►
From YouTube: Octant Community Meeting - July 14th, 2021
Description
Octant community meeting is held weekly. We discuss and talk about the current state and future of Octant, demo upcoming features and releases, and preview new ideas we are considering for Octant.
Meeting agenda: https://hackmd.io/CzaPxtmXT_SW8nEpdwvGzw?view
B
A
Start
us
off
hello.
Everyone
welcome
to
the
july.
14Th
2021
opted
community
meeting.
I've
done
any
of
these
in
a
while.
So
it's
going
to
be
exciting,
got
a
fresh
release
out
the
oven,
0.2
and
we've
got
a
few
discussion
points
today.
So
we'll
jump
right
into
it
all
right.
Let's
kick
this
off
desktop
one
share,
all
the
things
all
right.
A
All
right
here
we
go
so
odot's
youtube,
release,
we've
got
attendees
and
affiliation
feel
free
to
add
these
on
the
agenda
if
you
want
to-
or
if
you
just
want
to
hang
out
totally
cool,
mostly
vmware
folks,
we've
got
our
status
update
0.22.
A
So
this
was
somewhat
waited
on
for
a
little
bit,
and
mostly
my
take
on
this
release
was
that
it
solves
a
lot
of
the
existing
access
error
checks
that
pops
up
the
crds
and
sometimes
when
these
accessories
kicks
in
octane,
doesn't
show
anything
at
all,
and
so
it's
always
been
a
back
and
forth
between
these.
But
hopefully
this
is
the
last
time
and
if
not
we'll
keep
we'll
keep
working
on
it.
A
We've
got
a
couple
of
issues
here,
they're,
mostly
internal,
but
if
you
pick
these
up,
for
example,
these
create
a
websocket
streaming
interface
and
exported
config
dash
and
related
interfaces.
These
are
mostly
done
by
wayne
and
some
internal
vehicle
folks
to
support
some
products
going
on.
It
should
not
really
affect
anyone,
but
if
you
are
consuming
these
public
interfaces,
I
would
like
to
know
about
it.
A
It
would
be
a
very
interesting
use
case,
but
essentially,
if
anyone's
interested
it's
using
opt-in
as
a
websocket
manager,
so
it's
kind
of
you're
just
importing,
often
and
you're,
just
relying
on
it
to
become
a
server
and
it's
not
even
running
it
headlessly.
I
don't
think
that's
an
accurate
description
of
it,
but
you're
kind
of
using
opt-ins
as
like.
I
don't
know
it's
kind
of
a
weird.
A
General
users
to
be
using.
We've
got
a
few
other
things,
namespace
lists
to
be
expandable.
I
believe
I
have
a
thing
of
logged
in
running
right
now
I
do
so.
I
think
this
is
just
yeah.
This
just
looks
like
a
show
more
type
of
thing.
If
you
have
a
lot
of
name
spaces,
of
course
we
don't
have
a
designer
so,
but
we
do
have
show
more
and
show
less
up
and
running.
A
Let's
see
anything
else,
that's
interesting,
updated,
broken
yeah,
some
link,
fixes
and
then
yeah
and
then
there's
some
form
stuff.
I
think
there
was
yeah.
Two
minor
form
fixes
so
we're
still
tankering
up
with
some
straggling
things,
after
the
huge
form
updates
that
we've
done
a
few
releases
ago,
and
so
these
are
going
to
continue
trickling
in
so,
if
you
don't
see
your
form
fix
in
here,
it
will
come
soon
make
some
noise
about
it
cool
so
that
I
think
about
covers
the
release.
C
Yeah,
I
want
a
big
couple
points.
I
think
this
was
vikram's
first
release
right,
so
congrats
vikram
I'm
doing
that.
Thank
you
and
also,
I
think
this.
This
one
beats
some
records
because
it's
I
I
checked
the
time
step.
I
think
it
turned
out
nine
minutes
before
the
community
meeting.
So
that's
that's
a
pretty
high
for
set
right
here.
D
B
Yeah
I
just
wanted
to
yeah
well.
Thank
you.
Thank
you
for
for
the
news
release
and
regarding
the
typescript
bug
fix.
You
know
how
how
to
communicate
this
to
our
extended
community
of
looking
authors.
Now,
the
experience
of
creating
the
or
developing
plugins
and
typescript
for
often
to
be
easier,
should
be
possible.
Now.
E
Sorry
yeah,
so
the
issue
was
previously:
there
were
issues
with
typescript,
so
first
the
components
were
outdated,
then
the
components
were
updated,
but
it
turns
out
there
are
some
errors
with
the
reflection
and
so
on
that
were
preventing
the
generated
components
from
being
correct.
So
now
everything
should
work
correctly.
E
We
can
test
the
latest
typescript
code
in
the
plugin
library
for
octant
repo
and
then
once
that
checks
out,
we
can
release
the
plug-in
library
version
0.22
so
that
it
corresponds
with
oxygen
0.22
and
we
can
be
sorted
and
the
latest
code
will
be
on
npm
and
everyone
can
use
it
from
there.
So
that's
really
good
and
I
mean
it
kind
of
brings
things
to
where
they
should
be.
I
think.
A
Yep
and
to
answer
the
second
half
of
the
question:
how
do
we
communicate
that?
Really,
if
someone's
trying
to
build
a
plug-in
and
it's
looking
for
a
component
that
doesn't
exist,
but
it
happens
to
exist
on
master
of
octane
right
now,
we'll
just
tell
them
to
upgrade.
A
Cool
and
for
folks
who
don't
know
about
this
other
repo,
it's
a
plug-in
library
for
auction,
so
we
use
something
called
a
yeoman
generator.
It
does
a
lot
of
the
scaffolding
and
helper
libraries
to
create
these
components
for
their
typescript
plugins,
so
yeah.
This
is
going
to
be
pretty
much
the
tooling
of
choice
if
you're
a
building
a
typescript
polygon
right.
Well,
let's
see
that
is
not
the
agenda.
This
is
the
agenda
cool
so
that
about
wraps
up
the
release.
We
have
one
discussion
point
of
the
agenda.
A
We
want
to
talk
about
an
0.23
sprint,
we're
changing
the
way
we
work
just
ever
so
slightly,
because
we've
done
some
introspection
about
the
way
we've
been
working,
the
last
couple
of
sprints
and
we
wanted
to
try
something
different
see
if
we
could
do
better
and
we
have
this
new
projects
layout
board
and
it's
not
new,
but
they
describe
things
a
little
differently
than
we've
done
in
past
releases.
A
So
here
we
are.
We
have
an
example
of
0.2
release
where
we
did
list
out
a
few
goals,
and
we've
got
essentially
a
very
generalized
description
of
what
this
release
does
and
for
our
d3.
I
believe
we
are
going
to
try
to
take
some
time
out
of
this
community
meeting
to
figure
out
what
this
release
is
going
to
be.
D
Yeah,
that
is
the
plan
yeah
sam,
if
you
click
into
the
odot
23
project
that
will
have
a
in
progress
kind
of
checklist
there
we
started,
brings
brainstorming
various
ideas
in
the
to-do
notes.
Yesterday
I
created
this
in-progress
checklist
of
the
items
we
needed
to
go
through
to
be
able
to
kick
off
the
sprint.
D
So
one
of
those
the
one
of
those
first
items
was
setting
the
sprint
theme
we
had.
We
had
a
good
discussion
around
around
various
themes
and
I
also
took
the
time
to
reorganize
the
backlog
project,
which
is
the
one
that
kind
of
handles
all
of
there
are
outstanding
issues
you
all
might
have
saw
a
message
in
the
octant
slack
channel
around
a
new
column
called
themes
inside,
so
I
originally
tried
to
like
create
a
column
for
each
theme.
That
was,
that
was
unreal.
D
D,
like
the
the
horizontal
scrolling,
was
just
way
too
much.
So
we
have
this
big
vertical
scrolling
thing
for
now.
If
the
mode
of
operating
the
mode
of
working
with
themes
works
really
well,
we'll
adjust
our
labels
to
reflect
that,
and
then
this
kind
of
will
go
away
and
we'll
use
labels
in
the
future.
But
I
didn't
want
to
go
like
and
go
and
redo
all
the
labels
and
then
after
two
sprints
we're
like
we
don't
like
this,
so
we're
gonna
try
it
this
way.
First,
I
know
it's
a
lot
of
vertical
scrolling.
D
I
apologize
for
that,
but
within
this
themes
column
I
started
to
break
out.
I
used
notes
to
kind
of
divide
it
up.
So
under
each
note
there
is
that's.
The
note
is
like
a
theme
and
then
there's
the
issues
that
kind
of
align.
With
that
theme
we
can
also
add
a
little
more
detail
to
the
notes
should
we
need
to,
but
some
of
them
are.
D
You
know
self-descriptive,
others
are
not
like
the
applications.
One
here,
for
example,
probably
needs
a
description
that
says
this
is
about
the
applications
view
of
octant.
You
know
not
the
like
the
very
specific
module.
I
believe
we
call
it
workloads
right
now.
We
call
it
applications.
Maybe
we
change
the
name.
I
don't
remember,
but
so
anyway,
I
wanted
to
get
folks
feedback
on
themes.
They've
been
thinking
about
since
our
discussion
yesterday.
D
Maybe
we
could
spend
some
time
going
through
the
themes
here
in
this
column,
if
that's
helpful,
yeah
kind
of
kind
of
take
it
from
there.
So
what
do
folks
think
would
be.
A
A
Kicking
off
the
discussion
here,
my
take
is
that
there's
a
lot
of
low-hanging
fruit
around
the
space
of
remote
debugging.
A
So
in
other
words,
you
have
the
infrastructure
layer,
whether
or
not
your
deployments,
config
maps
and
secrets
are
wired
up
correctly
in
a
way
that
lets
your
applications
to
run,
but
also,
let's
say
your
application
is
not
behaving
correctly
on
kubernetes
and
you've
configured
everything,
at
least
on
the
infrastructure
side
correctly.
How
do
you
start
debugging
that-
and
there
are
already
a
whole
multitude
of
tools
available
in
the
ecosystem
for
this,
so
the
next
step
is
to
figure
out
how
we
can
use
octane
as
a
bridge
to
make
those
tools
discoverable
or
start
re-implementing.
C
What's
the
difference?
Well,
debugging
by
you
know
when
you
say
debugging,
I'm
thinking
that
battling
the
code,
essentially
a
code
running
somewhere,
but
if
you
troubleshooting
is
you
know
trying
to
figure
out
why
something
is
not
working
and
then
you
need
to
go
and
check
the
statuses
and
go
from
one
object
to
another
and
so
on.
C
So
it's
a
slightly
different
task,
and
that
and
and
thinking
about
those
two,
a
little
more
clearly,
I
think
troubleshooting
is
the
one
that
we
that
we
are
trying
to
to
get
focused
as
much
as
we
can
basically
trying
to
help
our
users
figure
out.
What's
going
on,
you
know
something
is
wrong.
I
have
no
idea
what
it
is.
Let's,
let's
find
the
best
way
to
to
help
them
do
that
where
debugging
is
like,
oh,
I
know,
there's
a
problem
in
this
component.
B
Yeah,
because
you
know
so
far,
I
see
often
more
focus
on
the
infrastructure
side
of
the
application,
but
even
with
that,
we
still,
you
know,
there's
still
room
for
for
improvement.
You
know,
for
example,
I
was.
I
was
running
a
scenario
this
morning
trying
to
see
you
know
how
to
troubleshoot
and
multi
container
pod
issue.
B
You
know
doing
it
just
using
the
regular
cube,
pedal
tools
and
using
octane
and
with
octane.
I
completed
the
troubleshooting
in
the
half
of
the
steps,
so
that's
good,
but
it
could
be
less
than
that
if
everything
in
all
the
relationships
will
be,
for
example,
in
the
resource
viewer,
I
see
resource
viewer
like
the.
B
I
don't
know
if
the
default
view,
but
if
all
the
relationships,
for
example,
you
know
all
the
all
the
containers
inside
a
pot
right
now
that
is
not
visible
in
resource
viewer
or
or
a
more
clear
relationship
or
diagram
between
all
the
pots
that
are
running
or
scheduled,
to
run.
In
a
specific
note,
for
example,
if
if
we
complete
that
you
know
infrastructure
visibility,
relationship
side
of
the
things,
I
think
at
least
for
the
infrastructure
resource
beer
will
be
really
helpful
for
troubleshooting
and
yeah.
I
agree
with
milan.
B
I
don't
know
if
that's
the
academic
definition
for
that
or
not,
but
I
see
I
understand
the
vlogging
more
for
code.
It
will
be
really
awesome,
but
I
see
that
as
a
further
step.
You
know
first
just
getting
the
infrastructure
troubleshooting
right,
reliable,
clear
sample.
So
that's
those
are
my
two
cents
for
that.
A
D
D
Right
like
like
object
status,
is
very
broad.
So
so
I
would
I
would
like
just
saying
that
is
very
broad,
but
I
would
include
that
to
mean
the
the
actual
like
status
in
the
cluster
of
of
your
objects.
D
The
relationships
that
those
objects
have
with
other
objects
is
part
of
that
status.
The
actually,
if
you
go
to,
if
you
go
to
the
backlog,
theme
themes
lane
there
and
you
scroll
down,
there's
one
that
I
made
for
object
status,
which
I
think
will
show
just
kind
of
how
I'm
thinking
about
that.
D
But
now
you
kill
the
field
yeah
you're
just
gonna
have
to
scroll.
I
I
warned
you.
It
was
a
lot
of
a
lot
of
vertical
scrolling,
so
that's
object.
Updating,
so
here's
object
status,
so
yeah,
there's
just
a
lot
of
things
in
here
around
you
know
better
visibility
into
the
terminating
states.
D
Should
we
can
we
visualize
network
policies
and
topologies
the
one
that
david
just
created
recently,
which
around
you
know,
pvcs
shown
as
okay
when
they're
unbound
or
when
they're,
not
okay,
and
then
underneath
this
object
status
category
there's
just
a
pretty
wide
list
of
things
that
all
fall
into
this.
D
What
I
think
all
of
all
three
of
you
have
described
as
being
able
to
I'll
use
the
word
troubleshoot
being
able
to
troubleshoot
your
application
and
your
objects
running
in
a
kubernetes
cluster
and
along
the
lines
of
more
of
what
david-
and
I
think
milan
were
saying-
is
using
the
primitives
from
the
api
to
do
so,
using
without
extending
it
with
third-party
tooling,
which
I
think
is
more
along
the
lines
of
what
sam
is
is
suggesting.
D
I
think
getting
our
foundational
object
status
in
a
in
a
spot.
Word
is
really
well
defined
and
acting
appropriately,
having
some
sense
of
an
object
scorecard
to
represent
the
the
health
of
an
object
over
time
very
small
window
of
time,
but
just
over
time
once
we
have
that
in
place,
then
we
really
can
kind
of
open
the
floodgates
on
third
party
tooling
to
go
even
deeper,
but
I
I
I'm
hesitant
to
introduce
third-party
tooling
when
the
foundation
that
we
have
isn't
like
rock
solid
already.
F
I
mean
it
sounds
like
this
would
be
a
good,
a
good
candidate
for
this
release.
As
far
as
a
release
theme
goes,
and
it
was
one
of
the
ones
that
was
discussed
yesterday
and
is
in
the
note
for
for
this
release
that
sam
brought
up
yesterday.
F
So
unless
anyone
objects-
and
I
agree
that
you
know
making
sure
the
foundation
is-
is
solid-
makes
sense
before
we
do
anything
else
and
and
follows
along
the
line
of
that
note.
That
came
out
from
joe
bida
about
platform,
whiplash,
making
sure
the
foundation
is
strong
before
making
any
changes
on
higher
levels.
F
C
Yeah,
I
think
that's
a
good
one
and
I
absolutely
agree
with
you
said
wayne.
I
think,
there's
still
quite
a
bit.
We
can
get
improved
on
that
I
mean
even
even
like
conditions.
The
baby
show
conditions.
Now,
it's
probably
not
optimal.
We
use
table
for
that
and
it'd
be
nice
to
have
something
like
maybe
use
like
I
don't
know,
maybe
clarity,
tags
or
some
similar
controls
to
display
conditions.
So
there's
a
there's,
a
lot
of
yeah.
I
think
you
said
so
there's
a
lot.
I
mean
this
doesn't
need
to
be
table.
C
The
just
should
have
those
two
available,
progressing
fields
shown
and
everything
else,
displaying
full
tip
or
some
secondary
action.
So
there's
a
there's,
a
lot
of
improvements.
You
can
do
there
and
I
think
this
is
a
good
candidate
that
seems
to
be
very
requested
from
our
users.
Quite
a
bit.
D
Right,
okay,
so
yeah,
and
I
think
I
think,
just
picking
one
and
working
through
it,
the
the
first
sprint
is
probably
a
good
idea
anyway.
So
and
I
think
it'll
help
people
understand
as
we
do
it
in
the
future,
so
object
status
is
what
we're
going
to
call
our
sprint
theme.
So,
with
that
in
mind,
we
want
to
set
some
goals
for
the
sprint.
Essentially,
what
is
the?
What
is
the
success
outcome
of
of
this
two
weeks
of
working
on
object
status?
What
is
it
that
we
want
to
achieve?
D
You
know?
Is
it?
Is
it
reducing
the
number
of
steps
that
it
takes
to
debug
a
workload
using
the
pdf
that
david
had
sourced
as
how
to
troubleshoot
things
within
kubernetes?
Is
it
you
know
creating
a
better
view,
I'm
creating
a
better
default
view,
for
you
know
like
like
milan
was
just
pointing
out.
D
Certain
data
is
formatted
and
structured
in
certain
ways,
and
we
could
present
it
better
like
what
do
folks
see
as
some
of
the
the
the
outcomes
they
would
like
for
object
status
over
the
next
two
weeks
and
we
can
browse
through
the
the
column
of
the
object
status
theme
as
well
to
kind
of
spark
ideas
and
thoughts
there
too.
So,
but
if
anyone
has
any
top
of
mind,
feel
free
to
share
them.
B
Yeah
well
I'll
I'll.
Try
to
summarize
that,
at
least
for
my
opinion,
I
think
you
know,
all
the
important
information
is
is
there
right
often
is
is
collecting
everything
we
need
from
the
container
status
level.
Up
to
all,
you
know
all
the
objects
necessary,
at
least
for
from
the
kubernetes
infrastructure
point
of
view.
B
B
But
if
I
go
to
a
different
view,
I
think
is
the
overview
of
summary.
I
can
see
the
list
of
containers
there.
I
can
even
watch
the
logs
of
of
each
one
of
the
containers.
I
mean
everything
is
there,
but
a
default
condensed
view
that
you
know
gathers
all
the
information
will
speed
up
the
troubleshooting
use
case.
That's
at
least
my
reflection
so
far.
B
D
Yeah,
so
I
think,
there's
there's,
there's
there's
something
we
want
to
be
careful
of
with
the
goals,
and
this
might
be.
This
might
be
something
we
just
have
to
practice
and
get
better
at,
but
ideally
our
goals
would
be
objective
and
not
subjective.
D
Like
opinions
about
like
like
saying
something
like
make
resource
viewer
better,
that's
not
a
goal.
That's
that's
an
opinion
that
we
form
and
we
do
something
and
we
think
it's
better,
but
but
at
the
end
of
that
thinking,
it's
better
there's
some
goal
or
outcome
that
we
we
want
to
be
shooting
for
to
really
and
the
reason
this
is
important
when
you
set
a
theme
for
a
work
cycle,
is
that
these
goals
are
what
we
will
use
as
a
team
to
decide.
Did
we
achieve
the
the
outcome?
D
Did
we
achieve
and
accomplish
the
the
desired
outcome
of
this
theme
for
the
work,
and
so
just
just
to
help?
I'm
not
saying
this
is
one
of
them,
but
just
to
help
the
conversation,
a
theme
or
a
goal.
A
success
goal
for
this
might
be
all
objects
accurately
reflect
their
status
and
octant
as
to
match
statuses,
reflected
in
cube
ctl.
D
That's
exactly
what
I
wanted
to
say.
So,
thank
you.
Man
like
that's,
that's
that's
that'll,
be
a
goal
and
that
that's
testable.
It's
it's
it's
measurable
and
then,
and
we
can
say
hey,
we
did
that
this
is
great
and
then
also
that's
something
that
we
can
tell
users
hey.
We
did
this.
This
is
great
and,
like
you
will
no
longer
have
you
know
yeah.
You
know
how
impedance
mismatch
between
this
tool
and
this
tool
so
think
of
things
along
those
lines
and
yeah
just
ideas,
thoughts
and
really
anything.
D
Maybe
we
can
add
that
one
since
milan,
you
know
echoed
it
as
well,
but
I'd
like
to
at
least
have
three.
C
Yeah,
I
think
this
one
is
important.
You
know
I
mean
like
the
issue
that
they
would
open.
They've
cuddled
tells
something
is
wrong
and
we
say
it's
green,
that's
wrong,
but
that
shouldn't
happen
and
I've
seen
some
other
issues
where
you
know
reports
one
thing
and
we
do
another,
so
I
think
it
will
we
we
need
to
be
trusted
too,
and
we
need
to
be
accurate,
showing
this
it's
like
an
instrument.
You
know
if
accuracy
is
not
correct,
people
people
will
not
use
it
and
kind
of
same
for
the
for
the
octant.
B
Yeah,
I'm
I'm
not
sure
on
the
how
on
this
I
mean
we
are
not
solving
the
how
that
I
wants
of
the
help,
but
if
we
are
using
conditions
for
that,
I
don't
know
if
you
know,
if
that's
the
case
for
a
multi-container
part,
that
is,
that
is
running
the
the
response
from
the
controller
is
the
pod
is
running
right,
there's
a
problem
with
one
container,
but
the
quality
is
running.
So
I
don't
know
how.
B
A
So
it
sounds
like
there
are
two
different
concerns
here
when
we
talk
about
cube
control
right,
so
we
have
the
parity
between
when
something
is
shown
as
an
error
in
cube,
control
and
octa
does
not.
Are
there
any
considerations
for
the
other
side,
where
cube
control
says
it's
okay,
but
maybe
often
should
be
saying
it's
an
error.
D
Yeah,
I
think,
yeah.
I
think
that
touches
on
what
david
was
just
saying,
with
the
difference
between
the
the
status,
the
api
status
of
a
thing
you
know,
pending
running
failure
and
the
conditions
of
that
thing,
you
know
terminating,
you
know,
crash
loop
back
off
things
that
there
are
things
that
are.
There
are
actions
that
are
happening,
and
then
there
are.
D
There
are
the
the
the
state
of
that
thing
as
a
result
of
those
actions-
and
I
think
this
is
this
first
one
is
making
sure
that
if
something
says
it's
running
in
cuba
ctl,
it
says
it's
running
an
octant
or
something
says
it's
pending
it's
pending
or
suspended,
it's
suspended,
making
sure
that
all
of
those
statuses
are
reflected
appropriately
and
then
the
subjective
piece
of
what
do
we
do
with
conditions?
How
do
we
flag
things?
What
color
do
we
make
them?
D
Can
something
be
running
but
also
have
warnings
on
it
because
or
something
be
running
and
also
have
the
orange
or
yellow?
You
know
the
indicator
with
with
a
warning
icon
right.
That
is
a
to
me.
That's
a
different
different
thing
that
is
addressed
from
more
of
a
ux.
Like
object,
status,
ux
approach
versus
are
these
accurately
reflecting
the
status
the
the
status
field
without.
F
The
next
goal
might
be
something
along
the
lines
of
surfacing
the
status
of
objects
in
all
appropriate
places.
I
see
a
number
of
things
within
this
particular
theme
in
the
theme
column,
about
like
showing
pod
status
in
workloads,
pod
pods
view
there
are
a
number
just
about
like
showing
something
in
a
desired
place,
so
there
could
be
a
goal
around
just
surfacing
the
status
appropriately
and
then
we
can
select
which
tasks
to
tackle
within
that
goal.
D
Yeah,
I
think
the
yeah,
the
issues
that
I'm
seeing
that
you're
talking
about
are
definitely
they're
like
I'm
looking
at
it
here,
and
it
says
it's
right,
but
then
it's
not
right
over
here
and
I'm
looking
at
it
here
or
it
shows
an
error
in
this
place,
but
not
in
this
place.
F
Is
it
possible
there
are
places
where
it's
just
not
shown
where
the
stat?
Where
places
I
mean
it,
looks
like
which
one
was
I
looking
at.
G
D
Oh
yeah,
okay,
so
so
the
phase
yeah,
what's
interesting
about
that
one
is
it's
explicitly
calling
out
the
difference
between
the
phase
that
it's
in
running
or
pending
and
the
status
that's
happening
which
is
like
terminating
or
container
creating,
and
I
think
yeah
that
is.
This-
is
directly
one
of
the
ones
that
comes
up
a
lot
with
cube,
ctl
versus
octant.
F
If
there
aren't
more
that
that
it
looks,
I
thought
there
were
more
that
were
that
could
be
encompassed
by
the
idea
of
status
being
surfaced
in
more
areas.
But
if
that's
not
the
case,
then
it's
not
a
good
goal.
If
it's
just
this
one.
F
D
I
think
I
think,
having
something
around
conditions
would
be
good
there.
There
are
some
open
issues
around
showing
conditions
that
they're
detected,
showing
conditions
for
crds.
D
I
think
I
think,
having
a
way
to
surface
conditions
and
view
them.
You
know
object
status
should
take
conditions
into
account
right.
So
so,
there's
there's
there's
like
at
least
three
classes,
they're
aligned
with
conditions,
they
touch
different
areas,
but
they
definitely
are
a
part
of
object
status,
and
I
think
I
think
that
would
be.
A
good
goal
is
to
surface
object,
conditions.
A
D
A
D
C
So
so
currently
we
very
show
them.
We
show
them
as
a
table
and
because
I
don't
think,
there's
a
need
for
that.
I
mean
they
should
just
be
basically
like
a
set
of
labels
and
that
they
can
be
on
the
arbitrary
text,
so
they're
not
fixed,
but
you
know
we
can
show
them,
maybe
as
a
tags
or
something
similar
clarity
world
and
then
provide
additional
information
either
through
tooltip
or
or
you
know,
some
other
means
on
the
side.
D
Yeah,
so
I
don't
mean
to,
I
don't
mean
to
like
cut
you
off.
I
think
what
you're
saying
is
valid,
but
I
just
want
to
call
out
that
we
specifically
have
like
setting
the
goals
and
then
right
after
this
is
go
through.
The
issues
where
we
can
discuss
them
in
detail
get,
like
add
context,
add
opinions,
but
I
want
to
get
through
these
goals
and
then
we
can
go
on
to.
C
C
D
Sure
that
works
all
right
cool,
let's
move
on
and
then
for
number
three.
I
would
like
to
hear
from
anyone
else,
but
me
and
sam
and
milan,
so
I'm
we're
going
to
force
force,
luis
or
vikram
or
felipe
or
liam.
E
E
You
know
niche
behavior,
where
basically
I
was
trying
to
forward
a
port
that
was
not
active
yet,
and
I
forget
exactly
what
the
behavior
was,
but
maybe
something
that
would
be
useful,
especially
since
octan
is
like
a
graphical
user
interface
would
be
making
some
adjustments
to
prevent
users
from
trying
to
do
things
that
we
already
can
predetermine
and
know
that
will
like
will
not
work
correctly.
So,
basically
preventing
you
know
known
failures
from
occurring.
E
I
think
that
would
be
super
useful,
especially
for
someone
like
myself,
who
isn't
as
familiar
with
kubernetes
and
is
also
somewhat
using
often
to
learn
a
bit
more
about.
You
know
what's
going
on
and
so
with
a
specific
port
forwarding
example,
and
this
isn't
you
know,
documented
in
issues
or
anything,
but
I'm
pretty
sure
this
applies
beyond
that
as
well.
E
Just
you
know,
adding
maybe
indicators
of
like
this
resource
is
not
yet
done
loading,
so
you
can't
complete
these
actions
because
we
know
they
will
fail,
would
be
super
helpful,
yeah
a
way
to
say
this.
A
sentence
would
be
like
prevent
the
known
failures
from
occurring
that
it
might
be
useful.
D
Yeah,
I
I
I'm
liking
where
this
is
going
as
I
look
through
the
issues
under
our
object
status
theme
in
particular,
we
have
a
class
of
issues,
that's
that's
like
deleting
a
namespace
and
updating
this
thing
and
you're
not
getting
any
status
out
of
it
as
it's
happening.
It's
almost
like
when
objects
are
in
trend
in
transition
or
when
objects
are
changing
their
state.
We're
not
doing
a
great
job
of
representing
the
fact
that
we
might
want
to
hold
off
on
taking
action
while
they're
in
transition
or
while
they're
changing
state.
F
Yeah
there's
improved
external
ip
status,
looks
like
provide
better
details
when
the
external
ip
is
not
yet
set,
which
seems
to
fall
into
the
same
category.
D
Yep
yeah
there's
definitely
a
class
of
issue
here,
yeah
that
that
aligned
with
that,
I
think
the
wording
that
you
used
liam
was
was
was
close.
I
think
it
was
good.
I'm
trying
to
think.
C
Yeah
and
I
see
how
this
fits
in
the
status,
so
that's
the
excellent
point
I
mean
if
you
do
the
status
right.
This
is
going
to
be
the
benefits
because
you
know
in
order
to
to
know,
if
you
can
forward
reports
forward,
you
need
to
know
the
status.
So
if
we
business
status,
we
can
just
ask:
oh,
can
this
object
do
port
forward
and
I
don't
know
either
disable
the
button
and
do
some
other
ui
indication
that
that's
not
possible.
B
Yeah,
I
think
that
related
to
one
of
the
original
goals
is
that,
regarding
the
yamaha
editor
editor
to
make
it
make
it
object.
Aware
I
mean
I
I
also
tested,
because
I
I
created
an
object
and
I
went
to
yaml
editor
to
change
part
of
the
spec
that
I
know
cannot
be
changed
in
a
running
state.
B
I
hit
update
and
often
didn't
say
anything
right.
Obviously
the
change
didn't
happen.
I
went
to
a
different
view,
came
back
to
yaml
editor
and
the
line
wasn't
there.
So
if
you
are
a
new
user,
probably
you
can
start
hitting,
update
and
update
and
update
and
things
that
the
api
won't.
Let
you
to
update
that
field
so
yeah
in
line
with
what
liam
is,
is
mentioning
and
if
octan
is
also
a
tool
that
aims
to
lower
the
barriers
for
nuclear
united
users.
B
That
will
be
helpful
to
have,
or
you
know,
disable
user
actions
or
for
invalid
requests,
or
at
least
some.
You
know
better
error
reporting,
which
is
a
totally
different
category.
You
know
it's
a
category
on
its
own
er,
better
error
reporting.
I
think
it's
a
can
of
worms
in
itself
so
yeah.
I
think,
in
my
mind,
that's
related.
D
Okay,
okay,
great
okay!
So
that's
three
goals:
everyone
happy
with
this
or.
D
D
Going
through
the
backlog
and
starting
to
identify
issues
that
kind
of
align
with
this
theme,
the
the
other
thing
that
I
want
to
call
out
is
that
I
want
to
make
sure
that
each
each
sprint
includes
at
least
three
items
from
the
product,
excellent
column.
D
I
think
that
it's
important
that
we're
always
so
we
can
work
to
find
items
from
product
excellence
that
align
with
the
theme,
but
generally,
I
think
it's
important
that
we
are
also
fixing
bugs
reported
by
users
that
get
in
their
way,
so
that
that's
just
kind
of
the
the
like,
I
guess,
escape
hatch
there
of
like.
D
If
there
are
any
bug
issues
that
you
know
are
blocking
people
or
important,
we
will
try
to
always
get
three
kind
of
baked
into
the
sprint
from
the
start
and
they
can
be
swapped
out
and
changed
as
we
go,
but
just
like
with
any
anything,
but
just
wanted
to
call
that
out
explicitly
so
with
that
under
the
under
the
object
status
theme,
there
is,
as
we
were
scrolling
through
before.
D
There
are
a
lot
of
issues
and
I
guess
we
can
just
kind
of
work
top
to
bottom
or
we
won't
be
able
to
get
through
all
of
them
in
this
call.
So
what
I
want
to
do
is
just
get
the
process
started
so
that
we
can
do
this
asynchronously,
the
the
the
kickoff
of
the
actual
sprint
work
is
is
starting
on
monday,
so
we
do
have
the
rest
of
this
week
to
get
things
planned
and
get
things
in
here,
but
but
yeah.
This
should
be
our
primary
focus.
D
If
you
don't
have
an
active
issue
that
you're
working
on
our
primary
focus
should
be
looking
at
issues
and
commenting
on
them
getting
more
context
so
that
we
can
plan
them
into
our
r0.23,
which
is
what
we're
going
to
do.
We
have
another
meeting
set
up
tomorrow
that
we
can
record
and
post
to
youtube
that
we'll
use
to
continue
this
process
so
with
that
out
of
the
way-
let's
yeah-
I
guess,
let's
just
start
going
through
them
under
object
status,.
A
All
right,
so
we
have
first
item
here:
it's
increased
visibility
to
resources
stuck
in
a
terminating
state,
this
one's
kind
of
gnarly.
A
Essentially,
you
could
have
resources
when
you
try
to
delete
therapy
and
terminating,
but
they
use
something
called
a
finalizer
which
suggests
that
there
is
another
resource
that
is
currently
using
it
and
as
a
result,
it
becomes
stuck
in
terminating
there's
some
hacks.
You
can
pretty
much
just
edit
out
the
finalizer
or
patch
it
over
json.
A
Whether
or
not
that's
a
good
idea
is
debatable.
That
said,
we
could
probably
start
off
by.
We
don't
necessarily
have
to
solve
this
problem.
This
has
increased
visibility,
so
I
don't
know
what
that
necessarily
means
we're
going
to
invent
a
new
color,
maybe
create.
A
I
don't
know
we
have
an
object
filter
for
this.
Do
we
have
an
object
filter
for
terminating?
Maybe
we
do
phase
filter,
maybe
not-
or
I
think
this
might
be
one
of
those
phase
versus
status
things
so
maybe
maybe
there's
a
way
to
sort
this
stuff
out.
Maybe
the
first
thing
we
want
to
do
is
just
cr,
create
some
searchable
group
for
these
terminating
for
these
terminating
pods
or
objects.
D
Yeah,
I
think
what
we
do
know
is.
We
know
if
there's
a
finalizer
pending,
we
can
get
that
off
of
the
object.
We
can
surface
that
that
there
that
there's
a
finalizer
for
the
object.
So
so
we
know
that
it's
in
deleting
terminating
state
and
then
we
also
know
that
there's
a
finalizer
associated
with
it.
D
We
could
then
also
use
some
sort
of
opinion
about
what
is
a
reasonable
amount
of
time
for
an
object
to
be
deleting.
I
think
scott
andrews
raised
this
on
the
original
issue
of
like
how
do
you
distinguish
between
stuck
and
something
that
is
actually
cleaning
up?
D
I
don't
know
that
we
can
do
that
reliably,
but
we
can
at
least
give
insight
into
hey.
This
thing's
been
trying
to
delete
for
an
hour
two
hours,
three
hours
whatever
it
is,
but
like
at
some
point
when,
when
that
time
threshold
has
been
reached,
maybe
we
we
start
to
show
it
like.
There's
a
finalizer
on
this.
It's
been
trying
to
delete
for
the
last
three
hours,
here's
a
and
then
extend
the
actions
grid.
D
This
is
this
is
long
longer
term,
but
that
could
lead
to
okay
when
something
is
stuck
with
a
finalizer
for
a
set
period
of
time.
Add
to
the
actions
grid,
the
ability
to
patch
away
the
finalizer,
so
it'll
be
removed
right,
like
that's.
D
That's
kind
of
like
the
incremental
step,
so
so
yeah.
I
think
I
think
this
is
a
good
candidate.
I
actually
run
into
this
quite
a
bit,
especially
when
dealing
with
crds
and
and
like
external
tooling,
where
I'm
just
like.
G
Yeah,
I'm
trying
to
figure
out
how
to
move
this
thing
from
backlog
too
yeah,
so
it
should
be.
D
In
the
issue
there
you
scroll
down
or
on
projects
you
see
where
it
says
it
says,
projects
there,
click
on
the
little
gear
icon
and
then
it
will.
Oh
sorry,
yeah,
there's
multiple
projects
on
the
same
page
in
the
right
hand,
menu
there's
a
project's
that
yeah
and
then
you
just
23
and
off
of
off
of
backlog.
D
Cool
yeah
so
on
to
the
next.
I
think
the
next
one
is
a
larger
feature
of
like
a
larger
body
of
work
that
involves
some
research
and
design
and
is
probably
not
a
good
candidate
for
our
first
try
at
themed
based
sprints.
D
It
would
be
yeah
yeah,
there's
some
good
context
in
there.
I
think
it'd
be
a
great
candidate
for
someone
who
wanted
to
play
around
with
this
on
for
on
friday
and
and
use
use
like
half
of
their
day
on
on
fridays,
to
creep
this
forward,
so
good
one
to
call
out
it
could
potentially
be
a
very
fun
one.
D
B
D
D
So
what
I,
what
I
want
to
do
is
get
some
added
to
the
sprint
and
then
we'll
go
and
have
discussions
on
the
issue
that
is,
that
is,
that
is
the
action
item
for
everyone
on
the
team
after
this
community
meeting
is
to
look
at
the
issues
that
ended
up
in
0.23
and,
if
you
don't
understand
it
or
you
need
more
information,
ask
and
we're
going
to
fill
it
out
together.
We're
going
to
we're
going
to
we're
all
thinking
about
object
status,
this
sprint
a
little
less
context.
Switching,
it's
yeah!
F
Quick
question:
how
many
items?
How
do
we
scope
what
goes
into
this
release,
because
if
we
do
everything
from
within
this
object
status
column,
it's
gonna
be
too
much
right,
yeah,
I
skipped
one
of
them,
so
so
what
goes
in?
What
doesn't
and
then
should
we
not
complete
everything
within
this
release?
F
Does
it
just
get
tossed
back
into
the
backlog,
and
we
do
another
object
status
release
at
a
later
date?
Do
we
continue
like
do
we
roll
over
the
theme
and
I'm
sure
these
are?
I
know
these
are
things
that
we're
working
out,
but
I
thought
I'd
surface
them
anyway.
Yeah.
F
About
it,
should
we
do
another
object
status
release
following
it
since
we're
already
within
the
context
of
we're
all
thinking
about
it,
we've
been
working
on
it.
Do
we
do
another
one
and
try
and
wrap
it
up,
or
do
we
move
on
to
another
theme.
D
So
great
questions
for
what
work
goes
in,
I
would
say,
at
least
during
the
initial
experiment.
We
can
kind
of
use
our
ours
our
judgments
of
of
like
two
issues
per
person.
Some
are
small.
Some
are
big.
Some
are
medium
whatever
it
doesn't
matter,
we'll
just
say
two
per
person.
What
we
get
done,
we
get
done
when
we
don't
we
don't
to
like
establish
just
working
this
way.
I
think
that's
more
important
right
now.
D
D
So
the
goal
will
be
to
move
towards
sprints
that
are
completed
in
full,
but
as
we
establish
this
new
way
of
work,
I
think
we'll
err
on
the
side
of
less
so
that
we
can
complete
and
then
now
we
have
these
categories
to
pull
things
in
from,
whereas
before
it's
really
difficult
to
know
what
to
pull
in
so
so
I
think,
I
think,
just
getting
started
with
like
12
issues,
which
is
a
number
I
just
snatched
because
six
times
you
know
six
times
two
and
and
seeing
how
that
works
and
and
then,
if
we
get
them
all
done,
then
we
pull
in
more
and
then
as
for,
if
we're
going
to
continue
the
theme,
I
would
just
leave
that
as
a
decision
for
the
team
in
this
this,
the
next
version
of
this
exact
meeting
that
we
do
if
people
are
really
excited
about
wrapping
it
and
getting
all
of
them
done
then
yeah
we'll
just
keep
on
it
and
if
people
are
like
hey,
I
think
we
did
great,
but
I
think
there's
more
value
in
this
other
theme.
C
Yeah-
and
I
think
I
think
we
should
use
our
preset
calls
as
a
you
know,
guidance
when
we
make
those
decisions
both
when
we
decide
other
issues
going
into
the
sprint
and
at
the
end,
you
know
kind
of
measure
against
those
goals
where
we
are
and
make
decisions.
What
we're
going
to
do
with
the
remaining
work,
and
things
like
that.
D
Yeah,
I
agree
milan
and
that's
so
there's
a
as
I
mentioned
previously
we're
at
four
minutes
now.
So
this
is
this
meeting's
about
to
be
over.
We
do
have
an
hour
blocked
for
tomorrow
to
actually
go
through
the
issues
that
end
up
in
odot,
23
and
and
work
through
them
with
any
questions
that
have
been
asked
on
them
and
make
sure
they
have
all
the
context
and
everything
that
folks
need
to
feel
comfortable,
starting
those
issues
and
working
on
them.
D
So
that
we'll
record
that
and
make
sure
we
publish
that
as
well
to
the
same
octane,
playlist,
that's
on
youtube
and
so
yeah
that
that'll
be
the
process.
D
Yep,
that's
that's
yeah
arbitrary,
arbitrary
number,
15.
and
then
yeah,
we'll
iterate.
On
that
I
I
I
hesitate.
The
the
reason
I
hesitate
to
say
we're
going
to
do
any
type
of
story
pointing
or
scoring
is
because,
in
my
experience,
scores
always
trend
to
the
median
and
the
real
value
that
you
get
from
scoring
is
actually
just
getting
all
your
engineers
talking
about
an
issue.
F
D
Yep
absolutely
so
yeah
we
can.
Let's
see,
we've
got
two
minutes
left,
let's
go
ahead
and
just
wrap
up
the
the
community
meeting
with
the
last
two
minutes
and
we
will
continue
to
populate
0.23.
D
So,
if
you're,
if
you
want
to
you,
can
watch
it
happen
in
in
slow
time
as
we
as
we
move
stuff
into
there.
I
just
want
to
say
thank
you
to
everyone
for
participating
in
this
process.
I
know
it's
new
and
it
might
feel
like
there
it's
a
lot
or
that
it's
slowing
things
down
a
little
bit,
but
I
think
ultimately
will
result
in
us
having
a
better
way
of
working.
So
I
really
appreciate
everyone's
time
and
input.