►
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
Welcome
to
the
distribution
demo
for
june
16
2022
and
today
I
will
be
showing
some
of
the
work
I've
been
doing
on
the
hype,
the
ubi
pipeline
certification,
and
with
that,
let's
get
started
in
the
demo
notes.
I
have
links
to
the
the
merge
request,
which
is
modifying
the
ubi
pipeline,
along
with
the
link
to
the
preflight
repository
that
I
used
for
the
pipeline.
So
let
me
get
started
with
share
screen.
A
There
we
go
so
this
is
the
section
of
the
cng
pipeline.
It's
actually
one
of
the
sub
files,
the
dot
get
lab
and
let's
just
pull
this
up.
Dot
get
lab.
Ci
ubi
get
lab
ci
dot
here
ml
is
the
actual
file.
This
is
in.
A
This
is
the
the
new
certification
job
built
it
out
as
a
template,
and
then
we
just
reference
that
template
for
each
of
the
images
we
need
to
certify
fairly
standard.
I
mean
we,
we
have
log
into
the
since
we're
fully
within
red
hat
land,
we're
using
podband.
That's
what
the
is
in
the
preflight
binaries
podman's
installed
instead
of
docker,
so
we
log
into
the
get
lab
registry
and
then
basically,
we
just
call
preflight.
A
There
are
a
couple
of
additional
options
that
allow
us
to
submit
the
report
directly
to
red
hat
when
we're
done,
which
is
the
minus
by
submit,
and
you
need
to
give
it
a.
I
think
it's
at
the
beginning
here,
the
yeah.
You
need
to
give
it
to
the
red
hat
token,
to
be
able
to
submit
the
reports.
A
A
A
Although
hussein
has
a
couple
of
comments
and
it's
I'm
actually
going
to
be
making
a
change
this,
where
the
project
id
will
disappear,
because
we
already
have
the
project
id
and
a
file
in
the
repository,
so
we're
going
to
I've
just
been
adding
yq
to
the
preflight
image
and
with
yq
we'll
pull
that
project
id
out
up
here
in
the
script
and
use
it
in
the
script
area.
A
So
we'll
eliminate
the
need
to
maintain
the
project
id
here
or
at
least
multiple
places,
and
then
we
just
we
throw
the
needs
dependency
on
there.
That
way,
we
can
run
the
certification
job
as
soon
as
the
image.
A
The
final
image
has
been
cut
in
the
pipeline
and
then
that's
just
duplicated
for
all
the
images
all
11
of
them,
or
so
that
we
have
now
with
that.
What
I
can
do
is
basically
I'm
going
to
just
run
this
command
from
the
container,
just
without
all
the
additional
api
tokens
and
certifications
of,
or
the
submitting
certifications
and
so
forth
again.
This
only
affects
ubi
pipeline
does
not
affect
regular
cng
pipeline.
A
So
so
I'm
just
I've
got
the
pre-flight
docker
image
local,
it's
1.2.1,
which
is
the
latest
release
that
they
have
had.
It
calls
once
and
talker
images
up
the
road
we
call
preflight
check
container
and
we
give
it
an
image
name
and
that's
all
we
really
need
to
do.
A
This
actually
makes
our
certification
pipeline
a
heck
of
a
lot
easier
than
what
we
had
in
the
past,
because
now
it
just
fits
into
the
into
our
pipeline
versus
us,
trying
to
submit
stuff
back
to
red
hat,
trying
to
trigger
the
image
or
the
certification
run
on
their
side
and
everything
else.
It's
always
been
just
problematic.
A
The
other
thing
that
this
does
for
us
is
we
get
the
actual
certification
results
as
an
artifact
of
the
jobs.
So
if
we
want,
we
can
go
investigated,
look
into
those
certification,
json
outputs,
to
see
what
actually
is
going
on
and
we'll
see
that
here
in
a
minute.
So
it
takes
a
little
bit
to
run
because
it's
pulling
the
image
each
time.
It
runs
a
certification,
so
it
does
take
a
little
bit
of
to
get
it
going
once
it's
once.
It's
pulled
down
that
image.
B
Question
george,
even
considering
that's
a
really
tiny
job
by
all
means,
did
you
think
of
maybe
making
it
something
external
that
we
can
pull
from
all
the
projects.
So
we
have
one
job
definition
across
everything
that
tries
to
scan
with
red
hat's
preflight
so
that
whenever
the
pre-flight
calls
change,
we
only
have
one
place
to
change
it.
B
A
B
You're,
first
off,
like
you're,
doing
the
scanning
for
the
ubi
images
in
the
cng,
I'm
doing
the
same
thing
in
the
operator.
Yep
and
runner
folks
are
doing
it
in
their
own
repo
as
well
or
they're
doing
it
manually.
I
don't
know
anymore,
but
wouldn't
it
be
beneficial
to
kind
of
throw
that
into
a
template
that
could
be
included
from
multiple
pipelines.
B
A
Yeah
yeah,
that's
that's
possibility.
I
just
I
was
just
looking
at
partly
for
a
pre-flight.
I
was
starting
to
add
some
container
scanning
onto
that
and
we
have
in
the
pipeline
template
repo
the
one
that
red
hat's
excuse
me
that
get
labs
puts
on
puts
out
there.
That
was
there.
So
I've
been
adding
that
in.
I
could
see
where
we
might
put
that
in
there.
But
I
wonder
is
that
a
little
too
specific.
B
For
us,
their
majority
of
things
are
basically
the
podman
login
is
a
must,
if
you're
doing
third
party
correct
repo,
because
as
I
discovered
so
that's
one
thing
that
is
not
optional,
so
everybody
will
have
to
do
it
and
again
it
wasn't
quite
obvious
from
their
documentation
or
from
our
document
or
from
any
documentation
that
that
is
done.
So
having
that
as
a
common
template
for
everybody
to
use
could
be
of
use.
If
we
have
more
components
trying
to
follow
the
suit,
which
is,
I
don't
know
whether
that's
the
case.
C
A
I
like
the
idea
of
making
a
common
template,
it's
certainly
among
the
three
or
four
repos
that
are
using
it
today
or
they're,
going
to
be
using
it
today,
I'm
just
not
sure,
what's
the
best
way
to
distribute
that,
I'm
just
not
convinced
that
you
go
into
the
public
pipeline
templates
that
everybody
has
access
to
off.
Getlab.Com.
A
B
Yeah
yeah
they're,
like
you,
can
actually
include
things
from
a
specific
project.
If
you
need
to
so.
A
A
Think
that's
a
thought
too.
Okay,
let's,
let's
sort
of
I
gotta-
make
a
note
of
that.
But
let's
continue
discussing
that
figure
out.
What
and
again
we
should.
Probably
I
mean
you
and
I
will
you
could
use
it
easily.
We
should
probably
get
the
runner
team
involved
to
just
discuss
here.
Here's
a
template
you
can
use
and
see
what
their
requirements
are
too
to
see
if
yeah
what
we
want
to
do
exactly.
A
All
right,
sorry
again,
I
love
the
autocorrect
so
that
all
my
misspellings
get
corrected
all
right
come
on
there
we
go.
Let's
look
at
this
output
a
little
bit
more
all
right,
so
we
started
up
there's,
not
a
huge
number
of
tests
that
are
that
do
occur.
At
this
point,
I
suspect
this
will
grow
over
time
from
red
hat,
but
so
because
I
I
think
there
were
actually
more
tests
when
we
were
running
the
certification
on
red
hat's.
A
Certified
certification
pipeline
excuse
me,
but
so
in
this
case
everything
everything
pass
and
then
so
this
is
basically
standard
out.
Excuse
me
standard
error.
What
comes
out
of
the
container
this
is
standard
out,
which
is
I
redirect
into
a
file.
This
is
the
json
results
and
then
within
the
the
job.
A
We
specifically
look
at
this
value
right
here,
the
past,
if
it's
true
or
false,
and
if
it's
false,
that's
when
we
fail
the
job
that
way,
we
get
a
at
least
a
visual
indication
in
our
pipeline
that
hey,
we
had
a
had
a
had
a
image
that
didn't
certify
correctly.
Now
the
one
good
question
might
be-
and
maybe
I
should
put
this
into
the
mrs
may,
do
we
put
the
do
we
fail
the
pipeline.
A
If
we
fail
a
certification
or
do
we
do
the
allow
failure
and
just
so
it
warns
us,
that's
a
that's
a
question.
We
need
to
probably
figure
out
right
now,
we'll
fail
anyways,
because
we
have
and
I'll
demonstrate
that
here
in
a
second,
we
have
alpine
certificates,
that's
failing,
but
that's
failing
because
of
the
some
of
the
labels.
Some
of
the
darker
labels
are
blank.
B
D
A
And
I
know
dj
you're
here,
that's
great
because
you've,
you
do
a
lot
more
with
the
release,
stuff
and
so
forth.
So
you
probably
have
a
much
better
thought
or
feeling
as
to
whether
we
should,
if,
if
we
have
an
image
or
a
number
of
images
that
fail
certification,
do
we
go
ahead
and
just
break
the
ubi
pipeline
and
force
us
to
go
fix
it
or
do
we
sort
of
say
well,
the
images
are
cut
and
let
it
go
through
and
so
forth.
E
I
think
initially
like
for
the
first
pass.
We
probably
wanted
to
allow
failure
and
the
reason
for
that
is
in
the
release
pipelines.
If
the
pipeline
fails,
the
public,
like
the
automation,
may
not
allow
the
publish
final
steps
to
run
or,
like
the
later
steps
in
the
in
the
you
know,
release
tooling
may
not
run
right.
E
But
that
being
said
being
that
we
have
this
and
now
like
we
can
run
this
earlier.
You
know
we
would
like
to
be
able
to
catch
the
failures
earlier,
so
it
really
depends
on
how
like
stable.
This
becomes
right,
because
if
it's
really
showing
us
like
the
actual
issues
and
when
there's
an
issue,
it's
an
actual
issue
we
would
want
to
care
about,
it
would
make
sense
to
not
even
allow
it
to
fail.
A
How
stable
yeah
I
agree
and
given
our
history,
we've
we've
gone
to
the
allow
failure
for
all
this
stuff
because
of
the
amount
of
problems
we've
had
and
I
I
think
I
think
I
sort
of
agree
that,
yes,
initially,
we
do
allow
the
failure
just
to
test
the
waters
until
we
feel
a
lot
more
confident.
I
guess.
E
C
A
E
You'd,
have
I
I
don't
know,
actually
it
would
be
dependent
on
whether
the
allow
failure
accepts
like
conditionals
or
not,
which
probably
not,
I
don't
think
it
does.
I
think
it's
just
a
true
false
yeah,
most
likely
if.
B
E
B
If
it
was
accepting
the
parameter
in
the
variable
you
could,
then
we
can
have
that
variable
default
to
fall
to
allow
failure
true
and
then
the
release
tools
will
supply
false.
But
I
don't
know
yeah.
E
A
E
A
A
Okay,
all
right,
so
that's
the
that's
a
complete
success.
Let's
I've
got.
I've
got
queued
up
here,
the
command
for
the
alpine
certificates,
which
is
the
only
container
that's
failing
today.
A
A
There
we
go
it's
a
smaller
excuse,
it's
a
smaller
image,
it
comes
faster
and
here's
the
one
that
fails,
the
whoops
scrolls
off
the
page,
it's
the
has
required
label
and
so
and
looking
at
that
and
then
of
course
we
get
the
the
nice
thing
is
we
see
past
failed
or
false.
So
that's
where
we
can
trigger
in
our
pipeline
job
to
say
it
failed
and
fail.
The.
B
D
A
A
A
And,
of
course,
if
you
run
and
I've
done
this
a
couple
times
because
I
forgot
to
put
the
dash
ubi
aid
on
my
branch,
if
you
write
it
against
regular
images,
it
fails
left
and
right
all
over
the
place.
A
Right
here,
the
labels-
you
know
everything
looks
good,
they
the
release
and
the
version
everything
else
is
labeled
up,
and
if
you
compare
these
labels
against
an
image,
that's
passing,
those
are
the
only
differences
is
that
these
are
empty.
So
that's
the
only
only
thing
I
can
figure.
So
I
think
once
we
get
something
into
those
spots,
I
think
the
alpine
certificates
image
will
start
passing.
A
So
I
have
a
mr
opened
up.
Another
mmr
an
issue
opened
up
about
that,
should
be
easy
to
fix.
I
just
we
currently
do
try
to
put
the
get
lab
version
into
those
labels
which
is
what's
used
on
a
number
of
other
images,
but
they're
not
going
there.
A
I
think
the
only
thing-
and
I
haven't
looked
at
it
extensively
I
think,
what's
going
on-
is
that
with
the
other
images
we
probably
have
a
script
that
we're
running
to
determine
what
versions
or
what
version
numbers
we
should
use,
and
we
may
not
be
running
that
for
the
alpine
certificates.
A
C
I
noticed
that
they
red
hat,
publishes
a
docker
file
for
the
openshift
pre-flight
themselves.
Is
there
a
reason
that
we
use
our
own
other
than
getting
jq
on
it?
Is
there
a?
Is
there
like
a
general
strategy
that
we
don't
wanted
to
depend
on
upstream
images,
because
it's
pretty
simple
to
build
our
own
here
or.
C
A
Sorry
I
got
a
contact
move
around
on
me,
but
originally
originally,
I
was
not
finding
anything
an
image
for
pre-flight
at
all.
It
was
just
the
binary
out
there
out
of
their
repo
now
whether
whether
that
docker
file
was
out
there,
I
just
wasn't
seeing
it
or
whether
it
appeared
after
we
were
doing
it.
A
I
don't
know
so
it
was
easy
enough
for
me
to
just
grab
the
image,
throw
it
into
a
container
for
us
to
do
some,
pre-testing
and
so
forth,
because
they
had
been
prominent
promising
us
pre-flight
for
two
or
three
months
before
they
actually
finally
released
it
as
a
beta
and
so
as
soon
as
as
soon
as
we
basically
were
get
able
to
get
our
hands
on.
A
I
want
to
start
playing
with
it,
because
that
was
going
to
solve
a
lot
of
our
problems
and
they
were
pretty
much
actually
almost
the
time
they
released
it.
They
turned
off
all
our
other
certification
methods
too.
So
it
broke
us
a
bit
so
and
today
we're
still
broken
until
we
get
this
finalized
and
out
to
the
pipeline.
A
The
other
advantage
we
have
is
that,
yes,
we
can.
We
do
put
in
jq
and
yq
now,
I'm
just
adding
yq
today,
so
that
we
can
process
specifically
pull
these
project
ids
out
of
the
existing
gamble
file.
That's
in
the
repository
today,
I
don't
expect
we're
going
to
be
putting
a
whole
lot
more
into
the
pre-fight
binary.
I
don't
think
we
need
a
lot
more.
A
I
don't
want
to
make
it
big
and,
and
so
the
other
requirement
there
becomes
when
they
start
releasing
and
they've
done
this
once
recently,
as
they
did
a
release
and
I
revved
from
the
110,
which
was
the
original
release
they
had
to
the
one
two
one,
which
is
what
we're
using
today,
which
fixes
a
couple
of
balls.
We
just
have
to
keep
an
eye
out
there
for
a
while
or
every
once
a
while
to
see
if
there's
a
new
release.
If
so,
we
go
rebuild
our
copy,
so
that's
I
mean
that's
principally
wise.
A
A
I
mean
that
that
is
actually
so.
One
of
the
questions
that
I've
sort
of
had
a
little
bit
is
right.
Now,
the
pre-flight
repo
is
in
my
my
area,
my
my
personal
area.
Do
we
think
I'm
should
we
think
of
moving
that
to
one
of
the
other
build
areas?
Just
so
it's
more
visible,
I
don't
know,
maybe
not
initially,
maybe
it's
okay,
just
where
it
is,
but
I
think
eventually,
maybe
we
need
to
put
it
where
it's
a
little
bit
more
supported.
A
A
B
But
speaking
of
rebuilding
container,
that's
where
I
see
that
having
a
common
template
that
everybody's
going
to
inherit
is
going
to
be
even
better
because
as
soon
as
we
need
to
update
the
container,
we
don't
have
to
go
and
chase
it
across
the
board
to
make
sure
that
the
fight
is
working
properly.
So
I'm.
A
The
the
template
in
the
preflight
repo-
so
it's
there
it's
you
know
everything
all
the
collaterals
right
there
and
then
we
just
reference
it
from
that
repo.
I
still
like
that
idea.
D
A
A
B
B
A
So
that
that
will
be,
that
will
be
the
plan.
Let's
I'm
gonna,
I'm
gonna
keep
probably
just
saying.
I'm
gonna
go
to
turn
my
video
off
for
a
second.
While
I
get
my
contact
out.
It
just
keeps
poking
me
in
the
eye.
A
So
yeah
I'll
just
put
that
to
the
side
for
the
moment
and
put
it
in
later,
but
for
now
I'll
keep
pushing
the
cmg
mr
forward
to
get
the
certification
in
there
we'll
start
making
within
the
next
week
or
so
get
the
preflight
repo
moved
over
to
distribution
and
start
getting
the
template
in
there
and
then,
once
all
those
pieces
are
in
place,
we
can
go
back
and
just
put
a
quick,
mr
together
to
move
the
template
into
cng
and
then
that's
available
for
you
on
in
the
operator
dimitro
and
meanwhile
we
can
start
having
discussions
with
runner
team
as
to
how
do
they
do
it
and
see
if
what
we've
got
satisfies
them
or
if
we
can
help
them
with
our
template.
A
Yeah
exactly
so
all
right
good.
I
think
I
think
it's
been
a
good
demo.
Then
it
looks
like
we're
about
time
anyways.