►
From YouTube: 2023-06-13 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
Morning,
I
remember
shortcuts,
so
there's
Alan,
I'm
gonna
share
my
screen.
Hopefully
everyone
can
see
my
browser
and
zoom
is
doing
a
great
job,
always
blocking
everywhere.
I
need
to
click,
but
anyways
we'll
go
ahead
and
get
started
with
the
build
board.
So
yeah
logging,
this
one's,
been
here
for
a
while.
It's
actually
been
a
topic
that
Knowles
and
I
have
been
talking
about
in
terms
of
how
we
can
actually
make
sure
that
we're
securely
storing
them
so
for
context.
A
We
have
a
couple
options
in
terms
of
like:
could
we
potentially
use
encrypted
volumes
to
do
the
logging
there,
or
can
we
ship
them
to
an
elk
stack,
elasticsearch,
log,
stash
and
kibana,
which
we
currently
do
for
existing
Production
Services
and
then
go
that
way?
A
Now,
as
I
know,
you
were
talking
with
someone
on
on
in
your
organization
up
to
possible
options
there,
especially
with
what
we
can
do
with
kubernetes
clusters,
not
to
put
you
on
the
spot,
but
is
there
anything
that
you'd
want
to
call
out
as
far
as
our
options
with
vlogging,
or
should
we
just
kind
of
just
move
on
from
this
topic?.
B
As
far
as
our
options
for
logging,
specifically,
then
I
think
we
probably
want
to
have
all
of
that
data
for
the
application
and
the
and
the
components
services
that
we
want
to
have
logging
for
and
be
able
to
review
put
into
another
elk
ins
index
just
like
we
have
for
all
of
our
other
Production
Services.
So
as
far
as
whether
or
not
we
want
to
retain
any
kind
of
or
any
sort
of
like
customer
event,
interaction
logging
in
those
indices,
I,
don't
think
so.
I,
don't
think.
B
I
know
that
we
don't
have
an
extensive
amount
of
customer
data
in
our
production
logs
other
than
things
like
accounts
other
than
things
like
project
names
other
than
things
like
specific
actions
that
they've
taken
right
to
interact
with
our
API
for
tracing
purposes
and
diagnostic
purposes,
troubleshooting
purposes
and
so
on,
and
so
I
think
that,
after
your
application
wants
to
like
exclude
any
sort
of
specific
like
project
or
business
customer
data
from
those
logs-
and
that's
great-
so
you
know
I
think
that's
just
a
matter
of
being
on
top
of
it
being
aware
of
that
sort
of
limitation
and
that
that's
that's
pretty
much
I,
don't
see
any
blockers
to
getting
another
index
in
Elk.
A
Gotcha,
okay,
yeah
we
can,
we
can
better
Define
anything
we
really
want
to
get
out
of
the
logs
for
each
application.
There
was
a
there
was
a
discussion
at
one
point
about
how
potentially
using
logging
to
actually
reconcile
how
many
events
we're
collecting
per
customer
but
I
think
there's
a
way
to.
We
can
just
query
that
across
clickhouse
clusters
to
kind
of
do
that,
reconciling
for
for
billing
and
not
actually
use
logging
for
that
purpose.
B
So
absolutely
you
might
even
want
to
consider
having
some
kind
of
a
like
a
an
archive
database
that
reflects
clickhouse
to
a
certain
degree
of
synchronicity,
so
that
you'd
be
able
to
have
a
sort
of
a
read-only
access
from
billing
for
that
data,
and
that
way
it
wouldn't
interfere
with
production
operations.
That's
what
we
do
for
our
gitlab
main
database.
They
seem
to
enjoy
that
Arrangement
as
far
as
encrypted
volumes
go
data
encrypted
at
rest.
B
It
turns
out,
according
to
our
our
my
delivery
counterpart
engineer,
we
have
a
all
of
our
gke
clusters
are
all
deployed
with
that
have
any
kind
of
volumes
attached
to
them.
Those
volumes
are
always
encrypted
by
Google
by
default.
Okay,.
A
So
that
would
that
would
satisfy
that
that
Criterium
then
correct
cool
awesome,
then
I
will
go
ahead
and
kind
of
like
this.
That
issue
needs
a
lot
of
refreshing,
anyways,
but
I
think
there's
a
couple
of
different
kind
of
offshoot
issues
that
we
can
kind
of
create
from
there,
but
we're
obviously
able
to
kind
of
move
that
topic
along
a
little
bit
more
cool
we've
got
adding
error,
States
panels,
I
think
this
one's
still
and
ready
for
design
so
nothing
there.
Yet
unless.
D
It's
actually
in
design.
If
you
update
the
board,
you
might
see.
D
Whatever
Goods
we've
been
like
discussing
this
with
yeah
on
the
on
the
issue,
so
it's
it's
following
its
course.
We're
probably
need
a
bit
of
copy,
but
I
already
pinned
right
now
on
it.
Mm-Hmm.
A
Cool
up
next
we
have
handling
customer
data
and
customer
facing
releases
and
access
to
production
cluster
Readiness.
This
is
a
confidential
issue,
but
we
are
making
improvements
to
Pardon
our
our
infrastructure,
and
there
is
a
topic
that
we're
discussing
in
terms
of
the
proper
deployment
to
be
able
to
handle
red
data,
and
we
continue
to
make
progress
there
thanks
to
nulls
and
Max
for
to
make
for
making
some
changes
there,
but
I
think
we're
the
long
story
short
and
we'll
probably
cover
it
in
a
couple.
A
Other
issues
as
well
is
that
we're
getting
a
little
closer
there
so
really
just
kind
of
a
handful
of
remaining
topics
to
kind
of
close
out
there
and
then
as
another
kind
of
previously
discussion,
which
is
going
to
be
a
handoff
I.
Think
in
terms
of
getting
the
true
production
deployment
going,
but
I
think,
given
our
current
setup
with
our
chart
and
things
like
that,
it
shouldn't
be
too
much
of
a
a
it
shouldn't,
be
a
large
effort
to
get
that
sorted.
A
This
one's
a
16-1
planning
issue,
so
nothing
I,
don't
think
to
be
said.
There
believe,
there's
still
a
little
more
clarification
needed
for
this
issue
to
update
the
developer
documentation
so
just
to
clarify,
and
then
perhaps
it's
done.
Writing
in
the
issue
itself
is.
This.
A
Is
the
user-facing
documentation
once
you've
completed
the
onboarding
process
so
that
it
shows
you,
you
know
which
SDK
to
use
and
how
to
actually
extend
events,
and
things
like
that
and
so
that
previously
is
or
currently
is,
set
up
to
work
with
jitsu's
or
the
Jitsu
flavor
of
our
browser
SDK?
So
we
just
need
to
make
sure
that
that's
been
updated
to
give
correct
documentation
the
correct
way
to
instrument
applications
using
the
browser
SDK,
that's
specific
for
snowplow,
so
yeah
Improvement,
improved
management
of
environment.
A
Specific
configurations
is
a
confidential
issue,
but
I
believe
well.
I.
Don't
want
to
open
the
issue
and
kind
of
clarify
that
but
I
know
we've
been
making
some
improvements
there
in
terms
of
being
able
to
deploy
to
specific
environments
and
having
specific
configurations.
A
As
far
as
the
cluster
configuration
of
the
sub
configuration
a
bunch
of
times
but
I
promise,
it
means
something
improve.
Clickhouse
credential
strategy
is
also
something
that's
related
to
our
secrets:
management
that
we've
been
making
improvements
on
a
couple
of
kind
of
remaining
items
there
to
make
sure
that
all
services
that
are
involved
with
with
that
interact
with
the
clickhouse
service
are
set
up
to
the
as
secure
as
possible.
A
That
issue
may
need
a
little
bit
more
misogyny
based
on
the
progress
we've
made
so
far.
We
have
an
issue
for
notifying
users
of
existing
prod
product
analytics
projects
to
update
to
xenopause
SDK.
This
hasn't
happened
yet
because
we
need
to
do
the
production
snowplow
switchover,
which
should
hopefully
be
we
should
be
able
to
get
to
this
week.
A
A
This
is
just
an
issue
for
all
that
needs
to
be
done
for
the
snowplow
switchover
plan,
which
I
believe
at
this
point
is
just
really
updating
the
environments,
so
staging
has
already
been
done
as
I
announced
last
week.
Dev
is
blocked
currently
because
Dev
is
just
running
off
of
an
older
version
of
the
gitlab,
and
then
production
is
just
really
pending
what
we're
going
to
do
for
the
actual
production
cluster
environment.
But
theoretically
it's
ready
to
go
I'll
still
plan
to
have
a
script
ready
to
drop
the
Jitsu
related
databases.
A
If
we
need
to
at
least
just
to
be
able
to
say,
we
are
at
least
running
off
of
snowplow
across
all
environments,
and
at
least
we
can
move
forward
with
removing
all
the
magics
occurred,
all
right,
Auntie,
workflow
design.
We
have
right
prototypes,
write
a
covered
error
state
to
panels,
so
we
have
right
prototypes,
creating
a
custom
dashboard
for
product
analytics
Kevin.
Is
there
anything
you
wanted
to
call
out
here.
D
No,
like
interviews
are
in
progress,
it's
mainly
with
internal
stakeholders,
so
we
might
want
to
like
run
others
with
external,
but
that's
about
it.
Gotcha.
A
Cool
over
on
blocked,
we
have
removing
just
your
related
front-end
code
from
prime
time.
With
this
feature
is
the
science
at
all
and
who's
not
here,
but
is
also
blocked
anyways
by
the
production,
snowplow
switchover
so
we'll
go
ahead
and
jump
over
to
indev.
We've
got
saving
visualization
to
a
file.
You
know
and
anything
you'd
like
to
call
out
here.
E
Yeah,
that's
just
recently
became
blocked
by
the
work
that
was
done
by
Alvin
to
make
the
visualization
designer
compatible
because
with
with
snowplow,
because
that's
the
first
thing
that
we'll
need
to
actually
save
the
visualizations
is
have
the
designer
be
fully
functional.
F
A
F
Any
update
your
point
down
please.
A
Create
a
follow-up
issue
for
the
proposed
navigation.
It's
been
some
discussion
and
I
think
we
can
just
kind
of
talk,
so
it's
in
progress,
connect,
Dev
and
staging
as
I
mentioned
before.
Staging
has
been
completed,
but
Dev
is
blocked
right
now.
So
this
will
just
kind
of
remain
here
until
we're
able
to
resolve
that
setting
up
Ingress
I
believe
is
specific
to
the
sorry
clickhouse
operator
work.
Is
there
anything
you'd
like
to
call
it.
C
A
C
Yeah
actually
I
think
we
can
put
it
to
block
because
yeah
yeah,
basically
they
have
like
their
debuted
ones,
but
we
don't
make
the
maintain
a
review
until
this.
F
Yeah
got
it
cool
awesome.
Let's.
A
Watch
later,
updating
documentation
to
focus
around
shared,
analytics
dashboards,
rob
you
have
this
issue
anything
calling
or
Notting
here.
G
Did
a
absolutely
amazing
review
of
this
yesterday?
Thank
you,
Jan,
and
it's
now
just
working
on
Lorena
to
have
some
time
to
do
an
additional
review,
because
there
were
so
many
edits
made
to
my
original
Mr
as
a
result
of
Beyond's
review.
So
once
that's
on
to
Lorraine
is
Happy,
and
then
he
changes
has
been
made
that
she
requests
and
we
should
be
good
to
go
great.
A
Cool
and
Ellen's
got
an
admin
setting
change
to
give
some
better
treatment
of
some
of
the
fields
that
are
containing
a
bit
more
sensitive
information.
We
encrypt
those
in
the
back
end,
but
we
wanted
to
train
change
the
visual
treatment
on
the
front
end.
A
So
it
looks
like
that's
another
view
as
well,
so
it's
confidential,
we'll
just
move
on
from
that
anyways
Alan
you've
got
standardizing
dashboard
date,
filter
presets
to
be
in
the
UTC
time
zone.
Anything
worth
noting
here.
H
No,
that
includes
the
functional
changes,
but
also
think
the
Gian,
the
there's
a
little
bit
of
ux,
noting
to
the
user
that
these
are
all
UTC
times.
So
he
added
that
and
processor
review.
So.
E
Yeah
we
we
have
the
new
icon
with
the
tooltip,
so
I
think,
since
this
just
needs
ux
review
and
then
we
can
prostitute
the
maintainer
all
right.
A
Sounds
good
all
right,
then,
we've
got
upgrading
live
clusters
with
snowplow
and
Kafka
analytics
stack.
This
is
actually
done,
they're
all
snowplow
ready.
We
just
have
to
kind
of
flip
the
switch
so
just
again
for
context.
A
Production
is
a
little
bit
tricky
because
it
already
has
existing
Jitsu
databases
for
the
snowplow
setup
to
actually
complete.
It
needs
to
create
new
ones.
So,
depending
on
whether
we,
what
will
likely
happen
is
we
will
use
the
existing
production
Buster,
that's
being
used
for
internal
projects,
at
least
and
flip
that
on
so
we'll
likely
move
forward
with
leading
all
the
Jitsu
databases,
switching
on
snowplow
support
and
reconfiguring
all
the
projects
to
use
snowplow
until
we
figure
out
what
we
want
to
do
to
actually
handle
customer
data.
A
But
that
said,
all
the
Clusters
are
updated
and
they're
ready
to
go,
so
we
just
have
to
connect
them
to.com
and
update
to
that
and
settings.
A
Cool
final
issues,
update
in
review,
that
is,
is
updating,
user-facing
project
analytics
docs
for
the
first
external
user,
face
learning
how
you
have
this
issues
or
anything
you
wanted
to
call
out
here.
F
So
I
still
I'm
still
catching
up
a
bit
after
holiday,
but
yeah
no
problem.
A
Great,
so
we've
got
two
issues
related
to
bring
in
your
own
cluster
and
connecting
them
at
the
project
level,
which
Alan
is
there
anything
you
want
to
call
it
here.
I
know
you
kind
of
starting
it
familiar
with
testing
this
out.
H
Yeah
I'm
verifying
it
right
now.
It's
still
in
progress.
A
No
problem
awesome
I,
get
to
see
that
and
thanks
to
everyone
for
your
feedback
on
being
able
to
verify
your
changes,
so
we
were
able
to
kind
of
reduce
the
number
of
issues
that
we're
waiting
to
be
verified.
A
All
right,
then
we
had
I
think
an
issue.
Maybe
I've
passed
over
it
I,
don't
think
I
have
first
onboarding
user
onboarding
issue
template
so
I'm
just
going
to
really
try
to
pretend
that
I'm
counting
all
these
that
quickly
but
I
think
we
have
about
seven
18
issues
closed.
So
this
will
be
a
fun
time.
A
A
There
we
go
cool,
we've
got
two
issues
from
Elwyn
that
were
already
closed
for
updating
the
product,
analysts
on
board,
onboarding
dashboard
flow
to
reflect
the
updated
snowplow
initialization
process
and
just
making
that
snowplow
compatible.
We've
also
ignored
the
Jitsu
q
and
snow
plus
support
is
enabled,
since
we
should
no
longer
respect
that,
instead
of
like
the
I
believe
the
or
condition
there
Rob
you've
got
the
next
issue.
That's
closed.
Anything
worth
noting.
G
The
we
had
to
close
this
out
as
a
won't
fix
because
the
flag
it
depends
on
is
Project
dependent
and
we
have
the
admin
settings
which
obviously
aren't
project
dependent.
So
we
can't
actually
hide
the
different
settings
in
both
scenarios,
which
would
be
super
confusing
yeah.
So
because
we're
going
to
be
swapping
up
to
snow
plow
pretty
soon
with
the
next
Milestone
or
two,
we
just
opted
that
it'd
be
too
much
work
to
try
and
get
around
that.
You
know
ux
UI
perspective
and.
A
And
soon
enough,
we'll
just
delete
it
entirely,
so
yeah,
exactly
yeah,
okay,
great
Ellen's,
also
got
an
issue
for
resolving
the
single
stat,
visualization
or
yeah
almost
said
widget.
That's
cool
shows
no
content
when
Q
returns
an
empty
array
so
great
that
we
kind
of
fixed
that
up
Alan
you've
got
the
next
three
issues
anything
worth
calling
out
for
bees.
Now
that
we've
closed
them
out.
H
No
once
once
no
plow
was
enabled
it
was
a
simple
thing
to
go
through
and
make
sure
it
was
actually
in
snow,
plow
mode
and
snow
plow,
except
for
the
input.
It's
all
good.
Nice.
A
All
Right
Moving,
On
Up
we've
got
two
front-end
issues:
Jan
Rob
is
it's
not
the
other
one,
but
anything
to
call
out
for
this
issue
here
for
events
over
time
showing
the
page
we
count.
E
Yeah
I
think
that
was
just
a
small.
Let
me
call
it
like
a
visualization
query
fix
so
yeah
just
wasn't
displaying
correctly.
Now
it
is.
E
Just
thanks
to
Rob
for
like
also
investigating
this
and
the
helping
me
out
and
yeah.
It
took
two
minds
to
figure
this
one
out,
because
I
think
it's
we've
been
trying
to
figure
this
out
for,
like
I,
don't
know
how
long
it
is
apparently
quite
a
common
issue
right
across
the
stack
or
right
across
the
application.
It's
been
showing
up
because
we've
turned
on
webpack
runtime
errors
that
show
up
on
a
pop-up
which
is
good
because
we
shouldn't
we
should
be
avoiding
generating
those
errors
but
yeah.
E
It's
now
fixed
Happy
Days
and
let
me
know
if
it
ever
shows
up
again.
A
Okay,
I
will
be
happy
to
not
have
to
close
that
out
now
allele,
you
have
an
issue
for
installing
a
clickhouse
operator
as
a
home
dependency.
Anything
you
wanted
to
note
here.
C
Not
not
really
yeah
I
mean
it's
also
related
to
the
clinicals
operator,
actually
like
I,
just
maybe
I
can
hit
it.
So,
instead
of
adding
a
hand,
dependency
actually
I
use
a
different
approach,
so
we
install
it
directly
to
the
cluster,
but
there
are
instructions
in
the
report.
A
Got
it
perfect
excited
to
check
that
out,
we've
got
tracking
aggregated
unique
users,
viewing
a
product
analytics
dashboard
max
anything
worth
noting
here.
I
Only
it's
merged
and
I
assume
will
take
a
little
bit
of
time
for
it.
Filter,
Solutions,
but
I
think
Jenkins
has
already
added
the
graph
that
it
will
generate
two
assessments,
so
we
should
start
seeing
data.
A
A
Up
and
then
updating
users
configurations
confidential,
but
holy
anything
new
is
worth
noting
here.
C
Trying
to
understand
yeah,
okay
yeah,
so
this
is
basically
like
kind
of
copying.
The
configuration
that
we
use
in
analytics,
Tech
and
in
the
separate
cluster
so
got
it.
A
Max
you've
done
some
more
care
to
Carry
On
from
what
Rob
is
doing
for
improving
Secrets
management
stack.
Anything
you
wanted
to
call
out
for
the
group.
I
Other
than
we're
kind
of
there.
Finally,
so
we
now
have
the
ability
to
store
Secrets,
not
in
plain
text
in
inside
the
infra
managed
vault,
which
is
very
exciting.
H
I
just
say
that
adds
a
lot
of
fields
from
the
various
browser
context,
plugins
that
we
have
enabled
on
the
snowplow
side
to
our
analytics,
no
plow
schema.
So
there
should
be
several
Fields
added
with
that.
One
muscle.
A
That'll
lend
us
off
to
being
available
on
the
data
Explorer
as
well,
so
that
would
be
really
cool.
Owen's
done
work,
as
we've
noted
previous
issues.
If
we're
ensuring
the
visualization
designer
is
notepad
compatible
and
then
allele,
you
have
the
next
two
issues
as
well.
Anything
worth
calling
out.
C
E
A
Great
yeah
I
think
that
was
about
18
issues
that
have
been
closer.
A
lot
of
good
progress
being
made
so
really
really
cool
to
see
that
and
We've
we've
effectively
used
the
entire
most
of
the
duration
of
this
meeting
to
do
the
Billboards,
which
changes
there's
a
lot
of
there's
a
lot
of
stuff
being
completed
and
a
lot
of
stuff
being
done.
So
we're
really
happy
to
see
that
James,
you
I
think
yeah.
This
is
all
just
that's
here,
James
you!
You
have
a
a
point
here
on
the
agenda.
A
J
I
wanted
to
share
with
the
team
I
drafted
an
MR
to
rename
the
group
and
the
category
and
add
a
category
to
better
split
visualization
and
data
storage
and
management.
A
little
bit.
There's
just
been
a
lot
of
confusion
when
I
talked
to
folks
about
what
product
analytics
is
and
what
we
do
so
open.
For
you
know,
review
suggestion,
edits,
feel
free
to
jump
in
there
and
Dennis
I
tried
to
answer
some
of
your
questions
to
clarify
things
that
I
didn't
get
in
my
initial
buy
statement.
F
A
I
understand
the
need
for
for
more
clarity
around
our
names.
I'm
just
I've
got
concerns
I.
Think
product
analytics
itself
is
as
as
I
mentioned.
This
is
an
industry
term,
so
I
just
wonder
if
that
will
cause
confusion
externally,
but
we
definitely
have
that
conversation
async
great
thanks
thanks
for
calling
that
out
and
announcing
that,
though
yeah
and
you
have
a
point.
E
Yes,
I
just
okay,
so
first
of
all,
I
want
to
say
great
work
like
Dennis.
Let's
just
mentioned
a
lot
of
work.
We
did
in
the
past
week
pretty
much
all
of
the
non-ops
or
no
definitely
all
of
the
non-ops
related
issues.
E
As
far
as
I
can
see
in
the
product
analytics
now
plus
switch
over
plan
like
all
the
stuff
we
identified
is
done
so
with
that
I
just
wanted
to
ask
if
what
everyone
thinks
about
scheduling
some
time
or
doing
it,
sync
or
async
to
basically
just
run
through
and
click
through
staging
to
make
sure
that
everything
is,
as
we
expected
to
be,
maybe
there's
some
last
minute
bugs
that
we
have
introduced
or
haven't
caught
yet
specifically,
I
have
one
there
that
I
mentioned
in
slack
I,
don't
know
if
this
is
expected
or
weird
Behavior.
A
Yeah
I
still
need
to
jump
into
the
data
side
of
things
to
see.
If
there's
any
reason,
why
that's
that's
showing
the
way
it
is
as
far
as
the
last
seven
days
but
I
think
it'd
be
useful.
It
is
important
to
distinguish
that
snowplow
support
doesn't
is,
is
separate
from
the
the
actual
customer
release.
There's
still
I'm.
A
Obviously
more
Ops
work
involved
around
that
in
terms
of
being
able
to
onboard
customers,
but
yeah
I
think
it'd
be
useful
just
to
kind
of
walk
through,
and
you
know
everyone
go
through
that
exercise
to
make
sure
that
we
think
this
is
an
appropriate
release
for
customers
or
are
we
missing
anything?
Are
there
any
more
bugs
since,
since
we've
identified
since
over
the
past
couple
weeks,
happy
to
like
you
know,
we
can
do
it?
Sync,
we
can
do
async
I'll,
leave
that
to
the
group
to
decide.
E
Oh
yeah
I'll
start
a
select,
pull
or
something.
Maybe
we
can
do
like
an
async
Workshop.
That's
good.
A
Cool
we
have
three
minutes
left
I
just
wanted
to
quickly
I'm,
not
sorry
with
Show
and
Tell,
or
is
there
anything
that
anyone
else
would
like
to
discuss
that
hasn't
had
a
chance
to
put
anything
on
the
agenda?
A
All
right,
then
I
will
stop
the
recording
for
second.