►
From YouTube: 2023-05-09 Product Analytics Group Sync
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
Hello,
everyone
and
welcome
to
the
May
9th
product
analytics
group.
Sync,
we
will
go
ahead
and
start
things
off
with
the
classic
billboard
review.
So
I
will
share
my
screen
and
get
started
over
there.
Hopefully,
everyone
can
see
the
billboard
seeing
some
nodding.
So
let's
go
ahead
and
get
started
a
couple.
Few
issues
actually
quite
a
lot
of
issues
in
open
still,
but
not
a
lot.
That's
assigned
to
people
that
I
mean
these
should
really
ideally
need.
A
Workflow
labels,
there's
34
bug
which
we
can
check
and
I
suppose
next
time,
I,
don't
think,
there's
anything!
It's
a
report
on
here,
no
so
quickly
browse
through
here.
A
Hopefully
we
can
start
kind
of
pulling
some
of
these
Jitsu
cleanup
issues
soon,
because
I
think
we're
about
close
to
wrapping
up
the
snow
cloud
stuff
yeah.
So
let's
just
go
ahead
and
get
into
issues
that
are
in
workflow
States.
So
upgrading
the
live
clusters
with
snowplow
and
Kafka
analytics
stack,
so
this
was
kind
of
blocked.
This
is
blocked
on
a
change
to
the
cube
production
deployments,
which
I
mentioned
yesterday.
I
got
some
good
progress
on
so
I'm
in
the
final
stages
of
wrapping.
A
A
A
As
mentioned
previously,
there
are
different
usages
of
quick
house
currently
in
gitlab
and
one
of
the
observability
one
of
the
other
groups.
This
actually
has
their
own
operator
to
to
manage
and
scale
clickhouse
cluster.
So
that's
something
we
may
want
to
look
into
we'll
have
to
determine
how
we
want
to
interact
with
that
and
what
our
options
are.
Of
course
we
have
to
be
mindful
of
how
we
can
package
this
for
self-managed,
as
well
as
how
we
might
handle
this
for.com.
But
it's
a
little
bit
more
context
around.
A
B
So
this
has
been
fun
so
Rob
just
to
bring
up
speed.
Rob
did
some
work
to
introduce
something
called
sealed
Secrets,
which
essentially
allows
us
to
encrypt
secret
values
in
our
Helm
charts.
B
It
turns
out
that
gitlab
uses
hashicor
Vault
to
store
secrets
and
that
we
can
use
that
so
there's
an
open
access
request.
I
know,
Dennis,
has
already
chased
on
to
get
us
access
to
that
essentially.
But
in
the
meantime
today
me
and
Rob
did
sort
of
a
mammoth
pairing
session
where
we
proved
the
concepts,
essentially
that
we
can
store
secrets
in
an
instance
of
azure
football
and
inject
them
at
runtime
into
our
own
kubernetes
pods.
B
There
is
a
90
minute,
rambling
screencast,
that
we
recorded
earlier
that
I
need
to
go
through
and
hide
all
the
secrets
from
which,
given
the
subject
matter
may
take
a
while
but
yeah
it's
interesting.
We
learned
a
lot.
We've
proven
it
can
be
done
once
a
success
request
is
approved.
Hopefully
we
can.
We
can
migrate
all
of
our
horrible
plain
tech
secrets
over
to
wonderfully
encrypted
secure
ones.
A
But
it
makes
sense,
perhaps
to
upload
that
just
to
Google
Drive,
knowing
that
or
are
we
actually
exposing,
because
it's
not
anything
like
production
right,
so
so
to
save
these
kind
of
like
censoring
things.
A
Given
I
think
the
issue
itself
is
confidential,
but
I
mean
given
that
the
sensitivity
of
handling
Secrets
I
think
there
was
never
any
goal
to
oh
publicly
share
this.
This
information,
I
guess
the
issue
is
confidential,
but
if
it's
easier
just
to
upload
it
privately,
then
I
would
say.
Let's
do
that
just
so,
we
can
do
at
least
knowledge
right
across
the
group,
but.
A
Sounds
good
awesome,
then
we've
got
adding
new
visualizations
as
panels
to
the
dashboard,
which
is
assigns
he
yawned,
but
it's
blocked
by
the
ad
dashboard
functionality.
So
you
can
hopefully
check
on
that
later
in
the
Board
review.
We've
also
got
fetching
dashboard
configurations
from
the
back
end,
which
is
one
of
the
final
pieces
to
be
able
to
switch
over
to
snowplow,
so
to
speak
out,
all
the
snowplow
specific
schemas
and
the
retrieval
of
that
happening
in
the
back
end
now,
which
is
in
development.
So
hopefully
we
can
see
what's
going
on
here.
A
No
later
point
same
time:
we've
got
the
updating
product
analytics
back
in
first
I'll
put
a
bit
compatibility.
Excuse
me
Alan
anything
to
call
out
here.
C
That's
actually
in
review
just
waiting
on
maintain
a
review
to
merge.
It
I
think
awesome.
A
Cool
so
I
think
between
that
and
one
of
the
issues
I
will
get
to
the
issue.
So
add:
instrumentation,
details
to
general
settings
analytics
settings
Ellen
anything
to
call
out
here.
D
A
B
Yeah,
that's
good
sorry,
four,
open,
Mrs
or
tracking
various
usage
things
through
which
are
now
merged
and
I
assume
in
production,
if
not
far
away,
and
one
is
currently
in
product
intelligence
review.
A
Awesome
and
then
we've
got
the
sorry
intelligence
is
now
analytics
instrumentation.
Just
another
signal
boost
for
that.
A
That's
that
is
true
and
then
Ellen
you've
got
updating
product
analytics,
onboarding
dashboard
flow
to
reflect
the
updated
snowplow
initialization
process,
which
is
currently
even
marked
as
blocked,
but
I.
Don't
believe
that
is
the
case
anymore.
C
A
D
Right
this
one's
just
waiting
on
a
conflict
to
be
resolved
and
then
it's
going
to
be
going
to
be
emerged
very
close.
A
Awesome
and
then
we've
got
the
put
an
ad
dashboard
button
on
the
analytics
listing
page,
which
is
effectively
the
dashboard
designer
experience
at
least
adding
an
empty
dashboard.
So
it's
in
review
not
sure
if
there's
anything
else
to
call
out
here,
but
we'll
just
check
the
comments
just
to
double
check.
A
No,
so
that
we
can
go
ahead
and
get
to
verification
issues.
We've
got
updating
the
funnels
API
to
return,
still
block
compatible
SQL
and
allowing
users
to
connect
practical,
like
stack
at
the
project
level.
All
you
have
both
of
these
anything
to
call
out
here.
C
And
I
know
they're
just
waiting
on
turning
on
snowplow
support
in
production,
so
they
could
verify
that
they
work.
A
Perfect,
all
right
so
hopefully
I
can
get
that
sorted
and
then,
as
the
for
those
reviewing,
the
recording
I
can
scroll
got
a
couple
final
pieces
here.
As
far
as
updating
the
back
end,
the
making
sure
the
onboarding
flow
on
the
front
end
as
well
as
fetching
the
dashboard
configurations
which
are
the
final
pieces
for
the
snow
plow
switch
over.
A
At
that
point,
we
can
enable
the
feature
flag
and
we
can
start
instrumenting
and
sending
events
to
snowplow,
which
I
will
have
the
snowplow
enabled
in
the
Clusters,
and
then
we
should
be
able
to
once
the
snow
once
the
feature
flag
is
enabled
use
the
correct,
dashboards
and
schemas
to
actually
be
able
to
view
the
events
collected
from
snowplow
on
the
dashboards.
So
getting
really
close
there
cool.
A
Then
we
have
our
bookmark
for
the
last
issue
is
updating
the
audience.
Behavior
dashboards
assistance
than
we've
had
probably
looks
like
nine
or
so
issues.
So
we
fixed
transitioning
pods
with
persistent
volume
claims.
There's
a
PTO
Handover
issue.
I've
pushed
a
change
to
improve
the
credential
usage
in
clickhouse
and
Cube
still
needs
to
still
would
definitely
benefit
from
the
secret
management
stuff.
A
From
hashcorp
Vault,
but
we'll
we'll
iterate
on
that
and
then
Ellen
you've
got
the
next
two
issues
that
are
closed
in
terms
of
instrumentation
SDK
example:
code
being
invalid
of
deep
linking
girls,
those
top
pages
in
The
behavior
dashboard.
Anything
to
call
out
here.
D
The
only
one
is
that
that
SDK
example
code
now
uses
a
feature
flag
to
switch
into
ancient,
so
in
snowplow
compatible
code.
Awesome.
B
Yes,
one
of
the
engineers
from
appsec
raised
a
priority
One
bug
on
definitely
code
that
I
know
I've
written
where
it
was
possible
to
craft
a
dashboard
definition
that
was
possibly
would
enable
you
to
access
files
outside
of
your
own
repository
I
think
it
would
have
been
very
difficult
to
exploit,
but
an
absolutely
good
catch
from
absec
anyway,
so
easily
fixed
and
yeah.
Luckily
got
it
into
the
same
release
that
it
was
introduced.
So
there's
no
back
ports
or
anything
any
release.
Any
security
release
issues
we've
got
to
deal
with.
A
Let's
get
there
they're
catching
these
now,
when
we're
still
getting
out
to
customers,
but
yeah
yeah,
good
question
and
quick
fix
as
well.
Since
we
can
work
in
canonical
cool
and
then
we've
got
two
more
improvements
to
our
stack.
That
Helio
has
in
terms
of
improving
the
development
of
persistent
volume
claims
using
staple
sets,
as
well
as
improving
the
usage
of
environment
variables
by
pulling
them
into
deployment
yaml
files.
A
So
just
a
lot
of
quality
of
life
usage
there
and
then
allows
us
to,
for
click
house
scale
that
better
in
terms
of
having
multiple
replicas
set
up
so
that
wraps
up
our
Board
review.
I
will
make
a
bookmark
and
then
we
can
go
on
with
well
as
a
read-only
item.
So
I'll
just
pause
for
to
see
if
anyone
else
has
anything
but
I'd
like
to
verbalize
or
talk
about.
A
Very
cool
well,
overall,
I
think
we're
in
a
good
spot.
We
have
we're
very
close
to
the
snowplow
integration
being
complete
and
being
able
to
clean
up
a
lot
of
that
Jitsu
code.
So
we
don't
have
to
worry
about
those
dual
paths
and
hopefully
that'll
be
a
little
bit
easier
to
manage.
At
the
same
time,
a
lot
of
progress
being
made
on
the
plus
stack
management
side
of
things
so
cube
is
now
very
ready
for
production.
No
security
issues
well
less
likely
to
have
security
issues.
A
I
don't
want
to
make
a
statement
that
I'll
kick
myself
later
on
for
making,
but
it's
running
according
to
their
production
checklist.
So
that's
looking
good,
so
yeah
cluster
updates
will
go
out.
I'll
just
create
the
infrastructure
issues
to
get
those
connected,
and
my
goal
after
that
is
to
really
kick
off
reboot
the
production
Readiness
review
to
see
where
we
can
go
in
terms
of
getting
this
deployed
out
in
production,
so
yeah,
exciting
stuff.
In
the
meantime,
we'll
continue
iterating
on
what
we're
working
on.