►
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
So
welcome
everyone
long
time,
no
see
I've
been
back
to
office
since
last
week
and
I
think
not
too
much
happened
in
the
past
meetings:
yeah,
not
enough
attendees
meeting
councils.
A
A
One
topic
is
this:
yeah,
quite
old
short
certificates
and
signing
pull
requests.
So
after
some
discussions
about
the.
A
A
Rsa
signing
key,
we
use
a
certificate
for
DOT
signing
which,
and
we
we
do
not
change
the
type
of
the
existing
credential,
because
this
would
likely
cause
a
lot
of
trouble
depending
on
where
Posh
stores
the
credential.
A
So
now
the
solution
looks
like
this:
we,
we
add
a
new
credential
type
certificate
and
configure
this
here
as
key
two
and
then
you
can
switch
over
from
the
the
old
credential
to
the
new
one
and
yeah.
There
are
unit
tests
for
that
and
your
okay
from
my
side,
this
just
looks
now:
okay,.
A
A
B
A
Fine,
then
Marcus
trailer
will
be
pleased
to
hear
that
next
thing
is:
oh
yeah.
We
had
a
lot
of
trouble
with
the
infrastructure
pipelines
because
they
all
were
interrupted
by
a
concourse
problem
in
the
yeah
worst
possible
moments.
That
is
when
you
start
to
Bubble
up
an
environment
and
the
state
changes
and
then
real
estate
is
not
pushed
so
State
loss.
A
That's
why
Carson
and
me
had
to
repair
most
of
the
environments,
but
I
think
we
fixed
this
all
in
the
meantime.
By
now
so
yeah
I
had
to
delete
the
experimental
environment
in
AWS
and
recreate
it
and
all
that.
But
yes,
I
think
this
is
fixed.
Now,
I,
don't
know
what
exactly
happened,
though
rocket
disappeared.
A
B
A
A
The
update
release
job
has
been
was
failing
for
quite
some
time,
but
this
was
due
to
a
stem
cell
issue.
There
was
already
an
open:
well,
not
this
one
that
was
already
an
open
ticket
on
the
Bosch
release
for
that.
A
A
The
only
problem
was
now
that
we
do
not
have
a
validation
for
a
postgres
update.
We
use
postgres
only
in
the
from
scratch
scenario.
A
So
if
I'm
not
wrong,
this
is
the
only
one
that
uses
postgres
risk
uses
MySQL.
So
there
was
no
test
of
the
update,
so
I
did
this
manually,
so
I
just
deployed
used
to
Dev
environment
deployed
the
latest,
released
sea
of
deployment
with
the
old
postgres
and
then
hacked
this
job
here,
which
takes
the
release
of
deployment
and
deploys
the
develop
branch
on
top.
A
So,
in
other
words,
it
updates
postgres
and
a
few
other
things
and
also
measures
the
down
times,
and
there
are
a
few
more
down
times,
of
course,
because
the
database
is
gone
for
some
time,
but
it's
it's
nothing
too
severe.
The
update
itself
also
succeeds.
So
there
are
no
problems,
so
we
can
I
would
say,
release
this
as
a
minor
release,
CF
deployment.
A
If
there
are
no
objections,
because
we
should
print
a
hint
attention,
new
postgres
update
may
take
some
time
depending
on
your
database
size,
but
in
general,
we've
done
the
same
here
at
sap,
but
with
an
RDS
managed
postgres.
That
was
also
a
major
version
update
and
to
you
know,.
B
Thanks
for
doing
this
quick
question,
it
looks
like
there's
a
failure,
one
on
the
screen
that
you're
looking
at
for
up
timer.
A
A
Don't
know
how
exactly
the
conquest
task
behaves,
yeah
well,
Indeed,
app
push
ability
is
is
has
not
the
highest
success
rate.
This
is
true,
I,
don't
know.
Four
attempts
failed
in
here.
Five.
A
We
should
point
this
out
in
the
release
notes,
but,
on
the
other
hand,
I
guess
there
is
not
too
much
we
can
do
about
this.
A
Well,
yes,
there
are
failures,
definitely
in
in.
If
you,
if
you
check
the
upgrade
scenario,
you
will
always
have
serious.
There
are
always
failures.
B
B
A
Thresholds
are
a
bit
questionables
or
here
is
I,
think
an
upgrade,
but
there's
almost
nothing
in,
but
still
you
see
a
few
failed
attempts
to
push
and
fetch
locks
can
be
because
of
a
single
node
database.
Single
node
blob
store,
and
this
can
have
a
lot
of
root
causes.
But
I
guess
for
a
major
database
update.
A
You
can
point
this
out
in
the
release.
Notes
sure
expects
somewhat
longer
downtime
if
you
use
the
push,
managed
postgres
I.
B
B
A
A
Okay,
then
it
will
be
a
minor
release
with
a
notice
and
then
then
that's
it.
A
Okay,
so
we
just
created
a
release,
I
think
the
beginning
of
the
week
and
there's
already
a
lot
of
new
changes
coming
in
all
the
build
packs,
which
is
good.
Of
course,
I
mean
this
is
good
to
see
that
everything
is
under
heavy
development.
A
Good,
so
that
would
be
all
topics
from
my
side,
any
other
business.
B
Yes,
good
I
have
a
PR
that
I
think
is
ready
to
go
to
PR's,
actually
to
put
the
cfd
version
into
the
copy,
I
guess
to
give
Cappy
knowledge
of
the
cfd
version
so
that
if
you
curl
the
V2
in
lower
V3
info
endpoints,
you
get
you
get
the
the
actual
version
instead
of
just
empty.
B
A
A
B
B
I've
tested
it
locally
to
some
extent,
but
I
haven't
run
it
so
I'm,
not
sure
that
it's
gonna
work,
but
it
should
work.
It's
this
one
yeah.
B
A
A
B
Could
it
doesn't
right
now,
but
it
could,
because
we
do
have
a
version
colon
and
if
the
in
in
releases
and
stuff
and
if
the
releases
are
actually
if,
if
any
of
those
releases
happen
to
be
just
one
number,
yeah.
A
B
I
hadn't
thought
about
it
too.
Hard
I
just
checked
that
it
didn't
do
anything
right
now,
but
I
would
love
ideas.
If
you
got
one
for
how
we
could
prevent
that
in
the
future,
we
could
use
Bosch
and
triple
interpolation.
That's
always
an
option
rather
than
this
weird
like
setting
stuff.
B
Yeah,
like
I,
mean
I
could
make
it
more
Super
brittle
and
check
for,
like
Name
colon
CF
deployments.
Return
like
return
signal,
build
like
white
space,
builds
thing,
white
space
version
thing
right,
and
that
would
only
match
that
most
likely,
but.
A
B
A
I
mean
I
mean
there
are
100
ways
to
do
this.
A
bit
more
crude
way
would
be
adding
a
unique
placeholder
here
at
Dollar
copy,
build
version
and
search
and
replace
this
one.
True.
B
A
The
next
it
is
unlikely
that
this
matches
anything
else
than
what
what's
what
is
intended,
but.
A
A
B
The
other
thing
I
wanted
to
call
out
was
that
the
I
added,
a
I
added
to
the
CF
smoke
tests,
release
a
new
task
that
bumps
the
windows
cfcli
blob.
Whenever
a
new
cfcli
release
comes
out
What.
B
Would
this
there's
a
weird
thing
where,
like
we,
obviously
the
CF
smoke
tests
release
packages,
the
packages,
the
cfcli
release
from
the
Bosch
package,
cfcli
release
repo,
but
they
that
repo
does
not
include
the
windows
CF
CLI,
for
whatever
reason
I
tried
to
make
an
issue
for
them
to
add
it,
and
they
said
that
it's
potentially
changing
hands
to
another
working
group,
so
they're
not
going
to
add
it
anytime
soon.
B
Okay,
the
last
time
this
came
up
was
when
we
we
had
bumped
see
at
the
CFC
live
within
smoke
tests
released
to
eight
and
then
some
months
later
realized
that
the
windows
one
was
on
B6,
and
that
was
like
a
year
ago.
A
B
Two
years
ago,
and
since
then
it's
just
been
a
best
effort
whenever
we,
whenever
we
remember
we'll,
just
go
and
bump
The
Blob
in
this
in
this
release,
to
whatever
the
latest
version
is,
and
it
got
to
the
point
where
I
felt
like
it
was
better
to
just
automate
it
so
I
added
a
task
to
runtime
CI
Dave
approved
it
so
I
merged
it
and
and
to
hear
which
Dave
also
approved
so
I
merged
it,
which
should
keep
an
eye
on
the
cfcli
releases
for
V8
and
then
automatically
bump.
B
The
Blob
inside
of
the
inside
of
the
yeah
inside
of
this
release.
Blobs
are
in
blobs,
are
in
config.
B
The
last
time
someone
complained
was
when
we
forgot
to
bump
it
to
V8,
and
it
was
stuck
on
V6
for
a
long
time
after
the
other
after
the
non.
That's
the
Linux
cfcli,
but
the
complaint
tint
I
guess
would
be
me
because
I
had
to
manually
bump
it
every
now
and
then,
which
was
annoying
so
automated
if
it
works,
but
just
another
potential
Vector
for
things
to
break
down
as
well.
It
would
be
really
nice
if
they
added
it
to
the
to
the
Bosch
package
release,
but
hadn't
happened
yet.
B
Oh
yeah,
but
the
the
main
reason
I
wanted
to
bring
this
up
was
because
the
CF
smoke
test
release
does
not
run
Windows
tests
right
now,
so
I'm
fairly
sure
that
this
works.
But
if
we
start
seeing
Windows
failures
in
Smoke
tests
know
that
it
could
be
related
to
some
weirdness
with
that
automation
because
it
hasn't
been
running
for
very
long,
possibly
mucking
up.