►
From YouTube: App Runtime Deployments Working Group [January 12, 2023]
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
And
then,
let's
get
started,
I
don't
have
too
much
on
the
agenda
just
one.
Hopefully,
file
left
over
from
the
infrastructure
migrations.
So
the
Hermione
environment
is
now
fully
migrated
to
a
new
AWS
account
that
works
perfectly
fine,
there's
just
the
green
grass
environment,
which
still
still
runs
on
a
VMware
account.
A
It
has
no
infrastructure
pipeline
for
setting
up
and
destroying
it's
mainly
used
for
the
stem
cell
validation.
So
can
can
we
just
migrate
this
and
create
a
pipeline
or
Carson?
Was
there
a
reason
that
there
is
no
pipeline.
B
I
think
because
of
the
very
specialized
use
case,
we
actually
recreate
it
when
we
use
it
so
I,
don't
think
it
exists
most
of
the
time
if
I
remember
correctly,
and
then
during
the
update
Linux
stem
cell,
we
would
spin
up
Green
Grass
and
then
use
it.
Okay,
yes,
we
can
migrate.
It
I
think
we
didn't
previously,
because
there
was
a
little
more
confusion
about
how
we
would
do
that.
It
wasn't
as
straightforward
as
some
of
the
other
ones,
but
we
should
we
should
go
and
migrate.
It.
A
I
hope
so
I
have
not
found
anything
else.
So
do
you
want
to
do
it
or
should
I.
A
Okay,
it
shouldn't
be
a
big
problem:
just
creating
a
new
service,
account,
okay,
good
and
then
quickly.
This
validates
certificate
job,
it's
still
red
because
of
this
Concourse
environment.
B
Yeah,
that's
the
old,
Concourse
right
I!
Think
that
there's
one
thing
that
we've
just
kept
saying
that
we'll
move
over
and
haven't
moved
over
and
that's
the
jump
box
pipeline
once
we
move
that
over
then
I
think
we
can
go
ahead
and
delete
that
okay,
that
environment
well,
a
stop
gap
for
now
might
be
to
have
the
job
ignore
that
specific
set
of
credentials
so
that
this
doesn't
go
red
and
then
Dave
and
I
can
deal
with
the
the
old
CI.
B
A
B
One
one
thing
to
know
about
this:
credential
job
is
that
I,
don't
think.
I
I
took
an
action
item
for
myself
at
one
point.
After
doing
a
little
research
to
make
it
actually
watch
the
load.
Balancer
certs,
all
of
the
I
think
most
of
the
new
environments
should
have
load
balancer
certs
with
around
three
years
of
of
time
to
live
and
I.
Don't
think
this
job
watches
those
certs
I
think
it
only
watches
the
Bosch
created
certs.
B
A
B
A
Okay,
I
can
also
check
this
if
it's
Watcher
certificates
and
yes,
it
should
do
so
because
Hermione
expired
well.
This
is
a
special
case,
but
it
expired
already
twice
and
okay
good,
so
we
ignore
Concourse
and
check
if
it
watches
the
load
balancer
certificates
yeah.
Apart
from
that
to
see
if
Linux
fs4
progress
is
quite
good.
So
there
are
just
these
four
build
packs
left
where
we
wait
for
fs4
stack
support.
C
Yeah
I
made
a
PR
to
cfx
substance
tests
with
regards
to
our
current
syslog
tests.
The
way
we're
currently
testing
syslog
relies
on
some
bit
of
internal
routing.
That
I
think
is
kind
of
irrelevant
to
whether
or
not
abscess
log
drains
work,
and
my
Hope
was
what
it
ends
up
doing
is
depending
on
what
what
settings
you
have
turned
on
within
your
CF.
C
It
actually
tests
different
parts
of
CFS
log
trains,
and
so
my
Hope
was
that
we
could
reorient
the
tests
to
be
focused
on
to
use
TCP
routing
to
test
syslog
drains,
and
then
we
can
also,
we
part
of
the
the
point
of
that
was
to
build
on
that
and
then
add
tests
for
our
new
Mutual
TLS
ability
within
syslog
drains.
C
C
I
think
they
both
are
not
that
big
I
want
to
say
this.
One
is
just
the
first
step,
which
is
switching
over
to
sorry
sorry,
switching
over
to
using
TCP
routes
to
receive
traffic,
and
then
we
added
another
comment
on
that
which
is.
C
What
is
this?
We
broke:
something:
okay,
yes,
adding
a
mtls
test
as
a
second
test,
and
that
sets
up
Mutual,
TLS
credentials
and
then
uses
that
to
accept
the
connection.
A
C
As
part
of
that,
we
did
change
the
default
of
include
TCP
to
false
and
that's
because
not
all
environment
support,
TCP,
load,
balancing.
A
Was
the
default
CF
deployment
does
support,
it
does
yeah.
So
that
means
we
could
switch
that
on
for
all
test
environments.
C
I,
just
the
app
syslog
TCP
Suite.
A
A
Okay,
good.
That
means
we
should
yeah,
maybe
create
a
new
Cuts
release
and
then
switch
that
on
in
our
environments.
Right.
B
A
B
A
C
This,
if,
if
we
accepted
this
PR,
as
is
it
would
switch
it
from
on
by
default
to
off
by
default.
So
fine,
if,
if
you
wanted
to
include
it,
then
you
could
include
true,
but
if
we
wanted
to
change
that,
then
yes,
you
would
have.
If,
if
your
environments
didn't
support,
TCP
routing
you'd
have
to
turn
that
Suite
off
okay.
C
B
A
C
I
I
think
they're.
Unless
we
wanted
to
change
the
the
default
I
think
they're
ready
for
my
perspective,
yeah.
A
D
I
think
it
has
no
problem
on
our
side
because
we
do
not
run
cats.
Let's
say
out
of
the
box.
We
have
it
more
as
an
additional
development
thing
here
and
there
we
can
configure
it.
That's
no,
no
problem.
C
Okay,
I
I
would
like
to
break
up.
I
do
believe
the
default
TCP
tests
are
off
no
I'm
wrong,
they're
on
by
default.
A
So
we
have
I
think
this
is
the
upgrade
scenario,
and
here,
for
example,
we
use
this
cat's.
Configuration
routing
through
TCP
is
also
included
yeah,
so
I
guess
this
is
true
for
most
environments.
A
A
D
I
have
one
more
small
thing
to
to
raise.
We
have.
Meanwhile,
you
have
Linux
fs4
support
as
experimental
Ops
file
and
your
build
pack
coverage
is
increasing
shouldn't.
We
somehow
promote
those
Ops
files
to
regular,
because
it's
I
mean
the
Sea
of
Linux.
Fs4
is
a
real
release.
It's
not
experimental
anymore
and
the
build
text
once
released.
They
are
also
yeah
productively
usable.
So
why
not
also
then
at
the
the
Ops
file?
That's
regular
ones.
A
D
And
then
say:
okay
in
in
April,
it
disappears
anyway
and
then
yeah
the
experiment
is
finished
and
at
a
certain
point
in
time
we
can
then
merge
this
additional
experimental
Ops
file
that
can
be
used
to
switch
the
Linux
fs4
as
it
is
deck,
because
that's
the
latest,
when
all
the
build
packs
are
there
yeah.
A
D
A
D
Updates
anymore
I
guess
we
have
to
remove
it
right
from
from
CF
deployment
and
if
somebody
yeah,
VMware
or
sap
wants
to
keep
it
longer,
then
we
have
to
deal
ourselves
with
additional
Ops
files.
I
mean
it's:
it's
yeah
no
rocket
science.
A
A
Yeah,
well,
it's
a
little
bit
of
effort
because
we
have
to
update
the
validation,
jobs
and
think
about
if
this
is
one
or
two
steps
to
roll
out
smoothly.
B
We
haven't
exactly
been
following
it
because
there's
been
some
yeah
well,
not
not
necessarily
for
this
specific
case,
there's
been
just
a
few
Ops
files
that
have
either
been
promoted
or
deprecated,
which
theoretically
we
could
have
removed
at
some
point
which
we've
never
removed,
and
in
the
past
we've
we've
promoted
stuff
by
just
directly
pulling
it
out
and
pushing
it
up
and
then
you're
supposed
to
sim
link.
It
apparently
I,
just
just
randomly,
walked
across
this
file.
B
B
D
We
also
need
to
think
if
we
say
we
could
even
deliver
CF
Linux
fs4,
let's
say
in
April
as
the
standard
stuff
and
CF
deployment.
If
all
the
build
packs
are
there
and
then
offer
additional
Ops
files
to
upload
CF
Linux
fs3.
In
addition,
that's
also
a
possible
possible
steps
that
we
could
do.
Yeah
yeah,
we
I
think
we
should
should
come
up
with
a
plan
and
then
also
announce
it
when
we
want
to
do
it,
it's
all
vague
as
long
as
we
don't
have
oil
build
packs.
But
let's.
A
A
A
I
think
I
would
leave
that
as
it
is
for
now
and
then
in
April
yeah,
a
little
bigger
change
and
the
major
release
and
yeah
provide
and
Ops
file
for
uploading,
fs3
yeah.
A
B
The
on
the
Ops
file
promotion
thing
I
had
made
a
PR
to
a
couple
PR's
to
like
try
and
clean
up
a
little
bit
of
our
Ops
file
situation,
I
actually
just
merged
one,
because
I
saw
that
David
approved
it
yeah.
So
there's
the
the
most
recently
closed
one
as
well,
and
this
one,
but
it's
just
this
one
all
it
does.
Is
it's
not
rebased
right
now,
so
what
I
did
in
that
PR
that
I
just
merged
is
I.
B
Had
the
pipeline
stop
using
the
experimental
files
that
have
been
promoted
and
start
using
the
promoted
files,
and
then
this
clearly
marks
the
Ops
files
that
hadn't
been
marked
for
as
promoted
or
deprecated
as
promoted
or
deprecated
in
the
readme
and
then
also
goes
and
deletes
the
Ops
files
that
had
been
marked
as
deprecated,
so
that
would
that
would
theoretically
lead
to
a
major
cut
of
cfd.
A
D
B
Is
all
this
yeah
that
just
says
we
should
stop
using
the
experimental
ones
in
favor
of
the
promoted
ones
in
the
pipeline?
This
one
would
be
a
major
I
think,
because
we
I
I
ripped
out
two
experimental
operations
files,
one
that
had
been
deprecated
and
doesn't
actually
have
any
stuff
in
it
anymore,
but
the
people
might
still
be
using.
B
And
the
other
that
says
that
was
promoted
and
said
said
it
was
promoted
in
the
readme
yeah.