►
From YouTube: 2021-03-02 Delivery team weekly rollbacks demo
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
A
C
That's
fair
and
it's
a
little
we've
got
an
incident
ongoing
at
the
moment.
A
C
So
is
there
something
useful
to
use
this
time
for
or
I
have
a
discussion
item,
but
it's
probably
not
very
long.
So
if
there's
anything
more
engaging
any
problem
solving,
we
could
go
through
together
or
anything
like
that.
You
could
do
that.
B
Yeah
seems
everything
seems
clear.
Maybe
we
should
try.
So
there
is
something
new
which
is
the
these
little
bags
are
manual
now.
So
it's
worth
testing
this
change.
If
we
want
to.
C
Is
this
also,
what
how
does
this
fit
in
with
our
kind
of
moving
to
production
plan
like
like
is
this?
Do
we
expect
to
have
big?
Is
this
I'm
gonna
rephrase
this?
Have
we
made
all
the
changes
we'd
like
to
make
to
this
pipeline
in
an
ideal
world?
If
it
all
works
ahead
of
our
production
test
or
other
other
steps,
we
would,
we
would
want
to
test
out.
B
I
I
don't
see
other
blockers,
I
mean
it's,
it
always
works
so
far.
We
are
more
confident
on
it.
So
yeah.
C
D
Wow
yeah
you're
surprising
me,
oh,
what
should
I
do
in
this
case,
so
I
guess
I
should
share
my
screen.
D
Okay,
great,
so
in
this
case
I
guess
I
would
need
to
look
for
some
run
book
for
to
run
a
rollback.
D
D
To
be
honest,
I
wouldn't
even
know
where
to
look
for
our
run
books
at
the
moment,
because
I
know
we
look
at
the
infrastructure
runbooks,
so
I
don't
even
have
them
in
my
browser.
History.
C
B
B
D
Oh
sorry,
go
ahead,
yeah
I
mean
I
just
would
normally.
I
would
no
just
go
through
the
you
know,
team
page
and
then
look
where
we
have
our
run
box
and
there
it
seems
to
be.
C
Yeah,
that's
good,
we
should
add
them
to
somewhere
in
slack
as
well.
I
always
look
in
the
releases
subject
title
and
I
always
think
it's
there.
We've
got
release
documentation.
It's
enters
the
handbook.
It's
like
a
roundabout
trip.
D
Okay,
now
I
start
reading
it.
D
So
I
can
run
this
shut
ups
command
and
which
channel
do
we
best
execute
this
an
announcement?
So
we
usually.
B
D
Okay,
I'll
try
it
there,
but
we
want
to
try
it
and
stretching
our
production.
We
wanted
to
try
production
today.
C
Did
you
set
that
up
alessio,
or
is
this
just
that
laze
deployment?
Okay,
that's
good!.
D
So
that's
nice
find
the
package
to
roll
back
open
following
my
browser.
Okay,.
D
I
see
oh
yeah,
okay,
I
need
to
replace
the
one
that
we
want
to
roll
back
to.
So
that
would
be
the
previous
one
that
I
see
on
the
setups
command
and
I
also
see
a
compare
link
now
there.
Okay,
then,
I
probably
would
try
to
use
that.
D
D
But
I'm
trying
also
to
use
the
alternative
fae.
D
So
there's
something
and
looking
at
the
other
one,
I
mostly
see
a
list
of
commits
there.
Okay,
that's
nice,
but
I
did
this
probably
before
already,
because
I
needed
to
first
figure
out
that
I
need
to
roll
back
because
of
this,
and
that,
mr,
so
probably
a
first,
the
this
one
here
to
see
what
changed
and
then
figured
out.
Okay,
I
need
to
roll
back
and
then
I
would
look
into
this
one.
D
D
D
D
D
D
What
else
do
we
see?
Okay,
that's
after
rollback,
section,
deployer
guide.
I
will
not
read
that
now
because
I'm
under
time
pressure,
I
guess,
after
all,
back
has
been
completed.
D
Nice,
migrations,
yeah.
Wasn't
there
something
about
checking
if
we
have
post
migrations
when
you
need
to
roll
back
or.
B
B
What
I
realized
right
now
is
that
we
need
to
update
this
documentation,
explaining
that
gitly
is
not
rolled
back
by
default,
so
there
is
a
manual
job
at
the
end,
and
if
we
want
to
roll
back
that
as
well,
we
need
to
play
it
about
through
to
play
the
play
button,
push
the
play
button,
yeah.
D
D
I'm
just
now
reading
the
details,
because
I
was
a
little
bit
in
a
hurry
when
I
did
this
right
now.
B
Is
not
sure
it
will
be
blocked,
so
in
theory,
yes,
would
be
wise
to
not
only
to
check
this
but
also
to
block.
I
was
reply
tagging
and
things
like
that,
because
you
don't
want
to
have
things
interfering
with
your
role
back,
but
just.
A
D
Yeah
I
mean
this:
this
sentence
safe
to
roll
back.
That
could
lead
to
me
and
other
people
to
believe
that
we
are
fine
in
every
case,
so
we
don't
need
to
check
for
anything
else
right
and
that's
one
of
the
first
things
that
we
do
to
check
so
maybe
either
right.
This
package
is
safe
to
hold
back
instead
of
saying
safety
rule
back
in
general
right.
B
There's
something
worth
mentioning
here
is
that,
even
if
you
say
safety
row
back
and
you
start
rolling
back
and
there
is
an
ongoing
deployment,
no
harm
will
be
done
because
the
deployment
would
just
stop
before
starting
because
when
is
is
checking
the
status
of
the
environment.
We
just
fail
on
the
first
job,
because
the
omnibus
check
will
tell
you.
B
A
B
B
A
D
So
this
is
about
migrations,
but
I
think
we
said
that
we
don't
do
anything
if
you
need
to
do
migrations
by
hand
right.
So
it's
not
really
the
case.
It's
more
an
emergency
thing
right,
yeah,.
B
D
C
D
After
rollback
it
I
need
to
do
something
here
right.
I
mean
following
this
this
part,
but
this
is
more
like
in
case
of
error
or
something
and
then
yeah.
So
these
are
the
special
cases
and
extra.
C
Information,
what
do
we
need
to
do
to
put
together
the
things
like
that
omnibus
point
like
so
we've
kind
of
focused
so
far
on?
Can
I
roll
back?
What
about
the
step
before
that
of
deciding?
C
Is
this
incident
something
that
we
can
recover
from
by
rolling
back
somebody
said,
so
not
everything
is
going
to
be
a
bug
that
we
could
roll
back
like,
for
example,
it
was
introduced
two
days
ago,
rolling
back's
not
going
to
help
you
or
it's
in.
B
C
B
C
Introduced,
that's
what
I'm
saying,
but
my
point
is
like
that,
like
how
do
there's
like
a
step
before
deciding?
Is
this
safe
to
roll
back,
which
is
is
this?
Is
this
incident
something
that
can
be
recovered
from
rolling
back
right,
which
is
where
we
work
out
like?
When
did
you
introduce
the
change
and
is
it
a
satellite
project,
or
you
know
all
of
those
things.
B
B
So
in
that
one
there's
everything
around
that
which
is
about
trying
to
figure
out
if
this
is
something
that
is
a
code
change,
so
they
could
be
rolled
back
or
not,
and
so
this
kind
of
gives
you
something
around
right.
So
you
don't
start
because
you
want
to
roll
back.
On
the
other
hand,
thinking
about
automating
this
so
our
end
goal
as
a
delivery
team.
B
I
would
say
that,
because
you
are
in
control
of
the
role
of
rolling
forward
and
rolling
back
and
you
are
checking
metrics
while
you
roll
out,
you
have
see
kind
of
a
certain
level
of
confidence
that
you
were
rolling
out.
Metrics
started
dropping,
so
the
problem
should
be
in
the
things
that
you're
rolling
out,
but
this
is
because
of
if
you
want
to
automate
this,
so
if
you
want
to
blindly
roll
back,
if
something
goes
wrong,
then
you
have
to
say
something
like
this.
B
D
I'm
wondering
if
and
not
also
would
try
to
look
back
in
the
chat
ups
history
here
to
see
what
was
the
last
deployment
right
directly
from
there
figure
out
what
I
want
to
roll
back
right,
because
looking
at
this,
for
instance,
if
you
run
autos
deploy
status,
I
would
expect
that
I
get
the
same
right
the
package
number,
but
this
is
not
the
one
that
we
wrote
back
to
now.
This
is
the
one
that
we
had
previously
right
now
we
can.
This.
A
B
B
So,
and
in
the
announcement
channel
you
should
be
able
to
scroll
back
to.
E
C
An
eye
on
the
deployment
on
the
road
back
when
we,
let's
do
that
so
first
item
or
the
only
open,
I
opened
up
an
issue,
so
we
can
start
putting
together
the
plan
to
test
on
production
henry.
I
think
you
mentioned
last
week
that
there's
a
is
it
did.
You
say
there
was
a
production
change
calendar.
We
should
add
this
to.
D
Yes,
there's
production:
how
is
that
called?
Let
me
check
that's
the
changes
in
production
calendar
this
one
so,
and
you
could
just
put
in
changes
in
this
one,
because
it's
anyway,
I'm
blocking
time
frames
where
it's
fine
to
do
s1
changes.
So
that's
why
we
always
have
it
marked
in
this
time
frame
with
no
production
traffic.
But
I
don't
know
if
you
use
this
for
anything
nowadays
and
if
you
put
in
important
changes,
then
it's
also
bringing
kind
of
awareness
to
to
the
infrastructure
department.
D
I
think
so
it's
always
good
to
do
it.
I'm
not
sure
who
really
is
actively
looking
into
that.
I
think
the
the
most
important
thing
is
really
that
the
eoc
knows
about
it
when
it's
when
it's
getting
started,
and
normally
we
would
do
this
by
having
a
change
request
issue
or
at
least
pinging
vsre
on
call.
C
D
B
And
has
canary
being
considered
in
this
rollback
exercise.
We
don't
have
real
canary
rollback
pipeline,
mostly
because
they're
not
suggesting
I'm
not
suggesting
canary
rollback.
What
I'm
trying
to
say
is
that
we
are
rollback.
We
are
doing
rollback
in
production.
B
How
does
that
actually
affect
canary,
given
that
we
have
a
separate
stage
to
deploy
to
canary?
So
are
we
rolling
back
canary
before
we
roll
back
production
or
do
we
drain
or
we
drain?
I
mean
if
we
follow
the
so
what
we
would
do
in
case
of
a
real
incident.
We
drain
cannery,
we
roll
back
production,
okay
and
then
we
so
we.
C
And
the
picking
of
like
the
package
to
do
the
to
do
the
roll
back
on.
How
are
we
going
to
schedule
this
because,
obviously
we
have
to
avoid
incidents
and
we
need
to
manage
it
around
our
deployments
like?
Should
we
take
a
best
guess
at
a
good
time
of
day
for
us
based
on
deployments
and
things,
and
then
we
have
to
just
postpone
if
there's
an
incident
or
is
there
a
better.
B
Scheduling
this
is
really
hard
because
there's
the
pipeline
takes
so
much
time
just
building
packaging
because
out
in
the
beginning,
I
was
thinking
something
like
we
pick
a
bug
fix
or
something,
and
we
pick
it
so
that
we
have
something
that
we
deployed
plus
something
very
simple.
It
will
trigger
a
new
package
we
deploy,
but
this
adds
four
to
six
hour,
so
it
it's
really
well.
I
don't
know.
B
C
B
So
as
as
myra
was
asking
before,
so
you
prefer
to
create
a
synthetic
change
so
that
we
can.
We
are
sure
that
we
can
safely
roll
back
the
first
time.
We
run
it
and
we
are
more
in
control
of
the
content
so
that
when
we
want
to
run
it,
unless
there
is
an
incident,
we
can
actually
roll
back.
Is
that
your
idea.
B
So
the
idea
was
that
if
you
create
a
merge
request
on
security
targeting
the
how
to
deploy
branch,
this
thing
will
end
up
being
packaged
and
be
part
of
something
we
can
roll
deploy
and
roll
back
without
kind
of
affecting
the
master
history.
Because
my
my
first
attempt
the
one
that
failed
badly
was
that
I
would
just
add
the
file
to
the
to
the
route.
So
just
something
slash,
this
will
be
rolled
back
so
that
it
would
just
generate
a
new
package
with
no
differences,
but
we
have
our
pipelines
are
too
clever.
B
E
Yeah
something
similar
happened
when
I
was
trying
or
testing
the
security
releases
commands
and,
yes,
our
pipelines
are
too
clever
and
if
you
only
add
like
a
documentation,
change
a
pipeline
is
not
generated
so
to
circumvent
that
what
we
end
up
doing
it
was
to
command
something
on
the
project
model
or
in
the
pipeline
model
for
the
pipeline.
To
think
that,
oh
there
is
a
change.
I'm
going
to
execute
the
whole
tests.
B
C
Just
just
changing
up
the
package
and
stuff
though,
like
I
guess
does
that
add
it
adds
obviously
a
load,
more
tasks
and
like
coordination
and
stuff
like
I
guess,
does
it
make
it?
B
C
C
What's
the
state
of
the
environment,
if
we
don't
do
the
manual
italy
step.
B
B
B
B
I
have
a
topic
I
would
like
to
discuss
just
looking
at
the
agenda.
Do
we
no
there's
nothing
else
so
so
far
we
focused
on
making
sure
the
road
back
pipeline
exists,
having
a
process
around
it
and
that's
it
and
yeah.
I
mean
next
step
in
this
direction
will
be
testing
in
production
the
pipeline,
because
I
would
like
to
avoid
us
entering
in
this
never-ending
loop
where
every
week
we
just
try
to
polish
the
run
book
and
never
tackle
the
real
goal,
which
is
writing
automation
around
this
right.
So.
B
I
mean
I
just
want
everyone
to
be
aware
about
this,
so
our
goal
is
to
try
to
connect
the
things
that
we
are
doing
manually
now
with
the
with
the
deployment
pipeline
and
we
are
not.
The
target
is
not
starting
a
rollback
auto
automatically
right
now,
but
is
figuring
out
if
something
is
good
is
going
wrong
before
we
reach
the
point
of
no
return
and
providing
contacts
on
release,
managers
and
the
on
call
syrian
call
about
how
they
can
drop
back.
B
C
That's
a
great
point:
yeah.
What
do
you
think
will
be
the
first
step
in
moving
towards
the
having
more
of
an
automated
rollback.
B
We
need
to
so
the
the
pipeline
trick,
the
deployment
triggering
that
we
discussed
this
week.
So
the
thing
that,
where
the
deploy
the
deployer
is
triggered
environment
by
environment,
it
is
one
of
the
the
first
step
because
it
gives
us
the
ability
to
break
the
deployment
into
smaller
chunks,
because
we
need
something
that
allows
us
to.
B
Let's
say
I
would
just
it
very
simply:
we
need
to
trigger
a
deployment
up
to
completing
the
fleet
so
basically
doing
the
world
map
doing
migrations,
doing
the
the
gitly
and
then
the
fleet,
and
then
stopping
at
that
point,
because
we
want
control
back
in
release
tools
where
we
can
inspect
the
content
of
the
package
compared
with
the
content
of
the
database
and
see
if
there
are
post-deployment
migration,
not
based
on
the
the
code
diff
but
based
on
what
database
thinks
is
happening
right
now,
and
there
is
the
junction
point:
when
we
need
to
be
clever
and
say:
no
things
are
going
south
we're
not
going
past
this
point.
B
B
C
B
B
Maybe
it's
better
to
go
back
what
we
should
aim
when
we
have
more
control
over
it
is
to
stop
earlier
so
because
we
do
batch
deployment,
we
say
10
of
the
fleet
or
one
zonal
cluster,
really
because
too
many
things
are
moving
at
the
same
time.
Right
now,
but
as
soon
as
we
can
validate
our
metrics
on
a
smaller
set
of
subsets
of
the
deployment,
we
should
consider
making
the
same
decision
so
right
now.
B
We
know
there
is
a
no
return
point
which
are
post-deployment
migration,
so
I
think
we
should
aim
for
that
and
once
we
know
that
the
our
automation
is
working,
we
are
more
confident
with
everything
we
can
start
being
more
clever
and
shifting
control
over
the
left
side,
so
detecting
earlier.
C
C
I
wonder
I'm
trying
to
think
off
top
my
head,
but
I
don't
know
the
answer.
What
notifications
the
releases
team
are
putting
on
their
auto
rollbacks
to
alert
the
fact
that
it
actually
happened,
because
that's
the
the
step
afterwards
right
is
that
you
have
some
visibility,
that
a
rollback
happened
and
like
here
was
the
event.
So
I'll
have
a
look
at
that
stuff.
A
B
Because
the
merge
request
tracking
works
this
way
we
have
previous
com
previous
deployment,
the
next
one.
It
gives
you
a
list
of
commit
in
that
and
then
we
scan
the
list
of
commits
for
merge
requests
when
you
reverse
the
the
search
button,
the
list
of
commits
is
empty,
so
basically,
you
are
marking
all
the
margins
included
in
that
empty
list,
and
so
basically
in
the
environment
page
you
see
another
deployment
but
on
the
merger,
because
there
are
no.
B
C
Yeah
I
haven't
seen
anything
like
that.
I
mean
it's
only
first
iteration
I
think
so.
I
think
it's
tricky
because
it's
technically
it's
a
deployment.
It's
all.
Aside
from
the
package
you
chose,
the
gitlab
rollback
is
moving
forwards.
C
What
would
we
do,
ideally
like
you'd
have
another
note
or
something
that
said
this
got
rolled
back.
B
B
Because
there's
also
an
extra
thing
when
you
deploy
again
it's
a
bit
trickier
than
just
this
right,
because
when
you
deploy
again,
you
well
interior
should
work,
because
the
previous
deployment
is
the
very
old
sha.
So
the
difference
between
the
deployed
sha
and
the
new
world
forward
package
should
include
everything
was
drilled
back,
plus
eventually
new
developments.
B
So
in
theory,
you
should
see
something
like
staging
cannery
production
production,
rollback
and
then
again
staging
cannery
production
kind
of
no,
maybe
just
production,
because
you
wrote
back
on
the
production.
I
don't
know,
I'm
just
thinking
about
the
the
commit
range
commit
range
will
report,
those
so
in
theory.
A
B
C
C
C
Do
you
have
any
idea
of
like
how
often
would
we
we
run
this
gitly
roll
back,
like
I'm
wondering
if
we
should
have
those
qa
test
stages
repeated
after
that,
because
the
moment
we
roll
back
the
fleet
test?
If
that's
so
successful,
we
do
all
the
tracking
and
finish.
But
then,
if
we
have
a,
we
kind
of
have
a
separate
rollback
phase
where
you
can
manually
roll
back
italy
and
prefect
like
should
we
have
something?
That's
checking
the
state
of
things
after
that.
B
B
Quick
okay,
so
this
is
the
pipeline
we
are
looking
at
so
the
representation
here
is
that
this
is
at
the
end.
But
this
is
not
true.
So
if
we
go
here,
there
is
a
dag
implementation.
So
basically
I
don't
know
how
to
look
at
this
but
yeah.
So
this
is
gstg
italy
rolled
back
and
gstg
prefect
rollback
just
only
requires.
Oh,
this
is
how
it
works.
Gstg
prepare
and
gsg
warm
up.
B
So
as
long
as
as
soon
as
those
two
are
are
green,
they
will
run
so.
This
means
that
yeah,
basically,
when
you
start
rolling
back
the
fleet,
you
say
no.
I
also
need
this,
so
you
just
play
button
and
by
the
time
you
run
qa
the
qa
will
be
completed
in
staging
we.
We
also
have
a
qa
that
runs
on
every
hour.
I
don't
remember
so
eventually,
if
you
roll
something
back,
the
next
qa
round
will
beat.
B
B
B
You're
muted,
I
was
I
muted
myself
before
speaking
yeah,
so
that
error
means
that
no
ansible
was
not
able
to
reach
the
that
machine
by
the
time
we
wanted
to
clean
it
up
right.
B
E
A
B
F
B
B
So,
if
I'm
just
thinking
aloud
right
here,
so
I'm
guessing
that
all
the
jobs
that
we're
running
well,
but
the
world
war
map
should
have
should
have
downloaded
package
also
there.
Let
me
see,
because
what
I'm
thinking
is
that
at
the
beginning,
we
do
the
warm
up
and
we
do
so
we
download.
So
we
reach
every
machine
which
every
machine
we
don't
load
a
package.
Then
the
fleet
deployment
act
only
on
a
slice
of
the
of
the
fleet.
F
Me
see
alessio,
I
just
queried
the
chef
server
and
there's
only
five
servers
registered.
So
I
wonder
if
you
retried
it
again
if
it
would
show
up
in
the
inventory
so.
B
B
I
hit
retry
and
but
in
the
meantime
I
want
to
see
if
we
downloaded.
Let
me
share
my
screen,
so
let's
do
this
together.
We
are
all
here,
so
I
retried
and
we
will
see
what
happened
here,
but
what
I'm
thinking
here
is
when
I,
if
I
go
to
the
prepare
and
then
I
go
to
the
worm
up
at
the
end,
there
should
be
a
list
of
hosts
yeah.
Here
it
is
so
it
was
there.
D
D
F
F
Ansible
has
that
configured?
I'm
guess.
B
B
So,
let's
take
a
look
together
at
the
pipeline,
just
to
make
sure
that
everything
is
fine.
I
think
everything
is
fine,
because
we
were
at
the
finish
stage.
This
is
the
thing
to
delete
packages
from
a
machine
and
that
machine
no
longer
exists
and
looking
at
the
output,
it
just
told
us
that
it
reached
out
all
the
machines
and
did
whatever
he
had
to
do,
except
for
this
one
that
we
know
no
longer
exists.
F
B
I
can't
because
I
can
only
keep
it
should
be
somewhere
else.
I
don't
remember.
B
B
So
the
we
are
using
so
artifacts,
so
the
thing
is,
if
you
are
using
the
cache,
because
there's
also
cache
feature
if
you're
using
the
cache
features,
your
job
should
be
able
to
handle
missing
cache.
So
if
the
cache
is
not,
there
should
be
able
to
generate
the
things.
B
If
you
want
to
be
sure
the
the
file
is
always
persisted
and
the
next
job
will
always
have
it
no
matter
what,
then
you
need
to
use
an
artifact
and
because
our
artifacts
are
not
real
artifacts,
but
more
just
things
that
get
passed
by
from
one
job
to
the
next
one.
We
implemented
the
expired
feature
so
that
you
can
treat
them
as
a
kind
of
a
cache
thing.
D
B
C
D
Yeah,
it
must
have
been
manual
because
there's
no
trace
interform
for
it.
Yeah
I
mean
we
can
ask
the
infrastructure
team.
If
anybody
did
it
or
just
have
a
look
in
the
gcloud
locks,
I
guess
we'll
try
to
do.
D
C
So
just
to
wrap
up
on
this
demo
so.
C
Should
we
still
we'll
go
ahead
and
plan
in
a
production
roll
back
for
next
week?
Sometime?
Does
that
still
seem
like
a
good
time
frame
to
aim
for.
C
Cool
okay
sounds
good,
I'll,
have
a
little
think
about
whether
any
what
the
timings
look
like
on
whether
we
would
maybe
be
able
to
put
in
a
special
package
or
or
whether
we
should
just
go
rogue
and
see
how
that
looks
because-
and
it's
probably
pros
and
cons
either
way.
Maybe
it
doesn't
actually
matter
that
much
it'd
be
good
just
to
be
able
to
test
one
out
so
I'll
play
around
some
timings
there
and
see
what
that
looks
like.
C
Cool,
if
I
chat
to
the
other
ems
about
getting
a
potential
rollback
date,
would
someone
be
willing
to
take
on
the
like
prep
for
for
the
rollback
like
putting
in
the
like
get
us
a
change,
request,
open
and
basically
do
the
kind
of
ad
mini
tasks
for
that.
B
C
B
I'm
scanning
quickly
through
the
actions
in
the
gender
node,
so
maybe
we
can.
We
could
have
someone
that
can
volunteer
for
res
polishing
a
bit
the
round
book,
but
just
then
we
you
say
it's
done
so
we
just
don't.
We
don't
keep
it
iterating
over
and
over
on
it,
but
it
say
just
give
it
define
a
final
shape.
That
includes
then
mention
that
this
regular
back
is
manual
and
trying
to
figure
out
what
emerged
today
and
then
I
mean
at
least
we
consider
close
the
the
issue
about
creating
the
thing
then
yeah.
C
I'm
happy
to
have
another
pass
on
it
and
based
on
what
we
saw
today
and
then
we
can
say
the
issue
is
closed
and
then,
as
always
right,
mrs
to
make
improvements
from
there.
So
yeah
I'm
happy
to
take
that.
C
Oh,
should
we
go
async
for
this
final
bit?
I
feel
like
I'm
I'm
watching
nells
attentively
to
see,
but
we
probably
don't
need
to
sit
on
a
theme
for
that
one.
Unless
he
comes
back
and
says
no
and
then
we
might
want
to
just
review
our
pipeline
so,
but
we
can
work
out
that,
but.
D
Yeah,
I'm
having
a
look
as
incorrectly
to
see
if
I
can
figure
out
what
happened
because
he's
trying
the
same.
But
I
will
have
a
look
for
that.
C
Perfect,
that's
great
thanks
for
demoing
henry
nice
work.