►
From YouTube: 20210223 SIG Arch Conformance
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
B
A
120
which
basically
broke
conformance
certification
for
everybody.
In
the
meantime,
it
has
been
fixed
by
a
pr
created
by
stephen.
B
A
C
That's
on
me,
unless
someone
else
wants
to
pick
it
up,
the
there
is
a
blockade,
prowl
plug-in
and
the
blockade
plug-in
can
have
some
folders
specified
and
the
folder
we
want
specified
is
the
tests
conformance,
ede
conformance
folder,
so
that
that
cannot
change,
but
we
only
want
to
blockade
it
on
release
branches,
particularly
release,
1,
20,
19
and
18.
For
the
last.
I
doubt
people
there's
nothing
going
to
change
further
back
and
it's
just
time
for.
A
In
the
meantime,
I'll.
A
That
wants
to
go
for
it
we'll
see
the
moment
it
breaks
as
we're
watching
the
conformance
but
doing
its
work.
So,
yes,
I'll
put
that
ai
on
high
importance,
low.
A
A
The
next
point
is
we
actually
need
john
and
aaron,
for
that
is
the
ineligible
endpoint
yaml,
that
we
need
that
pr
merged.
So
maybe
we
can.
We
had
it
in
conforms
our
architecture
meeting
and
everybody
was
happy
with
it.
We
created
the
pr
we
just
haven't
had
eyes
on
that.
Yet
so.
C
I
might
drop
that
specifically
to
bellamark
and
john
right
now
since
they're,
not
here
and
they're
the.
A
And
and
dems,
I
think,
also
showed
some
interest.
Maybe
you
want
to
just
answer
this
question?
Oh
yes,
there
was
actually
discussion
from
the
side
yeah.
I
think
he's
happy
with
the
answer
that
takes
us
to
ilana's
point.
B
Yes,
so
I
put
a
thing
on
this
agenda
and
there's
been
some
discussion
back
and
forth,
and
so
when
I
saw
that
this
was
still
on
the
agenda,
I
thought
oh,
I
should
go
to
the
meeting
today,
even
though
this
might
not
necessarily
be
the
one
and
only
place
to
discuss
this.
B
A
B
We
claim
we
support
this,
but
it's
not
actually
getting
tested
anywhere,
at
least
like
not
as
part
of
the
release.
Automation,
so
dawn
gave
a
really
good
overview
in
the
last
sig
node
meeting
of
at
least
like
the
history
of
the
n
minus
two
skew
tests.
B
But
to
me
this
seems
like
a
large
gap
and
like
I
was
I've
been
looking
into
this
just
because
there's,
I
think,
an
interest
in
being
able
to
do
like
n,
plus
two
upgrades
of
cubelet,
and
so
I'm
trying
to
see
if
there's
like
more
project-wide
interest
in
that
or
if
that's
kind
of
a
dead
in
the
water
thing,
because
nobody
wants
to
actually
write
the
tests
to
do
it.
So
those
tests.
B
Right
now,
I'm
not
really
sure
what
to
do
about
that.
I
think
that
don
said
at
the
time
it
was
supposed
to
live
with
sig
cluster
life
cycle,
maybe
yeah.
B
C
It's
in
addition
to
the
the
who
should
own
it
it's
having,
and
it
sounds
like
you're
the
the
champion
and
advocate
for
this.
It's
it's
finding
the
right
people.
B
Yeah-
and
it's
not
even
necessarily
like
I
am
the
champion
advocate
for
this-
I'm
totally
fine.
If
we
don't
do
this,
but
if
we
don't
do
it,
we
should
probably
have
messaging
project
wide
saying
we
don't
actually
support
this,
even
though
we
said
we
did,
and
you
know
get
that
updated
and
that
way
we
can
sort
of
adjust
expectations.
I'm.
A
C
C
He
actually,
he
replied
in
your
thread
to
go
to
sig
religious
architecture.
They
were
the
ones
that
architected
that
test,
I
believe,
and
it
affects
really.
The
release
will
be
probably
where
you
can
go
to
say.
We
need
to
have
a
a
release.
What
do
you?
What
are
you
saying
the?
If
we're
not
going
to
do
this,
we
need
to
have
in
the
release
notes
by
the
way
we
don't
support
upgrades
just
because
it's.
C
And
that,
and
that's
just
part
of
the
information
and
release
of
the
information,
hopefully
within
now
now
coming,
I
think,
sig
architecture's
in
a
couple
of
days,
I
would
put
the
agenda
item
there
before
like
today.
I
would
send
an
email
to
sig
architecture.
Just
noting
here
is
where
we
claim
that
we
support
sku
version
testing.
Here
is
the
last
time
we
did
this.
I
want
to
put
forth
a
motion
that
we
notify
people
that
this
is
not
what
we
do
or
we
do
what
we
say.
B
It
where
I'm
at
it's,
mostly
just
trying
to
find
the
right
people
to
do
the
right
things
in
case.
We
do
decide
that
we
want
to
own
this
and
make
it
work.
C
B
I
will
take
an
action
item
to
send
that
follow-up,
email
and
discuss
at
the
meeting
this
week.
C
A
Do
you
have
the
link
for
this?
I.
B
B
Oh,
let
me
let
me
grab
that
I
think
I
do
have
it
it's
just
on
my
account.
A
B
That
is,
I
think,
the
foundation
of
a
healthy,
open
source
project
so.
A
Who
exactly
so,
we
will
be
seeing
you
again
on
friday.
It's
also
another
meeting,
we're
pretty
happy.
A
Then,
which
takes
us
to
the
next
point
from
steven
we
had
a
flaky
test
pointed
out
from
120,
and
then
we
have
a
fix
from
stephen.
A
D
Most
of
the
test
seems
to
be
okay
under
a
minute
for
most
of
or
some
30
seconds.
It's
just
when
the
cluster
seems
to
be
under
a
bit
of
stress
generally,
when
there's
a
lot
of
people
trying
to
get
stuff
merged
and
everything
as
part
of
the
pre-checks,
it
seems
to
get
a
little
bit
slow
and
there
was
two
parts.
It
was
generally
the
creation,
as
well
as
to
do
with
fetching
a
little
bit
later
in
the
test.
So
I've
just
done
a
consistent.
D
System
throughout
the
really
interesting
thing
is
that
for
test
grid,
for
if
you
go
back
to
the
conversation
where
I'd
put
the
links
in
originally
the
the
test
grid
for
the
master
one
at
the
very
start,
it's
passing
perfectly
there's
not
an
issue.
It's
just
on
the
precepts
pre-submits.
C
D
A
Yep
there
is
an
issue
against
it,
so
we
must
really
try
and
get
some
eyes
on
this
to
get
it
approved.
So
we
can
get
it
solved
for
this
release,
avoiding.
C
A
Thank
you
very
much
for
that.
Thanks
stephen
for
that
fix,
I
think
that
will
sort
us
out
and
then
happy
news.
We
have
another
endpoint
merge
for
stateful
state
replicas
hit
scaling
that
actually
merged
which
thick
off
another
endpoint
of
our
open
endpoints.
A
Then
something
else
that
I
do
want
on
the
test
grid,
where
we
need
some
eyes
on
to
get
the
lgtm
and
improve
in
the
previous
sick
or
the
conformance
meeting
john
gave
some
good
feedback
to
stephen
about
exactly
what
to
change
around
status.
So
he
did
update
the
test.
Let
me
just
go
to
the
pull
request.
That
link
should
have
been
the
pull
request.
I'll
fix
that
later,
unless
somebody
else
will
alana.
C
B
A
B
A
B
A
Two
pr's
ready
to
merge
that
that
was
not
the
idea.
Let's
do
that.
A
Tastes
great
looking
good.
We
got
two
days
left
so
by
friday.
When
we
get
to
sick
arch
meeting,
we
could
ask
around
there
if
there's
anybody
willing
to
give
it
and
when
she
did
look
at
one
of
the
two
and
also
sick
abs
looked
at
the
one
of
the
two.
I
think
they
kind
of
missed
their
days
too,
because
they
were
so
similar.
A
So
by
friday.
I
start
start
tracing
up
people
to
try
and
get
that
across
the
line,
and
that
brings
us
to
some
more
happy
news.
If
we
go
to
api
sniff
and
we
look
at
sick
apps
when
we
started
in
the
beginning
of
this
release,
tested
was
sitting
at
30
or
48,
and
once
we
got
these
merch
we'll
actually
be
conformed
up
to
60.
So
we
basically
pushed
up
12
for
this
release,
which
is
a
very
happy
event
specifically
for
endpoints,
that's
so
readily
used.
So.
B
A
A
And
then
we
verify
the
parts
came
up,
we
scale
the
replica
set,
verify
that
and
then
we
patch
the
replica
set
verify
that
and
in
that
we
basically
touch
the
scale
and
the
confused
myself
now
the
scale
and
the
the
raw
batch
scale
and
patch
endpoint
for
this
with
this
test.
So
we
would
also
like
some
eyes
on
that
to
get
us
approval
to
write
the
test,
but
it
is,
I
think,
pretty
well
on
the
way
good
to
go,
and
that
is
what
I've
got.