►
From YouTube: Harbor Community Meeting - Aug 25, 2021
Description
CNCF Harbor's Community Zoom Meeting
A
Hi
everyone
hi
everyone
today
is
august.
25Th
2021
here
is
harvard
community
meeting.
We
have
a
few
agenda
for
today's
meeting.
The
first
one
zooming
will
share
some
information
about
the
hubble
operator,
1.2.
B
Okay,
so,
since
recently,
our
hub
operator
has
just
released
the
version
1.1.
B
1.1.00
and
0.6
in
this.
In
this
release,
we
have
some
enhancements,
such
as
support,
harbor
2.3
and
for
kubernetes
1.2,
21
and
14,
the
kubernetes
network.
If
I
working
with
one
and
we
also
improved
some
gravity
definition
and-
and
we
introduced
a
new
2
operation,
which
user
can
use,
crv
based
configuration
to
define
to
define
the
computer,
and
we
also
support
the
congress
and
the
load
balance
service.
B
B
B
Next,
we
will
in
excess,
more
storage
driving
drivers,
both
supporting
include
adr,
allen
and
the
google
cloud.
B
We
will
also
try
to
speak
an
issue
and
which,
which,
when,
when
users
are
updating
a
hubble
cluster
back
to
to
delete
some
components
such
as
tv
deletivity.
B
B
That
can
make
our
that
can
save
some
memory
in
order
to
make
make
it
better
in
in
edge
environment
and
our
and
review
and
target
fc
date
in
early
november
and
the
next.
We
also
have
some
backlog
chat,
such
as
we
intend
to
where
the
data
back
up
and
restore
in
our
operator
about
reason,
and
we
want
to
use
use
to
study
based
results
to
to
to
manage
harbor,
which
is
which
is
a
more
kubernetes
base,
and
we
will
add
more
data
operations
such
as
manage
the
hybrid
c
server.
B
B
B
B
And
we
want
to,
and
we
want
to
have
an
operator
and
we
want
to
merge
all
operator
dependency
operator
into
into
into
one
controller.
Currently
we
have
such
as
io
operator,
credit
operator
and
also
a
hub
operator.
B
We
want
to
motivate
them
into
our
filter
and,
and
we
we
also
invest,
they
are
also
investigating
if
we
can
support
no
pause
service,
no,
not
post
exporting
to
to
expose
the
hybrid
service
in
other
that
some
developer
may
may
want
to
use
this
to
improve,
to
develop
devices,
and-
and
maybe
we
also
want
to.
B
D
Yeah
this
this
is
this
is
a
problem.
It
may
change,
so
the
user
after
after
the
user
change
changes,
admin
password
should
update,
update
as
a
admin
secret
resource
in
the.
C
Yeah,
as
for
the
support
other
asset
providers
and
running
full
harbor
operator
in
a
single
process,
do
you
have
issues
open.
B
C
B
E
B
B
C
C
B
A
Yeah,
we
will
share
the
proposal
to
oh
stop
scanning
job.
This
is
a
request
from
community.
Some
of
you
are
positive
to
you.
F
F
Stopping
scan
job
of
a
particular
artifact,
as
we
all
know,
is
a
nice
ability
to
scan
images
to
identify
potential
vulnerabilities
to
avoid
a
malicious
attack,
and
currently
hubble
supports
fury
and
glare
as
to
scanning
your
f2
scanning
tools
to
achieve
this
purpose,
which
can
be
configured
during
hub
installation
and,
generally
speaking,
there
are
two
kinds
of
scanning:
one
is
scan.
All
images
and
the
other
one
is
scan
a
particular
image.
F
So
it's
worthwhile
to
have
scan,
stop
scan
job
in
harbor
to
terminate
scan
or
image
job
or
terminate
scan
job
of
a
particular
image.
F
So
this
is
the
community
issues
being
arrested,
talking
about
the
potential
resource
consuming
or
stock
for
days,
and
our
goal
for
this
proposal
is
really
simple
and
straightforward,
which
are
first
be
able
to
terminate
a
scan
or
job
at
the
system,
level
for
all
artifacts
and
second
be
able
to
cancel
or
stop
be
able
to
cancel
a
scan
job
of
a
particular
artifact
and
as
for
breaking
changes
and
the
compatibility
there
should
be
no
breaking
changes
to
existing
functionalities
and
no
compatibility
issues.
F
F
There
are
three
major
steps
and
first
defines
lego
api
for
both
stop,
scan
or
and
stop
scan
a
particular
artifact
respectively,
and
the
second
implement
the
api
handler
and
third,
do
you
should
do
should
stop
check
in
the
actual
raw
method
within
the
scan
job
file,
the
scan
job
file
implemented
job
service
interface,
and
this
is
and
let's
talk
about,
stop
scan
off
job
first,
and
this
is
the
api
specified
in
for
stop
scan
all,
and
essentially,
we
will
have
http
post
method
for
these
systems
that
scan
all
stop
url
path
and
the
handler
for
this
api
is
implemented
here
and
essentially
like
scan.
F
F
In
both
cases,
they
are
managed
by
execution.
So
we
need
to
find
the
execution
first.
Once
we
have
the
execution,
the
execution
manager
can
will
invoke,
stop
method,
passing
execution
extrusion
id
as
a
parameter
to
take
care
of
it.
Essentially,
it
will
find
a
list
of
tasks
managed
by
the
execution
id
either
with
them
one
by
one
and
stop
each
every
single
task.
F
F
So
then,
let's
talk
about
the
handler
of
of
this,
stop
scan
job
of
a
particular
artifact.
F
So,
as
we
can
see
in
the
api
definition,
we
will
have
some
parameter
called
project
name
repository
name
and
references,
and
essentially
we
will
first
get
the
current
artifact
by
reference
of
this
parameter
pass
through,
and
the
next
story
is
a
little
bit
different
because
scan
api
is
the
behavior
of
scan.
Api
is
defined
by
scan
controller
interface
and
we
don't
have
a
stop
method
in
the
scan
controller
interface.
F
Yet
we
will
need
to
introduce
a
new
method
called
staff
in
the
scan
controller
interface,
which
is
here,
and
the
stop
methods
will
be
implemented
in
this.
In
this
scan,
controller
will
be
implemented
like
the
below
like
the
following.
First,
it
will
now
get
the
execution
of
the
artifact
by
the
artifact
id
and
then
like
before.
F
The
execution
manager
will
invoke
stock
method
and
passing
execution
id
as
a
parameter
to
take
care
of
it,
and
so
far
we
have
talked
about
the
api
design
and
the
handle
of
the
api
for
both
stop
scan
or
and
stop
skin
job
of
a
particular
artifact.
F
The
last
piece
is
that
we
need
to
do
should
stop
check
in
the
actual
wrong
method
within
the
scan
job
file
and
because,
like
if
there's
a
running,
scan
job
right,
it
should
be
able
to
perceive
the
request
the
stop
requests
being
received
by
harbor
system
right,
so
it
can,
and
just
it
can
stop
its
currently
running
job.
So
to
do
that,
we
went
we
should
do.
F
We
would
do
a
shoe,
stop
check
in
the
wrong
method
within
the
scan
job
file,
which
is
the
actual,
which
is
the
stop
actually
actually
happened,
happened
like
shoe.
Stop
is
essentially
a
an
anonymous
function
to
be
added
in
the
wrong
method,
and,
and
the
should
stop
check.
Will
be
invoked
in
the
following
scenarios
at
first,
it
will
be
invoked
when
a
scanner
client
being
created
second,
like
when
the
before
the
client
actually
submitting
a
scan
request
and
and
third
and
before,
like
every
time,
the
scanner
one.
F
When,
when
the
scanner
trying
to
fetch
the
scan
report,
it
will
check.
Should
I
stop?
If
I
should
stop,
then
it
will
be
stopped
and-
and
the
fourth
is
that,
before
persistent
persisting
scan
report,
data
to
database
like
before
persistence,
persisting
any
scan
report
to
database,
it
will
check
the
should
I
stop.
If
it
is,
if
there's
a
stock
request
being
received,
then
it
will
stop
the
persisting
process
and
that's
essentially
my
proposal
for
stop
scan
job
and
thank
you.
C
F
F
C
F
D
D
The
same
similar
pass
to
stop
stop
this
job.
C
But
that
is,
there
may
be
a
few
issues.
For
example,
the
delete
action
normally
is
item
total,
which
means
you
delete
the
resource
multiple
times
you
receive
the
same
results,
but
if
you
use
it
to
stop
a
job
you,
it
cannot
be
item
token.
You
know
what
I
mean.
C
C
F
A
Yeah,
I
have
a
question
where
these
two
new
api
exposed
to
to
the
robot
account
I
mean,
do
the
I
mean
do
the
robot
account
can
use
this
two
new
api.
A
D
For
robot
account,
I
think,
maybe
we
need
a
other
new
policy
for
the
rural
account
for,
for
example,
laris
or
scan
a
scan
artifact
policy
to
store
to
stop
the
scans
of
oil,
maybe
lead
or
stop
stop
for
scan
artifact
resource
resource
for
the
policy
of
over
the
count
under
the
scan
and
the
under
this
full
scan
scan,
or
I
think
we
also
need
a
true
audio
resource
for
the
robot
robot
access
policy.
A
D
F
A
Okay,
let's
move
to
the
next
agenda:
yeah
we
have
a
alex,
have
a
exciting
agenda.
I
want
to
be
share
with
the
community
alex.
Can
you
talk
about
your
alex.
G
Hey
yeah
yeah,
so
we're
yeah.
I
mean
we're
really
happy
that
tnn
has
accepted
our
invitation
to
be
a
maintainer,
so
you
know
we
are
growing
and
the
project
I
feel
like
has
you
know
since
graduation.
I
think
we've
really
gotten
to
a
point
where
the
community
has
become
sort
of
self-serving.
G
I
see
a
lot
more
people
stepping
in
to
answer
questions
and
to
help
others
troubleshoot,
but
you
know
there's
only
so
much
we
can
do
in
in
one
of
our
releases
right.
We
have
four
releases
a
year
and
there
there's
still
a
lot
of
items
to
be
done,
and
so
we've
extended
an
invitation
to
tienen
gravi
to
be
a
maintainer
for
the
harbor
project.
G
Dana
has
been
with
the
project
for
a
really
long
time
he's
you
know,
he's
been
attending
the
the
office
hours
and
the
community
meetings,
the
u.s
community
back
when
we
had
them
and
now
he's
attending
the
the
asia
based
ones.
So
really
grateful
for
everything.
Tien
has
done,
and
I'm
super
excited
that
you
know
we
have
someone
with
that
level
of
expertise
joining
the
harvard
maintainers.
I
don't
know
if
you
want
to
say
a
few
words
tendon.
G
Yeah
absolutely
definitely
want
you
to
be
here,
so
thanks
so
much
that
was
really
my
announcement.
I'm
gonna
submit
a
pr
to
nominate
and
you
know
we'll
go
through
the
process
to
get
someone.
You
know
officially
added
to
the
project
as
a
maintainer,
so
I
don't
think
we're
gonna
you
know
designate.
I
think
we
created
a
special
role
right,
like
a
a
repository
based
maintainer
or
no.
G
That
just
means
that
that
maintainer
works
on
that
repo
within
harvard
project
will
be
a
full
full
maintainer
of
harper,
just
like
everyone
else.
G
G
He's
in
the
u.s
and
he's
closer
to
the
west
coast
so
he's.
A
Yeah
yeah
welcome
and
I
also
noticed
that
tanner
attend
harvard
community
in
took
us
with
us
every
time.
Almost
every
time
I
remember
yeah,
that's
really
impressive.
A
Okay,
great
so
that's
here
we
have
done
for
today's
meeting
and
thanks
for
all,
thank
you
all
for
your
journey
and
see
you
next
time.