►
From YouTube: GitLab 13.9 Kickoff - Enablement:Distribution
Description
Find out what we are working on for 13.9 in the Distribution team
A
A
In
the
foundational
tooling
and
getting
the
first
two
of
our
services
enabled
for
the
operator,
we
have
chosen
to
leverage
the
helm
chart
to
allow
us
to
leverage
the
learning
and
all
the
work
that
has
gone
into
that
and
avoid
having
to
rebuild
all
that
logic
and
the
operator
generator,
and
so
for
us.
We
are
actually
using
the
helm,
chart
to
generate
the
kubernetes
objects
and
then
the
operator
deploys
them,
there's
just
to
be
clear:
there's
no
helm
tiller
or
any
other
helm
services
involved,
other
than
simply
creating
the
objects
within
the
operator
itself.
A
A
From
there
we're
exploring
more
of
the
operational
details
of
running
gitlab
in
openshift,
for
example,
being
able
to
have
ssh
connectivity
to
your
gitlab
application.
The
openshift
typically
doesn't
really
rely
upon
ingresses
and
they
instead
prefer
routers
routers.
Don't
today
easily
support
binary
port
forwarding.
So
arbitrary
to
city
port
forwarding,
which
is
what
we
would
need
to
have
for
ssh
the
enginex
ingress,
quick,
lower,
does
have
this,
which
we
use
and
so
we're
trying
to
figure
out
what
our
options
here
are
and
what
the
best
path
forward
here
is.
A
You
can,
of
course,
use
a
router
with
just
plain
h,
gps,
but
we
would
like
to
have
ssh
as
an
option
to
our
customers,
and
so
we
are
exploring
this
issue
we're
also
working
in
a
few
other
areas.
We
are
working
on
having
some
of
the
helm,
chart,
resources
and
parameters
being
able
to
pass
through.
You
can
see
that
here
and
we're
also
working
to
document
some
of
the
security
models
that
we're
working
with
for
the
operator
as
well.
But
overall,
we've
made
a
lot
of
significant
progress
here
and
we're
hopeful.
A
We
can
get
this
launched
in
13.9,
but
if
not
we'll,
maybe
have
a
couple
last
items
in
13.10,
but
again
we're
really
targeting
here,
13
at
nine
and
hopefully
I'll
be
able
to
get
that
out,
but
if
not
it'll
be
13
at
10..
So
that's
a
big
item.
That's
going
to
consume
a
lot
of
our
time
here.
This
release,
but
we're
also
finishing
up
or
we're
going
to
try
and
finish
up
more
of
the
gitlab
pages
support
great
news,
a
lot
of
folks.
A
This
was
one
of
the
last
remaining
functionality
gaps
present
on
the
kubernetes
helm,
chart
versus
our
omnibus
service
and,
of
course,
pages
is
quite
popular.
It
allows
you
to
do
some
really
easy
static
site
hosting
services
within
within
gitlab,
and
so
a
lot
of
customers
do
do
want
this,
and
so
great
news
in
13.8.
A
Alongside
a
lot
of
great
work
from
the
pages
team,
we're
able
to
offer
both
the
basic
pages
chart,
and
so
along
that
service
within
the
get
lab
deployed
in
on
kubernetes,
but
also
we
added
support
for
custom
domains
remaining
is
support
for
access
control.
We
are
looking
at
bringing
that
into
13.9
and
potentially
just
for
configuration
and
maybe
not
provisioning
of
the
oauth
secrets,
but
we're
hoping
to
have
some
version
of
this
and
13.9
here
coming
soon.
A
The
remaining
kind
of
key
items
here
we're
looking
for,
are
largely
addressing
some
compatibility
with
other
services
that
we
find
that
customers
are
commonly
using
or
addressing
some
some,
some
bugs
or
some
other
short
kind
of
gaps.
If
you
will.
The
first
of
these
is
that
the
gitlab
shared
secrets
charts
fails
from
using
istio.
We
don't
deploy
istio
by
default.
A
We
don't
configure
gitlab
to
utilize
it,
but
a
number
of
our
customers
do
prefer
to
use
service
meshes,
and
so
we
want
to
try
and
make
sure
that
we
can
have
them
work
together,
gracefully
and
so
we're
looking
to
do
this.
The
main
work
here
is
making
sure
that
we
have
annotation
support
for
this
particular
chart.
The
rest
of
them
already
has
support,
and
so
it's
also
addressing
sort
of
that
gap
in
our
in
our
helm.
Chart
where
we're
not
allowing
annotations
on
this
particular
one.
A
We're
also
looking
to
address
some
prolific
logging.
Nothing
too
special
there
just
looking
to
hopefully
be
able
to
reduce
some
of
the
noise
and
we're
also
working
to
address
some
negative
behavior
on
the
gitlab
side.
But
we
think
it's
due
to
some
changes
in
how
the
red
hat
pod
man
service
runs.
So
it
turns
out.
We
have
quite
a
lot
of
folks
using
podband
to
run
containers
and
for
those
folks
who
are
using
it
in
the
last
few
releases.
They
started
to
see
some
problems.
A
There
is
a
document
to
work
around
in
this
issue,
but
we
hope
to
not
have
folks
required
to
do
that
and
it
should
just
work
and
so
we're
looking
at
ways
to
work
around
this
issue
with
pam.
Finally
of
the
items
I
want
to
highlight
here.
The
next
one
is
we're
also
working
to
better
really
to
make
sure
that
we
have
a
gitlab
service.
A
We
want
to
make
sure
that
all
of
the
images
that
the
helm
chart
utilizes
are
hosted
on
gitlab
infrastructure-
that's
probably
a
better
way
to
put
it.
The
reason
is,
is
because
we've
seen
some
uptime
challenges
with
some
of
the
variety
of
services
that
we
use
here
at
get
lab,
whether
it's
docker
or
quay
or
gcr
or
other
places,
and
so
we
wanted
to
make
sure
that
we
kind
of
owned
our
own
destiny.
Here.
A
So
it's
also
a
nice
way
for
us
to
help
address
that
request
and
also
provide
some
more
stability
for
our
customers
and
also
gitlab.com,
who
is
also
recently
run
into
this
challenge.
So
those
are
the
highlights
here
for
13.9.
We
are
working
on
a
whole
bunch
of
other
items.
Here
you
can
see
in
our
deliverable
board.
It
is
ranked
out
by
priority,
so
you
can
see
we're
also
working
on
some
red
hat
certification
and
some
other
aspects
of
of
our
of
our
deployments.