►
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
Yeah
so
today
the
jennifer
today
has
demoed
the
proxy
cache
proxy
cache.
Is
that
the
ability
for
hardware
to
act
as
a
pull
through
cache
for
an
upstream
registry,
something
we
deliver
something
we
plan
and
deliver
for
the
2.1
payload,
just
quick
announcements.
We
released
the
201
and
the
1104.
A
If
you
head
over
to
the
release
stage,
the
zero
one
has
quite
a
few
fixes
for
gratification
trading
scanner.
There's
the
problem
with
the
session.
Timeout
definitely
check
that
out.
If
you're
entertaining
these
issues
before
now
we're
starting
to
look
at
the
work
for
the
2.2
so
on
the
harvard
project
board.
A
If
you
go
into
the
main
harbor
though,
or
the
project
page
and
click
on
projects,
copy
target
board
and
start
looking
at
some
of
the
things
that
we're
thinking
about
for
2.2,
these
aren't
necessarily
locked
down
just
some
things
that
we've
different
priority
too.
A
There's
also
a
whole
list
of
features
in
the
backlog
here
and
if
you
see
any
requirements,
their
feature
requests
from
the
backlog.
If
it's
something
that's
not
captured
here
already.
Definitely
you
know
raise
the
ticket
to
us,
just
bring
it
up
on
the
community
meeting
here.
A
A
So
this
is
in
reference
to
adding
additional
granularity
to
harvard
multi-tenancy
beyond
what
we
have
with
just
projects
right
now
right.
So
this
is
actually
a
pretty
large,
pretty
ambitious
feature
request:
it's
not
just
asking
for
granular
access,
but
the
ability
to
to
create
teams
of
users
and
robots
that
you
can
then
attach
to
a
set
of
projects
for
just
notion
of
a
workspace
right.
A
A
You
can
see
that
it's
been
planned
for
quite
some
time
already
and
yeah.
It's
definitely
something
we
have
to
do.
Prometheus
has
pretty
good
integration
with
communities
already
so
deploying
fibroids.
The
glutamate's
cluster
can
easily
expose
from
the
core
surfaces
the
endpoints
pausing
the
rest.
We
have
service
discovery
to
promote
these
things,
so
that's
definitely
something
I'm
tackling
through
the
2.2.
The
2.2
is
the
2.1
comes
out
end
of
august,
so
the
2.2
is
in
the
year
roughly.
A
So
here
I
have
an
environment
with
the
proxy
cache.
So
first
thing
you
notice
is
that
it
looks
just
like
a
project,
so
we
we
thought
about
whether
to
add
constant
cash.
As
you
know,
a
separate
panel
or
separate
entity
parallel
to
projects,
but
we
ended
up
just
borrowing
from
the
projects
workflow.
A
So
basically,
when
you
create
a
project
when
you
play
your
project,
you
have
the
ability
to
set
this
as
a
positive
cache,
and
then
you
can
set
from
these
lists
of
endpoints
here
that
you've
configured
as
your
registry's
endpoints,
very
similar
to
how
you
deal
with
replication.
A
A
A
So
you
know
the
assumption
is
that
I
don't
have
connectivity
to
dr
hub
right.
I'm
trying
to
hit
proxy
cash
allows
you
to
hit
a
registry
that
is
otherwise
off
limits
to
me,
and
so
the
proxy
camps
will
pull
it
down
and
attach
it
and
then
push
it
down
to
the
time.
So,
assuming
I
don't
have
it,
I
don't
have
access
to
the
docker
hub.
A
A
So
what
this
should
be
doing
is
that
harvard
should
be
pulling
it
down
from
proxy
cache
from
docker
help,
and
then
caching
and
then
the
client
will
pull
it
from
the.
A
A
You
can
double
check
so
413,
that's
when
it
was
pulled
yeah.
So
that's
how
it
works.
Essentially,
a
couple
of
things
to
notice.
A
So
you
can
set
the
the
proxy
cache
projects.
Okay,
it's
just
like
normal
project.
You
have
access
to
all
the
other
features
here,
like
you
can
imagine,
members
use
hypertension
tightening
ability,
weapons
everything
so
right
now
there
is
no
default
high
definition
policy
set.
You
have
to
you
have
to
consider
your
complication
policy
as
you
would
for
a
regular
project,
but
we're
thinking
about
adding,
like
a
default
number
of
days,
for
for
images
to
be
able
to
resist
in
the
possible
times
before,
exhausting
energy,
just
to
keep
it
off
the
tank.
A
That
depends
on
your
entitlement.
You
haven't
already
account
for
that.
So,
if
so,
look
at
how
we
create
a
project
match
right,
we
have.
If
you
look
at
how
you
create
a
registered
employment,
you
have
to
type
in
your
access
id
and
your
secret,
so
you
can
only
pull
the
projects
that
are
invisible
to
you
and
not
somebody
else's
right.
So
if
it's
a
private
project
for
one
user,
it's
not
available
to
another
getting
questions
so
far.
A
A
The
logs
are
a
little
spotty
right
now
it
doesn't
refresh,
and
if
there
is,
there
is
no
timestamp,
so
we
should
be
recording
all
the
actions
for
the
calls
from
the
processing
cache
right.
So
we
should
be
recording
the
the
poll
from
that
client
on
the
position
cache
or
we're
gonna
be.
We
should
be
recording
the
poll
from
the
upstream
registry
to
deposit
cash
as
well
as
push
down
to
the
client,
so
you
can
see
that.
A
And
then
the
behavior
is
such
that
every
time
you
pull
from
epoxy
cache
it'll
it'll
sync
with
the
upstream
to
make
sure
it
pulls
the
latest.
So
if
that
image
has
changed,
the
digest
has
changed.
A
So
that's
all
I
had
to
share
for
the
proxy
cache
any
any
questions
comments.
A
This
is
really
cool
thanks
for
doing
a
demo
of
it,
the
problem
yeah,
it's
something
that
a
lot
of
people
have
been
asking
right
now.
It's
gonna
be
limited
to
being
a
proxy
for
doctor
hall
in
harvard,
but
we'll
expand
that
list
of
epoxy
targeting
points
you
can
proxy
for
just
about
anything
right
now,
like
any
of
those
artifacts,
that's
supported
in
hardware,
but
that's
also
limited
by
what
is
supported
on
that
remote
or
upstream
registry
right.
A
A
Okay,
so
yeah
there
is
a
build
out.
I
will
post
it
on
the
on
the
units.