►
From YouTube: 2020 06 09 Database Team Weekly
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
All
right,
let's
jump
right
into
it,
so
we
had
review
goals
from
last
week
be
honest:
you
had
your
scripts
to
anonymize
production
data
and
wrap
up
your
non-partitioning
efforts.
How
did
you
do
last
week.
B
B
Gcp
instance-
and
we
will
check
tomorrow
how
things
are
working.
I
have
also
one
follow-up
there
to
check
all
the
query
plans
that
we
generate
with
this
with
a
benchmark
so
that
we
against
production,
so
that
we
are
sure
100
sure
that
everything
we
generate
is
in
accordance
with
what
we
are
seeing
in
production.
B
So
yeah,
that's
why
we
did
not
work
on
the
anonymizer
script
and
maybe
we
will
postpone
it.
Maybe
maybe
we
will
work
on
it,
but
the
initial
decision
was
not
to
work
on
it
at
the
moment.
Just
focus
on
the
best
part.
B
Oh,
a
lot
of
a
lot
of
updates
and
mars-
I
must
have
already
been
merged
and
marks
have
been
reverted
and
march-
will
be
created
again,
but
yeah.
Overall,
everything
is
done.
Even
I
only
have
to
to
fix
when
mario
was
reverted
and
then
I'm
my
full
focus
will
be
on
on
our
partitioning
stuff.
Also,
an
update
on
on
our
robocops
has
been
is
almost
ready
to
be
merged
and
some
documentation
updates
have
been
done.
So
overall,
I'm
satisfied.
A
C
Yeah,
the
second
part
of
the
migration
helper
that
was
merged
the
end
of
last
week,
so
that
is
in
there.
The
third
part
is
not,
and
quite
I
thought
I
thought
I
would
have
that
wrapped
up
last
week.
Getting
quite
that
get
that
done
is
a
little
bit
more
complicated
than
I
was
initially
thinking,
but
it's
come
along
well
and
I
was
able
to
get
another
one
of
my
other
mr's
merch.
That's
been
hanging
out
there
for
a
while.
C
I
guess
I
kind
of
wanted
to
see
for
the
mrs
that
we
have.
I
guess
that,
like
usually
we
try
to
do
some
kind
of
like
production,
validation
that
the
mr
has
gone
in,
but
I
guess
a
lot
of
these
that
are
tooling.
We
don't
really
have
a
way
to
develop,
there's
really
nothing
to
validate
other
than
it's
been
merged.
So
I
guess
once
it's
been
merged,
we
can
just
say
the
issue
is
closed.
It's
that
we're
good
on
that.
A
C
A
Okay,
all
right,
then,
should
we
is
there
anything
else
to
talk
about
from
last
week,
anything
to
wrap
up.
B
No
yeah
everything
was
disgusting.
A
Okay,
so
goals
for
the
coming
week.
A
You
know
pat-
and
I
chatted
a
little
bit
yesterday
and
you
all
talked
last
thursday
when
I
couldn't
make
the
office
hours
so
sounded
like
the
goal
for
this
week
was
to
get
the
partitioned
audit
tables
running
in
parallel
in
production,
so
that
we
can
do
some
measurements
comparison
between
what
we
have
now
and
what
the
partition
tables
look
like.
Is
that
correct?
C
Well,
not
all
the
mrs
for
that
have
been
merged,
yet
so
there's
one.
So
for
me
at
least
I
would
say
my
goal
is
to
get
the
third
piece
of
that:
the
migration
helper
finished
and
reviewed
and
merged
in
so
then
we
can
move
forward
with
the
migration.
B
B
So,
are
we
going
to
set
the
triggers
patrick
so
that
we
can
check
how
the
triggers
work.
B
The
triggers
for
loading,
so
are
we
going
to
create
the
table,
the
partition
tables
and
set
the
triggers
that
load,
the
partition
tables
from
insert
and
updates.
C
Oh
yeah,
yeah
yeah,
I'm
not
sure
I
mean
we
could.
C
C
Initially
we
had
talked
about
just
rolling
the
migration
out
in
stages,
but
I
think
now
we're
close
enough
that
we
can
just
do
it
all
at
once,
but
yeah
hopefully
we'll
be
ready
to
andreas
already
has
a
mr
out
there
to
add
the
actual
migration
that
would
run
the
partitioning
job.
So
I
think
once
the
third
piece
gets
merged
in
then
he
can
update
his
mr
and
get
that
merged,
and
then
it
would
start
to
kick
off
the
migration
process.
A
B
The
generator
and
the
benchmark
works
correctly.
Maybe
we
run
a
test
on
gcp
with
the
generated
data
to
make
sure
that
the
migration
works
and
then
releases
production.
I'm
not
sure
if
you
agree.
A
A
Okay,
anything
else
for
goals.
For
this
week
we
have
a
lot
of
little
things
in
one.
B
Yeah
get
the
little
ones
out
and
also,
as
I
I
said,
that
before
we
we
are
still
we're
going
to
make
sure
that
and
evaluate
that
the
benchmark
and
generator
workflow
correctly.
So
we
have
to
do
those
checks
and
query
query.
B
C
A
Okay,
yeah,
if
y'all
could
take
a
moment
sometime
in
the
next
day
or
two
and
make
sure
the
health
status
on
all
those
are
up
to
date,
looks
like
they're,
not
all
right,
just
done
a
couple
that
I've
looked
at.
A
I
wish
there
was
a
way
to.
I
wish.
There's
a
quick
action
for
those
would
be
a
lot
easier,
all
right
anything
else
on
goals
for
next
week,
or
is
that
pretty?
Well
summarize
what
we
have?
We
can
get
I'll
chat
with
andreas
tomorrow.
A
Okay,
a
couple
follow-up
items
so
from
the
company
retro
or
the
engineering
wide
retro,
and
when
I
was
summarizing
database
retro.
I
think
the
two
of
you
actually
mentioned
that
you
had
a
couple,
mrs
that
were
bigger
than
you
had
expected.
A
So
I
was
thinking
it
might
be
a
good
time
for
us
to
do
an
iteration
retro.
Any
thoughts
on
that
and
any
questions
on
what
an
iteration
retro
is.
You
know
what
I
mean
when
I
say
that.
B
A
I
will
just
hijack
one
of
these
meetings-
maybe
yeah
we'll
do
it
after
let's
get
through
next
week,
so
I'll
schedule
it
for
the
week
of
the
22nd.
So
we're
not
interfering
with,
like
the
end
of
the
milestone
and
trying
to
scramble
to
get
things
done
for
the
release.
So.
A
Yeah
and
then
have
either
of
you
spent
any
time
with
postgres
ai.
C
I've
done
a
yeah
a
little
bit,
just
not
just
like
the
features
that
we
already
use,
slack
for
the
joe
stuff.
C
B
A
Okay,
yeah
I'll
have
to
set
up
a
recurring
feedback
loop
so
that
we
can
justify
whether
or
not
it's
worth
it
or
how
far
you
know
if
it
is
worth
it
for
individuals,
if
we
want
to
roll
it
out
to
the
larger
engineering
team
and
who
who
would
end
up
wanting?
Is
it's
not
super
expensive
on
an
individual
basis,
but
we
still
need
to
make
sure
that
it
provides
more
value
than
it
costs
right.
So
yeah.
B
A
A
I
could
see
it
providing
value
for
not
just
the
thin
clones,
but
the
session
sharing
could
provide
a
lot
of
value
for
folks
that
are
just
getting
used
to
postgres
and
need
to
learn
about
it
and
they
want
to
pair
with
someone
that's
more
senior,
so
fortunately,
we
have
six
months
to
trial
this
and
make
sure
it's
worth
worth
the
effort.
So
all
right,
thanks
for
that
early
feedback,
anything
else
we
need
to
cover
for
the
team
meeting
portion
of
this.