►
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
A
First
order
of
business
is
you
know
a
huge
congratulations
to
the
team.
Harbor
has
graduated
in
cncf,
so
this
is
a
huge,
huge
achievement.
It's
a
testament
to
the
effort,
the
community,
the
work
that
the
hardware
team
has
put
in
the
the
community
contributions,
our
partnerships
and
everything
else
that
basically
went
in
to
create
this
project
so
huge
kudos
to
the
team.
We're
not
satisfied,
we're
gonna,
keep
going.
A
Excuse
me
we're
gonna,
keep
going.
We
have
a
huge
vision,
a
roadmap
ahead
of
us,
so
I'm
gonna
be
working
on
on
basically
delivering
on
that
speaking
of
roadmap.
We
are
actually
updating
the
roadmap
documents
for
most
of
our
projects
and
I
believe
hardware's
roadmap
should
be
updated
today.
If
it's
not
it
should
get
in
there
tomorrow.
Let
me
go
in
there
right
now
to
check
nope,
it's
not
updated
yet,
but
it
should
be
updated
in
the
next
couple
of
days.
A
So
a
couple
of
a
couple
of
announcements
on
releases
version
2.0.1
release
candidate
is
out
that
includes
some
fixes
around
the
tree
scanner.
Some
replication
fixes
retention,
some
ui
stuff,
you
know
like
we
had
a
major
release
for
hardware
and
as
as
expected,
any
major
release,
there's
some
fit
and
finish
things
that
kind
of
fell
through
the
cracks.
A
So
so
the
team
kind
of
worked
very
quickly
to
to
address
those
and
and
make
sure
that
you
know
the
community,
you
know,
gets
a
stable
release
of
hard
work
that
they
can
plan
on
and
there's
a
total
of
46
issues.
Overall,
a
lot
of
them
are
just
ui
fit
and
finish.
A
We
also
did
a
couple
of
security
fixes,
for
example,
you
could
test
web
hooks
today.
So
if
you
go
to
a
web
hook-
and
you
create
one
there's
a
test
button,
but
the
test
button
allows
you
to
input
any
arbitrary
uri
in
there
or
or
ip
address,
and
someone
could
actually
use
that
for
port
scanning.
So
we
actually
addressed
that
by
removing
that
all
together.
A
So
now
you
cannot
test
web
hooks,
but
we
may
enable
that
later
on,
based
on
policy,
if
users
want
to
there's
also
a
couple
of
other
security
kind
of,
let's
call
it
security
hardening
issues
that
we
fixed
and
everyone
will
be
able
to
see
them
once
the
release
fully
comes
out.
But,
overall,
you
know
good
progress,
we're
hoping
to
ship
that
very
very
soon
we
also
shipped
1.10.3
last
week,
so
that
addressed
quite
a
few
issues
and
then
home
1.4.1
will
be
out
next
week.
A
A
If
you
all
remember,
2.1
is
a
major
release
of
hardware,
it's
slated
to
be
late
august
and
it's
going
to
have
capabilities
like
proxy
cache,
non-blocking
garbage
collection.
So
you
can.
Actually
you
know.
This
is
something
that
I
know
chris
and
your
team.
You
probably
have
been
waiting
for
that
as
your
sizes
of
hardwood
installations
are
coming,
bigger
and
bigger
the
ability
to
run
garbage
collection
reclaim
some
of
the
disk
without
really
blocking
hardware
operations,
it's
going
to
be
huge
and
then
the
last
part
is
p2p
integration.
A
So
those
are
the
four
big
features
of
2.1
and
we're
making
great
progress
on
that
on
the
proxy
cache,
there's
actually
a
spec.
So
if
you're
interested
in
that,
let
me
share
my
screen
really
quickly
right
now-
and
this
is
the
uri
for
the
spec
is
pr144.
A
Essentially
the
way
it's
going
to
work.
Is
you
know?
I
can't
really
show
you
through
the
spec
here,
but
I'm
going
to
show
you
by
going
to
the
demo.harbor.io,
now
be
aware
that
this
environment
of
hardware
does
not
have
proxy
caching
in,
but
I'm
just
going
to
walk
you
through
how
that
looks
like
the
first
thing
you
do.
Is
you
add
the
registry?
A
A
A
Once
you
do
that,
essentially,
you
can
now
create
a
new
project,
but
as
you'll
come
in
here
and
you're
saying,
I
would
like
a
new
project
and
you
enter
the
details
here,
like
project
name
and
access
level
and
storage
code
you're
going
to
see
a
new
button
here
that
says:
do
you
want
this
project
to
be
a
proxy
cache?
And
if
you
do
you
get
to
select
that
registry
that
you
just
added
as
a
target
and
then
immediately
you
can
proxy
anything.
You
had
access
in
that
account.
A
So
let
me
give
you
an
example
here.
So
in
this
case,
if
I
go
to
docker
hub
and
if
I
were
to
provide
my
credentials
for
docker
hub,
you
see
here,
I
have
a
few.
My
own
account
is
mitch
mike
think
of
this
account
as
the
equivalent
of
a
project
in
harbor.
This
is
a
logical
entity
under
that
I
have
lots
of
repositories
and
each
of
them
has
tags
and
versions.
Right,
like
you
know,
this
block
engine
has
lots
of
tags,
for
example,.
A
So
if
I
wanted
to,
if
I
added
the
registry
that
basically
associated
hardware
with
my
docker
hub
account,
I
could
actually
proxy
cache
any
of
these
projects.
I
could
proxy
cache
project
contour,
which
mic
mike
ielab
and
the
way
that
it
would
look
like
is
if
you,
when,
when
you're
looking
at
it
in
hardboard,
basically
you're
going
to
call
it
this
way.
It's
going
to
be
demo.coharbot.io,
that's
your
uri
for
hardware!
Then
the
next
part
will
be
your
project
name.
A
A
So
I
could
enter
project
contour
and
then
put
contour
and
basically
do
version
1.3,
for
example,
version
1.2
and
I'll
be
able
to
proxy
cache
these,
or
I
can
remove
this
and
put
me
to
mic
and
then
be
able
to
block
engine
and
be
able
to
proxy
car
blog
engine.
So
this
is
the
beauty
of
proxy
cache.
You
get
to
associate
one
project
with
an
account
you
have
and
anything
that
that
account
has
access
to
can
be
proxy
cached
through
hardware,
very
similar
to
how
replication
works
as
well.
B
Yep,
that's
the
process,
and
so
I
just
want
to
point
out
that
it's
only
gonna
be
the
assist
admin
that
can
do
this
because,
like
previously
or
currently,
any
project
admin
can
create
projects
right.
But
the
project
that
I'm
in
here
or
the
person
here
doing
the
the
creation
of
the
proxy
project-
you
know
by
assumption-
has
access
to
all
those
endpoints.
B
So
it
has
to
be
a
synonym,
not
just
any
use
so
when
he
hits
that
we're
still
working
on
it,
but
I
presume
when
he
selects
you
know,
for
this
project
to
function
as
a
property
cache,
then
his
his
text,
permissions
will
will
kick
in,
and
so
he
only
gets
to
enter
or
select
the
from
the
list
of
target
registries.
If
he
is
assist,
man.
A
And
obviously
you
know
we
may
not
get
100
right
this
feature
the
first
go.
I
mean
this
is
the
first
time
we're
trying
this
out,
there's
not
that
many
registers
out
there
that
do
proxy
caching
or
mirroring
so
we'll
be
trading
and
getting
customer
feedback
and
trying
things
out.
A
So
as
folks
I
mean
this
is
going
to
be
super
useful
for
folks
that
want
to
be
in
semi-air-gapped
environments
or
they
want
like
a
local
mirror
of
the
registries,
so
we're
going
to
get
a
lot
of
feedback
from
those
companies
and
users
and
see
what
it
looks
like
like
alex.
I
guess
one
of
the
key
things
that
we
should
try
is
what,
if
the
target
registry
is
another
hardware
instance,
and
you
are
entered
system
administrator
credentials
for
hardware,
they
should
be
able
to
proxy
cache
anything
from
hardware.
In
that
case
right,
we
should.
A
Yeah,
that's
it
for
today.
That's
all
the
items
we
had
any
questions
concerns
ideas,
thoughts.
C
A
We
are
actually
in
a
bit
of
a
pickle
because
we
have
to
support
all
of
them.
Now
we
support
the
helm
chart
and
that's
why
we're
releasing
a
new
version
we.
So
we
have
like
five
different
deployment
methods
we
have
the
helm
chart.
We
have
the
operator
that
we're
trying
to
drive
towards
a
version
1.0
and
we're
making
great
progress
there.
A
We
have
the
tile,
that's
heavily
involved
and
heavily
deployed
in
enterprise,
pks
or
tons
of
kubernetes
grid
environments
of
vmware
ships,
and
then
we
maintain
that
and
then
we
also
creating
there's
a
a
an
effort
to
produce
more
simple
yaml
files
that
enable
the
deployment
and
packaging
of
hardware
using
ytt
and
cab
and
we're
also
working
on
delivering
those
as
well.
A
I
think,
over
time,
when
I
say
convergence,
I
want
the
home
charts,
white
team
cab
all
to
use
the
operator,
so
we
maintain
the
operator
and
everything
else
depends
of
it
and
then
over
time,
kind
of
defunct.
The
tile
right,
just
you
know,
get
it
to
a
point
where
the
tile
is
no
longer
a
needed
or
a
maintained
component.
A
So
the
operator
will
be
the
main
thing
that
we
have
the
cluster
operator
and
the
and
the
core
operator,
and
then
everything
else
would
just
basically
come
on
top
of
that
and
layer.
On
top
of
that
now
will
it
take
us
three
months
to
get
there
or
six
months?
I
don't
know
we
have
a
lot
of
work
to
do.
We
haven't
gotten
the
operators
in
version
1.0,
but
that's
a
long-term
vision.
Until
we
get
there,
we
have
to
support
them
all.
A
Cool
all
right,
there's
nothing
else.
Everybody
have
a
great
rest
of
your
day.
Thank
you
all
for
attending.
We
had
low
attendance
today.
Maybe
we
graduated
and
everybody
stopped
caring
about
hardware
by
the
way
on
the
graduation,
the
announcements
are
gonna
go
out
next,
tuesday
from
cncf
you're
gonna
see
tweets
a
press
release
everything
and
so
you'll
be
able
to
retweet
it,
and
you
know
again
many
thanks
to
the
community
and
the
users
contributors
of
hardware
you're,
the
ones
that
got
us
here-
cool.
Thank
you
all
have
a
great
rest
of
your
day.