►
From YouTube: Velero Community Meeting - August 20, 2019
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).
B
C
B
Say
observer
saying
now:
I
can
no
longer
Baker
myself
Carlita
Campos
and
that's
all
we
have
for
our
team
and
also
some
of
our
community
members
on
the
agenda.
Today
we
are
going
to
talk
about
as
always
what
we
have
been
working
on,
I
think
Steve
can
start.
We
talking
about
the
V,
1.1,
Bader,
bug,
fixes
and
part
of
pace.
A
Yeah
absolutely
so
yeah
over
the
past
week
or
two
since
we
put
out
the
first
beta
for
1.1,
we've
been
just
working
through
testing
and
doing
some
bug
fixes
so
definitely
wanted
to
say
a
big
thank
you
to
everyone.
Who's
been
testing
the
betas.
It's
been
really
helpful
for
finding
bugs
and
surfacing
from
there.
Some
scenarios
that
you
know
the
core
team
didn't
cover
themselves,
so
I
think
we're
in
pretty
good
shape.
A
A
So
we've
I've
been
working
on
this
issue:
number
192,
which
I've
talked
about
in
previous
community
meetings
and
sort
of
shared
the
design
for,
but
there's
an
issue
around
not
being
able
to
essentially
create
a
copy
of
a
persistent
volume
if
you're
doing
a
restore
that
involves
remapping,
a
namespace,
and
so
this
would.
This
could
occur
if
you're,
if
you
have
a
namespace
in
one
cluster
and
you're,
trying
to
create
a
second
copy
of
that
namespace
either
for
testing
purposes
or
for
whatever
other
reason.
A
But
if
that,
if
that
original
namespace
is
using
a
persistent
volume,
then
when
you
try
to
create
a
copy
of
the
namespace
velaro
previously
could
not
actually
create
a
copy
of
the
persistent
volume
as
well
so
I
started
doing
the
development
work
on
that
I
have
a
some
working
code,
so
I
just
wanted
to
do
kind
of
a
quick
demo
of
that
hasn't
been
merged
yet
so
this
is
development
bits
only,
but
let
me
switch
to
my
terminal
and
I
can
I
can
show
that
off
real
quick,
all
right.
It's
a
size.
A
A
A
Back
into
it
and
go
to
the
nginx
log
directory.
So
if
we
look
at
what's
in
here,
we
have
a
handful
of
different
files,
but
I've
added
this
file
called
prod
NS,
which
just
says
phew.
So
that's
that's.
What's
in
the
persistent
volume
right
now
so
first
thing
I'm
going
to
do
is
just
create
a
backup
and
I'm.
Just
gonna
include
the
nginx
example
namespace.
A
A
We'll
see
that
we
got
one
persistent
volume
snapshot
so
so
far
so
good.
So
now
what
I
want
to
do
is
I
actually
want
to
restore
from
that
backup,
but
I
want
to
restore
into
different
namespace,
so
I'll,
say:
Blair,
restore
create
from
backup
demo
and
then
I'll
use
the
namespace
mappings
flag
and
I'll
map
from
the
next
example
into
you
know:
nginx
demo,
so
I'll
just
go
ahead
and
run
this
actually
before
I
run
it
on
this
bottom
window.
Here
I'm
just
gonna
put
a
watch
on
PVCs
in
that
namespace.
A
A
Sorry
about
that.
So
if
you
look
at
the
watch
for
the
PV,
basically
our
old
PV
that
was
claimed
by
the
nginx
example
namespace
stuck
around,
and
then
we
got
a
new
PV
with
a
different
name
that
was
created,
became
available
and
then
was
claimed
by
the
nginx
demo
namespace
by
the
nginx
logs,
persistent
volume
claim.
And
if
you
look
on
the
bottom,
you
have
kind
of
the
PVC
side
of
the
story.
A
A
A
So
that
basically
covers
the
feature.
So
previously
you
wouldn't
be
able
to
get
a
second
copy
of
this
names
of
this
persistent
volume
created
from
the
snapshot.
But
with
these
code
changes
you
can
now
create
a
full
clone
of
the
name
space,
including
the
persistent
volumes.
So
I
know
that's
a
feature
that
many
of
you
out
there
have
been
anxiously
waiting
for
so
I'm
pretty
excited
to
bring
this
to
a
conclusion.
So
any
questions
on
that.
C
A
A
The
the
code
in
the
restore
package
for
persistent
volumes
has
a
bunch
of
different
branches
based
on
whether
you're
doing
a
regular
snapshot
restore
or
a
rustic
restore,
or
neither,
and
so
basically,
when
we
and
we
add
the
CSI
support,
we'll
need
to
make
sure
that
it
falls
into
the
appropriate
case.
There.
C
A
Yeah
so
that'll
definitely
be
an
interesting
thing
to
think
about
down
the
road,
and
you
know,
as
some
of
these
features,
that
we've
had
in
bolero
for
a
while
or
that
worked
for
traditional
velaro
snapshots
get
exposed
through
CSI.
How
do
we
kind
of
maintain
feature
parity
there?
So
it's
you
know
it
is
entirely
possible
that
when
we,
you
know
down
the
road
when
we
look
at
adding
support
for
something
like
this
for
CSI
snapshots,
specifically
that
we
would
actually
use
the
stream
clone
feature
rather
than
kind
of
having
our
own
implementation
of
it.
B
A
A
This
is
actually
an
area
that
I
still
need
to
figure
some
things
out
with
so
so,
typically
with
a
dynamically
provisioned
PV,
then
the
name
of
the
PV
is
PVC,
followed
by
the
UID
of
the
PVC
that
it's
provisioned
for,
but
we
can't
really
do
that
for
TVs
that
we
are
restoring
from
snapshot,
because
we
actually
explicitly
create
the
PV
first,
and
so
we
don't
yet
know
what
the
PVC
is.
Uid
will
be
so
for
now,
I
just
have
it
putting
a
random
UID
in
there.
A
But
if
you
look
at
the
you
idea
of
the
PVC
that
ends
up
claiming
it,
it
doesn't
actually
match
so
I'm
thinking
we
may
want
to
actually
not
use.
This
format
add
something
that
indicates
like
velaro,
created
this
PV
and
still
give
it.
You
know
some
some
random
component
of
the
name,
but
but
not
make
it
appear
like
it's.
A
dynamically
provisioned,
PV,
yeah
I.
Think
that
make
sense
is.
A
D
A
D
A
C
Yeah,
so
something
that
we
found
with
the
1.1
beta
one
actually
Scott
and
Dylan
over
at
Red
Hat
found
this
was
when
you're
doing
a
rustic
restore
with
1.1
beta
1,
the
the
rustic
CR
DS
that
we
use
we're
not
actually
being
properly
mapped
to
the
right
namespace,
and
this
was
a
bug
that
basically
stemmed
from
a
disconnect
in
how
valario,
server
and
Valera's
client
got
a
custom
name
space.
So
we've
unified
that
code
and
made
that
more
consistent
in
valero
1.1
beta
2.
C
C
C
C
Yep
and
as
part
I
mean
I
have
think
I'm
the
PR
open
here
and
as
part
of
that,
we
included
an
upgrade
dock
that
has
a
cube
control
patch
command.
That
will
do
that,
for
you
yeah,
so
yeah.
That's
really
the
only
upgrade
stuff
that
we
have
for
one
point:
one
everything
else
should
be
in
place,
but
if
you're
using
a
custom
namespace-
and
you
have
an
existing
install
that
you
want
to
upgrade-
you
will
have
to
run
run
this
patch
command,
which
will
go
add
that
downward
API
environment
variable
for
you.
B
If
is
there
any
comments
or
questions?
No
one?
Okay,
so
I'll
move
on
next
last
week,
I
think
it's
worth
mentioning.
I
was
out
on
Monday,
because
I
was
doing
a
mentoring
of
VMware
called
House
events.
It's
an
event
that
where
VMware
takes
a
group
of
20
women
and
it's
basically
a
hackathon
and
they
get
some
of
us
to
mentor
them
through
through
it,
and
this
year
the
team
was
doing
an
app
that
would
foster
diversity
and
inclusion
and
deployed
that
upon
communities,
and
it
was.
B
It
was
amazing,
great
experience
and
besides
that,
the
most
significant
thing
I
worked
on
was
writing
an
article
about
Valero.
There
is
a
we
don't
know
where
we're
going
to
publish
it
yet,
but
maybe
on
something
like
new
stack
and
I'm
co-authoring
it
with
the
product
project
manager.
And
basically,
this
article
covers
the
highlights
the
need
for
kubernetes
clusters
to
still
have
backup
in
that
department
that
things
don't
change.
B
Most
applications
are
for
apps
in
productions,
and
you
still
need
to
back
things
out
basically,
and
it
goes
into
heavily
allows
that
and
also
Mike
help
lets
you
migrate
data.
If
you
need
to
move
from
one
provider
to
another
and
with
any
question
about
that,
pretty
straightforward
stuff,
so
what
is
coming
up
is
what
we're
going
to
talk
next
and
I
will
hand
off
to
Steve.
A
Sure
I
can
cover
this
pretty
quickly.
So,
first
of
all
we're
you
know
planning
on
releasing
1.1
very
soon,
so
the
the
current
plan
is
Thursday,
certainly
a
potential
that
that
would
move
by
a
day
or
two,
but
but
unless
anything
we
should
be
releasing
that
on
Thursday,
so
keypads
up
for
the
announcement
there
and
then
beyond
that
we're
in
the
progress
of
defining
exactly
what
we're
gonna
tackle
for
1.2
we've
been
talking
a
lot
about
a
lot
of
these
things
already.
A
A
We're
planning
to
start
work
on
the
kind
of
the
official
CSI
snapshot
integration.
So,
as
you
probably
know,
Nolan
in
particular
worked
a
lot
on
prototyping
some
integration
with
CSI
snapshots
during
the
100
release
cycle
and
those
were
implemented
as
out
of
tree
plugins.
But
we
do
have
a
design,
that's
just
about
ready
to
be
approved
for
kind
of
the
official
integration,
and
so
we
will
start
work
on
that.
A
We're
also
looking
at
taking
the
existing
cloud
provider
plugins
for
AWS
gke
in
Azure
and
actually
moving
them
out
of
trees
so
that
they
kind
of
live
on
equal
footing
with
any
external
plugins
and
making
some
improvements
to
things
like
velaro
installed,
to
make
it
easier
to
install
falero
with
plugins
from
the
start.
But
this
is
is
mostly
kind
of
a
behind
the
scenes.
Refactoring
that
shouldn't
have
a
lot
of
of
user
impact.
A
You
know,
certainly,
if,
if
folks
are
interested
in
contributing
there
are
either
issues
within
those
those
bullet
points
or
additional
issues
that
are
in
the
backlog
that
we
we'd
love
to
have
help
working
on.
So
come
talk
to
us
if
you're
interested
in
that
and
then
I
guess.
The
last
thing
to
cover
here
is:
we've
decided
to
move
to
monthly
community
meeting.
So
we've
been
doing
two
a
month
for
the
last
few
months
and
I
think
we
feel
that
we
can.
A
We
can
kind
of
cover
all
of
the
content
that
we
want
to
and
still
have
a
fairly
regular
cadence
by
having
one
meeting
a
month.
So
our
current
plan
is
to
do
this
on
the
first
Tuesday
of
every
month,
and
so
the
next
meeting
would
not
be
until
I
guess
the
beginning
of
September
and
from
there
we'd
move
to
one
per
month
going
forward.
The
schedule
is
we're
still
making
sure
that
we
have
the
best
best
schedule
possible
for
everyone.
C
B
B
D
So
hopefully
this
is
big
enough
for
you
see
yes,
I've
been
working
on
this
new
from
schedule,
flag
in
Valera,
backup
great.
So
if
you
have
a
schedule,
here's
an
example
of
one
and
you
want
to
run
that
manually
at
any
point,
you
can
now
use
that
our
backup
create
from
schedule
and
give
it
a
name
under
schedule
and
it
will
go
and
create
the
exact
same
backup.
D
D
C
D
B
Think
we
are
ready
to
wrap
it
up
as
always
I
you
can
we
wanted
to
let
you
know
you
can
join
us
in
kubernetes
slack.
The
cello
is
valera,
and
I
want
to
remind
everyone
that
we,
our
survey,
is
running.
You
will
be
closing
soon.
So
if
you
haven't
had
a
chance
to,
let
us
know
usual
arrow
or
even
if
you
don't
use
forever,
but
what
your
the
profile
of
your
workloads
are.
B
We
want
to
give
a
shout
out
to
this
user
beta
one
beta
with
two
T's
for
including
object,
storage,
perfect
sin,
backup,
storage,
location
may
up
of
config
Tara
Nicolas
called
when
ski
for
a
documentation
fix
for
helm,
install
commands
user
code,
Chris
for
Docs
fix
for
port
forwarding
commands
and
finally,
Dylan
Marais
in
Scotts
Iago
for
testing
beta
1
and
finding
bugs.
Thank
you
so
much
for
the
O's
and
with
this
we'll
wrap
it
up,
stop
recording
and
we'll
see
you
the
first
week,
first
Tuesday
of
next
month,
goodbye
everybody.