►
From YouTube: 13.1 Release Kickoff
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
All
right,
it's
nine
o'clock,
at
least
in
the
Pacific
coast
in
California,
hello.
Everyone
welcome
to
the
gitlab
monthly
release,
kickoff
call
I'm
the
vice
president
of
product
management
at
KITT
lab.
This
is
my
first
time
hosting
this
call
and
I
look
forward
to
feedback
recall
that
13
dot.
Our
major
release
is
planned
shipped
this
month
on
the
22nd.
This
call
is
not
about
$13.
It
is
instead
about
13.1
a
release
planned
for
June
of
this
month
this
year.
This
is
a
first
minor
release
on
13
dot.
A
A
First,
we
have
a
kickoff
page
on
the
KITT
lab
website
about
get
lab,
comm
slash
direction,
slash
kick
off
if
you
would
like
to
follow
along
or
use
it
as
a
reference
later
you'll
find
links
to
issues
of
all
the
planned
functionality.
There
you're
all
invited
to
jump
into
these
and
provide
inputs
and
feedback.
We
welcome
all
your
contribution.
A
Secondly,
keep
in
mind
that
we
do
plan
ambitiously
and
some
of
the
things
we
talked
about
here
may
not
make
it
into
13.1,
although
we
do
intend
and
try
hard
to
ship
them.
Please
recognize
that
our
plans
are
there
for
dynamic
and
subject
to
change
that
I'd
like
to
hand
it
over
to
you
Josh.
Take
it
away.
B
Thanks
to
do
we're
working
on
three
major
buckets
of
things
for
this
release
in
the
Naval
mint.
The
first
is
that
we
are
switching
away
from
some
delivering
some
really
big
projects
like
the
minimum
reversion
of
posters
required,
namely
PG,
11
and
13
net
0,
as
well
as
also
delivering
Puma
as
the
deal
application
server.
B
During
the
course
of
the
PG
11
upgrade,
we
identified
some
improvements,
which
we
didn't
have
time
for
in
the
PG
eleven
upgrade
that
we
would
like
to
improve
on
for
PG
12,
the
next
one.
So
you
can
see
a
number
of
them
here
which
are
focused
on
which
will
help
make
this
process
easier
for
our
users
going
forward.
B
This
is
important
because
we
want
to
make
sure
PG
12
is
the
next
minimum
version
required
next
year,
the
next
May,
and
so
we
want
to
get
the
startup
process
pretty
quickly
here
with
30
met,
3
+,
PG
12,
so
people
have
time
to
upgrade
and
roll
over
and
test
it,
leading
up
to
that
kind
of
may
new
minimum
requirements
to
ensure
we
stay
on
the
cadence
of
PG
a--
as
well
itself.
So
this
is
really
important.
How
much
that
process
go
smoothly?
B
This
way
will
support
PG
12,
but
will
also
align
with
how
I
comm
currently
runs
it's
post-grad
services
as
well,
and
so
we
always
want
to
make
sure
calm
as
reflective
as
possible
of
our
customers,
and
so
we'll
go
ahead
and
get
this
done.
It's
largely
finished
already.
We
have
a
number
a
couple
things
we're
working
wrapping
up
here,
but
again
it
works
and
we're
just
driving
to
delivery
of
this
here
and
13
that
or,
if
not
surely,
thereafter
here
at
13.2
to
prepare
the
way
for
a
13.3
with
p12.
B
So
that's
some
of
the
improvements
are
taking
to
continue
on
the
Postgres
iteration
strategy,
we're
also
working
on
ways
to
improve
how
we
store
and
protect
on
us
configuration
secrets
right
now
they
are
stored
in
some
configuration
files,
but
ya
know
get.
There
would
be
a
couple
others,
and
we
have
heard
from
folks
that
they
want
those
out
of
those
configuration
of
files
and
into
a
different
storage
solution
that
is
more
protected.
So
we're
working
on
that
here
and
we're
going
through
a
couple
different,
perfect
concepts
to
test
these
out.
B
B
So
that's
some
of
the
new
projects
we
have
going
on
here
in
this
release,
and
now
we
can
cushion
to
the
performance
improvement
section
first
off
on
the
database
side.
Partitioning
is
when
I'm
working
on
for
some
time
here
and
we're
going
ahead
and
trying
to
iterate
on
a
proof
of
concept
for
the
audit
events
table.
This
is
a
good,
relatively
simple
table
right
now
that
we
can
kind
of
use
as
a
test
case
and
establish
some
best
practices
and
test
out
some
different
options.
B
We
also
want
to
improve
some
of
our
worst
performing
endpoints
in
particular,
two
of
them
one
is
the
projects
controller
which
right
now
it
takes
between
one
and
a
half
to
three
seconds
to
actually
return.
This
is
quite
a
long
time
and
we
want
to
bring
this
down
to
under
five
hundred
milliseconds.
This
is
a
pretty
popular
workflow,
and
so
this
is
have
a
pretty
high
impact
to
our
users
by
being
able
to
make
this
two
or
three
times
more
performance.
So
we're
excited
about
this
and
working
on
this
with
a
memory
team.
B
Similarly,
the
blame
controller
and
the
blame
API
are
also
quite
slow
in
this
case
significantly.
So
under
certain
condition,
as
you
can
see
her,
it
takes
over
17
seconds
to
return,
so
we
want
to
go
ahead
and
we
want
to
make
this
significantly
faster
to
help
make
sure
these
pages
respond
in
a
time
from
our
users
might
expect,
because
currently
they
might
think
it's
potentially
broken
so
looking
forward
to
that
here
and
the
Blaine
API
as
well.
B
Next
up,
we
also
want
to
have
a
better
understanding
of
how
our
users
and
our
self
managed
installations
actually
run.
To
date,
we've
been
using
get
low,
calm
as
a
proxy
for
a
lot
of
the
sort
of
pain,
points
and
performance
challenges
of
get
lab
itself,
and
but
that
might
not
necessarily
be
exactly
what
our
end
users
are
seeing.
They
have
different
configurations,
different
load
patterns,
and
so
we
can
get
some
better
telemetry
and
so
that
our
knowledge
of
kind
of
what
our
customers
are
seeing
out
there
in
the
field.
B
We
can
include
that
inner
part
ization
here
and
make
sure
we
address
those
concerns
which
you
might
not
be
seeing
as
potentially
impactful
and
github.com,
and
so
we're
going
to
try
and
get
some
of
those
things
in
the
users.
Ping
come
the
operational
data
to
help
make
better
decisions
going
forward,
and
so
that's
some
of
the
performance
assets
working
on
here
in
this
release,
and
we
can
move
over
to
these
simplification
of
things.
B
The
first
is
that
we've
been
working
on
the
self
service
framework
for
geo
as
we
go
through
and
add
additional
data
types
support
for
geo
itself.
The
idea
here
is
that
we
can
reduce
the
time
it
takes
to
add
a
new
data
type
to
the
replication
framework,
with
the
goal
being
that,
ultimately,
when
a
team
of
gitlab
implements
a
new
data
type
itself,
they
themselves
have
a
really
quick
and
easy
pass
to
simply
add
the
replication
to
it
from
the
beginning.
B
Right
now,
we
rely
upon
foreign
data
wrappers
a
lot,
and
this
has
a
lot
of
complexity
to
geo,
and
we
want
to
try
and
get
rid
of
this
and
simplify
its
development
again,
to
make
it
easier
for
us
to
build
and
also
operate
and
maintain
your
going
forward
and
also
on
the
simply
on
the
experience
side
of
things.
We
want
to
go
ahead
and
make
sure
that
promoting
a
secondary
should
be
as
as
simple
as
possible
and
there's
a
number
of
checks,
and
we
kind
of
asked
you
to
go
through
during
the
process.
B
We
want
to
go
through
and
actually
automate
as
many
of
these
as
you
possibly
can,
and
so
the
process
of
promoting
a
secondary
to
our
primary,
which
could
potentially
be
in
a
relatively
stressful
situation,
is
as
easily
and
with
as
much
little
as
risk
as
possible,
and
so
we're
go
through
and
make
that
process
simple
and
next
up
and
last
up.
Here
is
the
search
team
and
we
want
to
go
through
and
continue
to
refine
how
we
build
our
indexes.
B
We
had
some
significant
reductions
in
the
index
size
and
a
previous
release
of
around
40%,
and
with
this
change
we
think
we
can
get
another
30%
reduction
in
index
size
as
well
by
changing
how
we
currently
index
the
words
by
going
to
an
edge
and
Graham
filter.
This
does
have
some
impact
on
how
search
terms
will
match
right
now
we
match
partial
words,
so
we
type
in
the
first
couple.
Letters
of
a
word
will
match
the
rest
of
them.
B
This
is,
might
be
unexpected
actually,
but
you
can
see
here
our
competitors
currently
don't
support,
partial
matching
and
so
we're
learning
to
them,
but
lost,
but
I
recognize
some
significant
benefits
to
reduce
cost
and
reduce
index
size
on
the
elastic
elastic
search
side
of
things,
and
so
our
crimp
proposal
is
to
users
can
have
an
Asterix
here
if
they
still
want
to
retain
that
same
behavior,
but
for
now
we'll
also
potentially
reduce
the
number
of
potentially
confusing
results,
as
you
might
not
get
the
exact
term
looking
for
backing
the
search
results.
So.
D
But
today
I'm
going
to
do
something
a
little
bit
different,
which
is
kind
of
a
line
all
of
our
stages
together
into
three
themes,
and
the
three
themes
are
Enterprise
readiness,
iteration
on
improving
the
user
experience
and
some
tear
moves,
which
is
always
exciting
and
fun
to
announce.
Let's
go
ahead
and
dive
into
enterprise
readiness,
which
is
really
our
theme
around
making
the
enterprise
experience
better
for
galeb
comm,
as
well
as
giving
the
right
controls
an
audit
audit
ability
for
all
users,
whether
they're
on
calm
or
self-managed.
D
So
the
first
thing
I
want
to
highlight
it
is
a
feature
to
prevent
forking
outside
of
a
private
group.
So
this
will
be
a
feature
that
will
allow
group
owners
to
toggle
whether
or
not
Forks
can
be
created
outside
of
the
group.
This
will
help
with
protecting
you
know
very
sensitive,
IP
or
just
private
projects
that
you
don't
want
forked
outside
of
any
any
group
other
than
the
one
that
it
lives
in
the
second
one.
I've
talked
about
this
one
before,
but
it's
really
important
to
highlight
again,
which
is
optionally,
disabling
group
membership
inheritance.
D
This
is
when
I
talked
about
last
release,
but
it's
gonna
land
in
13.1,
but
this
is
important
because,
if
you're,
a
member
of
a
top-level
group,
you
automatically
are
a
member
of
every
subgroup
in
that
group,
but
forget
Montcalm.
We
know
most
organizations
manage
their
entire
company
at
a
single
top-level
group,
so
by
allowing
group
owners
to
optionally
disable
group
membership
inheritance.
This
allows
them
to
structure
their
organizations
as
they
need
to,
or
at
least
have
the
option
of
the
structure
them
as
they
need
to
on
github.com.
D
So
a
nice
improvement
here
that
will
really
help
enterprise
writing.
This
son
I'm
Caleb
comm.
The
second
kind
of
big
thing
within
enterprise
readiness
is
obviously
compliance
and
auditability
we're
going
to
iterate
on
a
compliance
feature
that
we
released
two
releases
ago,
which
is
these
compliance
framework
tags.
We
are
allowing
our
users
to
put
on
projects
and
we're
going
to
use
these
tags
by
allowing
admins
to
scope,
merge,
request
approval
settings
to
projects
that
have
these
compliance
tags
right
now.
D
You
can
set
this
at
the
instance
level,
but
we
want
to
provide
a
little
bit
more
configure
ability
so
that
only
these
rules
get
set
for
those
projects
where
they
really
need
to
be
set.
So
a
nice
way
to
use
these
tags
that
we
released
and
we'll
continue
iterating
here
as
well
as
always
we're
continuing
to
bolster
our
audit
log
and
in
13.1
we're
going
to
include
changes
to
approval
groups.
D
So
if
someone
is
added
or
removed
from
an
mr
peru
group,
this
will
now
be
surfaced
in
our
audit
log
and
to
help
our
user
understand
who
has
permissions
to
specific
groups
and
projects,
we're
also
adding
an
API
endpoint
to
help
surface
all
of
the
permissions
of
users
for
all
groups
and
projects.
This
is
something
that
we've
heard
pretty
consistently
in
our
customer
advisory
board
meetings
and
from
users
that
they
just
need
a
better
way
to
audit
who
has
access
to
what
side
of
gitlab.
D
So
we're
excited
to
add
this
I'm
gonna
talk
about
an
analytics
feature
very
quickly
and
the
reason
this
is
in
the
enterprise
readiness
theme
is
because
we
have
a
great
feature
called
value
stream
analytics,
but
it
lives
at
the
group
level
because
it
lives
at
the
group
level.
If
you
have
a
large
group
such
as
we
do
here
at
get
lob,
all
your
data
is
aggregated
and
it
can
be
difficult
to
glean
insights
from
this
feature.
D
So
we're
going
to
add
filtering
to
this
feature,
so
you
can
filter
via
label
or
other
keyword,
and
this
should
really
help
individual
teams
that
are
all
inside
of
a
single
group
really
get
better
insights
from
the
value
stream
analytics
feature.
It's
a
really
excited
for
filtering
and
I
think
this
will
really
enhance
the
functionality
and
the
usability
of
this
feature
as
well.
D
In
Bob
1210,
we
released
our
MVC
of
requirements,
management
and
our
next
step
here
is
to
be
able
to
trace
those
requirements
to
test
and
so
excited
to
see,
mark
and
the
team
iterate
here
and
see
where
we
can
go
with
requirements,
management
and
then
our
last
Enterprise
readiness
feature.
Of
course
is
giddily.
Cluster
Italy
cluster
is
what
we've
recently
renamed
get
elite
high
availability
and
we'll
be
working
towards
strong
consistency
in
13.0,
which
is
in
four
days
we're
going
to
release
the
first
general
availability
release
of
get
elite
cluster.
D
But
it
comes
with
an
eventual
consistency
model
which
means
that
there's
potential
data
loss
in
the
case
where
a
giddily
node
goes
down
before
data
is
replicated
to
it.
We
know
that
that's
a
great
first
step,
but
it's
not
where
we
need
to
be
long
term,
and
so
we're
going
to
be
iterating
towards
get
early
cluster
strong
consistency
via
a
three
phase
commit,
and
our
first
step
here
is
introducing
some
new
get
hooked
to
make
this
a
possibility.
D
Our
next
theme
iterations,
improving
the
user
experience.
Gonna
highlight
a
few
things
that
seem
super
minor,
but
I
think
are
really
going
to
help
the
experience
here.
The
first
one
is
making
it
easier
to
remove
labels
from
issues.
If
you
use
labels
a
lot
in
issues
or
mr-s,
you
know
that
it
takes
several
clicks
and
likely
some
typing
to
remove
a
label.
D
D
Basically,
their
experience
looks
like
what
my
what
my
screen
is
sharing
right
now,
you
don't
have
any
context
into
the
issue,
but
by
simply
sticking
the
issue,
title
you'll
be
able
to
see
this
at
any
point
in
the
issue
which
will
really
help
to
provide
that
context.
When
you
get
deep
linked
into
a
longer
issue,
we're
going
to
continue
to
iterate
on
boards
and
start
working
on
a
long
requested
feature
which
is
providing
epic
swim
lanes
into
boards.
So
really
excited
to
see
this
one
finally
activated
upon
and
to
see
this
one
lands.
D
If
you'd
like
there's
a
ton
of
designs
down
in
the
bottom
of
this
issue
that
highlight
all
the
various
states,
so
would
love
your
feedback
if
you
want
to
take
a
look
at
those
we're
also
going
to
allow
comment
pins
on
designs
to
be
resolvable.
So
today
you
can
comment
on
a
design
and
I'll
show
a
little
pin
on
that
design.
But
the
problem
is:
if
you
address
the
feedback,
the
pin
doesn't
necessarily
go
away.
D
So
we
want
this
to
be
resolvable
so
that
this
pin
goes
away,
especially
in
the
case
where
a
new
designs
been
uploaded
and
that
comment
has
been
taken.
Care
of
in
13.0
we're
releasing
version
snippets
and
our
next
iteration
here
is
specifics
with
multiple
files.
Here's
a
sneak
peek
of
what
this
might
look
like.
We're
gonna
allow
up
to
ten
files
in
a
single
snippet
which
will
be
really
great
functionality
and
a
great
enhancement
for
snippets
and
then,
lastly,
for
iteration,
we
are
going
to
allow
inserting
an
image
using
static
site
editor
and
thirteen
auto.
D
D
Currently,
this
is
an
ultimate
feature
and
we
think
this
really
aligns
better
with
a
director
level
and
so
we're
going
to
move
this
down
to
premium
for
our
buyer
tier
framework
and
really
excited
to
offer
this
value
to
our
premium
customers,
and
our
second
tier
announcement
is
that
we
are
going
to
move,
merge,
request,
reviews
down
to
core.
We
think
that
reviewing
merge
requests
is
something
that
an
individual
contributor
would
do
on
their
own
work,
and
this
this
functionality
is
really
about
batch
comments.
D
We're
excited
to
bring
this
down
to
core,
as
it
aligns
really
to
the
individual
contributor
on
our
buyer
framework.
That's
a
quick
highlight
of
all
the
things
coming
in
devil,
of
course,
there's
a
ton
more
things,
please
view
the
kickoff
page
and
the
videos
and
would
love
your
feedback
in
your
comments.
I
appreciate
you
all
and
back
to
you,
a
noob.
C
Thanks
a
new,
my
name
is
Kenny
Johnson
I'm,
the
senior
director
covering
the
ops
section.
The
ops
section
comprises
verified
package,
release,
configure
and
monitor
stages.
It's
nine
total
teams.
I'm
gonna
quickly
highlight
a
number
of
items
through
this
journey
and
really
focus
on
one
specific
stage.
C
That
I
think
is
doing
really
amazing
work
that
I
wanted
to
call
out
here
the
the
way
that
I'd
like
to
walk
through
this
is
to
think
about
the
user
adoption
journey,
and
we
know
from
the
experience
of
our
users
that
their
adoption
of
our
single
application
goes
through
different
phases
and
the
phases
that
I'd
really
like
to
focus
on
are
the
adoption
of
verify
our
CI
tool
as
existing
source
control
users
and
then.
Lastly,
the
adoption
of
monitoring,
I'm
gonna,
really
highlight
those
specific
to
in
CI.
C
We
are
working
on
doing
public
testing
for
Jenkins
wrapper.
This
is
an
important
tool
for
organizations
moving
from
Jenkins
to
get
lab
to
and
enable
them
to
kind
of
quickly
subvert
some
of
the
reconfiguration
they
might
have
to
do
up
there
Jenkins
pipeline
definitions
and
have
that
Jenkins
pipeline
run
natively
and
gabab
we're
looking
for
public
beta
testers
here.
This
issue
is
linked
to
in
the
kickoff
page,
feel
free
to
jump
in
and
contribute
and
utilize.
C
C
It's
really
important
that
you
and
the
administrators
of
your
CI
system
have
the
ability
to
manage
the
runners
at
a
singular
group
level
so
that
they
can
ensure
that
the
proper
runners
are
available
to
development
teams
when
they
need
them
and
we're
also
adding
in
the
CI
group,
a
new
capability
to
more
easily
create
a
kind
of
matrix
of
CI
jobs
by
defining
that
in
your
CI
mo.
So
the
kind
of
canonical
example
is
doing
different,
build
jobs
across
many
different
stacks.
You
can
do
that
really
quickly
in
a
CI
ml.
C
Instead
of
having
to
create
a
bunch
of
kind
of
static
pipeline
definitions,
we
generate
those
dynamically
in
this
matrix
format.
The
next
one
I
wanted
to
highlight
really
quickly
in
that
adoption
journey.
C
We're
also
adding
a
really
cool
feature
that
integrates
both
release
and
monitor
where,
when
you're
doing
a
incremental
rollout,
you
could
see
the
error
rate
from
your
that
that
might
have
happened
as
a
result
of
your
rollout
and
why
we
might
have
rolled
that
back.
We
would
automatically
rollback
if
an
right
hit
a
threshold
now
right
here
in
this
deploy
board.
You
could
see
that
error
rate
and
understand
why
a
rollback
was
happening,
we're
also
adding
in
the
configure
stage,
improvements
to
how
you
can
manage
kubernetes
clusters.
C
So
in
four
days,
we'll
be
announcing
13.0
and
that
will
include
our
minimal
nbc4
alert
management.
We're
going
to
continue
to
build
on
that
MVC,
adding
alert
identification
which
allows
you
to
do
alert
deduplication,
adding
the
ability
to
set
the
status
of
those
alerts.
So
you
can
say
whether
they
are
being
triage.
They
are
acknowledged
or
have
been
resolved
as
well
as
filtering
those
those
new
assignment
statuses.
C
This
is
especially
useful
for
teams
elaborating
across
time
zones
adding
the
ability
to
connect
individual
charts
to
other
charts,
so
you
can
drill
down
and
keep
your
context
when
you're
when
you're
firefighting
an
issue
as
well
as
have
a
kind
of
linkage
between
dashboards
themselves.
So
if
there's
a
common
dashboard
that
you
might
have
a
starting
point
for
your
triage
process,
it
can
link
out
to
well-known
and
kind
of
established
patterns
for
other
dashboards
that
will
help
you
triage.
C
Those
are
all
again
in
service
of
get
lab
becoming
a
singular
collaboration
point
for
your
dev
and
ops
teams.
I
would
encourage
any
team
that
is
pointing
their
their
alerts
to
a
separate
system
to
consider
pointing
them
to
to
get
lab
so
that
they
can
enable
this
kind
of
collaboration
between
their
two
teams.
C
So
again,
that
was
a
review
of
us
thinking
about
our
upcoming
features
from
a
perspective
of
our
users
adopting
get
lab
with
a
primary
focus
on
moving
from
create
to
verify,
as
well
as,
if
you're,
using
get
labs
planned
tools
using
get
lab,
monitor
tools
for
incident
management.
What's
that
I
will
hand
it
back
to
you
and
you
Thanks
Thank.
E
Okay,
I
guess
zoom
is
not
as
responsible
when
you
hit
unmute
as
it
used
to
be
so
yeah,
so
my
name
is
David
de
santé,
director
of
product
for
the
secure
in
defense,
section
I'm,
gonna,
walk
through
the
different
groups
related
to
our
security
offerings
and
we'll
first
start
off
with
static
analysis.
This
team's
working
on
two
very
exciting
things.
E
E
What
that
will
look
like
you'll
be
able
to
see
an
ace
are
gonna
scans
option
off
the
security
and
compliance
you'll
be
able
to
set
the
information
for
the
scan.
This
game
will
show
up
as
running
within
your
pipeline
view,
but
you'll
see
here
in
the
comments
it
says:
there's
no
commit
that
was
related
to
the
scan,
and
then
the
results
will
be
on
the
pipeline
security
report
and
you'll
notice
highlighted
here.
There's
no
filter!
Oh
that's
because
you're
only
seeing
vast
results
so
there's
no
digital
and
the
results
that
aren't
there.
E
So
this
is
a
nice
big
step
forward
working
across
cleverly
with
Kenny's
section
to
be
able
to
begin
to
offer
scans
outside
of
a
commit.
The
final
group
for
secure
is
composition,
analysis
again,
focusing
on
application
security.
Testing
leadership
will
be
exposing
the
license
check
approval
within
the
lines
complot
license
compliance
section
of
the
UI,
so
you'll
see
here
in
this
mock-up.
E
We're
able
to
then
also
make
the
filters
persistent
as
well,
and
the
final
group
for
defend
is
container
security.
They're,
focusing
very
much
driving.
Adoption
of
the
features
are
responsible
for
the
main
item.
I
want
to
highlight,
for
you
is
security
policies
and
being
able
to
view
them
in
the
UI
and
that
all
look
kind
of
like
this,
so
off
the
threat.
E
Monitoring
tab
where
you
can
see
the
overview
of
the
traffic
and,
what's
being
mitigated,
will
now
be
a
policies
option
where
you
can
see
which
policy
you've
rolled
out
they're
also
beginning
to
make
it
available
for
you
to
add
and
remove
policies.
I
don't
have
time
to
go
through
all
that
today,
because
I'm
actually,
at
my
time
but
I'll
hand
that
back
to
Nuuk
what
I'll
say
is
if
you
have
any
interest
in
this,
please
don't
hesitate
to
ping
the
team
on
the
issues
as
well
as
collaborate
with
them
on
any
additional
items.