►
From YouTube: Velero Community Meeting - April 4, 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
Foreign,
so
why
don't
you
take
it
away?
Scott.
B
You
said
one
PR
merged
over
the
past
week,
I
had
a
bug
fix,
found
relating
to
backup
deletion
and
the
new
phases,
specifically
with
the
finalized
phase.
That's
been
merged.
B
I
still
have
that
example
plug-in
for
back
about
action
from
action
B2,
and
the
main
thing
we
need
is
this
for
other
than
obviously
letting
other
people
know
how
to
write.
These
plugins
is
also
the
Indian
test
is
going
to
rely
on
this,
so
we
need
to
get
that
reviewed
and
merged
as
well
and
I
have
started
working
on
the
intest
code
as
well,
but
there's
no
PR
yet
for
that
one
hope
to.
B
C
Yeah
because
we
are
trying
to
tag
rc1
around
the
end
of
next
week
or
early
I
mean
end
of
this
week
or
early
next
week.
So
so
do
you
think
so
so
I
think
it
will
be
quite
challenging
for
intern
to
merge
before
the
RC
right,
yeah.
B
I
I'm
not
sure
about
that
part
of
the
issue
is
again:
we
don't
have
the
the
Prevail
plugin
yet
I'm
working
on
it.
I,
don't
know
you
know
whether
it'll
be
done
by
then.
If,
if
we
get
that
in
the
time
frame
for
rc2
is
that
I
mean?
Obviously
we
want
the
testing
you
know
done
before
release,
so
that
would
be
a
good
thing
to
get
in,
but
yeah
it's
not
100
clear,
just
because
also
I
haven't
done
a
lot
of
work
with
the
Indian
test.
C
B
And
all
that,
but.
C
C
With
grammar,
so
so,
maybe
next
week
in
the
a
community
meeting
Beijing
time
zone
we're
going
to
discuss
whether
we
wanna
you
know,
delay
the
Indian
test
host.
The
ga
I
think
that
that
will
also
work
for
us.
Okay,.
B
I
mean
either
way
I
need
to
get
it
done,
but,
like
you
said,
merging
it
right
before
ga
only
to
fail
the
test
and
delay.
Ga
may
not
be
the
best
way
of
handling
it,
because
but
yeah
we'll
see
where
we
are
on
Tuesday
or
next
week
at
the
meeting,
either
way
at
a
very
minimum.
I
should
have
a
draft
PR
App
by
then,
hopefully
you
know
done,
but
you
know,
obviously,
if
we
have
issues
with
the
passing,
you
know
all
that
you
know
we'll
have
to
come
back
to
that.
B
A
Okay,
so
how
about
Saturday?
Do
you
want
to
go.
B
Oh
I
had
a
quick
question
since
Daniel
mentioned
with
ga
sorry,
the
the
rc1.
So
at
this
point
the
target
is
end
of
this
week
for
the
rc.1.
C
Yeah
yeah
we
are,
we
are,
we
have
been
running,
I
mean
the
post
FC
test
or
a
bug
bashing
for
over
a
week
now
in
Beijing.
So
in
today
we
had
a
SIM
card
and
we
believed
that
we're
gonna
Target,
the
rc1
around
Friday,
we're
gonna
cut
the
branch.
D
E
Okay,
Daniel
should
I
go
next.
This
is
Satya
yeah
cool
Daniel,
it
looks
like
the
screen
share
is
disabled.
Is
it
possible
to
enable
screen
shares.
E
I'm
able
to
share
now
perfect
so
again,
just
a
quick,
intro
I
know
some
of
you
were
in
there
in
the
in
the
community
community
call
where
I
kind
of
walked
through
what
we
were
working
on
alongside
Valero.
E
So
just
just
a
quick
overview
is
I
think
we're
essentially
looking
to
release
a
product
around
Valero.
It's
a
service
around
Valero
which
will
allow
customers
to
essentially,
you
know,
manage
Valero
instances
across
multiple
clusters.
E
In
one
interface,
it
will
be
delivered
as
a
service
and
a
customer
would
be
able
to
use
an
existing
Valero
install
if,
if
they
already
have
Valero
installed,
they
can
still
utilize
the
the
service
and
we'll
simply
essentially
catalog
all
of
the
backups
resources
all
of
the
CRS
and
effectively
allow
them
to
manage
their
Valero
instances
through
our
our
SAS
portal
and
and
of
course
they
can
create
new
resources
as
well.
So
you
can
do
all
of
the
current
operations
on
the
various
CRS
through
our
SAS
portal.
E
So
just
a
quick
introduction
again
on
on
what
we
are
looking
to.
Do
we're
actually
going
to
release
this
publicly
within
the
next
week.
So
by
next
Tuesday.
The
announcement
would
be
out
and-
and
we
are
actually
demonstrating-
the
service
at
kubecon,
Amsterdam
and
I-
just
wanted
to
kind
of
introduce
the
service
to
the
the
community,
as
well
as
the
maintain
us,
primarily
because
I
think,
as
I
mentioned
on.
The
last
call
you'll
see
us
kind
of
participating
actively
in
forums
and
slack
and
and
so
on,
going
forward.
E
So
what
we,
what
we,
what
a
customer
gets,
is
essentially
multi-cluster
management
they'll,
be
able
to
do
real-time
monitoring
of
of
backup
and
restore
operations
that
are
running
in
Valero
through
the
service
and
and
they
can
again
do
all
the
crud
operations
for
the
Valero
CRS
through
the
service,
and
we
will
also
have
you
know,
guided
recovery
and
and
and
have
some
Advanced
capabilities,
even
on
top
of
Valero,
that
we
will
enable
for
existing
Valero
recovery
points,
and
that
would
allow
us
to
create
a
cluster
not
just
restore
to
an
existing
cluster
right.
E
So
we're
adding
those
pieces
in
in
customers
will
have.
The
ability
to
you
know
live
chat
with
our
with
our
team,
as
well,
as
you
know,
select
logs
for
all
the
Valero
backups
in
in
one
location
and
and
this
just
improves
the
speed
of
what
they
are
doing
with
the
solution.
Was
there
a
question?
E
Okay,
if
not
I'll
keep
moving
and
and
sorry
for
the
bills.
I
blame
the
marketing
team
on
it
and
and
again
the
goal
is
to
not
really
disrupt
anything
that
they
are
doing.
We
don't
we.
This
is
a
SAS.
They
don't
have
to
install
anything
any
infrastructure
element.
They
can
continue
doing
what
they
are
doing.
All
of
the
Valero
scripts.
All
of
the
the
CI
CD
pieces
all
will
work.
The
existing
recovery
points
will
continue
to
work,
we're
not
adding
any
lock-in.
E
There
is
no
proprietary
data,
we
are,
you
know,
locking
customers
from
if
they
remove
the
service
Valero
would
work
just
as
it
was
working
before.
There
is
absolutely
no
difference
other
than
the
value
add.
We
are
we're
doing
so.
I
think
the
service
we
really
built
is
to
be
completely
compatible
with
Valero
and
I'll
walk
you
through.
So
this
is
a
quick
view
of
everything
we
capture
and
and
we're
going
to
allow
customers
to
manage
and
and
everything
will
allow
a
customer
to
do
through
our
service.
E
On
top
of
you
know
all
of
the
CRS
and
activities
we
manage
any
any
questions
so
far
and
I
can
quickly
do
a
quick
demo
and
and
I
promise
to
not
take
more
than
10
minutes.
C
So
you
add
a
lot
of
components
in
addition
to
valeroes
who
support
like
the
full
stack
cluster
recovery
right.
E
Yeah,
so
the
the
I
mean
a
lot.
Less
is
all
relative,
but
I
think
this
is
our
first
release,
so
I
have
modest
expectations
on
what
we
do
on
top
of
Valero,
but
what
we
are
trying
to
do
is
essentially
become
open
source
compatible
right,
which
means
somebody
that
is
using
Valero
should
be
able
to
get
value
ad
from
our
service.
We're
not
looking
to
get
Valero
replaced,
but
they
can
take
an
existing
recovery
Point
that
is
created
with
Valero,
so
a
backup
that
is
run
with
Valero.
E
We
can
take
that
particular
backup
and
use
that
to
spin
up
a
cloud
cluster,
whether
it's
in
eks,
AKs
or
gke.
Those
are
the
big
three
Cloud
providers
we
support
today,
but
you
can
take
any
recovery
point
and
say:
hey
spin,
it
up
as
a
cluster
for
me
in
in
one
of
these
three
big
three
platforms.
Of
course,
we'll
we'll
try
to
do
that
with
the
openshift
ACM,
as
well
as
a
Rancher
as
well
in
future.
E
E
You
know
speed
up
a
lot
of
that
and
also
the
key
value
is
you
have
to
create
the
cluster
you're
restoring
to
so
you
can
only
do
restores
to
a
functional
cluster
today
right
and
what
we
are
essentially
saying
is
hey.
Creating
a
cluster
is
very
possible
in
the
cloud
with
just
a
few
API
calls
and
we
will
speed
up
a
cluster.
E
We
will
download,
install
Valero,
we
will
configure
the
BSL
and
we
will
kick
off
the
restore
so
that
the
time
taken
before
you
could
actually
perform
you
know
a
backup
or
restore
is
is
extremely
low
right.
So
so
that's
that's
the
extra
piece
we
add
and
we
also
make
some
of
the
little
things
like.
You
know,
alerting
easy
for
people,
monitoring
easy
for
people
and
they'll
be
able
to
do
an
adoc
run.
E
So
you
can
go
to
any
backup
and
say:
okay,
go
run
this
backup
now,
instead
of
having
to
create
and
and
and
create
a
command
structure,
so
on
I
think
today,
just
if
you
want
to
run
a
one
backup
right
now,
you
still
have
to
provide
a
decent
amount
of
information.
In
our
case,
it's
really
just
one
click
on
an
existing
backup
and
we
would
automatically
run
a
new
backup
based
on
that,
okay,
so
I'll,
quickly
kind
of
show
you
a
a
demo
again.
E
This
is
a
gke
cluster
and,
and
you
could
see
that
I've
got
Valero
already
installed,
there
is
no
Cloud
class
element
currently
and
and
and
you
should
be
able
to
run
and
see
kind
of
Valero
configurations
everywhere
people.
Let
me
know
if
you
can't
see
my
screen,
you
can
see
kind
of
the
backups
already
configured
on
this
cluster.
What
we
are
actually
going
to
do
is
just
add
this
cluster
to
Cloud.
Casa
cloudcaster
is
something
you
can
sign
up
for
with
just
an
email
address.
E
You
don't
need
credit
card,
nothing
just
literally
get
in
and
and
and
you
can
sign
up
here
and
I'm
gonna
call
it
Community
demo
and
I'm
gonna
manage
an
existing
Valero
instance.
Instead
of
creating
our
own
agents
to
perform
this
I'm
going
to
register
and
when
I
register
it's
going
to
give
me
a
a
cube,
cuddle
command
to
run
I
copy.
This
I
go
here
and
I
apply.
This.
E
And
what's
going
to
happen
here
is
that
this
particular
cluster
is
going
to
get
registered,
it's
going
to
initiate
connection
to
the
service,
and
you
will
see
this
community
demo
going
from
register
to
inventory
to
actually
becoming
active
in
about
a
minute
right,
so
this
cluster
will
will
essentially
establish
a
connection
to
the
service
and
and
it
will
essentially
turn
into
an
active
state.
So
I'll
wait
for
this
to
turn
into
an
active
State.
E
It's
already
detected
the
kubernetes
version,
the
agent
version
and
it's
gonna
begin
an
inventory
process
very
quickly.
While
we
do
that-
and
you
can
also
see
that
if
there
is
a
cluster
where
I
said,
hey
I
want
to
manage
Valero
but
turns
out,
there
is
no
Valero
installed
in
this
cluster.
It's
it's
gonna,
warn
you
that
hey!
There
is
no
Valero
found
on
this
cluster,
so.
A
E
This
case
there
is:
where
error
we
just
turned
active,
and
now
you
can
actually
CC
all
of
the
backups
that
are
defined
in
this
all
of
the
restores
that
are
defined
as
I.
Don't
think,
there's
anything
and
all
of
the
recovery
points
that
have
been
created
on
this
cluster
again.
This
is
all
done
outside
of
cloudgasa
before
we
got
into
this
right
and
you
can
also
see
all
the
job
activities
that
are
kept
on
the
cluster.
E
You
can
see
all
the
jobs
that
ran
and
and
whether
they've
been
successful
or
or
not,
and
you
can
also
see
the
various
settings
as
well
yeah
that
our
system
is
a
little
slammed
and
you
can
see
the
the
variables
and
the
plugins
that
are
installed
everything
from
right
here
right.
So
if,
if
someone
wanted
to
really
just
go
edit,
this
backup
job,
it's
it's
really
as
simple
as
go
run
and
edit
in
and
edit,
and
you
can
see
you
know
what
have
you
selected
I've
selected
just
to
name
space.
E
You
can
also
Define
what
type
of
backup
you
want
to
perform.
Where
do
you
want
to
send
this
backup?
Is
there
a
vsl
Associated?
E
What's
the
schedule,
retention,
app
works
and,
and
you
gotta
see
how
you've
configured
your
backup
job
in
in
and
our
spec
is
very
much
compatible
with
the
Valero,
spec
and
and
you'll
be
able
to
essentially
go
through
a
wizard
and
specify
you
know
what
you
want
to
backup
right
and-
and
there
is
value
in
that,
because
a
lot
of
people
don't
know
what
various
options
mean
and-
and
this
is
just
allowing
people
to
be
able
to
do
that
and
for
the
advanced
user.
E
There
is
again
you
can
edit
as
yaml
and
and
make
those
edits
here
and
directly
as
well
and
and
as
I
mentioned
it
it's
going
to
automatically
pick
up
all
of
the
activities.
So
all
of
your
jobs
that
were
previously
run,
you
could
see
what
the
backup
looks
like
and
you
can
actually
download
the
logs
as
well,
and
you
can
see
what
PB
details
where,
where
there
on
the
snapshot
or
copy
it
in
this
case,
is
a
copy
job.
So
you
can
see
how
much
data
was
was
transferred.
E
While
it's
live,
you're
gonna
see
some
red
throughput,
which
is
really
just
how
much
is
being
read.
Of
course,
once
it's
transferred,
that's
going
to
be
the
amount
that
actually
matters
in
in
this
case
right
so
so
far,
we've
actually
done
nothing.
It's
just
existing
piece
and
and
we've
gone
ahead
and
and
inventory
and
everything,
and
we
we
just
start
essentially
creating
a
dashboard
of
view
into
a
Valero
install.
Of
course
you
can
do
Define
a
backup
from
scratch.
E
You
choose
the
cluster,
you
want
to
backup
and
and
select
all
of
these
options
as
well.
The
the
second
piece
I,
I
and
also
anytime
there
is
there,
are
problems
you're
going
to
see
you
know,
alerts
generated
and-
and
that's
that's
another
key
feature.
If
we
see
there
is
some
problem
with
some
of
the
things
that
you're
you're
performing
you're
going
to
see
some
alerts
generated
as
well
and
and
I
I
mentioned
that
we
we
actually
have
some
restore
capability.
That
is
on
top
of
what
Valero
does.
E
So
it's
not
just
a
UI
for
Valero.
When,
when
we
do
that,
let
me
see
if
I
can
quickly
show
you
a
restore
right.
So
here
is
an
example:
I
mean
I'm.
I
have
a
this.
Is
a
Valero
cluster
and
a
Valero
recovery
point
I'm,
picking
kind
of
that
recovery,
point
and
saying:
hey,
restore
all
of
the
namespaces
and
and
instead
of
just
restoring
to
an
existing
cluster,
which
is
really
what
I
think
Valero
CRS
do
we're.
E
Actually,
in
this
case,
saying
Hey
I
want
to
restore
it
as
an
eks
cluster
AKs
cluster
GK
cluster.
In
this
case
it's
a
GK
cluster
you're
going
to
select
the
the
cloud
account
to
use.
You
can
also
ask
for
that.
Every
time
this
job
runs.
You
can
clean
up
the
last
instance
of
this
cluster
and
create
a
new
one
and
and
within
gke
you
can.
You
can
set
up
all
of
the
options
that
are
necessary
for
gke,
like
region.
What
networking
should
I
use?
What
cluster
add-ons
should
I
use?
E
You
can
see
it's
a
zonal
or
Regional
type
cluster,
which
zone
what
network
should
I
use.
If
there
are
multiple,
you
can
actually
select
the
right,
VPC
and,
and
you
can
select
any
of
the
subnets
and
what
node
pool
how
should
I
size
this
cluster
you
may
want
to
have
you
know
minimum
and
maximum
note
counts
defined
in
this
and
also
various
cluster
add-ons.
Do
you
want
to
add
on
these
clusters
when
you're
creating
the
jobs
and,
of
course
you
can
select
the
Valero
options
as
well?
E
If
you
want
to
rename
a
namespace,
you
just
pick
the
app
and
say
not
to
do
right,
so
doesn't
take
Caps
or
probably
will
enforce
anyway,
and
and
you
can
perform
everything
that
you
do
in
Valero,
but
on
top
of
it,
we
have
prevent
up,
provide
all
of
the
options
that
allows
you
to
create
the
aksc
case
and
gke
clusters
as
well.
So
what
happens
when
you
run
it?
E
Let
me
see
if
I
can
quickly
show
you
a
job
that
is
completed
so,
when
you're
on
the
job,
you're
gonna
see
that
it's
it's
creating
a
new
AKs
cluster
in
this.
In
this
particular
case,
it
just
created
one,
it
waits
for
it
to
come
online.
That
takes
a
few
minutes.
Once
that
cluster
comes
online,
we
install
our
agent
first
to
perform
the
things
that
we
need
to
perform,
and
then
we
install
the
Valero
agent.
We
we
configure
the
Valero
agent
successfully.
E
This
is
just
creating
the
BSL
automatically
and
once
the
BSL
is
created
now
we're
waiting
for
Valero
to
sync
the
contents
of
the
BSL,
and
now
we
have
everything
we
need
to
perform
a
Valero
restore
and
that's
what
it
does
at
the
end
of
the
day.
But
this
entire
process
is
is
also
automated
for
a
user
again.
This
is
on
any
recovery
point
that
was
previously
created
with
Valero
Cloud
plus.
That
need
not
be
in
the
picture
to
get
all
of
this
value
when
you
are
doing
backups,
I
think
when
you're
doing
restores.
E
Of
course,
we
need
to
be
in
the
picture
to
do
all
of
this,
and
this
is
something
that
we
add
on
top
of
Valero
as
well
again.
That
I
think
in
a
nutshell,
captures
everything
we
essentially
inventory
what
is
already
on
Valero
and
we
provide
an
interface
where
they
can
perform
a
lot
of
things.
I
can
basically
to
show
me
Valero
clusters.
It's
going
to
show
me
these
clusters.
You
can
already
see
everything.
That's
there.
E
We
today
allow
people
to
change
or
edit
bsls,
bsls,
backups
and
restores,
and
the
recovery
points
and
the
activity
and
the
settings
are
all
kind
of
view
only
today
there
is
no
changes
you
can
make.
You
cannot
update
Valero
through
our
interface,
but
that
may
be
something
we
change
pretty
soon,
but
essentially
become
a
multi-cluster
Valero
manager
for
these
customers
and,
as
you
can
see,
there
are
multiple
Valero
clusters
here
that
we
manage
them
all.
E
You
know
right
here
from
our
interface
and
and
gives
you
the
monitoring
and
alerting,
as
well
as
the
configuration
capability
on
top
of
that.
Okay,
so
just
wanted
to
quickly
introduce
the
maintainers
to
what
we
are
working
on.
I.
Think
we
are
the
first
of
its
kind.
There
are
other
solutions
that
use
Valero,
but
you
kind
of
have
to
start
with
their
solution
at
first.
This
isn't
a
solution
where
you're
forced
to
use
or
are
service
up
front.
E
B
Yeah
I
had
a
couple
quick
questions,
so
it
sounds
like
I
guess
from
the
from
the
last
part
of
the
demo,
with
restore
you
do
support
cases
where
Cloud
cost
is
actually
installing
Valero
in
the
cluster.
Do
you
also,
but
you
also
mentioned
for
existing
Valero
users.
So
so
are
there
cases
where
some
clusters,
Cloud
Casa,
installed,
Valero
and
other
clusters
that
already
existed.
E
Yeah,
so
in
in
so
let
me
answer
it
differently.
So
when,
before
we
created
this
open
source
compatible
service
right,
we
were
leveraging
a
fork
of
Valero
in
our
in
our
agents
already,
so
we
were
installing
a
4K
Valero
it
it.
The
primary
difference
we
had
Scott
is
that
we
perform
backups
on
the
snapshots.
So,
unlike
you
have
rustic
and
copy
are
running
on
the
live
system.
We
actually
take
the
CSI
snapshot,
we
mount
it
and
then
we
back
up
from
the
CSI
snapshot.
That's
one
of
the
primary
differences.
E
We
also
deliver
something
called
Agent
plus
backups,
so
you
don't
have
to
actually
add
the
cluster
one
by
one
in
our
product.
You
can
essentially
go
to
Cloud
accounts
and
you
add
your
cloud
account
and
we
will
automatically
ingest
all
of
the
eks
AKs
and
gke
clusters
and
inject
our
age
automatically
inside,
and
we
start
backing
up
immediately
right.
So
so
we
had
all
of
these
benefits,
but
it
was
effectively
replacing
an
existing
Valero,
install
right
and
and
over
a
period
of
time
we
just
basically
said
hey.
E
We
have
no
reason
to
force
a
user
to
switch
out
a
Valero.
We
can
essentially
add
a
lot
of
value
on
top
of
Valero
itself
and
and
that's
really
the
two
differences.
If
you
used
the
traditional
I
mean
what
we've
called
cloudcaster
so
far
for
Enterprises.
That
includes
an
installation
of
an
arrow.
It's
not
the
mainstream
Valero.
We
we
actually
use
a
fork
and
in
in
this
case
we
will
just
walk
with
any
Valero
versions.
E
B
E
The
answer
is
all
of
the
above,
so
we
do
have
a
way
to
Auto,
detect
where
Valero
is
installed,
but
you
can
also
choose
which
namespace
to
monitor.
If
you
have
multiple
Valero
instances.
B
E
We
can
only
monitor
one
Valero
install
per
cluster,
and-
and
so,
if
somebody
with
multiple
Valero
names
in
installations
across
multiple
name
spaces,
you
do
kind
of
have
to
pick
one
for
us
to
Monitor
and
manage,
but
we
Auto
detect.
If
you
want
us
to
force
to
a
specific
namespace
to
manage,
we
will
do
that
as
well,
and
then
you
can
specify
that
right
here.
B
B
So
so,
for
example,
Red
Hat
we
have
oadp,
which
is
our
backup
product
for
openshift,
that
includes
Valero
as
a
component.
So
if
you
had
no
ADP
installation
in
a
cluster
in
an
openshift
cluster,
that
would
show
up
as
an
actual
valerical,
you
know
it.
It
is
an
actual
Valero
way.
Again.
We
have
a
fork
with
some
minor
changes,
but
it's
a
Valero
install
in
a
namespace
that
could
probably
interface
with
this
as
well.
Yeah.
E
E
Are
also
working
in
our
case.
D
E
Mean
if
it's
based
on
online,
the
good
thing
is.
The
next
version
will
also
be
supported
when
you
move
to
110
and-
and
we
verify
one
and
one-
ten
works.
Yeah.
B
And
actually
our
next
major
version
is
going
to
be
based
on
111.
Once
111
comes
out,
you
know
in
a
few
weeks
well.
C
Thank
you
so
so
I
think
you
may
find
you
know
dealing
with
different
version
of
Valero
little
tricky
because
we
keep
adding
you
know
making
change
I
mean
adding
new
attributes
to
the
API
yeah.
E
And
it
has
been
I
think
we
probably
underestimated
when
we
started
the
work,
because
I
think
the
specs
are
different
or
the
options
supported
are
very
different
in
each
version
right.
So
we've
that
has
been
a
problem
for
us
in
terms
of
integration,
but
look
I
I,
it's
it's
part
of
any
growing
Community
people
want
options
and,
and
what
we
have
found
is
the
options
that
are
supported
today
were
in
supported
in
the
previous
one.
E
So
it's
not
exactly
Backward
Compatible
to
you
know,
have
an
option
there,
but
just
Valero
ignore
that
that
doesn't
seem
to
be
the
case.
So
we've
had
to
manage
and
maintain
that
Raghu
is
on
the
call
and
Raghu
is
actually
the
the
Chief
Architect
for
the
solution
and
and
he
could
probably
share
how
painful
some
of
those
have
been,
but
yeah.
B
And
again
that
goes
more
in
One
Direction
than
the
other.
You
know
when
we
add
new
attributes
and
new
spec
fields.
For
example,
we've
generally
tried
to
be
careful
to
make
sure
that
a
backup
created
from
before
that
was
really
valid
in
the
newer
version,
but
the
reverse
won't
be
true
if
you
take
a
1.11
backup,
for
example,
with
some
of
the
new
fields
in
it
that
try
to
install
you
know
create
that
in
a
older
cluster,
it's
probably
not
going
to
work
for
you,
yeah.
B
E
Mean
to
ignore
the
ignore
the
settings
you
don't
understand,
but
but
look
you
can't
change
the
past
so
so
how
it
works
is
how
we
will
make
it
work.
D
F
F
Hi,
so
just
a
quick
introduction,
like
Satya,
said
on
the
architecture.
My
name
is
so
anyway,
so
about
the
version.
Compatibility
right,
I
I
would
like
to
say
that
so
far,
between
1.9
and
1.10,
it's
been
generally
good.
F
I
mean
most
of
the
fields
are
are
same,
for
example
in
the
backup
of
spec
right
I
haven't
found
many
differences,
but
I
know
that
in
the
upcoming
versions
a
lot
of
changes
have
been
talked
about
in
the
spec,
and
we
have
to
really
you
know
be
up
to
date
on
those
and
to
the
point,
I
think
somebody's
card,
I
guess
about
the
newer
versions
working
with
the
older
backups.
That
is
true.
F
So
when
you
define
a
restore,
for
example,
from
a
Valor
or
recovery
point
right,
so
we
we
know
what
is
the
version
that
people
use
it
for
the
backup.
So
we
would
let
users
pick
a
version
starting
from
that
and
anything
in
future
for.
F
F
It's
really
not
reasonable,
expect
I'm
going
back
all
the
way.
You
know
beyond
that,
and
probably
people
most
people
upgrade.
F
Idea,
like
the
community,
for
example,
do
people
still
use
130,
for
example?
Is
there
any
any.
A
B
I
I
can
give
it
give
us,
for
example,
on
the
oadp
side.
Oadp
1.0
was
based
on
Valero
1.7
and
then
1.1
was
based
on
1.9
and
then
we're
going
to
be
going
to
111..
So
we
certainly
have
people
currently
using
Valero
1.7
and
no
idea
who
won
zero
context
right
now.
I
I
don't
actually
know
what
what
point,
though,
what
the
official
Moes
could
probably
fill
in.
B
You
know
exactly
when
that
goes
out
of
support,
but
but
because
we
have
a
tendency,
at
least
in
the
past,
we
sometimes
we
skip
a
Valero
version,
so
the
N
minus
two
May.
Well
just
be
you
know:
n
minus
one
from
an
ADP
point
of
view,
but
again.
B
B
F
Yeah,
we
may
not
immediately
update
because
we
made
some
changes
to,
for
example,
like
Satya
said
some
Fork
right,
so
we
made
some
changes
to
the
CSA
plugin
and
some
of
these
things
we
already
created
PRS
for
so.
F
1.10
has
has
some
changes
so
we're
still
in
the
process
of
updating
1.10
in
our
internal
distribution,
but
going
forward.
I
think
this.
This
actually
brings
up
another
good
point.
So
whatever
few
changes
that
we
have
made
in
some
of
these
Valero
four
could
repos
I
would
really
want
to.
You
know,
contribute
it
back
to
to
Valero.
So
you
might,
you
might
see
some
some
PRS
coming
out
of
cloud
Casa
and
even
even
the
one
big
difference
that
Satya
mentioned
for
the
copy
right.
So
we
we
do.
F
F
Right
do
the
backups
on
Snapshot,
so
this
is
maybe
one
area
where
we
can
definitely
work
together
and
we
can.
We
can
contribute
that
you
know
that
knowledge
and
even
that
code,
if
required
back
to
the
community.
So
some
of
these
things,
yeah
you'll,
definitely
see
more
activity
from
us,
giving
back
whatever
we
have
in
our
Folk.
C
Yeah
that
would
be
really
helpful
thanks
thanks
in
advance
and
by
the
way
I
would
like
also
like
to
point
out.
In
future.
There
may
be
a
break
change
introduced
to
the
API,
because
there
are
some
attributes
here
in
the
fields
we
we
consider
Tech,
that's
so
sooner
or
later
we
we
I
I,
think
we
need
to
clean
them
up,
but
I'll
make
sure
you
know
that's
communicated
to
the
community
and
we
have
enough
time
to
prepare
and
review
the
process.
Yeah.
B
C
C
F
F
All
the
data
with
the
recovery
point,
for
example
right
once
I,
have
a
backup,
CR
I
shouldn't
have
to
go
to
anything
else.
I
mean
I
may
not
have
exact
details
right
now,
but
we
notice
a
few
things,
so
we
will
try
to
bring
them
up
for
Community
discussion
in
order
to
improve
in
in
a
good
sense
right,
improve,
be
useful
for
everybody
else.
F
C
F
E
So
let
us
know
how
we
can
help
and,
and
we
will
we
will
certainly
get
our
feet
wet
and
and
start
contributing,
I
think
we
you
see
that
already
in
slack
and
and
forums,
a
lot
of
us
are
starting
to
respond
and
and
we'll
we'll
a.
A
Hey
cool,
so
I
guess
next
we
go
to
Daniel.
C
Yeah,
just
I
want
to
share
with
you
guys
we
are
planning
the
release
of
1.12,
so
we
don't
have.
We
didn't
have
a
chance
to
discuss
the
timeline
at
the
moment
because
we
were
quite
occupied.
Testing
one
die
to
make
for
1.11
can
go
out
so,
but
now
pradeep
has
been.
You
know,
adding
labels
as
112
candidates
to
the
issues
and
in
the
next
one
or
two
weeks,
we're
gonna
label
more
of
them
and
we
start
sharing
with
candidates
and
decide
a
payload
for
112..
C
So
the
one
anchor
feature
of
112
will
be
the
data
mover.
We're
gonna
introduce
this
plugable
framework
so
that
a
data
like
this
snapshot
can
be
moved
to
the
object
store.
So
so
we're
gonna
make
sure
1.12.
We
have
a
good
end-to-end
for
the
data
mover
and
as
for
now,
if
you
have
any
features
you
wanna
work
on
112,
please
make
sure
add
the
labels.
C
If
you
do
not
have
the
permission,
just
ping
us
and
discuss
with
us
we're
gonna,
you
know,
take
them
into
consideration
in
the
planning.
Yeah
I.
Think
that
that's
my
update
and
just
let
you
know
let
you
guys
know
that
we
are
planning
112
and
the
window
is
open.
If
there's
any
issues,
if
you
want
to
be,
you
know,
included
in
112,
make
sure
bring
that
into
our
radar.
Thank
you.
Daniel.
C
A
C
Yeah
I
haven't
thought
about
this
process.
I
think
it
will
only
be
me
editing
this
week,
but
if
you
want
to
add
anything,
I
think
first
thing
you
want
to
do
is
add
the
labels
to
the
issues
you
wanna
work
home
in
112.
We
want
to
include
in
112
and
the
wheelchair
them
internally
and
we
can
also
charge
them
together
on
the
community
meeting
or
via
slack
and
later
you
know
we
can
work
on
together
on
the
laptop
that
we
keep.
You
think
that
will
work.
G
I
yeah
I
mean
I.
Think
that's
fine,
for
you
know:
dollar
next
release
I,
think
I'll
also
chat
with
pradeep.
Maybe
keep
con
and
see
if
we
can
get
high
level
ideas
for
113
114
115
whatever.
So
we
can
have
a
little
high
level
direction
of
where
we're
driving
the
project
to
in
the
future,
but
yeah
I
think
that'll
work,
okay,
Daniel.
C
I
think
that
will
be
helpful
that
we
can,
you
know,
work
on
like
a
future
roadmap,
wiki
page
or
MD
file,
so
we
have
yeah
a
relatively
concrete
workload
that
we
know
what
we're
going
to
do
in
future.
A
Okay,
I
think
that's
everyone,
you
guys
have
any
have
anything
else.
You
want
to
add
to
this
meeting.
D
Actually,
if
folks
have
a
few
minutes,
I
have
a
very
quick
question
about
Valero
upgrade.
This
came
across
an
issue
before
this
meeting.
If
everyone
is
okay,
maybe
I'll
just
take
two
minutes.
D
So,
okay,
so
basically
like
I
guess,
my
question
is
around
like.
Is
there
any
general
guideline
around
like
how
to
handle
like
in
progress
or
running
back
up
jobs
during
like
a
Valero
upgrade
like
you,
send
me
like
advice?
Do
we
cancel
it?
Do
we
let
it
go
and
do
its
thing
and
let
it
finish.
B
I
mean
I
I,
don't
know
if
we
have
a
general
guide,
if
you
have
an
in-progress
backup
or
restore
the
only
way
for
it
to
succeed
would
be
to
let
it
finish
before
you
did
anything
that
would
restart
Valero,
so
an
upgrade
would
result
in.
You
know
shutting
down
the
Pod,
so
I
think
you.
If
you
want
that
backup
to
not
have
to
not
have
to
repeat
the
you
know
recreate
it,
then
you
would
want
it
to
finish.
B
Well,
I
mean
if
the
Valero
pod
restarts
or
you
shut,
that
shut
it
down.
You
know
whether
or
not
there's
an
upgrade
being
involved
and
in
progress
backup.
You
know
right.
Currently,
the
Valero
backup
and
restore
controllers
are
not
designed
to
pick
up
and
end
progress,
backup,
so
just
in
general,
even
outside
of
upgrades,
if
you,
if
something
causes
your
Valero
pod
to
bounce
out
of
memory,
whatever
any
backups
or
restores
that
were
in
progress,
will
not
complete
those
those
are
going
to
fail.
B
No
I
won't
resume.
We
do
have
a
controller
on
restart.
What
we
do
is,
if
there's
a
backup
that
was
previously
in
progress.
We
Market
failed
because
right
now
the
backup
and
restore
controllers
are
not
designed
to
pick
up
a
backup
that
was
in
the
middle.
So
if
it
doesn't
complete,
then
it
you
know
it.
It
can't
resume.
D
Okay,
so
I
guess
the
general
guidelines
for
users
would
be,
you
know
either
make
sure
they
are
no
backups
running
or
be
ready
to
expect.
That
is
that
the
the
backups
are
going
to
fail.
Yes,
okay,
all
right
got
it
thanks.
A
And
with
that
there's
nothing
else,
I'll
stop
the
recording.
You
guys
have
anything
else.
A
All
right,
cool.