►
From YouTube: CDEvents Working Group (EMEA/APAC) - Aug 28, 2023
Description
For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/
A
A
So
yes,
yes
before
we
start
this
afternoon
or
later
today,
I
would
say
there
will
be
a
presentation
on
the
association
proposal
by
jalander
in
the
sick
events
meeting
later
on
today.
So
that's
coming
up,
but.
B
Yeah
I
can
show
you
something
that
I've
been
working
on
sure
we're
just
about
to
get
to
the
directions
now.
Actually
so
this
is
just
one
experimentation
I
did
this
week,
so
I
instead
sorry
I'm
using
the
City
events
for
incident
reported
with
respect
and
we
set
it
up.
We
were
testing
for
we
can
have
a
GitHub
issue
or
we
can
have
a
Giro
issue
as
well.
B
We
can
have
a
juror
support
desk
issue
so
essentially
when
on
the
support
for
queue
when
an
issue
gets
created
instead
of
having
you
can
have
an
atlassian
one
as
well.
So
you
can
have
like
a
lesson:
Json
format,
but
I
made
a
custom,
CD
event
spec
s.
So
using
this
custom
data
I
can,
then
you
know
form
a
payload
that
that
adheres
to
the
incident
reporters
and
then
I
can
have
some
custom
database
in
it
as
well.
B
That's
working
pretty
nicely,
so
that's
one
thing
we're
doing
to
try
and
standardize
on
different
issues
being
raised.
Like
you
know,
if
it's
a
customer
that
has
a
problem
as
well
and
then
I
can
also
see
it,
I'm
also
firing
that
as
a
metric
and
Prometheus
as
well,
but
and
then
storing
all
the
events
in
mongodb.
B
Just
for
one
day,
when
I
have
more
use
cases,
I
can
pull
them
back
and
I
might
be
able
to
do
different
reporting
or
just
like
audit
log
or
something
like
this
I,
don't
even
know
why
I'm
doing
it,
I
just
know
one
day,
I
might
have
a
use
case
for
it,
so
I'm
storing
them
in
and
then
firing
the
makeup
as
well
right.
A
B
A
B
So
what
I'm
boredom
sort
of
sort
of
spiking
at
the
moment
is
I
want
to
sort
of
follow
the
the
life
cycle
of
that
issue.
So
the
common
idea
is
the
jira.
A
B
So
that
that's
in
everything,
that's
in
our
source
control,
so
I
want
to
see,
like
you
know,
when
a
incident
has
been
made
following
the
process
of
you
know
getting
that
back
into
production
fixed
as
well,
so
I'm
just
experimenting
of
following
that
idea
across
for
the
different
events
that
will
happen
within
that
life
cycle.
A
So
why
is
why
I'm
asking
is
that,
when
we
Define
that
event
type
we're
quite
a
discussion
on
whether
we
should
call
this
an
issue,
a
defined
event
or
a
issue
reported
events,
and
we
we
concluded
that
probably
it
would
be
good
to
be
able
to
State
well
more
in
line
with
with
the
Dora
thinking.
Maybe
that
this
is.
A
A
B
This,
if
we
look
at
if
I
say
the
events.
B
B
A
B
So
my
one
in
this
case
is
reported,
but
it
would
be
cool
as
well
to
run
some
test
cases
and
then,
if
we
see
that
something's
detected
instance
Detective,
then
that
could
make
the
incident
instead
of
reported.
Sorry
and
then
goes
through
development
so
but
could
possibly
be
yeah
change,
managed
yeah.
So
just
seeing
the
seeing
how
that's
going
to
work,
I.
A
B
Really
know
don't
just
experimenting
like
I,
don't
know
how
it's
all
going
to
work,
but
it's
pretty
cool
you
know,
even
just
having
this
to
be
able
to
use
the
same.
Payload
for
different
systems
is
even
a
alone
is
pretty
valuable.
B
B
I'm
actually
I'm
actually
working
with
my
team
or
not.
Now,
because
everybody
went
a
bit
quiet
for
a
while,
so
yeah
yeah
I
decided
to
just
just
get
it
I
just
wanted
V1.
Even
if
it's
not
that
good.
Just
you
know,
if
I
have
a
V1,
then
at
least
we
can
use
it,
and
then
we
can
just
keep
making
it
better.
More
people
can
contribute
to
it,
yeah,
so
just
trying
to
get
a
V1
working
and
then
should
be
pretty
cool.
A
So
I'm
not
sure
Adam.
If
you
have
any
input,
though,
but
I'm
pretty
sure
that
jalander
has
also
done
some
GitHub
actions
connection
with
with
cereal
events
yep,
he
has
as
far
as
I
know.
Yeah
do
you
know
any
details
there
or
I,
wouldn't
know
too
much
about
that.
Sorry,
okay,
yeah,
because
I
I
remember
him
saying
something
that
we
had
some
kind
of
remote.
What
is
it
called
now
remote
worker?
What
is
it
called?
Self-Hosted,
Runner,
self-hosted,
Runner
right,
yeah,
yeah,
so
yeah,
it's
an
external
run
of
them.
A
That
is
called
from
the
GitHub
action
through
a
web
vocabulary.
This
is
that
how
it's
done
radar.
B
Yeah
for
for
the
Southwest
Runners,
if
you
run
them
on
kubernetes,
it
has
a
controller
and
that
will
that
will
keep
talking
to
get
happens.
They
do
have
a
job,
do
have
a
job.
Do
you
have
a
job?
You
just
have
a
job,
and
then
that
will
run
inside
the
cluster,
which
is
pretty
cool.
We
use
that
as
well,
so
you
can
use
it.
B
Well,
the
the
controller
does
the
the
the
it's
called
GitHub
actions:
Runner
gitlab!
Does
it
as
well,
but
very,
very
cool
yeah
we're
going
to
do
so.
We
run
our
runners
in
kubernetes
as
well,
but
it
opens
up
doors
because
you
can
send
the
internal
requests
because
your
you
know
the
Pod
is
running
in
your
cluster,
so
you
can
interact
with
internal
services.
So
we
we
want
to
play
around
with
it
as
well
and
start
seeing
some
City
events
to
other
tools.
A
B
Can
have
it,
you
can
have
a
listener
as
well
to
push
you
can
you
can
pull
a
push?
It
depends
how
you
want
to
set
it
up.
We
we
just,
we
just
keep
hitting
it.
A
A
A
B
Start
that
work
up
again
and
yeah
and
right
in
the
group
Channel
when,
when
it's.
A
A
You
can
register
that
if
you
wanted
as
well,
can
I
do
afterwards,
good
yeah,
so
that's
that's
cool.
So
what
about
what
Yolanda
was
doing
there
I
believe
is
in
the
visualization
demo
he
would
have
elected
today.
He
would
also
I
guess,
show
this
guitar
actions
connection
with
the
cell
phone
said
Runner
that
is
prepared.
B
B
A
Know
that
there
was
some
issue
with
the
automation
there,
because
I
think
there
will
be
they
are
supposed
to
be
automatically
uploaded
there
by
some
script.
That
Andrea
takes
care
of
now.
Is
it's
not
clear
this
week
and
I
will
come
back
next
week,
I
believe,
but
he
had
some
hiccups
sometime
there,
so
my
automation
might
not
work,
but
in
a
week
or
so
I
think
it
should
be.
There.
B
Yep,
no
just
write
a
note
here
as
well.
The
test
Cube
folks
wanna
they
keen
on
the
get
direction
as
well
so
yeah.
A
B
And
there's
an
open
there's,
open,
Telemetry
and
CD
events.
Is
there.
A
I'm
not
sure
if
there
is
much
progress.
Actually,
this
issue
here
is
hasn't
been
updated
for
quite
a
long
time
now,
so
I
think
it
was
maybe
Andrea
who
added
it
to
the
coming
meeting
proposal
there.
So
it
seems
to
adoptom
is
aiming
to
come
to
our
meeting
next
week.
That
is
the
City
events
meeting
next
Tuesday
and
I
believe,
which
you
might
not
be
able
to
join
them.
A
B
B
A
Yeah
I
think
one
use
case
would
be
that
if
they
have
a
CO
pipeline
or
cocd
pipeline,
that
somewhere
fails
and
you
want
to
drill
down
to
see
what
server,
what
was
it
run
on
and
what
was
the
the
performance
of
the
load
on
that
server.
For
example,
then
they
could
be
open,
Telemetry
metrics,
showing
such
issues,
and
then
it
will
be
good
to
correlate
them.
A
A
B
Think
personally,
I
think
the
the
parent
span
of
the
open,
Telemetry
Trace,
would
be
beneficial
in
the
city
event
because
you
could
go
Port
out
because
it's
hard
to
yeah,
because
you
can
correlate
metrics
to
logs
and
open
Telemetry.
So
that
part's
done.
So,
if
you
can.
A
Yeah,
so
that.
B
A
I
think
that's
quite
in
line
with
I
mean
there
is
a
proposal
also
on
this
connected
events
which
Ben
from
Apple
is
working
on
and
that
the
focus
in
that
pull
request.
That
is
there
right
now.
B
I
think
another
use
case
as
well
is
you
can
persist
the
open
the
span
start
time
as
the
span
progresses?
You
know
the
child's
friends
Etc.
You
can
persist
the
timestamps
in
the
CD
event,
whereas
some
tools
you
couldn't
because,
if
you're,
if
you're
taking
a
while
to
do
something
right.
So
let's
say
for
an
example
that
you
kick
off
a
test.
B
Your
test
tool
runs.
The
city
event
persists
the
time
that
that
test
started
or
that
event
transmitted.
B
And
then
once
you
know,
obviously
you
could
go
through
multiple
tools
or
places
that
don't
necessarily
propagate
the
the
timestamp
of
when
the
span
started,
because
it's
you,
you
need
the
start
time
minus
end
time
for
duration,
etc.
For
the
span.
So
I
could
see
a
use
case
of
CD
events
persisting
the
timestamps
to
make
the
spans,
whereas
in
some
tools
they
wouldn't
be
able
to
or
processes,
does
that
make
sense.
A
Capturing
metrics
from
a
running
system
like
open,
Telemetry,
that's
more:
how
does
a
process
execute
something
might
not
be
explicitly
connected
to
a
pipeline
in
the
City
events
notifies
about
so
I
mean
they
could
be
potentially
multiple
pipeline
runs
contained
within
the
same
process
that
one
Series
in
Spanish
is
monitoring
and
the
other
way
around.
Of
course,
as
well.
One
one
pipeline
could
trigger
multiple
activities
which
are
in
different
cities.
A
Sorry
different,
open,
strumentaries
found
sorry
in
the
tracing
context,
so
there
won't
be
a
one-to-one
mapping,
yeah
I'm,
pretty
sure,
and
that's
also
something
to
look
into
more
where
what
is
the
most
valuable
thing
and
I
mean
what
use
case
are
we
trying
to
solve?
This
is
the
one
that
I
talked
about
before
that.
Will
we
want
to
more
establish
it?
Why
did
this
build
or
test
fail?
Yeah?
It
was
because
this
server
was
loaded
or
something,
but
that's
one
use
case,
and
it
could
be
others
saying
like
well
on
this
server.
A
B
A
One
thing
is
obviously
that
when
discussing
the
scope
of
the
city
and
spec
itself,
we
we
need
to
know
so
that
we
don't
go
into
the
open,
Telemetry
area
or
scope.
If
we
shouldn't
and
we
could,
you
should
make
sure
that
we
notify
about
the
right
things
in
the
right
way,
using
open
Telemetry
for
the
things
that
they
are
that's
best
suited
for
and
open
give
sincerity.
That's
one
of
the
things
that
it's
suited
for
and
I
think
the
the
limit
there,
the
boundary
isn't
really
set
in
stone.
A
It
so
that
is
something
that
will,
of
course,
continue
for
quite
some
time,
but
will
be
interesting
next
to
it
there
for
the
discussion
on
the
topic:
yeah,
yeah,
okay,
okay,.
B
No,
that's
I
just
thought.
I'd
come
check
here
tonight.
I've
been
at
one
of
these
meetings
for
a
while,
because
I've
been
pretty
busy
with
the
kubernetes
release,
but
that's
all
finished
now
so
yeah.
This
is
like
you
know
one
of
my
bigger
priorities
now.
So
you
see
me
again.
B
Yes,
I
yeah
I
talked
to
them
once
a
week
as
well.
They,
they
I'm,
not
sure
if
they've
been
working
on
City
events,
but
they
can
definitely
handle
them.
I
mean
their
products
very
cool.
They
have
different
sort
of
entry
points
for
they're.
B
Cloud
events
they're
very
generic,
it's
very
cool,
but
we
were
on
the
talking
on
the
captain
staring
committee
last
week
or
the
week
before,
and
we
brought
up
City
events
again
so
I'm
not
sure
if
you
know,
but
there's
a
V1
and
a
V2,
so
V1
Captain
was
very
much
event
based
and
V2
is
more
Cloud
native.
It's
it!
It's
really.
You
know
using
the
kubernetes
scheduler
and
it's
it.
It
didn't
really
have
the
capability
of
events,
but
we're
getting
a
lot
of
feedback
that
people
want
that
again.
B
So
we're
starting
to
talk
about
the
adoption
of
CD
events
as
well.
So
that's
in
discussion
at
the
moment,
which
is
pretty
cool,
yeah.
B
We're
coming
yeah,
it
was,
you
know
it.
It
was
sort
of
a
donor,
Trace
decision,
but
they
probably
should
have
had
a
feature
parity
before
they
because
everybody
still
wants
their
events.
It's
still
a
very
cool
thing,
so,
okay
I
can
update
you
on
that
as
well.
B
A
It
was
before
the
summer,
okay,
yeah,
so
Adam
do
you
have
anything
you
want
to
bring
up
or
nope
I,
don't
have
anything
to
add
anyway.
Okay,
that's
all
right!