►
From YouTube: Kubernetes SIG Node 20201102
Description
Meeting Agenda:
https://docs.google.com/document/d/1j3vrG6BgE0hUDs2e-1ZUegKN4W4Adb1B6oJ6j-4kyPU
A
Hello,
it's
november,
2nd
new
month
started
and
it's
a
sig
note,
ci
group
meeting
we
have
seven
and
attendees
on
a
call,
welcome
everybody.
A
We
have
very
short
agenda
for
today,
so
I
wanted
to
start
with
a
short
reminder
that
we
need
to
get
our
prs
merged.
I'm
still
working
to
understand
how
like
who
can
I
ask
to
approve
iprs?
We
don't
have
many
approvers
in
this
area,
so
I
will
keep
working
on
that.
We
have
six
lgtm
tprs.
Actually,
there
are
14,
but
six
are
related
to
our
team
effort.
A
I
will
keep
pushing
if
you
approve
of
14,
it's
even
better
and
for
non-approved
non-lgt
mprs.
We
have
45,
not
all
of
them
related.
I
didn't
have
energy
to
count
which
ones
are
related
which
not.
There
are
two
pages
of
them.
So
if
you
have
time,
please
go
ahead
and
review
all
these
pr's.
A
It's
it's
nice
to
have
progress
going
so
yeah
and
somebody
else
had
another
lincoln
project
file.
Who
was
it.
A
B
Item
yeah,
so
this
one
is
essentially
along
the
similar
lines
of
what
you
were
just
saying
when
the
team
got
started,
we
created
this
a
project
board
and-
and
I
guess
at
this
point-
it's
also
an
open
question:
do
we
want
to
keep
the
project
board?
B
A
Productive
boards
are
very
handful
handy
when
we
assign
specific
people
for
items,
so
we
can
like
at
a
glance,
see
in
progress
item
who
is
working
on
who
is
pushing
them
forward.
A
We
managed
to
do
that
successfully
over
time
with
just
assigning
it
this
way
so
like
as
action
items
yeah.
So
I
mean
whenever
there
is
an
update
somebody
coming
up
and
explain,
update
on
like
on
action.
So
yeah
I
don't
know,
project
work
may
be
very
handy
if
you,
if
you
start
doing
it,
start
making
assigning
responsibility
for
driving
some
items
forward.
A
B
Yeah
yeah
and
actually
that's
a
good
comment.
If
anybody
here
is
looking
for
a
place
to
contribute
is
so
essentially
we
have
the
project
board
and
a
handful
of
people.
Add
issues,
add
add
issues
and
pr's
to
the
project
board
every
now
and
then,
which
is
separate
from
a
query
aprs
and
issues
directly
through
labels
in
kubernetes.
B
C
A
The
issues
containing
the
people
switched
from
cri
repository
to
container
d
repository
to
defining
their
test
yaml,
and
it
doesn't
seem
to
be
working
well
and
derek
from
container
d
organization
prom
like
wanted
to
follow
up.
So
I
will
keep
track
on
this
item
and
well
another
test
trailer
that
matt
was
planning
to
look
at
and
I
don't
think
she
met.
Oh
my
you
you're
here
right.
D
Yeah,
I'm
here
yeah,
there's,
there's
been
some
conversations
on
that
so
me
and
morgan
and
jorge
spent
some
time
looking
at
it,
but
still
still
ongoing,
jorge
posted
a
link
to
a
change
in
port
which
might
explain
why
it's
not
accessing
the
end
point.
So
I'm
trying
to
figure
out.
You
know
if
there's
something
missing
in
terms
of
connecting,
but
anyway
it's
a
work
in
progress.
D
Well
it
it
might,
it
might
be
something
related
to
test
configuration,
so
there's
either
a
problem
with
the
infrastructure
which
I'm
not
really
sure
how
to
debug
or
there.
There
might
be
some
test
configuration
because
there
was
a
change
from
8080
to
6443,
for
like
the
port
for
the
api,
server
and
hostport,
and
I'm
just
trying
to
figure
out
if
something
was
missed
in
terms
of
docker.
The
port
is
blocked
or
something,
but
I
haven't
been
able
to
find
anything
too
positive.
Yet.
B
And
related
to
that
one
also,
I
also
apologize
for
not
really
not
responding
to
things
promptly,
I'm
just
getting
back
to
kubernetes
things
this
week,
but
that
one
day
brief.
The
other
brief
detail
that
I
can
give
is
essentially
the
test
suite
that
things
actually
runs
it
we
test
naughty
tweet
test
is
actually
my
name.
B
We
think
over
everything
kkk
and
the
cha,
the
patch
that
matt
just
mentioned,
that
was
actually
done
in
the
test
suite
the
thing
failing
is
that
before
the
test
suite,
that
is
a
gingko
that
is
a
that
is
a
gingko
thing,
so
there
is
also
the
possibility
that
is
so
a
node
we
test.
The
additional
step
is:
whenever
we
are
a,
we
have
allocated
resources
in
ci.
We
actually
use
the
gcp
cli
to
create
bms
and
ssh
into
them,
so
it
might
pos
it
is.
E
I
have
not
attempted
to
run
it
on
the
infrastructure
myself
yet
yeah,
I
don't.
I
don't
know
how
to
do.
B
A
Yeah,
I'm
sorry,
I
lost
the
context.
So
are
you
talking
about
how
to
investigate
test
failures.
A
A
See
so
the
suggestion
is
then
to
follow
up
on
slack
to
understand
how
like
whether
documentation
is
okay,
so
we
can
try
it
out.
B
I
I
think
first
thing
to
try.
I
think
I
think
couple
things
to
try
out
is
to
to
review
the
pr
that
was
merged
around
the
same
time
that
the
job
started
began
to
fail,
which
is
the
thing
that
matt
mentioned
around
the
switching
from
port,
a
something
to
6443.
B
That
one
and
I
think
another,
a
manual,
a
manual
way
of
betting,
whether
that
br
is
good
or
not,
is
to
actually
is
to
actually
manually
reproduce
the
ci
runs.
I
think
that
if
we
can
use
the
test,
suites
remote
runner
and
go
from
the
com
the
kk
commit
before
the
apr
was
merged,
and
then,
after
that
pr
was
merged.
I
think
comparing
the
the
results
from
those
two
could
be
also
a
good
check.
C
A
Okay,
so
I
know
for
pneuma
I
really
following
up
internally.
I
don't
have
any
updates
here,
yeah
any
other.
A
A
Okay,
then,
if
we
okay,
let's
wrap
it
up
happy
monday
and
happy
have
a
productive
week.
Bye-Bye.