
Oops
then,
interspersed
in
here
you
will
start
to
see
failures
so
like
this
is
recent
log
failures
and
then,
if
you
go
all
the
way
to
the
bottom,
you'll
see
a
summary
of
how
many
failures
there
were
so
up
a
little
bit.
This
is
like
the
cleanup
output,
so
there
should
be
a
table
right
here
here
that
green
section.
It
will
tell
you
how
many
requests
it
made
during
the
upgrade
and
how
many
failed.
B
Now,
in
this
case,
I
think
this
is
a
fairly
small
upgrade,
but
we
could
go
back
and
find
which
let
me
see
if
I
can
find
it,
it
will
be
a
ways
back
now,
but
I
can
try
and
find
the
upgrade
for
18
and
we
can
see
what
those
numbers
were
like.
A
A
D
B
B
So
if
we
look
at
that
same
section
here,
this
was
pretty
much
the
entirety
of
so
that
that
went
from
17.1,
which
means
that
it
was
upgrading
to
what
became
18..
So
if
we
go
down
and
look
at
bosch,
deploy
cf
develop
again,
which
is
near
the
bottom
and
just
scroll
all
the
way
to
the
bottom.
So
we
saw
zero
app
availability
failures.
We
saw
one
push:
failure.
B
B
C
A
C
B
B
It's
just
never
been,
I
guess
it's
just
never
been
a
priority
for
the
for
release.
Integration.
D
Let's
assume
we
would
have
found
out
here
that
we
have
problems
with
the
lock
accessibility.
B
Okay
right,
if
any
of
these
metrics,
so
you
can
see
on
the
right,
it's
got
the
past
percentages
and
we
then
have
actually,
I
think
the
thresholds
are
based
on
on
number
of
failures
rather
than
percentages,
so
it
says,
like
one
failed
out
of
ten,
so
if
there
were
more
than
ten
push
failures,
then
this
job
would
go
red
and
it
wouldn't
allow
the
upgrade,
and
then
we
would
go
back
to
the
we'd
look
at
okay.
B
B
D
We
ran,
we
ran
into
the
same
trap
with
our
standard
tests
and
yeah.
So,
but
that's
maybe
something
we
could
at
least
put
on
on
the
task
list
to
to
think
about
whether
we
want
to
have
one
test
that
runs
on
a
let's
say:
landscapes
that
is
scaled
out.
Don't
know
what
that
means
for
the
costs,
so
we
should
have
here
then
a
temporary
landscape.
That
is
not
a
long
living.
One.
B
Yeah,
I
mean
all
of
these
environments
other
than
cats
are
temporary.
That's
why
it
does
the
deploy
the
latest
release
and
then
deploy,
develop
and
then
tear
everything
down
again
so
yeah.
We
could
definitely
do
something
as
long
as
we
have
a
gcp
project
or
aws
or
whatever
we
choose
to
run
it.
That
has
the
capacity
to
to
spin
up
more
vms.
You
know
I,
I
don't
think
it
would
be
hard
to
set
up.
D
C
A
D
Find
a
workaround
to
minimize,
let's
say
see
the
effect
on
customers.
B
A
A
B
A
D
I
have
one
more
question.
C
D
And
then
open
box
and
let's
say
for
the
smoke
tests,
the
typical
findings
is
your
golang
version
is
slightly
too
old
and
I
think
we
we
had
recently
a
conversation
about
that
and
my
question
is
now:
is
the
patch
version
updated
automatically
for
smoke
tests.
B
Any
change
there
would
completely
flow
through
and
end
up
with
a
a
release
being
cut
if
everything
goes
green.
B
So
all
that
we
need
to
worry
about,
there
is
just
merging
prs
and
the
changes
will
just
flow
directly
through
and
kind
of
release
at
the
end
of
it.
D
And
the
next
release
would
then
automatically
include
golang.
117
9,
I
think,
is
the
latest
one,
or
at
least
that's
one.
I
don't.
B
So
there
is
a
process
here
for
the
the
packages
it
uses,
the
the
bosch
vendor
package
process.
B
B
D
D
B
I
think
it
would
just
be
a
case
of
it
needing
the
the
blob
store
credentials.
B
Yes,
it
must
do
it,
it
must
still
be
updated
uploaded
to
the
web
store
so
yeah
as
long
as
we.
If
we
put
those
in
ci
yeah,
let's
grab
the
s2
key
for
the
for
the
the
blobs.
D
D
D
Or
something
like
that?
Okay-
maybe
that's
another,
just
yeah
it
has
to
plan-
I
mean
it's,
it's
not
really
urgent.
We
can
do
it
also.
You
know
manually
if
we
get
the
s3
key
somewhere
and
then,
but
that
would
be
something
that
comes
up
basically
after
every
security
scan
and
you
can
do
whatever
you
want.
D
A
Yeah,
so
that
would
be
everything
on
the
agenda
for
now.
Stefan,
you
have
more
no
questions,
suggestions.
A
If
there
is
nothing
else,
I
will
compile
meeting
notes
tomorrow
and
yeah
the
next
meeting.