►
From YouTube: Foundational Infrastructure Working Group [May 25, 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).
B
A
Because
why
not
right?
There
is
all
sorts
of
new
features
that
you
nowadays
can
enable.
A
So
there's
Dependable
alerts,
code
scanning
alerts
and
secret
scanning
alerts,
I've
enabled
the
three
of
those.
So
the
people
that
I
don't
know
who
actually
can
see
this
I,
don't
think
it's
public,
so
I
think
you
need
at
least
some
permissions
on
this
repo
to
be
able
to
see
it,
but
yeah
some
Dependable
stuff
all
right
found
things
in
some
sort
of
The
Blob
store
load
tests.
C
C
A
Really
a
problem
more
of
a
it's
nice
like
once,
we
have
three
hours
these
things,
because
we
should
at
least
try
if
these
keys
are
disabled,
I
assume
so
because
they
are
really
old,
but
let's,
let's
make
sure
that
they
are
but
then
like
it
will.
Also,
if
you
try
to
push
something
with
a
key
in
it,
it
will
prevent
you
from
pushing.
So
that
is
actually
quite
a
nice
feature.
A
A
C
No,
it
was
yesterday,
and
most
of
them
are.
C
Yeah-
and
there
was
I
created
in,
is
an
issue
in
the
from
the
eagle
rainbow
dude
and
he
yeah
and
and
yeah.
D
B
C
Yeah
I
haven't
tested
it
see
if
he
says
it's
tested
and
I
believe
him.
C
A
One
this
one
I
had
already
reviewed
and
I
had
some
command
comments
like
I
was
wondering
like
why,
if
we're
gonna
do
this,
why
not
just
do
it
like
the.
A
C
Not
specific,
but.
A
A
So
that's
why
you
can
also
sometimes
set
it
through
Cloud
properties
that
that
is
a
bit
of
a
mess,
but
like
I,
don't
see
a
reason
why
the
director,
like
the
tags
that
are
supported
through
the
Manifest
right,
the
top
level
ones
and
the-
and
this
is
about,
like
instance,
level
ones,
I
think
it
makes
sense
to
do
it
properly,
like
there's
no
reason
to
take
a
shortcut
and
try
to
put
it
in
Bosch
EMV
right
conceptually,
it
doesn't
make
sense,
maybe
it's
the
easiest
way
to
implement
it.
A
That
is
the
reason
I
could
see
to
to
do
it
this
way,
but
yeah
I
mean
like
if
we
do
it
this
way,
it
will
stay
that
way
right.
We
never
have
a
chance
to
yeah.
A
A
Yeah
I,
it
is
probably
a
bit
more
work
right
and
because
that,
like
the
the
instance
planner
is
or
the
default
a
deployment
plan
code,
it's
it's
quite
complex,
but
I
think
it
would
be
the
the
better
place
to
put
it.
So
we.
A
No,
no
I'm
not
going
to
do
it
myself,
no
yeah
I
don't
have
the
time,
but
and
if
people
really
get
stuck,
maybe
it's
also
a
nice,
maybe
reaching
out
to
Joseph
Palermo
would
be
a
good
opportunity
to
do
some
bearing
or
something
like
a
bit
when
you
really
get
stuck
because
I
think
he
is
the
most
familiar
with
that.
A
That
piece
of
the
director
of
code
base
okay,
yeah,
and
he
is
also
a
tech
lead
in
the
open
source
like
Bayonne
and
I,
so
he's
just
at
a
different
time
zone
so
makes
it
a
bit
different,
more
difficult.
So
there
there
is
that
up
that
option.
A
I
believe
we
had
a
discussion
about
this
right.
Yes,
yes,.
A
I
think
the
the
place
to
like
this
should
so
on
the
agent
side.
We
we
should
allow
setting
a
regex
or
something
for
getting
rid
of
a
prefix,
and
then
we
can
inject
that
through
the
agent
settings
into
a
different
stem
cell
targets
right
different,
so
we
basically
make
it
is
basically
specific
because
I
understand
the
problem,
but,
like
my
main
concern
is
with
like
the
fact
that
this
works
when
you're
splitting
it
on
a
specific
character
that
might
change
in
the
future
right
and
like
the
yeah
it
is
it
is.
A
A
But
that
is
an
is
specific
thing
right.
It
happens
to
be
that
on
both
Ali
cloud
and
AWS
splitting
on
a
dash
or
something
fixed,
it
was
it
oh,
no,
it
was
an
underscore.
Was
it
an
underscore?
That
was
actually
the
thing.
No.
E
It's
it's
always
the
prefix
with.
A
Yeah,
but
that
like
it
is
it
took
us
a
while
to
get
to
like
the
actual
problem,
or
at
least
me
it
took
me
a
while
right,
because
the
the
code
that's
written,
is
really
abstract.
Like
you,
you
don't
know
a
lot
of
like
what
what
you
actually
encounter
on
the
in
the
is,
and
then
you
have
to
go
through
this
issue
to
find
that
so
yeah.
That
will
be
a
bit
of
a
maintenance
issue
in
the
future.
So
by
making
it
is
specific
right.
A
A
And
it's
something
different
on
AWS
right.
D
A
Yeah
I
think
this
is
actually
the
thing
you
want
to
get
rid
of
right.
You
want
to
get
rid
of
the
D
Dash
that
that
is
the
thing
right.
So
yeah.
B
A
Want
to
get
rid
of
so
that
the
Globe
will
match
these
things.
Oh.
A
All
right
yeah
so,
but
allowing
flexibility.
So
if
we
can
Define
a
regex
per
ayas
and
put
that
in
this
file
right,
so
you
have
Porsche
Linux
stem
cell
Builder
and
then
you
have
say
something
agent,
settings,
I,
think
agents.
A
So,
for
example,
AWS
agent
settings
apply
right
here.
You
already
have
the
device
pod
resolution
type
all
right.
So
if
we
all
would
also
have
a
property
here
to
define
the
device
part
resolution,
prefix
or
something
or
I,
don't
know
a
better
name
but
like
something
right
and
then
that
could
in
this
case
just
be
Vol
Dash
right.
B
A
I
mean
the
the
coaching
agents
would
it
would
be
pretty
like
similar,
or
it's
more
about
like
where
you
put
the
responsibility
right
like
the
agent
at
that
point,
will
will
still
be
like
is
agnostic
in
that
sense,
and
the
is
specific
things
will
be
in
the
stem
cell
Builder.
That
would
make
me
way
more
comfortable
with
this
change.
A
To
explain
that
yeah,
but
I
like
we
have
many
things
to
go
over,
so
I
will
do
that
afterwards.
Okay,.
A
Yeah
I
think
that
is
because,
in
the
end,
it
was
also
like
it
is
probably
something
that
was
also
has
an
OS
aspect
to
it,
maybe
like.
If
you
update
to
a
new
version
of
Ubuntu,
these
names
might
change
again.
A
E
D
E
A
I,
don't
know,
let's
see
yeah
the
thing
is
it's
such
a
specific
thing
that
you
need
to
wrap
your
head
around
like
so
by
the
time
we
find
another
one
of
these
issues.
We
probably
forget
that
we
have
documentation
about
it.
So
yeah
maybe
add
a
note
somewhere
in
the
CPI
I,
don't
even
know
what
the
appropriate
place
for
that
would
be.
We
don't
have
docs
on
the
stem
cell
Builder.
D
A
C
A
D
I
would
I
would.
E
E
Is
maybe
you
can
add
something
quite
quick,
open
source
we
support
by?
We
don't
test
the
open
store
stuff.
The
openstack
stuff
should
be
at
the
comment
that
this
is
fine,
but
we
cannot
validate
it.
A
Is
using
openstack
and
bubble.
A
B
B
A
B
B
A
A
Yeah
he
wants
to
install
some
sort
of
security
agent,
which
is
sometimes
taking
this
hostage.
So
he
wants
to
have
a
dream
to
be
able
to
resolve
that.
But
if
you're
gonna
go
through
that
level
of
effort,
you
better
off
creating
brush
release.
C
C
A
Yeah
but
I
don't
like
we
have
the
the
pre-start,
which
is
the.
C
A
More
space
right,
I
know,
but
we
don't
like
at
some
point.
It
starts
to
be
it's
better
for
you
to
create
a
release
instead
of
using
OS
confax
right.
So
that
was
the
discussion
we
had
last
time.
Okay,
where
we
said
like
when
they
asked
for
a
pre-stock
stop
right.
We
don't
want
to
expose
all
the
lifecycle
hooks,
because
it
is
basically
a
really
like
a
big
goat
smell
right.
A
If
you
start
having
a
need
for
those
types
of
things,
then
you
that's
probably
a
sign
that
you
should
start
creating
a
brush
release
because
it's
not
worthy
the
effort
of
just
hacking.
All
your
shell
scripts
in
your
manifest
I
mean
it
would
play
be
way
easier
to
just
create
a
bush
release
and
also
I
mean
like
Benjamin,
is
from
like
any
nines.
He
maintains
some
Bosch
releases
in
the
open
source,
so
I
mean
he
knows
how
to
do
it
right.
So.
C
Yeah
I
know
we
had
some
a
long
time
ago.
We
removed
all
also
like
Chi,
Ruby
stuff.
C
C
A
Yeah
but
they
still
use
see
a
Dre
Ruby
on
Windows
right.
B
A
A
A
A
C
C
C
C
B
C
Ones
are
just
go,
Mega
changes
and
linking
changes,
but
there
are
a
few
functions
that
were
removed
and
I
was
like.
Should
that
be
removed.
A
A
Yes,
which
one
was
it
the.
B
A
E
And
this
popping.
A
B
A
B
F
D
A
D
B
C
Yeah
I'm
still
waiting
for
leftovers
to
be
fixed
and
then
I
can
push
some
big
fixes.
C
Always
Azure
AWS
building
it
for
arm
building
the
CLI
for
arm
architectures.
A
Yeah,
so
I
will
wait
with
enabling
our
Security
checks
on
this
repo.
B
A
Yeah
but
yeah,
it
seems
like
we
just
want
this
one
to
be
reviewed
right
yeah.
It
would
have
been
nice
if
it
referenced
the
other
one
I.
B
A
Yeah,
because
now
we
can
leave
the
other
one
open
what
we
should
focus
on
the
other
on
this
one,
there's
one
other
three
viewers.
A
Let's
leave
it
open
for
discussion,
but
not
assign
viewers.
A
E
So
I
yeah
I
forwarded
your
the
information
and
they
are
organizing
themselves
to
have
lunch.
I'm,
not
sure
whether
they
will
be
able
to
get
some
budget
for
the
launch,
but
they
have.
Anyone
can
have
the
ones.
A
See
I
think
the
people
on
the
VMware
site
got
approval
because
we
all
they
already
have
like
a
team
budget
for
like
to
do
things
like
every
month.
So
they
could
just
use
that
for
that
foreign.
A
A
Yes,
so
this
is
prioritized
as
in
Johannes
working
on
it
right.
F
Yeah,
that's
also
from
us.
We
we
saw
that
four
very
old
stem
cells,
the
the
regions,
are
missing.
C
Yeah
and
that's
correct
for
these
for
golf
West.
D
B
C
I
would
not
use
them
anyways
because
they
have
some
critical
kernel
CVS
in
it.
A
You
know
and
the
heavy
stem
cells
can
be
used
right
if
there
are
some
needs.
A
F
A
C
I
think
we
announced
it
in
Slack.
That
was
it.
C
No,
no,
it
was
in
the
gym
specific
because
it
was
not
The
Bionic
one
that
was
cleaning
up
was
the
Jimmy
one
that
was
executed
and
then
so
stopped
at
cleaning
job
up
immediately,
because
Joseph
noticed
it
immediately
when
he
was
running
the
task
and
then
I
also
noticed
that
we
don't
have
proper,
proper
tagging
for
the
China,
the
ali-based
stem
cells
as
well.
F
C
Yes,
yes,
those
it's
only
about
the
AWS
Amis
cleanup
yeah,.
C
C
A
B
B
A
B
A
Go
there
no
wait.
No
there
index.
C
C
A
A
Yeah
maybe
but
I
mean
like
that.
That
feels
like
not
ideal
like
that,
would
potentially
hide
other
issues,
true
like
because,
if,
if
it
turns
out
that
killing
your
worker
process,
while
it's
doing
a
task
and
not
draining
it
actually
causes
this
to
happen
right,
we
might
want
to
find
a
way
to
reschedule
those
jobs.
A
A
Task,
but
actually
like
like,
if
you
reschedule
or
deploy
like
you,
would
run
into
a
deployment
log.
A
E
A
Yeah,
so
you
okay,
so
then
you
don't
want
to
do
that,
which
is
also
fine.
But
then
you
want
to
at
least
long
that
that
it's
happening
right,
because
you
still
have
the
problem
that
you
could
be
like
halfway
through
a
diff
creation
and
when
you
didn't
persist
to
do
the
database,
but
you
did
create
a
disk
I'd
see
basically
of
orphaned
resources,
oh
yeah,
so
these
are
probably
things
that
you
want
to
log
or
alert
about
and
then
like
require
a
human
to
investigate.
Yeah.
F
A
F
A
Yeah
because,
like
the
fact
that
I
don't
know
what
the
interfall
is
yeah,
so
it
happens
multiple
times.
Oh
actually,
this
is
like.
F
F
A
Yeah,
but
it
happened
in
like
these
look
like
oh
it's
all.
The
workers
are
being
chaos
at
the
same
time
or
something
right
like,
for
example,
this
right.
So
all
workers
at
the
same
time
have
an
issue.
A
But
would
it
be
possible
to
do
a
bit
of
Investigation
to
see
if,
if
it
can
be
correlated
to
worker
processes
dying.
F
F
Also
already
done,
but
yeah
I
think
you
you
mentioned
it,
it's
easy
to
test
right.
F
A
Okay,
that
will
be
perfect,
I
think
like.
If
we
have
that,
then
we
can
like
decide
on
like
how
we
want
to
deal
with
those
cases
right.
If
we
can
confirm
that
that
is
indeed
costing
it,
and
then
we
can
think
about
like
how
we
want
to
add
those
to
the
cleanup
right.
Maybe
we
want
to
I,
don't
know
just
move
them
into
a
state
of
error
or
something
so
they
get
cleaned
up.
B
A
I
have
to
drop
now.
Thanks
all
for
attending
see
you
all
next
week
see.