►
From YouTube: Kubernetes SIG Scheduling Meeting - 2019-10-03
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
A
So
one
one
thing
that
I
that
I
figured
while
working
on
the
translation
layer
is
that
we
allow
duplicate,
custom,
predicates
and
priorities
to
be
specified
in
the
policy.
We
don't
have
support
for
such
semantics
and
the
component
config
plug-in
API
and
so
I
created
an
issue
explaining
the
dis,
dis,
semantics
and
I
was
wondering
if
anyone
is
aware
of
how,
whether
whether
there
are
any
actual
deployments
out
there
that
that
benefits.
For
that
only
case
where
I
think
that
this
is
useful,
I
would
like
to
create,
for
example,
service,
affiliate,
multiple
service
affinity.
A
A
A
A
C
A
A
You
know,
so
they
don't
really
create
something.
They
don't
create
a
plug-in
on
the
fly
sort
of
the
plugin
exists.
They
just
have
the
ability
to
configure
it
in
the
current
world.
They
they
actually
create
a
plug,
a
predicate
on
the
fly
and
they
give
it
a
name
and
it
gets
registered,
etc,
etc.
So
that's
way
more
flexible
than
what
I
would
like
it
to
be.
C
A
C
A
D
D
I
would
suggest
away
galaxy
bands
first
day
for
them
or
if
they're
you
know
really,
there
are
no
feedback
on
just
feel
people
using
that.
So
we
can
speed
that
probably
yeah
I
know
if
there
are
lots
of
people
we
need
to
get
their
case,
and
maybe
we
get
a
you
know
more
time
to
to
do
that
great
sounds.
A
So
the
second
item
on
the
agenda
again
so
I
have
the
framework
regression
therapy
completely
ready.
I
have
the
first
predicate
as
a
filter
under
view
I'm
working
on
the
first
priority
as
a
filter
as
a
as
a
score
in
progress
right
now,
once
I
merged.
Those
I
will
create
issues
for
folks
to
do
the
rest
of
predicates
and
and
and
priorities
I
just
want
to
set
up
the
like
presidents
or
how
we
how
we
think
we
should
be
doing
this
and
then
and
then
we
create
issues
for
the
rest
of
the
predicates
and
priorities.
A
A
We
thought
the
that
thing
is
a
little
bit
lacking
in
terms
of
metrics
and
bitter
logging,
et
cetera,
so
he
in
the
past,
like
couple
of
weeks,
he
was
looking
deeper
into
the
schedule
and
see
what
how
we
can
improve
that
and
he
came
up
with
a
list
of
possible
metrics
to
add
also
in
the
context
of
the
new
framework
where
we
can
add
more
metrics
to
track
a
different
extension
points,
etc.
So,
please
take
a
look
at
the
umbrella
issue.
A
I
posted
in
the
agenda
give
feedback
any
concerns
or
suggestions
how
to
improve
it
as
well.
We
want
to
be
able
to
you
know
us,
as
you
know,
like
cloud
providers
be
able
to
support
the
scheduler
the
best
we
can
and
the
only
way
we
can
do.
That
is
by
having
more
insight
and
ability
to
debug
running
running
clusters.
Basically.
C
The
background,
I
I
know
is
they're
guessing
from
when
we're
was
working
on
there.
It
hasn't
been
active
for
for
pretty
long
time,
so
yeah
you'd
better
check
with
justing
to
check
the
latest
status.
If
you
want
to
take
over
yeah.
It's
also
good
to
talk
to
him.
I
think
Brent
ground
was
also
involved
into
some
discussion
with
Bobby
and
yes
in
the
last
year's
Capcom
in
in
Seattle
I,
don't
know
the
latest
at
status,
so
you
better
check
with
yeah.
Yes,
okay,.