►
From YouTube: Foundational Infrastructure Working Group [Feb 3, 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
Let's
go
over
the
open,
vrs
a
lot
got
merged
this
week,
so
nothing
is
spending
a
merge.
There
will
be
3.1
vr.
B
Correct
somehow
the
commits
we
have
now
too
many
commits
constantly
pushed
something
too
much,
but
yes,
yeah.
A
B
But
overall
the
change
looks
good
yeah
I
just
commit
message
was
a
little
bit.
I
wasn't
sure
whether
it
is
something
finished
or
not.
A
A
A
A
A
A
No,
why
yeah
I
mean
we
can
do
we
want
to
discuss
that
at
the
toc
level,
or
I
mean
it
is
basically
tech
that
right
at
this
point.
A
Like
I
have
to
discuss
these
types
of
things
at
the
toc
level,
I
mean,
if
you
feel
like
we
should,
I
think,
having
a
discussion
within
our
working
group.
I
think
should
be
enough
and
this
threat
can
can
work
for
that,
we're
just
removing
code
that
has
been
deprecated
for
a
long
time.
A
A
D
Could
this
go
to
the
bosch
fashion,
300
or
something
that
that
can
be
marketed
as
a
major
step
in
bush
lifetime?
Could
we
could
we
possibly
prepare
a
bosch
fashion,
300
like
to
be
a
kind
of
boss,
d3.
A
D
D
A
A
A
No,
we
don't
have
a
release
and-
and
it's
a
stack
depth
right.
So
it's
stuff-
that's
not
prioritized
that
high
this
works.
So
if
there's
no
other
stuff,
people
are
working
on
this.
But
it's
we
like
looking
at
the
suggested
timeline
here
right.
This
could
be
a
few
months,
and
that
would
mean
that
we
have
prs
that
are
open
for
a
long
time
and
then
we
will
be
difficult
to
get
them
merged
because
you
would
have
to
rebase
a
lot
so
yeah.
I
don't
prefer
that
solution.
A
D
Oh
okay,
I
see
so
we
match
that
one
and
then
I
apply
the
patch
to
complete
the
work
right.
We
let's
do
this
or
or.
A
Is
it
I
oh
yeah,
because
it's
a
separate
branch
right
I
mean
you
could
also
just
take
his
branch
and
then
work
from
there
right
and
yeah.
His
commits
will
still
be
in
your
branch
and
he
will
get
credit
that
way.
A
D
We
need
to
move
on
and
keep
keep
some
pace
yeah.
Otherwise,
it's
so
so
painful
when
the
pace
is
it's
not
yeah.
This
is
too
slow,
yeah,
okay,.
B
A
A
A
D
A
I
mean
the
agent
already
starts.
No,
this
is
not
about
money.
This
is
about.
We
we
have
the
this
is
the
net
stuff,
so
this
is
just
iptables
firewall
rules
to
block
traffic
to.
A
A
A
Yes,
so
we
would
still
need
to
have
the
brush
agent
in
some
c
group
when
it
starts,
so
that
would
be
the
stem
cell
change.
But
then
the
logic
around
getting
the
director
ip
address
or
the
net
ip
address
and
creating
the
ip
table
rules
would
be
in
the
agent,
because
I
think
that's
currently
what
the
problem,
the.
A
The
difficulty
is
around
getting
the
net
ip
address
because
there's
like
this
whole
in
this
process
around
like
getting
the
initial
config
and
then
bootstrapping,
the
setting
of
the
network
interfaces
to
be
able
to
talk
to
the
metadata,
endpoint
and
then
like
and
that's
all,
is
specific
right.
So
there's
different
ways
of
getting
that
initial
agent
configuration,
and
so
you
would
have
to
wait
in
your
script
till
the
agent
basically
has
gone
through
its
configuration
and
then
parse
that
agent
configuration.
A
D
As
has
a
different
ip
and
all
the
agents
have
to
update
to
reach
out
to
the
correct
ide,
we
had
not
sort
of
this
dependency.
I
I,
but
it
doesn't
no
link
with
what
we
are
talking
about,
but.
B
B
Already
commented
down,
there
is
a
comment
which
explains
this.
B
A
C
A
A
A
A
B
A
B
D
I
don't
have
his
email.
Oh
his
status
on
slack
is
out
sick.
D
Not
responsible,
we,
we
just
ask
them
to
test
and
validate
its
work.
It's
culture
working.
E
E
Oh,
I
I
thought
benjamin
yeah
benjamin
committed.
A
Contacts
on
this
this
had
to
do
with
the
is
native
disk
resize
right
so
that
feature
broke.
The
soft
layer
thing
check,
check,
check.
A
A
A
Yeah,
we
also
have
a
story
in
our
backlog
to
to
standardize
a
bit
of
our
release
processes,
because
our
pipelines
are
a
bit
different
like
some,
you
have
to
double
version
and
some
you
just
have
to
click
a
box
and
somewhere
you
will
get
a
release
and
then
some
there
isn't
so
I've
added
this
as
well.
There
apparently
there's
a
github
api
to
so
when
you,
when
you
create
a
release
through
the
github
ui,
there's
a
built-in
to
generate
release,
notes,
and
apparently
you
can
also
do
that
through
the
api.
E
Yeah,
the
only
thing
we
do
now
is
like
we
generate
everything
for
the
usn
logs.
We
just
put
it
in
there
and
yeah
and
the
bosch
agent
bumps
and
that's
what's
auto-generated,
and
then
there
is
a
manual
step
that
we
still
do
well,
I
still
do,
and
that
is
for
like
every
time
we
do
a
manual
I
will
check
which
pr's
we
have
included
from
the
master
into
the
to
the
to
the
line,
builds
and
then
manually
put
it
in
there,
and
I
don't
want
to
do
that
anymore.
A
Yeah,
this
maybe
would
solve
that.
I
don't
know,
I
don't
know
if
it's
gonna
be
too
noisy
right,
because
of
when,
when
you
generate
this
on
this
stem
cell
ci,
oh
no,
not
on
on
the
bosch
linux
stem
cell
builder,
you
might
get
all
those
commits
from
the
bomb
os
image
and
stuff
right.
There's
those
types
of
commits
a
lot
of
them.
E
A
I'm
just
gonna
sign
it
to
you
all
right
right.
This
is
this
is
more
of
a
reminder
right,
so
I'm
gonna
put
it
in.
A
D
D
Yeah,
the
the
workflow,
the
the
workflow
with
stop
vms
is
quite
has
many
corner
cases.
We.
We
have
hit
an
issue
today
where
an
insta
concourse
worker
instance
has
been
stopped
apparently,
and
the
deployment
has
been
updated,
removing
some
turbulence
release
and
then
bosch
thinks
the
turbulence
release
is
not
attached
to
any
deployment,
but
in
the
database
the
release
is
still
attached.
It's
still
linked
to
the
stopped
instance
that
has
not
been
updated
and
then,
when
you
try
to
delete
the
release
from
bosch
server,
you
get
a
foreign
key
violation.
D
E
D
E
D
That's
the
the
question
we
need
to
to
give
proper
semantics
to
this
and
and
something
consistent
with
the
rest.
D
We
need.
We
need
to
decide
what
what
is
the
most
consistent
thing
to
do
and
yeah,
and
then
the
implementation
will
follow.
Naturally,.
E
E
Basically
we
can.
We
can
then
do
what
felix
suggested.
Is
that,
like
a
force
like
it
stopped,
and
then
you
can
put
an
extra
flag,
do
it
anyways,
but
it's
really
hacky,
don't
know
if
you
really
want
that
here.
E
B
But
wheels
try
to
start
again
because
you
can
do.
I
think
there
are
errands
that
do
like
a
money
to
stop
monet.
Stop.
Okay,
yeah!
It
can
execute
money,
stop
yeah,
but.
B
E
A
Because
the
thing
is,
ideally,
you
wouldn't
also
need
to
handle
in
your
errands
some
the
case
where
bar
stop
has
not
been
executed
right,
so
you
would
need
to
error
out
on
the
errand
saying
you
know.
First,
one
bush
stop,
so
you
have
a
dependency
on
bus
stop,
so
why
why
they're
not
run
money?
Stop,
I
guess.
A
Yeah
I
mean
let's
just
wait
for
some
more.
E
C
D
But
what
happens
when
you're
running
around
on
a
stopped
instance?
Previously,
it
restarts
the
incense
a
lot.
B
Yeah,
it
was
to
avoid
use
cases
which
you
described.
The
benjamin
to
have
from
vmware
haven't
been
updated.
You
know
you
try
to
execute
the
errand
and
you
know
don't
know,
what's
what
is
the
state?
Yes,
yeah.
A
E
C
D
Yeah,
the
the
main
problem
is
that
a
stopped
instance.
You
cannot
test
an
update
on
it.
So
do
you
consider
it
is
supposed
to
be
updated
and
run
into
the
problems
into
any
problems
when
it
is
restarted,
or
do
you
do
like
currently
saying?
No,
it
just
keeps
as
is,
and
it's
not
updated,
so
we
cannot,
we
don't
consider
it
has
been
updated.
D
I
think
I
I
have
edited
his
description
that
yeah
so
that
we
can
read
the
logs
and
stuff
otherwise
yeah
there.
We
are
missing
backticks
to
find
that
properly.
D
So
it
seems
that
the
generated
sql
query
is
in
valid
because
there
is
an
end
cloud
property
is
nothing
else
inside
the
parenthesis.
We
are
missing
the
rest,
but
I
remember
seeing
in
the
logs
that
you
have
parentheses
cloud
properties
and
that's
it
nothing
not
equals
anything.
E
E
A
If
it's
a
subnet
right,
then
you
have
it's
an
array
and
then
it
has
a
range
and
it
this
is.
These
are
the
azs
in
which
you
can
use
that
subnet.
So
then
it
makes
sense.
A
A
D
Yeah,
it's
not
the
problem
here.
E
A
D
D
E
B
E
D
B
C
D
F
A
A
A
D
Oh,
I
I
sorry
I
I
was
to
try
to
reproduce
this
and
I
did
not
find
the
time
to
do
this,
but
I
should
definitely
try
so
yeah.
It's
still
relevant.
C
A
A
Yeah,
I'm
looking
at
last
activity
column,
so
everything
since
the
last
time
that
we
talked
so
this
one,
oh
yeah,
that
one
was
already
we
talked
about
and
I
think
the
rest
is
yeah.
We,
oh
this
one.
Oh
this
is
dumb
yeah,
so
that's
it.
That
was
everything
that
was
cool
any
reason.
A
A
Yeah
I
mean,
I
think
I
left
some
comments
in
the
document
and
I
think
the
next
step
would
be
after
we
have
that
discussion
in
the
document.
We
would
create
rfc
all
right
to
discuss
other
plc.
So
if
that
there
is
currently
a
document
that
benjamin
started,
where
we're
talking
about
deprecating,
monet
or
removing
wallets.
D
Yeah,
okay,
all
right!
I
do
this.