►
From YouTube: Distribution Group Conversation
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
B
A
C
Yeah
products
opinion
on
some
of
these
to
see
like
hey,
which
ones
are
actually
more
important.
What
ones
are
we
seeing
customer
demand
for
versus
ones
that
are
just
a
gap
in
the
product
that
we
that
we've
identified,
but
don't
really
know
if
folks,
are,
would
be
willing
to
to
use
that
feature?
I,
don't
I'm
sure
if
it
cut
any
product
Eric
we've
got
on
the
call.
We
can
talk
about
that
offline.
Yes,.
D
I
think
I
think
we
should
kind
of
disentangle
the
prioritization
from
like
the
documentation
of
of
the
issue.
I
think
it's
as
Sid
mentions
like.
We
should
absolutely
have
a
link
to
every
single
issue
that
is
a
red
X.
On
this
chart.
There
should
be
at
least
an
issue
that
documents
that
we
have
a
desire
or
plan
to
turn
this
whole
chart
green
sounds.
C
E
Particular
example
comes
to
PG
bouncer
right
now,
we're
actually
relying
on
other
systems
to
actually
provide
the
H,
a
and
redirects
for
Postgres.
We
can
look
at
bringing
those
in,
but
some
investigation
on
those
will
be
needed
say
the
least
node
exporter
in
one
is
one
in
particular
that
we've
discussed
in
the
past.
It's
a
scenario
where
we
actually
be
asking
for
the
node
level
information
of
each
individual
caperna
DS
node,
which
we
can
already
pull
from
the
previous
data
in
place
so
actually
running
node
exporter
doesn't
do
us
anything
additional.
E
D
Yeah,
that's
probably
worth
some
value
in
using
a
different
icon
for
the
paths
that
actually
don't
make
sense,
I.
Think
of
it
a
big
you
know,
digital
red
X,
you
know,
can
mean
hey,
that's
not
a
valid
path.
It
could
also
mean
it
doesn't
actually
exist
as
functionality,
so
I
think
I
think
we
should
actually
use
a
different
icon
to
kind
of
disentangle
those
two
things.
So
you
know
it'd
we'll
make
that
change
so.
B
B
E
So
mini
cube
while
effectively
relevant
at
all
times.
It
is
a
large
amount
of
resources
to
consume
our
chart
in
particular,
actually
requires
you
go
well
beyond
the
default
resources
of
what
money
Cube
wants
to
use
out
of
the
box.
It
wants
to
use
to
CPU
and
four
gigs
of
ram.
The
minimal
configuration
right
now
is
three
CPU
and
eight
gigs
of
ram.
E
Now,
that
being
said,
that
cuts
it
from
a
normal
requirement
of
between
750
Meg's
and
one
gig
of
ram
just
for
kubernetes
to
operate
on
a
node
to
as
little
as
256
to
512
Meg's
of
RAM.
So
that's
a
significant
reduction.
However,
at
that
same
time,
you
do
actually
run
into
issues
because
there's
some
functionalities
that
are
not
present
in
k3s,
that's
how
k3s
gets
its
name
is
k8s
is
a
shortened
form
for
the
length
of
the
word
kubernetes
and
they've,
taken
five
components
and
replace
or
remove
them,
and
that's
how
they
came
with
k3s.
E
So
we
could
possibly
do
a
lot
of
rework
on
the
gek
to
make
it
actually
operate
in
a
kubernetes
fashion,
and
maybe
k3s
will
give
us
a
better
option
there,
but
most
of
our
developers
will
still
need
to
actually
have
a
Linux
VM
that
has
anywhere
between
three
CPU,
and
you
know,
eight
gigs
of
RAM
will
reduce
the
overhead,
but
still
have
the
primary
set
of
constraints
in
terms
of
resource
requirements.
I
think
leaving
the
GDK,
as
is
for
the
time
being,
understanding
please
that
I
don't
have
a
full
understanding
of
it.
E
B
Been
involved
in
maybe
made
and
maybe
is
like
more
of
a
theoretical
idea
of
mine,
then
something
else
I
just
I
think
gonna
get
on.
Kubernetes
is
probably
our
future
and
then
I
thought
for
the
developers
to
kind
of
be
familiar
with
that
as
well.
But
maybe
this
is
not
the
way
to
do
it
and
we
lost
our
priorities
for
12
to
12,
to
1,
1203,
2
or
3.
So
a
great
work
and
tom
has
the
next
question.
A
E
I
can
I
can
happily
let
Ian
talk
about
the
PostgreSQL
and
the
investigation
that
he
has
done.
I
do
want
to
point
out
that
in
the
past
month,
what
people
have
asked
why
the
charge
don't
follow
the
gitlab
matching
release
numbers.
This
right
here
is
one
of
the
primary
reasons
we
have
not
made
the
switch
so
that
our
charts
are
version
identically
to
the
gitlab
version.
E
The
way
we
do
it
with
omnibus,
because
we
have
major
components
that
are
still
under
heavy
evolution
that
we
have
not
yet
gotten
into
a
lockstep
behavior
with,
so
we
were
100%
targeting
to
get
PostgreSQL
and
Redis
updated
in
time
for
12-point
Oh.
However,
that
didn't
happen
after
further
investigation
showed
that
we
didn't
have
a
way
to
do
any
automation
around
it
and
we
needed
to
get
good,
solid
migration
data
and
that
it's
going
to
be
a
breaking
change.
E
E
G
Our
current
road
map
around
to
publishing
am
ice
with
with
no
EE
license
is
going
to
be
in
the
future.
I
know
that
there
was
some
changes
that
happen
around
that
with
the
marketplace
and
I
know:
we've
extended
that
support,
but
we
do
have
a
number
of
enterprise
customers
we're
trying
to
get
to
to
move
to
a
different,
install
type,
but
still
rely
on
that
today.
So
just
wanted
to
get
some
background
on
where
that's
at
and
where
it's
headed
from
here
yeah.
D
D
We
took
the
existing
EE
ami
and
applied
a
license
to
it
and
then
used
it
for
the
marketplace
image
and
what
we
should
have
done
was
created
a
separate
ami
with
a
license
and
left
the
one
without
a
license
alone,
and
so
you
know
when
we
realized
that
we
clicked
quickly
made
an
issue,
it's
linked
there
and
six,
a
probably
linked
it.
Thank
you
believe,
and
so
I
believe
that
actually
is,
is
already
available,
Brian
and
so
the
your
customers
that
are
using
that
ami
can
go
ahead
and
reuse.
D
H
A
We
probably
so,
even
though
it
says
over
the
rudder,
if
you
look
at
the
issue
the
proposals,
don't
really
have
it
going
through
the
rudder
directly.
So
we
have
a
couple
different
proposals.
What
we're
probably
going
to
end
up
doing
is
supporting
multiple
ways
of
doing
it
and
having
talks.
Asian
is
probably
what
will
be
the
results
of
that
that
issue.
The
reason
is
the
router
itself
only
has
80
and
443
available,
but
so
so
one
of
I
think
one
of
the
proposals
does
involve
on
node
port.
A
When
we
wrote
the
issue
we
were
looking
at
seeing
if
we
could
do
it
through
the
router
I
think
probably
we
won't
be
doing
it
through
the
rudder
directly,
but
through
some
other
some
other
method
that
will
be
documented
yeah.
The
issue
is
is
to
fit
to
to
nail
that
down
and
then
doc
document
how
we
do
it
in
ensure
we're
supporting
supporting
it
all.
H
A
For
the
12.2
and
12.3
issues
here
they
were
linked
somewhat
based
on
our
okay,
ours
and
also
based
on
just
the
things
we've
been
iterating
on
doing
one
piece
of,
for
example,
of
this
past
release
with
opener
shift.
We
got
the
registry
working
or
it
wasn't
working
the
chart
before
so
they're.
Not.
We
haven't
hard
scheduled
limb
to
a
milestone.
They
were
thrown
in
here
more
as
like
follow-up
to
the
work,
we're
continuing
to
do
so.
D
D
E
Of
customers
interested
in
OpenShift
for
some
time
and
we've
never
neglected
them
intentionally
or
in
any
way,
we've
answered
every
question.
We
can
and
look
into
everything
that
they
have
asked
just
to
quick
touch
base
on
what
DJ
was
talking
about
is.
We
do
actually
already
have
a
method
to
configure
SSH,
not
using
an
ingress,
which
is
what
open
ships
router
replaces
by
the
way
we
have
most
of
that
documented.
E
What
we're
going
to
have
to
do
is
specifically
call
out
that
this
is
one
of
the
paths
that
you
actually
have
to
take
when
using
open
ship
at
this
time.
There
are
much
larger
concerns
that
are
in
that
issue
list
when
it
comes
to
open
shift,
and
we
actually
have
developers
of
interest
in
Red
Hat,
and
we
have
a
couple
of
team
members
that
we're
adding
to
the
company
that
are
actually
going
to
have
OpenShift
experience.
We
have
a
couple
people
that
we
brought
on
recently
and
other
teams
that
actually
came
from
OpenShift
implementation
engineers.
E
I
Thank
you
so
much.
First
of
all
for
all
of
the
work
of
the
team
for
enabling
geo
and
M
charts.
This
is
super
appreciated.
That's
just
a
shout
out
to
the
team,
so
the
question
I
realize
it's
a
little
bit
open-ended,
but
as
sort
of
a
new
person
I
think
it's
amazing
to
see
you
can
install
get
lab
in
many.
I
E
Currently,
we
still
primarily
say
that
most
people
do
use
the
omnibus
we're
not
saying
that
most
people
will
that
use
the
omnibus,
but
do
its
the
tried
and
true
tested
that
most
people
fully
understand
how
to
install
in
a
VM
wherever
it
is
to
figure
it
as
necessary.
It's
also
what
our
support
currently
has
the
most
experience
with
that
being
said,
I'm
going
to
touch
on
the
subject
of.
Can
this
be
used
for
get
lab
comm?
The
answer
is
not
only
yes,
but
we're
beginning
that
migration.
E
Now,
okay,
the
there's
a
certain
amount
of
opinionation
when
it
comes
to
what
the
omnibus
is
and
what
it
will
or
will
not
be
in
the
future.
In
our
mind,
there
will
be
both
methods
for
a
long
portion
of
the
foreseeable
future.
We
want
to
get
the
cloud
native
charts
to
be
as
capable
or
more
than
the
omnibus
and
we'd
like
to
see
more
large
customers
using
that,
because
it
does
specifically
as
designed,
scale
better.
E
E
So
if
we
have
to
be
very
careful
that
we
don't
push
everybody
to
use
kubernetes
before
the
background,
knowledge
in
the
industry
is
in
place
and
that
when
we
do
start
pushing
people
that
way,
we've
made
it
as
reliable
and
straightforward
and
well-documented
as
we
have
the
Omnibus,
and
the
big
difference
here
is.
The
Omnibus
has
been
around
for
five
plus
years
with
all
of
the
refinement
and
time
that's
put
into
it.
E
The
charts
have
been
around
officially
for
less
than
a
year
on
GA
we've
been
working
on
them
for
a
year
and
a
half,
but
there's
a
big
difference
in
age
and
a
small
difference
in
maturity.
We've
come
a
very
long
way,
but
so
has
the
ecosystem
that
has
released
every
quarter
that
we're
continuing
to
try
to
do
this
work.
G
I
can
jump
in
just
real
quick
I.
Think
I
think
there's
another
point
to
be
made
on
that,
as
well
as
in
the
pre
sales
perspective,
especially
in
the
enterprise
space.
You
know
omnibuses
still
our
recommended
approach.
Of
course,
as
Jason
was
just
mentioning,
and
but
one
of
the
things
that
we
tell
them
is
that
you
know
as
part
of
this,
we
are
also
working
on
a
migration
path
so
that
in
the
future,
when
it
feels
like
you're
able
to
you
know
take
that
on,
we
have
geo
support.
G
We
have
a
Shea
for
giddily
all
those
things
that
are
required
in
the
enterprise
space.
They'll
have
a
migration
path
that
they
can
move
over
to
the
cloud
native
charts
in
the
future,
which
will
make
that
easy
for
them
to
do
that.
So
that
is
kind
of
the
pathway
that
we're
working
with
this
as
it
starts
to
get
more
matured
as
Jason
with
vegetable
right.
E
Thank
you
for
for
making
sure
that
that
was
said.
While
we
did
this,
that's
actually
a
part
of
one
of
our
okay
ours
coming
up
is
to
make
sure
that
it's
actually
easy
to
transit
between
the
two
methods.
It's
there,
it's
functional
as
it
is,
but
there's
a
lot
of
refinment
to
be
done
to
it
to
make
it
much
more
seamless.
E
F
I
jump
in
very
briefly
from
the
support
side
and
I'm
super
happy
to
hear
all
this
as
well.
I'm
super
excited
about
what
Jason
and
team
are
doing
with
cloud
negative
stuff.
It's
gonna
be
amazing.
It
is
amazing,
really
happy
to
hear
that
sales
of
promoting
omnibuses
the
default
method
just
for
now
for
anybody
that
isn't
super
experienced
with
kubernetes
and
really
excited
for
the
future
when
I
imagine
the
cloud
native
option
will
be
our
default
is
gonna,
be
a
wonderful,
wonderful
thing.
J
I
want
to
speak
up
and
say
one
other
thing,
though
cloud
native
is
default,
it's
wonderful
for
large
orbs,
but
we
should
also
be
present
in
the
concept
that
omnibus
is
better
for
small
to
medium
business,
because
the
complexity
of
kubernetes
and
how
many
systems
it
would
take.
That's
a
big
lift
for
smaller
businesses.
When
you
say
I
know
for
me,
if
I
was
in
Sepik
operating
system,
I'd
say
minimum
for
machines
docked
out
on
the
bus.
J
J
But
if
we
ignore
those,
it's
a
large
part
of
the
market
and
also
like
the
hobbyist
to
pick
it
up
so
I
know
when
I
was
making
decisions
in
my
company,
I
picked
up
the
small
thing
to
see
if
I
wanted
to
fit
to
my
CIO
right,
I'm,
not
gonna
pitch
it
to
him,
if
I
can't
run
it
at
home,
and
that
is
a
big
lift.
So
there's
a
lot
of
stuff
there.
We
want
to
make
sure
we
have
that
and
I'm
gonna
call
times
is
five
till
and
we
have
another
meeting
here
in
five.