►
From YouTube: Velero Community Meeting - Jan 10, 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).
A
All
right,
hello,
everyone
today
is
January
10th
and
that's
the
official
community
meeting
for
Valero,
my
name
isilif
and
I'm,
the
community
manager
for
Valero.
Please
follow
the
code
of
conduct.
That's
just
be
nice
to
each
other
I've
pasted
in
the
chat,
the
link
to
today's
and
every
other
community
meeting
notes.
A
So
please
add
yourself
or
add
audio
topics
that
you
want
to
discuss
with
that.
I'm
gonna
share
my
screen
hope
everyone
can
see.
Well,
so
let's
refresh
that
so,
first
of
all
happy
New
Year
for
those
of
you
who
I'm
seeing
for
first
time
this
year,
wish
you
all
the
best
and
we
got
Scott
I-
think
you're.
First
in
the
list
with
some
stuff
to.
B
Okay
update
so
the
backup
item,
action,
V2
implementation,
PR
that
I
had
put
together
late,
December,
finally
got
the
acts
necessary
after
review
and
that's
then
merged
as
of
this
morning.
So
thanks
for
that
everybody
I
submitted
a
new
PR
yesterday,
also
relating
to
the
overall
epic
here
for
the
async
operations,
and
this
is
the
one
to
define
the
item
operations,
Json
format
and
updated
download
request
API
to
use
this
on
the
backup
and
the
restore
side.
B
There's
also
some
housekeeping
relating
in
that
PR
as
well,
because
Dave's
original
design
that
this
is
kind
of
taking
over
for
is
had
a
new
plug-in
type
called
item
snapshotter,
but
the
new
design,
instead
of
doing
that,
just
modifies
backup
out
of
action
with
the
V2
with
that
operation,
so
I
didn't
remove
the
plugin
itself.
B
That'll
be
a
separate
PR,
but
I
did
remove
the
since
I'm
already
working
on
the
download
request,
I
kind
of
removed
it
from
the
download
request
area
that
PR
or
rather
that
item
operation,
plug-in
type
exists,
but
there's
no
controller
logic
that
touches
it.
That
does
anything
with
it,
because
that
that
was
an
Open
PI.
That
was
never.
C
B
So
that's
a
PR
or
a
plug-in
type
that
will
be
safe
to
remove
as
well,
because
there's
nothing
in
Valero
that
uses
it,
but
this
this
PR
doesn't
do
the
whole
thing.
It
just
removes
it
from
the
download
request
API
since
I
was
changing
it
anyway,
so
it
kind
of
replaces
the
backup
item
action,
V2
field,
or
rather
the
Json
instead
of
the
one
for
the
the
item
snapshotter,
so
that
PR
is
out
ready
to
review.
B
I
have
two
other
PRS
from
before
that
were
they
were
draft
PRS
because
they
depended
on
the
item.
Action
V2
changes,
so
one
of
them
is
the
restore
I.
Have
action,
B2,
API
changes
I
need
to
rebase
that
retest
it
and
then
merge
that?
Oh
sorry
rather
and
then
update
that
for
ready
for
review
and
I
also
have
in
the
plugin
examples,
backup,
atom
action.
B
V2
updates
I
need
to
rebase
that
again
and
then
that
will
be
ready
to
review
and
that
basically
adds
a
second
plugin
that
implements
the
new
API.
So
we
can
use
the
plugin
examples
to
show
that
V1
and
V2
plugins
can
coexist.
B
It
doesn't
yet
actually
return
in
the
operation
IDs
because
we
don't
have
the
controller
logic
yet
so
I
will
be
updating
the
Valero
plugin
example
as
a
way
of
testing
that
as
well
eventually,
but
right
now,
we're
just
testing
that
you
can
register
a
V2
plugin
nv1
plug-in
from
the
same
plugin
image
and
Valeria
will
call
both
of
them
through
the
adapter
for
the
V1.
So
those
will
be
ready
to
review
later
this
week,
but
the
item
operations
gsm1
is
ready
to
review
right
now.
B
Would
say
the
the
5752
that
one
is
ready
to
review?
That
would
be
great
to
get
attention
on
that
one
this
week.
If
we
can
get
that
merged,
then
I
can
move
on
because
once
that's
merged
and
because
we
already
have
the
plugins
site
merged,
then
I
can
start
working
on
the
actual
controller
logic,
which
is
kind
of
the
the
biggest
part
of
this.
A
B
We'll
do
yeah
I,
actually
think
of
this.
One
I
I
actually
made
a
mistake
and
painting
on
the
user's
Channel
instead
of
the
dev
Channel,
but
but
yeah,
but
I
did
I
did
Ping
for
both
of
these
as
well.
I'll,
probably
update
that
once
I
get
those
other
PRS
ready
to
merge,
I'll
ping,
the
Channel
with
all
three
M
again
as
well.
A
B
I
would
say
of
these
three:
the
item
operations
Json,
which
is
the
5752-
that's
the
highest
priority,
because
that
one
is
needed
to
do
the
next
step,
which
will
be
the
backup
controller
logic,
because
the
restore
side
will
be
the
kind
of
second
phase
here,
so
those
other
two
PR's
that
will
be
ready
to
review
later
this
week,
that'll
be
great
to
get
them
in,
but
those
are
lower
priority
right
now,
because
the
the
item
operations,
Json
PR-
will
become
a
blocking
thing
sooner
than
the
others.
A
Sure,
right,
if
you
need
that
many
assistance
from
us
hit
again
just
drop
a
line:
okay,
okay,
all
right
all
right
thanks!
So.
B
Yeah
5752
is
ready
to
review
now
that
is
the
highest
Priority
One
right
now,
the
5569
and
the
six
six
under
the
plug-in
examples.
First
of
all,
those
aren't
ready
to
review
yet
anyway,
I
need
to
update
those
from
draft
after
I
rebase
them,
but
also
those
won't
really
block
my
work
until
I
get
to
the
restore
side.
B
The
the
this
one
here.
The
item
operations
includes
backup
and
restore
related
changes
needed
to
download
requests,
and
this
this
provides
the
ability
to
take
that
list
of
item
operations
with
their
current
progress,
upload
those
to
object,
storage
and
download
them.
That's
what
this
part
that's
the
purpose
of
this,
which
will
be
needed
once
we
get
to
the
controller
exchanges.
D
Okay,
oh,
could
you
please
also
add
you
or
ask
the
reviewer
because
he
will
be.
You
know,
looking
over
all
these
piarab2.
D
Yeah,
it
will
be
better
if
you
can
add
in
the
you
know
better
panel
yeah.
That
will
be
even
there.
Yeah
thanks
yeah.
B
D
Think
his
could
have
Handler
is
Linden
l-y
and
yeah.
D
B
D
Yeah
I
I
joined
this
meeting
just
to
give
a
heads
up
regarding
the
latest
update
of
1.11
planning.
We
have
triage
all
the
Wonder
gallon
candidates,
I
think
there's
only
one
remains
others
priorities
Garden
some.
We
have
decided
to
move
out
of
one
another
and
the
others
we
put
them
in
the
1.11
Milestone
and
in
order
to
meet
the
theater
freeze,
Milestone,
which
is
targeting
the
end
of
the
week.
So
the
only
one
remaining
with
the
issue.
D
B
Direction
I
I
think
so
so
so
that's.
B
Action
API
PR
includes
the
API
needs
for
that,
and
the
other
thing
is
that
I,
actually
I
already
have
a
PR
from
about
a
year
ago.
That
implemented
this.
So
I
do
need
to
rework
that
and
refactor
that
to
work
with
the
current
code
base,
but
I
mean
I'm
not
starting
over
with
the
feature.
I
already
have
a
PR
that
I
just
need
to
update,
so
that
shouldn't
be
a
problem
to
get
in,
because
the
building
blocks
are
already
there.
I
just
need
to
kind
of
put
them
together.
In
the
current
code
base.
B
D
Unfortunately,
not
because
we
just
you
know,
go
through
those
issues
and
based
on
the
discussion,
remove
the
ones,
but
mainly
those
1.11
candidates
were
added
by
pradeep
and
we
confirmed
with
him.
Those
are
I
mean
all
the
remaining
candidates
he
added.
It's
not
must.
D
D
Me
yeah
just
in
case,
if
you
you
can
check
the
Milestone
v1.11
and
if
there's
any
issue
your
you
know
interested
not
in
the
Milestone
just
feel
free
to.
Let
me
know-
and
we
can
you
know
rediscuss
it,
but
bear
in
mind
that
this
is
a
relatively
short
release.
D
In
order
to
meet
the
future
completed,
we
may
miss
it.
We
may
delay
it
for
a
couple
of
weeks
and
we
may
also
move
a
few
issues
out
of
the
Milestone.
If
we
want
to
you
know,
we
don't
want
to
delay
too
much
so
that
may
happen
just
I'll
give.
You
has
asked
on
that:
okay,
okay,
yeah
I!
Think
that's
regarding
the
1.11
planning
and
I
I
have
one
discussion
item:
shall
we
do
that
earlier
or
in
and
I
I
may
drop
early,
since
it's.
D
Thank
you,
so
I
wanna
do
a
quick
discussion
with
Scott
regarding
the
pr
five
three
three
three
yeah.
A
A
D
Yeah
this
week,
I
I
spent
some
time
to
discuss
with
Shin
regarding
it
yeah
I
wanna.
Let
me
share
my
screen
yeah
here,
so
this
one
on
the
important
use
case.
Let's
discuss
this
Scott
and
the
current
default
Behavior
yeah.
C
B
Based
on
the
way,
this
proposal
is
written
right
now
in
the
pr
that
this
part
doesn't
change.
Basically,
right
now,
there's
three
there's
three
values
for
the
way
this
is
set
up,
you
set
it
to
false
and
nothing
cluster
scoped
is
included.
We
completely
skip
anything.
A
C
B
It
to
true,
then
that
means
we.
We
include
everything
cluster
scoped
subject
to
includes
excludes
so
that
so,
if
you
set
it
to
true
and
then
you
list
the
resources
out
there,
you
want,
then
those
are
included
in
nothing
else.
If
you
set
it
to
true-
and
you
know-
and
you
include
all
cluster
scope,
but
then
you
have
exclude
resources
set
or
rest
exclude
cluster
Skype
resources
that
then,
then
those
are
removed.
The
the
part
that
would
be
missing
if
we
remove
this
field
completely.
B
Is
that
there's
the
default
value
where
it's
set
to
nail
or
or
absent,
and
this
is
a
special
setting
that
tells
Valero
pull
in
specific
cluster
resources
as
needed.
This
is
actually
important
for
the
for
the
PVC
use
case,
because
what
that
does
is
basically,
if
you,
if
it's
set
to
nil,
then
the
only
cluster
scope
resources
we
include
and
again
everything
subject
to
includes
excludes,
but
the
only
ones
that
we
include
are
going
to
be,
for
example,
TVs
that
are
related
to
people.
You
know
we
include
the
PV
yeah
yeah.
D
Yeah
yeah,
yes,.
B
Same
with
same
with
CRVs,
you
back
up
a
bunch
of
CRS
and
we
in
fact
we
have
another
issue.
There
I
noticed
that
a
user
is
saying,
hey
I,
don't
want
to
include
all
the
PVS
in
the
cluster
when
I'm
backing
up
one
namespace,
because
it's
a
bunch
of
things
that
you
know
we
don't
need.
D
Yeah
I
understand
so,
but
but
my
thought
is
because
the
philosophy
of
Valero
from
my
understanding
that
it
should
always
try
to
make
sure
the
workload
work
after
the
restore.
So
we
just
make
it
always
the
default
Behavior.
Even
after
we
remove
this
flag.
Wouldn't
that
work.
B
B
B
Something
I
guess:
here's
the
part
that
this
to
me
about,
if
we
get
rid
of
another
example,
how
between
a
use
case
where,
where
the
user
says
I
want
to
include
all
the
PVS,
so
you
list
PV,
as
you
know,
versus
a
use
case
where
we
say
I
only
want
to
include
the
relevant
PVS.
You
know,
how
do
you
tell
Valero
to
do
one
versus
the
other.
D
Yeah
yeah
I
think
that's
possible.
I
won't
submit
the
comment,
but
let
me
type
here
I
think
to
include
all
the
PVS
regardless,
if
necessary
or
not.
We
just
add
this
one,
so
we
explicitly
tell
Valero
I
want
TVs.
Okay.
If
we
remove
that
one
we
just,
we
just
include
the
necessary
ones
and.
B
B
So
so
yeah,
so
the
one
change
that
would
be-
and-
and
this
may
be-
okay
but
I
I-
want
to
see
if
anyone
else
here
can
see
a
use
case
where
this
is
wrong.
What
this
would
mean
is
that
it
would
be
impossible
to
run
a
backup
where
you,
where
you
back
up
a
PVC,
but
you
don't
include
the
bound
PV.
That
may
be
fine
because
it
may
be
useless
to
back
up,
but.
B
D
Yeah
first
I
think
what
you
mentioned
is
a
really
rare
case.
I,
don't
think
anyone
really
wants
that,
but
I
also
want
to
point
out.
We
will
not
lose
that
ability.
If
we,
you
know
we
just
let
user
you
know
set
this
flag
sell
flag
like
this,
so
so
so
the
idea
is
in
backup.
We
exclude
attribute,
has
higher
priority.
So
if
I
explicit,
if
explicitly
tell
malaria
to
to
exclude
this
one
and
the
certain
resource
will
always
be
excluded.
D
Therefore,
in
in
in
your
use
case,
we
just
tell
Valero
to
explicitly
tell
Valero
to
exclude
some
TV
so
that.
B
Okay,
so
so
so
that
also
means
there's
here's
another
thing,
that's
different
in
the
current
logic,
but
both
the
wavelier
works.
Now
with
that,
you
know
with
just
the
the
simpler
include
expert
resources,
but
also
in
the
proposal
in
this
PR,
as
it
stated,
is
that
Valero
considers
a
missing
resource
and
the
includes
as
equivalent
to
listing
it
in
the
excludes.
This
would
be
a
change
from
that
that
might
require
some
changes
to
the
excludes
model,
because
right.
D
B
C
B
That
current
model
there's
no
way
that
Valero
could
distinguish
between
PV
in
the
exclude
list
or
PV
missing
from
the
include
list.
If
we
intend
to
make
that
distinction
relevant,
then
we're
gonna,
we're
gonna,
have
to
change
them
and
the
internal
model
change
apis,
but
we'll
have
to
change
the
internal
model.
So,
instead
of
excludes
just
removing
items
from
the
include
list
excludes
is
going
to
have
to
actually
be
maintained
as
a
separate
list
on
the
call
stack
and
then
the
then
the
lodge
and
then
we'll
need.
B
You
know
help
our
mouth
is
to
basically
distinguish
because
right
now,
there's
two
categories
included
and
excluded:
we're
adding
a
third
category,
included,
nothing
and
excluded,
and
that
that
middle
default
category
is
only
relevant
for
cluster
scope
resources.
That
middle
category
says
included
as
if
relevant,
for
namespace
resources.
B
That's
going
to
be
the
same
as
exclude
so
we're
going
to
probably
need
it
when
we
do
this
implementation,
we're
going
to
have
to
keep
that
model
in
mind
to
say
what
what
I?
Basically
we're.
We
still
need
that
concept
of.
We
have
three
states.
You
know
it's
not
quite
as
simple.
As
you
know,
a
Boolean,
it's
not
included
for
his
exclude,
and
it's
explicitly
included
this.
You
know
default
and
explicitly
excluded
yeah.
B
And
basically,
for
cluster
scope
resources,
instead
of
doing
that
at
a
top
level
of
all
cluster
scope
reasons,
we're
doing
this
on
a
resource
level,
so
explicitly
included
means
we
always
include
it
explicitly.
Excluded
means
we
never
included,
and
that
default
level
for
cluster
scope.
Resources
means
we
only
included
as
needed,
following
the
same
logic
that
we
currently
do
with
the
include
cluster.
D
Resources
right
right.
B
B
Well,
except
if
we,
except
by
if
we
say,
include
pods
and
deployments,
for
example-
and
we
don't
list
anything
else,
current
Valera
Behavior
means
only
include
those
things
include
everything
else,
so
you're
right
default
is
include.
If
we
don't
say
anything,
but
if
we
have
a
if
we
have
an
explicit
includes
list.
C
B
C
B
And
I
think
the
same
would
be
true
in
this
new
model,
where
you
say
include
namespace
resources
deployments.
Anything
not
in
that
list
is
excluded
because
there's.
D
B
Think
that
middle
ground
doesn't
exist,
you're
right
the
default
is,
you
know
if
you
say,
include
namespace
resources
star,
because
there's
our
comments
about
why
boy
that's
needed
so
for
consistency
with
the
new
resources
but
but
I
think
I
think
we
can
get
rid
of
that
field.
If.
B
Again,
if
we
make
use
of
the
you
know,
if
we
treat
the
if
we
change
the
model
so
that
that
middle,
you
know
not
excluded,
not
included
for
cluster
resources,
has
that
specific
meaning?
That's
equivalent
to
this?
You
know
Legacy
parameter,
but
for
namespace
resources.
That
notion
of
include
as
relevant
doesn't
really
mean
anything
because
there's
nothing
in
Valero.
That
has.
D
B
So
the
only
other
thing
is,
it
would
mean
I
I,
don't
know.
If
that
would
mean
we
would
remove
the
field
or
we
would
include
the
field
with
deprecation
because
again,
there's
there's
the
of
the
whole
point
of
removing
it
would
be
to
get
rid
of
the
confusing
logic
around
that
yeah.
D
D
Field
is
younger,
yeah
I,
think
I
think
this
one
will
be.
You
know
these
two
Fields
I
mean,
ideally
in
my
opinion.
After
we
introduce
this
a
new
flag
or
attribute
these
two
attributes
will
be
mutually
exclusive,
so
this
one
is
sad.
You
know.
C
B
Know
we
have
the
old
include
resources.
We've
already
stated
in
this
design
that
include
resources.
E
B
Include
you
know,
name,
statistics
and
in
other
words
you
need
to
use
either
the
old
included,
actually
resources
or
the
new
split
resources.
I
think
in
that
case
we
still
keep
this
field,
but
we
use
that
those
that
go
together.
In
other
words,
instead
of
just
saying,
include
exclude,
you
know
Legacy
or
exclusive
exclusive
with
the
new
Fields.
It's
all
three
of
the
Old
Fields
they're
mutually
exclusive,
because
we
have
three
current
fields
and
then
we
have
four
new
fields
and
basically
you
have
to
either
use
only
new
fields
or
only
Old,
Fields
and.
A
B
To
say
oh
you're,
using
the
new
Fields,
you
can't
use
the
old
ones
anymore,
or
vice
versa.
That
would
probably
work.
Then.
B
If
we
do
that
that
this
is
going
to
require
some
changes
to
the
internal
structs
around
includes
excludes
because
we're
gonna
I
wish
we
had
to
change
anyway.
Obviously,
this
whole,
this
whole
proposal
is
going
to
involve
a
lot
of
changes
to
that.
So
that's
fine
we're
already
changing
code
anyway,
but
we
need
to
basically
track
all
three
categories
for
cluster
scope.
Resources
of
explicitly
included,
yeah,
you
know
include,
is
necessary
and
excluded,
but
for.
D
B
And
also
I
just
there's
another
issue
that
was
someone
submitted
that
you
commented
that
this
proposal
would
fix
their
issue,
and
you
know
I
responded
to
that
earlier
today
saying
it's
right:
here's
the
case
that
didn't
handle
it
I
think
what
we're
talking
about
here
probably
does
solve
that,
because
the
the
the
the
the
the
challenge
there
was
they
wanted
to
include
cluster
roles,
all
of
them,
but.
B
D
D
Okay,
okay,
so
so,
if
you're,
okay
with
that
I
will,
you
know,
work
with
student
to
refine
this
proposal.
B
Yeah
exactly
if
you
can
get
those
updates
to
hear
kind
of
listed
in
the
design.
Clearly,
then
I
can
review
that
and
we
can
kind
of
go
iterate
over
that
to
make
sure
we're
clear
on
what
the
requirements
are.
I'm
all
right,
mostly.
B
What
I'm
trying
to
make
sure
we
do
here
is
that
I
know
we're
trying
to
provide
new
functionality
that
will
help
users
to
be
a
little
more
flexible
in
their
using
Valero,
but
I
just
want
to
make
sure
that
we're
not
losing
any
real
world
use
cases
I
mean
if
it's
an
edge
case
that
no
one
cares
about
like
you
know
that
we're
fine,
but
let's
just
make
sure
that
we're
not
losing
a
use
case
that
we
actually
have
users
that
want
and
I
think
what
we're
talking
about
here,
we'll
do
that.
B
We
just
have
to
be
careful
to
make
sure
you
know
that
we're
handling
all
those
edge
cases
that
we
care
about.
F
Quickly,
I
want
to
kind
of
share
one
use
case,
and
probably
you
can
see
if
that
works.
For
example,
I
have
an
application
which
I
have
a
PVC
and
then-
and
there
is
a
PV
which
stores
the
logs
only
so
now
when
I'm
restoring
that
application,
I
don't
want
to
hold
the
log
I
still
want
empty
PV
so
that
when
application
starts
again
from
the
backup,
it
has
a
volume
where
it
can
dump
the
new
lock
but
I,
don't
know
the
whole
lock.
So
I
still
want.
D
B
B
For
your
backup,
you
could
exclude
the
PV
from
the
copia
backup
to
say:
don't
back
up
this
PVE
content,
but
the
actual
kubernetes
resources
are
still
there.
So
it's
a
little
tricky
because
normally
Valero.
C
B
You
tell
it
to
use
copy
Elise,
copia
or
breastic.
If
you
tell,
if
you
tell
it
not
to
it'll,
try
to
do
a
native
snapshot,
but
it
if
that
log
happens
to
be
on
a
volume
that
you're
not
configured
for
snapshot.
For
example,
you
know,
if
you
don't
have
a
volume
snapshot,
location
figured
then
if
you
tell
it,
don't
use
copia
for
this
volume
and.
B
Using
snapshots
and
you're
not
using
CSI,
then
it
won't
be
backed
up
if
you're,
using
CSI
I,
don't
know
if
there's
a
way
right
now
in
the
CSI
plugin
to
say
for
this
one
volume
don't
back
up
to
contents.
F
B
But
I
think
the
scope
of
where
we
need
to
talk
about
the
question
of
I
want
to,
for
this
particular
volume,
not
back
up
the
contents,
just
back
up
the
fact
that
it
exists
I
think
that's
that's
in
the
scope
of
CSI
plug
and
copia
rustic,
not
in
the
scope
of
or
conclude
resources.
Okay,.
B
D
Do
it
yeah
yeah
that
that
would
be
an
additional
enhancement?
You
know
to
the
to
the
one
you.
B
On
telling
Valero
I
want
to
use
rustic
copia
for
this
volume
versus
I
want
to
use
CSI
snapshot
for
this
volume
and
right
now,
Valera
treats
that
as
a
Boolean
of
you
either
use
file
system,
backup
or
you
use
some
sort
of
snapshot-based,
backup
and
Valera
doesn't
have
a
notion
of
I
want
none
of
the
above.
That
may
be
something
we
need
to
add
to
say:
hey
I,
want
the
the
resource
backed
up,
but
don't
use
a
file
system.
It's
not
shouting
backup
with
the
contents.
B
That's
something
I
think
Miller
is
missing
right
now
that
we
need
to
think
about
how
to
solve
it
in
the
context
of
the
fact
that
we're
it's
got
to
work
with
CSI
and
snapshots
and
rustic
and
possibly
data
movers.
So
it's
probably
relatively
easy
to
implement,
but
I
think
the
requirements
of
edge
cases
we're
going
to
require
some
discussion
about
how
to
handle
it
in
a
way.
That's
easy
for
a
user
to
understand.
A
Thanks
for
that,
okay
I'll
try
to
share
again
that
thing.
What
is
it.
A
All
right,
I've
added
few
stuff
to
discuss
like
the
release
process
in
documentation.
This
one
can
we
drive
this
forward
place
and
I
have
asked
for
pradeep.
A
Can
you
suggest
an
update
about
the
the
Milestone
usage
when
we
have
the
candidates,
how
we
prioritize
them
and
how
we
add
them
to
the
milestone
for
for
the
next
release?
That's
not
in
that
document.
For
now.
F
Okay,
I'll
look
at
this,
but
at
my
view,
like
anybody
can
say
this
is
the
candidate
I
want
and
then
the
candidates
get
selected
based
on
the
team
wounds.
Some
places,
maintainer
may
have
a
say
yes
or
no,
depending
upon
the
kind
of
current
functionality.
Would
it
break
something
and
all?
But
as
long
as
there
is
no
challenge
with
the
kind
of
current
functionality,
it's
a
new
functionality.
If
you
have
funding
your
own,
then
it's
chances
of
very
high.
F
A
Yeah
thanks.
My
next
thing
was
I've
opened
not
too
long
ago
code
owners
issued
to
discuss
if
we
want
to
have
code
owners
or
all
to
assignees,
because
right
now
we
have
autos
and
East
for
for
PS,
which
is
assigning
the
author
of
The.
The
pr
is
that
okay
for
everyone
or
we
need
some.
My
idea
was
to
have
more
visibility
when
new
PR
comes
in.
A
B
B
List
for
different
directories-
I
I-
think
excuse
me
obviously
for
for
plugins
and
for
you
know
the
different
in
the
different
repos.
You
know
we
we
have
that
the
ability
but
I
think
within.
B
I
I
don't
know
that
there's
enough
specialization
among
current
maintainers
that
that
really
makes
a
whole
lot
of
sense,
I
think,
especially
not
at
the
directory
level.
It's
not
like.
Okay
I'm,
you
know,
I'm
the
backup
expert
and
you
know
shubham
is
a
restore
expert
or
whatever
I
think
so,
I
think,
right
now
and
when
there
is
a
logical
person,
that's
maybe
not
the
default
it
more
has
to
do
with.
Oh,
this
person
is
interested
in
this
particular
feature,
but
the
feature
crosses.
C
B
Times
like
the
feature
I'm
working
on
with
the
async
operations,
you
know
it's
it's
plugins,
it's
it's
crds,
it
kind
of
crosses
the
the
the
controllers.
You
know
it.
Our
features
tend
not
to
be
isolated,
so
directories
just
the
way
the
project
is
architected.
A
Great
I
was
just
wondering
if
we
can
make
the
the
process
of
assigning
folks
to
review
and
stuff
a
bit
more
like
not
the
physical,
but
that's
all
perfectly
fine.
A
My
next
thing
was
the
honoring,
the
restore
PV
flag,
which
is
and
show
topic
and
I'm
gonna
leave
it
for
him,
and
I
have
two
very
social
ones.
Anyone
going
to
force
them
in.
B
B
I
may
be
there.
I
may
not
partially
depends
on
the
talk
that
we've
submitted.
I
was
talking
internally
the
way,
but
this
this
possibility
I
could
be
there
either
way.
I'm
just
not
sure
at
this
point.
A
Yeah
yeah
I'm
I'm
asking
because
we
chatted
with
us
on
the
background
to
do
some
kind
of
co-located
Valero
event,
maybe
at
kubecon
yeah.
So
we
can
gather
and
have
drinks
and
chats
not.
A
B
If
I,
if
my
talks
approved
or
redhead
approves
my
budget,
then
you
know
I'm
there,
but
I,
obviously
nothing's
affirmed
at
this
point.
A
B
D
E
Yeah,
so
we're
all
fighting
that
battle
sounds
like
do
you
guys
have
any
suspicions
of
when
you
might
know
for
sure
who
from
Valero
VMware
is
going
or
not
going
like?
Are
we
weeks
or.
A
The
first
concrete
date
will
be
when
the
the
cfp
announcement
is
out,
which
means,
at
least
for
VMware
means,
if
you
have
a
talk,
you're.
B
A
Right
and
then
on,
the
next
phase
is
like
with
the
acquisition.
Now
it's
total
Madness.
Okay,
maybe
I'm
not
supposed
to
share
that.
But
it's
like
another
round
of
discussions
who
is
like
eligible
to
go
or
not
sure
sure,
so
that
will
take
another
few
weeks.
G
A
But
but
I'm
trying
internally
to
get
the
idea
for
the
whole
project
who's
willing
to
travel.
That's
the
first
thing,
because
there's
two
folks
that
are
not
willing
to
travel,
and
the
second
thing
is:
if
they're
willing
are
they
allowed
in
terms
of
budget
to
travel,
so
I
hope
I
can
have
some
some
numbers
soon.
So
we
can.
E
A
A
I'm
related
to
that
I'm,
working
on
the
background
to
get
some
budget
for.
For
that
event,
even
like
a
dinner
or
something
like
that,
so
we
can
thing
around
a
bit
yeah
perfect,
which
includes,
by
the
way
swag
for
for
kubecon
and
some
other
stuff,
yeah,
all
right.
A
Anyone
else
any
topics
on
this
one,
if
you
feel
like
it,
you
can
add
a
comment
over
here
if
you're
planning,
because
there
are
two
different
things
right,
your
willingness
to
travel
and
your
planning
and
you're
if
you're
gonna
be
allowed
or
not
in
terms
of
money.
A
G
So
just
hoping
to
get
some
traction
on
both
of
these
PRS.
We
have
been
pending
for
a
few
months
now
for
the
restore
previous
plan.
I
think
we
discussed
in
previous
meetings
as
well
and
I'm,
not
sure
if
there's
any
conclusion
from
you
folks
yet
on
this
bisc
shubham
is
also
in
the
call
today.
So
hopefully
we.
B
Can
explain
this
I
can
provide
a
little
bit
of
effort
on
the
first
one.
I
was
actually
talking
to
Shiba
about
that
earlier
today.
My
last
comment
that
you've
probably
seen
basically
to
restate
the
issue
right
now
we
have
the
restore
PVS
flag
and
then
there's
the
snapshot
volume
flag.
The
snapshot
volume
flag
determines
whether
the
volumes
are
snapshotted
on
the
backup
side,
and
there
was
RPV.
Slag
is
the
equivalence
on
the
restore
side
both
of
those
are
currently
used
for
the
native
snapshot
for
CSI
snapshots.
B
The
backup
side
uses
that
same
flag
as
native
snapshots.
The
restore
side
does
not
so,
as
I
was
pointing
out
in
my
comment.
We
have
two
two
options
because
either
way
we
need
this
functionality.
We
either
need
to
continue
with
the
current
process
of
you
know,
piggybacking
on
the
existing
internal
fields
to
use
for
the
CSI
plug.
In
same
way
we
do
for
the
native
snapshots.
The
alternative
would
be
to
stop
using
snapshot.
Volumes
for
CSI
come
up
with
the
CSI
specific
annotation
on
the
backup
to
say
for
the
CSI
plugin.
B
Here's
how
we
control
this
and
again,
because
it's
a
plug-in
we'd
use
an
annotation,
not
a
native
field
and
I
was
talking
to
you
about
my
thing,
and
maybe
you
can
give
your
opinion
here,
but
I
think
Shiba
was
kind
of
the
opinion
that,
between
those
two
options
being
explicit
about
using
the
the
kind
of
separate
annotation
for
the
CSI
plugin
probably
makes
sense,
because
for
one
thing
that
I
that
that
insulates,
the
CSI
plug-in
from
changes
to
the
wave
Valero
handles
native
snapshots.
So
that
might
be
a
little
more
stable
for
CSI.
B
On
the
restore
side
as
your
PR
does,
it
might
be
better
to
remove
that
connection
on
the
backup
side
and
in
the
same
PR
implements
using
annotation
for
both
backup
and
restore
to
do
what
you
need
so
I
guess.
First,
you
about!
Is
that
accurate
as
to
what
you
were
suggesting
yeah.
C
So
my
appointment
I
agree
with
the
changes,
but
if
you
even
see
the
API
descriptions,
the
snapshot
volumes
and
restore
PVS
Flags
cater
to
Native
volume
snapshots,
not
the
CSI
ones.
So
maybe
during
the
implementation
of
CSI
plugin,
it
was
an
oversight.
So
we
should
remove
that
and
I
use
another
approach,
specifically
The
annotation
one
if
possible.
Like
that's
my
opinion,.
D
I
I'm,
not
I,
I,
don't
quite
follow
the
idea.
I
I,
don't
think
I
mean
from
user
experience
perspective.
Maybe
users
shouldn't
care
if
it's
CSI
or
or
not
right.
They
just
want
to
take
snapshot.
If
you
follow
that
principle
or
idea,
it's
probably
okay,
to
reuse
that
and
extend
from
the
Native
snapshot
to
CSI.
Do
you
see
a
problem
with
that
football.
C
B
Approach
would
be
more
work
from
a
user
point
of
view.
I,
don't
know
what
you
I
mean.
It
might
be
easier
to
add
the
field,
because
it's
there
it's
well
documented.
The
problem
with
annotations
is
they're
a
little
harder
to
document
the
only
risk
here-
and
maybe
this
is
a
slow,
a
low
risk.
Is
that
kind
of
ties
the
CSI
plug-in
to
the
internal
implementation
of
the
way
we
handle
internal
snapshots,
and
so
that
means,
if
we
later
change
the
way
snapshots
work
in
Valero
that
affects
the
CSI
plug-in.
G
B
Happen
that
may
be
okay,
so
you
know
the
The
annotation
approach
kind
of
makes
the
two
independent
of
each
other,
which
means
one
is
less
likely
to
affect
the
other.
But
the
shared
approach
is,
from
a
user
point
of
view,
probably
a
lot
easier,
because
the
same
Fields
work
in
both
cases,
I
I,
think
either
approach
is
fine.
B
G
C
The
same
thoughts
from
my
side
as
well,
it's
either
or
of
those
things,
those
two
things.
G
Yeah
I
would
like
I
like
to
make
the
case
mainly
from
the
user
point
of
view
right.
Most
of
the
cloud
providers
are
migrating
to
CSI
and
that
might
be
the
path
going
forward
and
not
the
Valero
way
of
snapshotting
for
more
scenarios.
So
I
think
if
anyways,
that
becomes
a
default
in
the
long
run
and
like
the
AP
in
the
API,
it's
already
there
in
the
CI
light.
It's
already
there
right.
Why
not
do
that?
B
B
G
That
was
the
first
PR
and
the
second
one
was
around
backup
results
that
ER
is
reviewed
by
Scott
and
I.
Believe
I.
Think
Daniel
also
added
it
to
the
Milestone,
but
just
wanted
to
bring
some
traction
for
review
on
that
particular
PR.
I'll.
G
B
B
It
used
to
be
the
case
where
back
up
on
the
backup
errors
and
restorers
were
in
the
log
file,
but
not
in
the
described
results
and
on
the
restore
side.
They
were
in
the
describe
results,
but
not
in
the
log
file.
So
it
was
completely
consistent.
B
My
original
intent
when
I
went
to
fix
that
was
to
to
make
it
consistent
and
show
up
in
both
places.
I
got
halfway.
There
I
implemented
a
PR
that
added
restore
warnings
and
errors
to
the
log
file,
because
that
was
kind
of
our
pain
point
I
mean
we
would
get
customers
that
would
provide
us
the
logs,
but
they
didn't
include
the
areas
and
warnings
because
they
were
undescribed
only
so
that
was
kind
of
the
bigger
pain
point
for
us
at
the
time
implemented
that
that
was
marriage.
So
we
have
it
now.
B
I
never
got
around
to
doing
the
second
part,
which
was
also
making
the
described
results
consistent
across
both.
So
you
get
warnings
and
errors
on
the
describe
side
for
backups,
and
this
pair
does
that
so
that's
kind
of
where
we
are
with
that.
So
I'm
glad
you
put
this
in
place
because
you
know
I
never
got
around
to
doing
it
myself
earlier.
So
we
just
need
one
more
reviewer
to
this.
A
Thank
you
without.
Let
me
refer
just
one
one
more
time
if
someone
else
edit
anything
anyone
else,
I
think
we
exhausted
the
list
of
topics
added
for
today,
but
still.
A
Anyone
no
thank
you
very
much.
Everyone
have
a
great
rest
of
the
day
and
possibly
evening
Daniel
for
you
and
for
me
and
see
you
next
time
then.
Thank
you.
Thank
you
bye.
Thank
you.