►
From YouTube: Support Release Review Party - 12.5
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
C
D
B
A
B
B
So
I
can
everybody
should
see
my
screen
yep?
Yes,
okay,
so
the
first
thing
we
are
going
to
look
at
is
getting
a
list
of
the
feature
flags.
Now
all
of
these
examples
are
taken
directly
from
the
documentation
except
the
last
one
which
is
for
deleting
a
features
bag,
but
so
you
can
also
try
this
on
your
own.
I
can
also
put
them
in
the
dock,
so
I'm
just
gonna
go
to
my
terminal
and
and
run
this
one.
B
So
this
is
on
my
own
test
get
level
instance:
I,
don't
have
admin
access
to
get
live-dot-com
and
neither
would
I
want
to
turn
off
or
on
feature
flags
from
gearlive.com.
So
when
I
run,
this
API
call
I
get
back
to
feature
flags,
one
is
called
the
CI
dag
limit
needs
and
the
other
one
is
called
CI
tag.
Support
I
turned
on
these
two
feature:
flags
for
one
of
our
previous
release,
parties
where
I
was
demonstrating
the
CI
dad
feature.
B
B
B
B
So
here's
the
API
call
so
here's
what
I'm
here,
what
I'm
doing,
if
it's
a
post
request
actually
and
what
I'm
doing,
is
that
I'm
going
to
be
creating
a
new
feature
flag
on
my
gate
level
instance-
and
this
is
called
see,
I'd,
add
support.
That's
the
same
one
I
deleted
previously
and
its
value
can
be
false
or
true,
either
way
that
feature
flag
is
gonna
get
created
on
the
instance.
So,
let's
start
with
false
and
I
do
this
I
will
return
whatever
feature
flag
just
got
created
now,
I
can
go
and
run.
B
B
Now,
if
I
lift,
it
still
sure
we
can,
it
set
the
true
and
the
state
is
now
on
so
previously
when
the
value,
what
solves
the
state
was
off.
So
now
that
the
state
is
our
first
Avicii
friends
active
and
that's
pretty
much
it
for
the
features
last
API,
we
can
also
turn
the
feature
flags
on
for
different
users,
and
these
are
groups
and
so
on.
So
this
extra
parameters
in
the
API,
so
that's
all
from
my
side.
D
D
D
D
D
D
So
the
AWS
account
ID
tells
gitlab
where
to
find
the
user
that
it's
got
going
to
assume.
The
role
for
external
ID
is
a
unique
identifier.
This
prevents
it
from
being
a
global
configuration,
so
you
can
imagine
on
github.com
and
if
you
didn't
have
external
ID
and
everyone
was
using
a
semi
an
ideal,
I
could
all
access
each
other's
projects
and
the
cost
is
associated
with
that
project.
D
Provision.
Roll
a
are
in
that's
something
that
AWS
gives
us,
so
we
put
that
in
here
and
that
will
allow
us
to
assume
that
role.
So
once
you've
reached
this
screen,
you
want
to
go
to
iron
management
in
AWS.
For
those
services,
I
am
now
the
count
I'm
using
is
the
one
from
one
password.
It's
the
AWS
underscore
support,
underscore
testing
account.
So
if
you
need
to
reproduce
issues
or
whatever
you
can
use
that
account
to
do
this,
I'll
go
into
roles.
D
D
D
D
You
go
be
careful
not
to
click.
It
gives
a
big
click
it.
It
goes
to
the
policy
screen
and
you
can't
sit
back
and
then
you
go
start
over
again.
So
I
just
want
to
check
the
box
go
to
tags.
This
doesn't
need
any
tax.
What
we're
doing
you
can
tag
it
if
you,
if
you
are
setting
up
cluster,
you
probably
want
to
tag
it.
D
D
D
D
D
There
instances
number
of
nodes
once
this
is
all
filled
in
you'll,
create
your
Cuban
Inez
cluster
I
won't
do
it
now
because
it'll
take
10-15
minutes
to
create
and
then
we'll
create
a
whole
bunch
of
different
machines
in
AWS,
which
will
have
to
go
and
clean
up,
which
is
actually
something
else
to
note.
This
will
not,
if
you
delete
your
cluster,
your
integration
of
won't
delete
the
machines
from
kubernetes
from
out
of
your
sir,
which
is
a
real
pain,
in
fact,
I
think
it's
documented
somewhere
in
the
post.
D
There
was
a
warning
about
it
some
way.
So
unless
you
want
to
go
in
and
remove
all
those
machines
manually
and
all
of
their
associated
services,
I,
don't
recommend
creating
a
cluster
just
for
kicks
unless
you've
got
an
empty
AWS
account
and
it's
easy
to
find
everything
and
remove
it
and
not
gonna.
Interrupt
other
people
or
some
machines
or
anything
now
until
now,
does
anyone
have
any
questions
about
integration
with
github
back
on.
D
D
One
is
this
feature
flag.
This
is
required.
Well
said:
Oh
was
required,
however.
Id
actually
didn't
have
I
didn't
require
massive
mana,
so
the
blog
posts
says
for
self
manage
installs.
You
must
enable
the
feature
because
it's
behind
the
feature
flow
one
was
already
enabled
so
don't
really
know
what's
going
on
there
yet
and
the
other
thing
to
be
aware
of
is
when
working
with
self
manage.
You
need
to
set
up
your
own
aw
count
globally.
So
this
is
my
self
or
self
manage
instance.
D
D
D
D
Ran
into
an
issue
doing
this
and
self
managed
I
set
it
all
up
according
to
the
documentation,
it
still
wasn't
working
when
I
looked
at
the
logs,
it
was
during
a
four
hundred
and
I
think
the
reason
it's
throwing
the
four
hundred,
even
though
it
didn't
tell
me
this
I
pretty
sure.
The
reason
is
that
on
you
can
save
myself
first
instance
he's
actually
using
HTTP
and
it
doesn't
have
a
domain
name
Alaba,
so
I
think
AWS
just
doesn't
trust
it.
So
that's
something
that's
likely
to
come
up
in
support.
D
D
D
F
Adam
related
to
the
I
am
role.
I
was
working
on
the
ticket
where
customer
was
facing
issue
when
they
are
sitting
on.
The
I
am
goal
for
auto-scaling
their
runner,
so
is
the
configuration
for
I
am
role
the
same
as
this.
D
The
I,
the
user,
so
it
needs
to
decrease
and
it
needs
to
use
user
credentials
on
one
a
bit.
A
little
use
account
to
access
resources
on
the
other
to
access
those
resources.
It
will
assume
the
role
so
there's
two
reasons
to
use
ir
roles.
One
is
if
you're
using
multiple
accounts,
you
know
WS,
so
you
can
access
resources
between
those
accounts.
You
can
share
resources
for
Monica
now
to
another
you'll
see,
there's
a
lot.
D
If,
for
example,
someone's
got
a
number
of
clients
and
different
Ava
WIC
clients
accounts,
then
those
clients
can
share
these
resources
with
the
main
account
and
then
they
will
have
access
between
each
other
and
the
other
reason
is
to
limit
scope.
So
you
can
set
the
permissions
on
the
world
and
then
the
scope
that
way
so
I'm,
not
too
sure
how
that
applies
to
your
question,
but
it's
actually
quite
simple.
Once
you
start
playing
around
with
them,
you'll
sort
of
just
get
a
feel
for
what
they're
used
for
cool
thanks.
G
Yeah,
unfortunately,
I
don't
have
much
to
show
my
points
anyway,
so
exception
and
bite.
Racers
are
now
in
API
loss
in
production,
Jason,
Knox
I
believe
this
was
because
of
an
issue
created
by
Kody
and
thence,
then
sort
of
took
it
and
ran
with
it.
So
now
we
do
have
those
in
the
API
and
Jason
locks,
one
less
reason
to
look
at
the
mess
that
is
production,
the
law.
G
We
have
a
number
of
omnibus
improvements
as
well.
The
main
one
to
cloud
is
that,
right
now
we
are
doing
automatic
backups
of
the
configuration
folder
on
upgrades
and
the
default
option
is
actually
just
config
underscore
backup
in
the
configuration
folder.
Interestingly,
there
was
actually
a
Gillett
CTL
command
to
back
up
the
configuration
for
the
which
was
introduced
in
Chavan
tree,
which
I
had
no
idea
about.
So
that's
something
to
know.
G
Container
registry
is
now
enabled
by
default,
so
one
less
need
for
administrators
to
actually
do
something.
Eli
Bobby,
to
get
things
to
work,
give
like
metamours.
Now
she
was
just
a
key
Italian.
I
can
show
this
one
just
very
briefly,
although
I
didn't
manage
to
get
it
fully
working
for
some
reason,
but
let
me
share
my
screen
yeah.
G
G
G
Yep,
this
is
the
documentation
for
the
met
the
most
plugin.
It's
not
actually
meantime,
as
we
just
decided
ship.
It
so
register
configure
the
plug-in
alright.
So
once
that's
done,
you
can
be
working
at
connect.
You
actually
bring
you
to
a
page
to
connect
I've
already
done
it.
So
that's
that
and
once
you
have
that
you
can
actually
set
up.
You
actually
have
access
to
slash
for
months.
So
I
think
the
most
interesting
one
for
me
was.
G
Session
once
I
could
get
documentary
as
well,
but
yeah
there's
a
tool
which
you
can
use
this
year
to
do
this,
which
is
nice
and
there's
also
a
subscribe,
which
will
actually
send
notifications
of
relevant
project
updates
to
the
channel
itself.
So
you
do
this
by
doing
slash
he
likes
subscribe
and
then
clean
the.
G
But
the
important
thing
to
know
here
is
that
you
can
set
up
a
instance
level
work
for
this.
You
have
to
set
up
a
project
level
web
book
forever,
using
a
project
that
you
want
to
integrate
with
in
this
way.
So
you
do
have
to
you
know,
go
to
every
single
project
and
then
do
that
I
think
that's
it
for
metals.
I've
just
found
interesting,
because
I
very
seldom
look
at
metals
issues.
So
if
you're
serious
yeah,
this
is
what
metamours
looks
like
any.
G
That's
a
good
question:
I'm,
not
sure
it's
actually
mentioned
in
the
release
post.
Let
me
just
show
you
the
release
post,
so
the
really
suppose
starts
mentioned
this
documentation
right
here,
but
as
far
as
whether
it's
actually
including
dogs,
I'm,
not
very
sure,
I,
probably
looked
at
a
Flickr
cool.
A
second
fine
thank.
F
G
G
Okay,
I
think
that's
it
for
the
meta
moss.
So
the
last
one
is
something
I
haven't
really
looked
into,
but
I
think
if
you
have
relevance
for
support
of
hey
Shay,
so
Khonsu,
which
we
use
for
service
discovery
and
other
things
for
Postgres
has
been
upgraded
from
zero
point.
Nine
to
one
point
six
point:
one
I'm
not
sure
other
implications
are
but
I
think
that's
good
to
coop.
G
It's
actually
very
interesting.
It
is
actually
used
by
sis
or
infrastructure.
It
means
to
define
managed
services
as
available
resources
and
then
application
admins
can
actually
provision
these
services
by
claiming
those
resources
in
Altru
using
cube,
CTL
yammer
configuration
files.
So
it's
very
similar
to
the
idea
of
per
system
volume
claims
and
persistent
volumes,
except
in
this
case
was
she
doing
things
like
manage.
Postgres
Redis,
s3
buckets
that
kind
of
thing.
G
The
interesting
thing
here
as
far
as
active
abusers
crossplane.
In
fact
our
EEOC-
has
glowing
testimony
about
Crossley
and
on
the
crossing
website
might
be
interesting
that
I
put
in
links
to
documentation.
So
some
of
the
implementation
details
I
found
interesting
was
that
we
are
currently
installing
zero
point.
Four
point,
one
using
the
chats.
The
latest
version
is
0.5,
so
there's
something
to
watch
out
for
in
case
we
look
up,
look
up
documentation
when
we
are
trying
to
debug
this.
G
You
currently,
we
only
support
adding
a
manage
database
because
of
the
way
we
install
the
app
so
to
install
when
we
install.
Yet
we
do
need
to
provide
permissions
to
the
cross
plains
of
this
account
in
order
to
actually
provision
these
resources
and
right
now,
the
cluster
row
and
cluster
row
binding
settings.
We
have
a
limited
to
the
database
tour
cross
plane,
the
I/o
API
group.
G
So
what
this
means
is
that
we
can
only
actually
or
rather
see
a
means
and
application
owners
can
only
actually
request
for
a
database
right
now,
if
you
are
using
our
manager.
So
here's
some
of
the
stuff
I
ran
into
trying
to
get
it
going
on
my
own
instance
and
clusters.
So
if
you
are
trying
to
do
this
with
a
pre-existing
project
where
you
already
have
an
environment
where
you
already
have
F
deployed
environment,
the
existing
service
accounts
will
not
be
updated
to
use
the
new
crossplane
permissions.
G
So
I
think
what
you
actually
need
to
do
is
to
delete
and
recreate
in
space
right
now.
You
have
to
do
this
via
voice
console
and
cube
cgl
beginning
from
shell
point.
Six.
We're
actually
gonna
make
this
easier,
because
you
can
do
this
from
the
UI
by
clicking
on
the
clear
cluster
cache
button.
I
have
a
link
to
the
Emma,
so
this
will
get
easier
from
next
month
and
I
believe.
G
G
Another
interesting
thing
I
came
across
while
trying
to
get
this
whole
thing
to
work,
which
turns
out
not
to
vote.
For
me,
because
I
didn't
have
sufficient
permissions
on
our
Gila
gtp
account,
but
yes,
that's
actually
a
way
to
get
help
to
connect
to
the
key
like
manage,
Killah
and
actually
use
days
to
travel.
Sure
the
issue
where
the
chat
was
complaining
that
there
was
no
available
releases
even
though-
and
there
was
so
check
out
that
issuing
right
there
so
yeah
any
questions
about
cross
lane.
I
think
this.
G
D
G
Believe
so,
yes,
so
as
long
as
you,
so
the
idea
of
crossplane
is
a
snotty
cloud.
You're
not
limited
to
a
certain
provider
in
terms
of
where
you
want
to
provision
your
resources.
So
in
theory
you
could
run
your
app
notes
off,
perhaps
like
kubernetes
in
Azure,
and
if
you
really
wanted
to
you
can
prove
you
can
define
and
provision
the
database
in
yeah
in,
like
you
know,
RDS
or
something
like
it,
WS
or
Google,
I
guess
so.
It's
really
dependent
on
so
there's
two
steps
to
this.
G
The
first
step
is
that
if
structure
admin
has
to
define
this
services
as
a
available
resources,
so
you
could
that
mean,
could
define
that
you
know
we
have
a
manage
database
available
in
AWS.
We
have
a
manage
database
available
in
GCP
and
then
it's
up
to
the
application
owner
to
actually
define
what
resources
you
want
to
claim.
Using
QC
terayama
configuration
paths
so
I
think
the
limitation
to
the
current
killer,
managed
apps
crossplane
install
is
that
we
only
support
one
stack
provider
and
not
per
install.
G
D
G
G
Resources
by
using
cube,
CTL,
apply,
dash,
F
and
then
you're
providing
a
config
file.
So
you
probably
change
that
file,
but
not
the
entire
CI
file
I'm
itself,
if
I'm
not
wrong.
This
is
currently
designed
heavily
around
use
with
Auto
DevOps.
So
if
you're
trying
to
do
something
custom,
a
bit
I'm,
not
sure
if
you
stand
up
as
well.
G
A
Thanks
very
much
guys,
Nathan
I'm
curious
to
hear
from
you.
What
did
you
think
of
this?
As
someone
that's
coming
in
from
an
outside
perspective
that
hasn't
seen
one
of
these
released
review
parties
in
a
pack
and
bearing
in
mind
this
is
only
the
third
one
we've
done
was
it
useful?
Was
it
averaged
any
constructive
feedback,
something
that
you
see?
Other
regions
could
be
taking
part
in
give
us
your
thoughts.