►
Description
August 4, 2020
Status Updates
ashish
Community support this week
Have a list of PRs to review
New issues that need triage
Working on restore hooks
PR 2787 to implement the init container restore hooks.
Need to run this through tests before removing WIP.
nrb
Out sick late last week
Working on DeleteItemAction design. Please take a look!
Investigating options on project management and release cadence to get issues handled more quickly, as well as get fixes out more regularly.
carlisia
PTO!
Working on migrating ServerStatusRequest to kubebuilder
Discussion Topics
Let us know what Velero code you depend on
Andrew Reed - need any help with exec restore hook?
How do we mitigate people being on PTO? Slackening the reviewer constaints?
Storj persistent volume support
Goal: back up any Kubernetes PV to Storj object store
rclone/restic support?
treating volume as a block device?
vSphere plugin currently has a data mover - https://github.com/vmware-tanzu/velero-plugin-for-vsphere
github.com/openebs/velero-plugin has a data mover too
Let’s start a thread on the Google Group
Contributor Shoutouts
@jenting: Add a flag cacert for velero backup-location create
@jenting: fix: helm upgrade did not work correctly on BSL and VSL configuration
@MrFreezeex: Fix Helm 2 CRD generation
@alikhtag: [impovement] Add podLabels to the helm chart
@kyounger: fix: Restic daemonset SA always set, like velero deployment
@fredgate: Add log configuration to Restic
@cesarokuti: add schedule labels
@ckotzbauer: add affinity option to velero deployment
A
A
A
Oh
okay,
so
for
me
I
was
outside
late
last
week,
so
I
don't
have
a
ton
to
talk
about,
but
my
big
thing
is
working
on
the
delete
item,
action
design
which
I
can
show
here.
I
think
this
was
a
big
topic
two
weeks
ago
in
the
community
meeting.
The
thing
to
share
here
is
I've
changed
I've.
Originally
I
had
this
delete
deletion,
plugins
and
then
I've
now
redone
it
with
a
delete
item
action
plug-in
type.
A
A
So
that's
been
a
big
driver
for
what
I've
been
working
on.
Also
something
we're
looking
at
is
looking
at
different
things.
We
can
do
for
our
product
project
management
and
release
cadence
so
that
we
can
handle
github
issues
a
little
bit
quicker
more
quickly
and
get
fixes
out
more
regularly,
but
we
don't
have
really
any
any
concrete
updates
to
share
there
other
than
that
we're
talking
about
it.
So
that's
what
I
have-
and
I
also
wanted
to
mention
that
carlesia
is
on
pto
this
week
and
she
was
working
on
migrating.
A
B
Okay
looks
like
they're
still
not
done,
but
hopefully
then
I
shouldn't
be
too
bad
for
me.
I
am
on
community
support
this
week.
I
have
a
list
of
vrs
that
I
have
to
review
and
some
newly
new
issues
that
came
in
that
neat
triad.
So
I'm
I
didn't
do
that
yesterday,
so
I'm
going
to
do
some
of
some
of
it
today
and
hopefully
make
up
for
the
stuff
that
came
in
yesterday.
Other
than
that
I
have
been
working
on
the
restore
hooks
part.
B
I
have
a
pr
which
is
kind
of
in
is
in
progress
at
the
moment
to
implement
the
init
container,
restore
hooks
I've,
I've
kind
of
in
the
pr.
I
have
bundled
it
up
with
an
existing
restore
item,
action
that
we
have
for
parts
and
now
I'm
thinking
that
it
might
make
sense
to
make
this
its
own
restore
item
action
plugin.
So
I
might
go
in
and
switch
that
out.
C
A
A
A
We
have
this
open
issue
at
the
top
of
our
issues.
List
number
2654
if
you've
got
any
valero
code
that
you
directly
depend
on
in
your
own
go
code,
please
let
us
know
by
a
comment
on
this
issue,
because
we
will
be
moving
a
lot
of
this
code
out
of
package
into
if
it's.
If
it's
not
explicitly
listed
here,
we
will
be
moving
it
from
package
into
an
internal
directory.
A
D
B
I
have
been
working
only
on
the
init
container
part
of
it.
So
once
once
I
have
this
done,
I
might
jump
on
the
restore
hook
that
we've
designed
so
so
far.
I'm
okay
doing
this,
but
if
you
can
pick
up
the
exact
restore
hook,
part
let
us
know-
and
I
also
added
you
on
the
pr.
So
if
you
want
to
do
a
quick
review
for
us,
that
will
also
be
great.
D
Yeah,
I
can
take
a
review
on
that.
I
mean,
do
you,
do
you
need
help
on
the
exact
restore
hook?
It
seems
like
you're
making.
You
know
pretty
quick
progress.
So
if
you
feel
like
you
know,
you're
good
to
just
do
it
on
your
own,
for
the
1.5
release,
that's
you
know
totally
cool,
but
if
you
feel
like
there
might
you
need
some
help
then
I
could
you
know
take
this
week
to
work
on
it.
A
I
was
gonna,
say
andrew
if
you're
offering,
if
you
want
to
take
2731,
I
would
say:
please
do
yeah
cool
yeah
that
that
way
like,
I
would
say,
let's
let's,
if
like,
if
you've
got
the
the
bandwidth
and
the
time,
please
please
take
2731
and
I
can
assign
that
to
you
that
way.
We
can
kind
of
keep
things
rolling.
We
can.
We
can
assign
that
to
you
and
if,
if
you
comment
on
it,
I
can
I
can
do
the
the
sign.
Okay
I'll,
do
that.
A
B
A
Yeah
and
if
that
needs
to
be
pulled
out
into
a
separate
pr
like
if
parts
of
of
the
the
this
parts
of
2787
maybe
need
to
still
be
tested,
but
the
scaffolding
would
be
helpful,
maybe
we
can
put
it
up,
but
if
you
do
want
to
coordinate
on
the
on
the
kubernetes
slack
for
that
and
ping
me
on
on
what
what
could
be
done
to
speed
this
up
or
not
to
beat
this
up
but
to
make
it
so
you
can
work
in
parallel.
I
can
help
accommodate
that.
C
A
A
It
will
look
similar
to
restore
item
actions,
so
hopefully
the
work
won't
be
too
too
dissimilar,
but
I
don't
know
that
for
sure.
Yet
I
need
to
flesh
out
that
dock
and
get
it
closer
to
merging
merging
this
week
may
be
a
little
slow
because
our
settings
on
github
do
require
two
two
maintainers
to
approve,
and
carl
you'd
see
us
on
pto
this
week,
so
just
just
giving
that
heads
up
to
everyone
that
there's
two
of
us
on
board.
A
B
E
B
A
Thank
you,
and
I,
I
guess
I'll
also
say
like
the
the
three
maintainer
thing
is
not
permanent.
Like
that's
not
a
that's,
not
a
situation.
Anyone
wants
to
be
in
right
now,
so
that's
going
to
be
changing,
but
it's
the
current
state
of
things
so
yeah
for
for
the
time
being,
unfortunately,
that's
how
it
is
and
and
like
she
said,
we
could
pro
if
it's
absolutely
necessary,
like
we
have.
A
F
Oh
perfect,
hi,
I'm
caleb
from
storage
labs,
and
we
are
you
know
recently
we
we
merged,
or
we
built
out
like
a
community
provider
for
storage
labs
and
it,
but
it
only
does
the
object,
store
portion
of
things
and
we
are
trying
to
figure
out
what
the
best
path
forward
for
building
out
a
volume.
Snapshotter
is,
and
we
had
considered
two
paths
so
far,
and
maybe
there
are
others
we
should
look
at.
F
The
first
was
to
try
to
so
we
have
an
integration
with
our
clone,
and
so
we
also
considered
using
your
restic
beta
as
a
way
to
accomplish
this.
That's
one
path.
The
other
path
we
were
looking
at
was
potentially
doing
something
where
we
treat
the
pv
as
like
a
block
device
and
take
a
just
like
a
snapshot
of
all
the
data
out
of
it
directly.
F
A
A
So
it's
it's
meant
to
say:
hey
you're,
the
platform,
that's
keeping
volume
and
I'm
going
to
use
your
native
apis
to
do
whatever
snapshotting
support
you
have
so
that
was
kind
of
their
intended,
their
intended
their
intended
use.
I
know
I
think
it's.
The
current
implementation
of
the
vsphere
plug-in
does
does
some
other
stuff
where
it
has
like
it.
Has
that
plug-in?
That
does
that,
but
it
also
has
a
daemon
set.
That's
hanging
off
the
side
so
like
on
the
node.
A
G
Basically
says:
hey,
you
know,
storage
system.
You
take
care
of
all
this
under
the
covers
and
we're
we're
doing
that
for
vsphere.
But
I
think
we
want
to
be
able
to
enable
the
same
kind
of
behavior
for
ebs
and
google
cloud
etc.
Where
valero
can
take
a
snapshot,
but
then
there's
actually
a
way
to
extract
the
data
and
move
it
somewhere
else.
G
I
think
we
we
need
to
talk
about
how
this
would
become
more
of
a
valera
thing
in
terms
of
data
movement
being
something
that
valero
actually
knows
about
and
manages
and
has
some
some
apis
on,
but
we
can
have
that
conversation
and
and
look
at
what
makes
sense.
I
mean
right
now.
One
thing
you
could
do
is
like
is
like:
we've
done,
we've
implemented
like
a
backup
item
plugin,
but
you'd
wind
up
having
to
call
either
the
volume,
snap
shadows
or
something
else
to
get
the
system.
Specific
snapshot,
behavior.
G
C
G
B
G
F
A
A
B
G
Is
what
you
would
do
with
that?
That's
fine!
We're
looking
at
we've
got
some
other
projects
underway
that
we
have
some
snapshotting
some
file
system,
snapshotting
technology
coming
down
the
the
road
from
the
vsphere
side,
okay,
but
it's
strictly
for
our
own,
our
own
implementation,
so
figuring
out
how
to
make
that
a
more
generic
setup
would
be
a
good
thing
and
we've
got
another
project
running
the
asteroid
project,
which
is
trying
to
become
more
of
a
generic
plug-in
api.
G
F
A
A
Yeah,
I
think
I
yeah
this
this.
I
think
this
is
to
kind
of
generify
this.
If
that's
a
word
is
we're
looking
at
maybe
a
data
movement,
some
sort
of
data
movement
in
valero
support
for
data
movement,
but
maybe
that's
not
right.
I
don't
know,
but
for
examples
right
now
it
looks
like
we've
got
the
vsphere
plug-in
and
the
ebs
plug-ins
have
their
own
data
movers.
The
ubs
plug-in
have
a
mover.
F
G
Yeah
our
implementation
has
it
as
a
as
a
standalone,
not
quite
standalone,
but
it's
it's
separate.
Okay
and
there's
also
the
you
might
be
interested
in
the
kubernetes
data
protection
working
group
because
we're
hashing
through
a
bunch
of
things
there
and
we're
looking
for
other
lots
of
people
to
contribute
and
join.