►
From YouTube: SIG Instrumentation 20220901
Description
SIG Instrumentation Bi-Weekly Meeting Sept 1st 2022
A
B
Yeah,
actually,
can
you
scroll
up
a
little
bit
yeah
sure,
okay,
great
I,
just
wanted
to
update
people.
We
have
new
dates.
126
dates,
I
have
updated
the
doc,
so
you
can
see
when
everything
is.
Everything
is
supposed
to
happen.
So
a
release
cycle
starts
in
like
four
days
and
enhancement.
Freeze
is
going
to
be
in
a
little
a
little
over
a
month
could
freeze
in
about
two
months,
so
that
said,
I
have
basically
two
caps
I'm
working
on
the
first
one
is
right.
B
B
B
B
A
B
And
then
there's
a
metrics
framework
cap,
basically
I,
have
I
need
to
update
the
cap.
I
will
do
that
by
next
meeting
and
I
will
send
it
out
to
you
guys
for
review
before
I
will
try
to
do
it
next
week
before
our
triage
meeting
and
I
will
share
it
during
the
triage
meeting,
but
I
have
basically
prototyped
the
stability
framework.
Stuff
I
have
the
static
analysis
working
on
like
a
far
broader
range
of
types,
so
that
basically
it
passes
almost
everything
now.
B
B
A
Sense,
yeah,
okay,
opinions,
true
and
then
I
guess
I
can
go
to
my
topic,
and
so
next
week
was
about,
like
the
event,
appear
in
regulation
that,
like
we
have
been
like
that
she's
around,
like
four
like
ages.
Now
so
essentially
like
we
created
this
new
events,
API,
that
is,
events
Slash
V1
like
even
so
I
use
V1.
A
That's
super
CD,
core
V1
API
and
the
issue
that,
although
the
cap
has
graduated
to
GA
and
the
API
is
stable
now
it
hasn't
been
spread
to
all
the
components
in
kubernetes
like
we
only
have
it
in
the
scheduler,
I
think
and
yeah.
Quite
recently,
like
I've
been
having
a
lot
of
issues
where
there
are
operators
that
kind
of
spam
GPA
server
is
like
the
same
event
over
and
over,
and
the
new
API
is
supposed
to
fix
that
issue,
because
it's
like
similar
events
will
be
merged
together.
A
Apparently
so
I
discussed
with
him
like
before,
and
and
we
were
able
to
find
like
a
couple
of
gaps
in
the
implementation
that
we
might
want
to
cover
before
doing
the
migration
like
the
the
first
one,
is
that
between
the
previous
API
in
the
new
one,
there
was
a
change
in
kind
of
the
user.
Experience
of
the
events
like
before,
since
when
an
event
was
created,
it
was
like,
was
emitted.
A
A
We
thought
about,
maybe
adding
a
backup
mechanism
that
essentially
say
that.
Well,
if
we
see
the
same
even
twice,
then
we
will
like
updates,
try
to
one
DBA
server.
If
it's
three
times
then
after
one
minute,
we
will
update
it
on
the
API
server
and
so
on
and
so
on
in
the
back
of
planner,
so
that
at
least
we
see
repetitions-
and
we
don't
spam
gpa7-
that
much
because
well
when
GPA
was
implemented
like
having
the
update
like
once,
every
30
minutes
might
be
a
bit
too
much
like.
A
It
was
making
the
UPSR
crash,
so
they
added
some
kind
of
rate
limiter
to
the
implementation,
and
today
I
don't
know
like
I,
know
one
type
of
30
meter,
but
I
don't
know
if
there
is
more
than
just
that,
because
they
have
a
queue
in
the
first
implementation
and
when
the
queue
has
more
than
a
thousand
objects
inside
of
it,
it
will
just
drop
the
new
incoming
ones
and
I,
don't
like
that's.
Definitely
not
in
the
new
implementation
that
there
is
something
else.
Then
maybe
it's
good
enough.
A
The
migration
and
please
hold
the
components
there
might
be
other
issues
that
Jordan
pointed
to
me,
but
I
haven't
yet
had
the
time
to
look
at
them,
but
yeah
at
least
like
we
have
a
pass.
I
would
say
now
it's
more
like
how
do
like
I'm
planning
to
do
a
cap
to
detail
like
that
fast
forward,
and
then
maybe
we
can
plan
a
5G,
actual
migration.