►
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
There
we
go
all
right.
Welcome
to
the
distribution
demo
for
october
november,
18
2021.
today
I'll
be
demoing.
The
recent
changes
we've
been
doing
what
I've
been
doing
for
adapting
the
gitlab
chart
to
use
the
new
nginx
ingress
controller.
A
Quite
simply
because
we
forked
the
chart,
this
was
a
little
bit
more
involved.
Otherwise,
we'd
just
be
able
to
do
an
update
to
the
requirements
and
allow
us
to
bring
in
the
new
chart.
Let
me
go
ahead
and
get
it
starting
to
kick
off
and
we
can
talk
a
little
bit
more
about
some
of
the
changes.
A
I
get
started,
that's
good.
Thank
you,
yep
all
right,
so
that's
gonna
take
about
two
minutes
to
install
and
get
going
all
right.
So
when
I
first
started
doing
this,
one
of
the
first
things
I
did
was
started
to
bring
in
the
new
chart
and
for
reasons
I
still
cannot
explain.
At
this
moment
there
was
a
number
of
helm
checks
that
were
failing
specifically
anything.
Let
me
find
one
here.
A
A
I
spent
a
bunch
of
time
debugging
that
finally
figuring
out
what
it
was.
I
taken
the
resource
off
fixed
that
and
then
actually
earlier
this
week
did
some
more
testing
and
it's
now
working
again.
So
I
have
no
as
to
why
that
happened,
but
something
we
should
at
least
keep
an
eye
on.
If,
if
anybody
else
experienced
it,
we
can
figure
out
exactly
what's
causing
that,
but
that
that
threw
me
for
quite
a
while.
A
Now
another
thing
that
caught
me
for
a
little
bit
is
specifically
under
1.20
and
1.21.
A
It
is
the
new
ingress
controller
does
not
allow
both
the
ingress
class
name
and
the
class
name,
an
annotation
that
exists
in
the
ingress.
If
it
does
it,
it
validates
the
ingress
and
will
never
process
that
ingress.
A
So
there's
a
bunch
of
pieces-
and
actually
I
just
brought
it
in
as
a
defined
last
night
or
yesterday
this
right
here.
A
If
we
have,
if
we're
using
the
ingress
class
resource,
then
we
go
ahead
and
do
not
deploy
with
the
annotation
so
anything
that
any
of
the
newer
versions
of
kubernetes
that
does
support
the
ingress
class.
We're
going
to
favor
that,
because
that
is
actually
what's
becoming
the
standard
becoming
what's
going
to
be
deployed,
the
ingress,
annotation
ingress
class
annotation
will
be
actually
deprecated
and
actually,
I
think
it
is
now
and
122
actually
deprecated,
but
it
will
disappear
and
I
think
1.25
so
now,
we've
actually
deployed.
A
A
Actually,
this
is
still
adding
the
parameters,
but
my
understanding
is,
I
think
I
still
have
that
in
my
values
file,
but
my
my
understand
is
from
what
I've
read
the
nginx
ingress
controller
does
not
use
the
parameter
section
at
all.
At
this
point,
it's
basically
completely
ignored.
The
only
thing
that's
used
in
the
spec
section
is
the
controller
to
identify
which,
which
instance
of
the
nginx
ingress,
is
going
to
be
used
or
more
specific,
that
it's
an
exchange
because
you
may
have
multiple
ingress
controllers
deployed
in
your
cluster.
C
A
If
I
understand
what
you
just
said,
yes,
the
the
the
ingress
class
resource
or
the
ingredient
yeah
the
ingredients
class
resource
here
that
we're
looking
at
actually
is
generated
by
the
nginx
ingress
controller
by
that
template,
and
so
really
this.
This
is
really
just
tie
in
the
specific
controller,
in
this
case,
the
one
that
was
created
to
the
class
by
this
name.
A
So
the
name
of
the
ingress
class
is
what
is
that
tie-in
record
between
the
class
name.
A
A
All
right
by
this
time,
we
should
have
everything
up
and
running
cool.
Why
are
we
crashing
here.
A
A
Yeah
we
can
try
that
real,
quick
I've
had
a
lot
of
problems,
sometimes
because
of
the
start
manager
where
nate
spaces
take
20
or
30
minutes
to
go
away.
But
let's
give
it
a
try.
C
A
Oh
another
thing
to
be
wary
of
bringing
the
new
ingress
controller
into
kubernetes
is
gonna.
Basically,
allow
only
allow
us
to
deploy
back
to
1.19
kubernetes
1.19
anything
prior
to
that
will
not
work
unless
we
go
back
to
the
old
ingress
controller.
A
Yep,
so
you
think
if
I
just
get
rid
of
the
pvcs,
we'll
be
back
to
semi
normal.
C
So
I'll
just
delete
those
three
pvcs
and
and
then
reinstall
the
home
chart.
A
Okay,
so
yeah
I've
got
some
documentation
I'll
be
finishing
up
today,
specifically
about
the
new
ingress
controller
and
the
jump
to
119.
C
The
perk
of
jumping
to
119,
too,
is
that
I,
if
I
remember
correctly,
119,
has
ingress
api
v1,
and
so
we
wouldn't
actually
need
to
do
any
juggling
of
ingress
api
versions
either
in
the
helm
chart
or
in
the
operator,
which
would
actually
make
things
much
simpler.
We
could
just
always
pick
that
one
ingress
version
yep,
which
would
be
pretty
nice.
So
that
said,
of
course,
the
new
required
minimum
version
is
a
pretty
big
jump,
so
I
think
we'd
probably
need
to
do
an
announcement
before
yes,.
A
Yep
yeah,
so
I
was
hoping
that
this
would
all
be
tied
up
a
little
bit
ago.
I
had
a
few
other
problems
pop
up
that
was
ugly,
but
so
it
won't
be
for
this
release,
but
next
release
so
we'll
have
to
start
prepping
now
to
get
that
notice
out,
but
I'm
wondering,
if
that's
the
case,
do
we
even
really
need
to
do
this
function
anymore?
Once
we
make
the
jump,
I
mean
because
we
won't
be
able
to
deploy
to
the
extensions,
be
e1
beta1
and
effectively.
This
is
gone
too.
Now.
C
Go
ahead,
yeah,
no
you're
right!
That
is
weird.
C
A
That
took
me
a
little
bit
took
some
digging
and
finally
found
some
github
issues
that
specifically
called
out
that
inversions
1.2
1.21,
that
the
ingress
class
name
annotation
and
the
english
class
name
attribute
or
resource
whatever
you
want
to
call
it
could
not
coexist.
A
C
A
It
was
created
if
you
looked
at
the
ingress
controller
log
output,
they
would
come
back
saying
basically
saying
invalid
ingress,
ignoring
it.
A
A
A
A
I
was
going
back
and
forth
on
like
several
different
clusters,
the
other
day
with
different
configurations
and
was
getting
the
wrong
url
to
run
git
clone
ur
url
to
all
right.
So
there.
A
So
well
now
we
now
we
know
the
ssh
is
working,
and
so
the
ingress
from
all
from
everything
I
could
tell
is
working
fine.
The
only
thing
I
didn't
test,
which
is
what
so
I
was
just
thinking
of
a
little
while
ago,
is
if
I
change
the
port
of
ssh.
Does
that
actually-
and
I
don't
expect
you'll-
have
a
problem
with
it.
A
I
I
want
to
test
that
a
little
bit
later
on
here,
just
verify
that
we
don't
have
any
inconsistencies
in
the
nginx
controller
comes
up,
but
does
not
map
to
the
new
ssh
port.
So
that's
about
the
only
other
thing
I
can
see.
That
might
be
an
issue
at
this
point.
A
And
hearing
none
yeah
silence
is
golden.
I
think
on
this.
It
seems
to
work
and
we're
at
time
now
too.
So
all
right,
bye,
that's
what
I've
got
and
we'll
get
this
updated
with
the
documentation
updates
and
pushed
out
later
today.
So
people
can
look
at
a
little
bit
closer
and
see
if
there's
anything
else,
that
I'm
missing.