►
From YouTube: 2021-03-24 Database Scalability Working Group 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
A
B
A
A
All
right,
let's
get
started,
I'm
not
going
to
share
the
board
today,
because
there's
only
five
issues
and
they're
all
the
blueprints
from
the
ca
build
stuff.
So
we
can
probably
skip
that
and
dive
into
the
specifics
to
him.
B
Yeah,
a
staffing
plan
has
been
consolidated
into
one
issue:
one
single
issue:
everything
in
one
place.
What
we
wanted
to
do
is
activate
the
db
icon
reset
as
soon
as
possible,
then
activate
the
staffing
for
sharding
in
about
one
or
two
weeks
after
the
the
blueprint,
mr
is
merged.
So
details
are
in
the
issue,
any
suggestions
or
questions.
C
Why
why
do
it
after
the
blueprint
is
merged?
Why
not
parallelize
them?
Because
you
know
people
need
time
to
get
settled
and
communicate
and
get
to
know
their
teammates
and
things
like
that.
B
D
And
that's
that's
documented
in
the
timeline,
basically
moving.
Some
people
immediately
as
beginning
of
this
week
versus
moving
people
at
the
basically
the
cut
over
the
release
on
the
19th
of
april.
B
Okay,
if
no
more
questions
we'll
work
offline
to
identify
the
the
people
who
are
going
to
join
these
two
staffing
plans
after
the
approval
jerry
over
to
you.
A
Yeah,
I
was
just
asking
to
see
if
there
was
any
movement
on
ci
builds,
but
writers
already
answered
that
were
consumed
by
rapid
action.
So.
A
E
E
A
I
was
the
the
first
one,
the
partitioning,
so
I
I
mentioned
later
that
while
we
should
keep
status
on
what's
going
on
in
rapid
actions,
I'm
not
going
to
make
it
an
item
in
this
working
group
because
it's
already
being
discussed
by
essentially
this
group
in
the
standup,
so
I'll
keep
an
eye
on
it
and
bring
up
any
items
that
I
that
I
think,
or
anybody
should
be
able
to
bring
up
any
out
of
any
items.
They
feel
are
relevant.
A
That
are
happening
on
rapid
action
to
this
group.
But
I
I
rather
not
spend
time
here
talking
about
that,
given
that
it's
already
been
discussed
in
the
standup.
B
Yeah
but
heads
up
to
andreas
and
janice,
probably
because
we
tend
to
combine
the
pk
migration
together
with
partitioning,
so
maybe
we
need
to
climb
together
accordingly,.
A
Yeah,
I
have
a
question
about
this
as
well.
A
little
later
in
the
discussion
section.
A
I
started
dmr
on
sharding
and
I
know
a
few
people
have
already
been
requesting
feedback.
I
have
a
bunch
of
things
that
I
need
to
apply
to
it,
but
anyway,
just
giving
awareness
of
that
because
it's
an
important
one.
My
question
to
eric
is:
does
this
address?
A
A
C
Oh,
I
see
well,
let
me
let
me
read
it
and
I
can
get
back
to
you.
E
B
Yeah
also,
I
pushed
up
mr
to
edit
the
glossary
table
and
also
including
the
sharding
in
the
in
the
scope.
So
please
help
to
add
your
comments
into
that.
Mr
yeah.
B
Thank
you
then,
jerry
over
to
you.
A
B
A
I
put
this
in
the
blocker
section
because
I
I
yeah
I'm
on
a
tour
as
a
walker.
We
don't
have
dates
for
when
these
blueprints
are
supposed
to
be
to
materialize.
That
doesn't
mean
we
cast
them
in
stone,
but
it
would
be
nice
to
have
something
to
measure
against.
A
So
I
was
sort
of
raising
my
hand
to
say:
can
we
assign
due
dates
to
these
things
and
then,
as
as
we
approach
them?
Obviously
we
can,
you
know,
move
them
as
we
see
fit,
given
all
the
other
work
that's
going
on,
but
right
now
I
just
feel
that
it's
sort
of
an
open-ended
thing
we're
going
to
continue
to
sort
of
talk
about
hey
we're
going
to
write
these
we're
going
to
write
these
we're
going
to
write
these,
but
we
never
will
so
does
that
sound,
reasonable.
F
To
be
honest,
I'm
I'm
still
unclear
about
the
scope
of
this,
so
I
was
going
to
figure
that
out
this
week
and
come
back
with
a
date.
If
that's
okay
right
now,
we
only
have
that
notion
in
the
handbook
I
think
about
what
is
what
we
consider
reading,
mostly
as
a
pattern,
and
I
wasn't
sure
if
that
is
sort
of
more
than
just
caching
small
tables
or
what
the
scope
of
this
is.
A
The
license
table
8
000,
reads
per
second
against
the
database.
I
don't
know
if
it
makes
sense
to
have
that
in
postgres,
and
you
know
force
postscript
like
it's
obviously
being
read
from
memory,
but
it's
also
forcing
postgres
to
execute
8
000
contact
switches
per
second,
so
maybe
we
can
move
it
to
ready's
on
boot
or
whatever.
So
that's,
that's
the
general
idea
that
there
are
some
tables
out
there.
A
You
know
we
need
a
place
to
to
put
this
data
and
we
chose
postvis
and
that's
totally
reasonable,
but
maybe
it's
time
to
maybe
keep
the
authoritative
source
in
postgres.
But
not
you
know
just
use
redis,
maybe
for
these
things
that
we're
reading
wholesale,
because
these
are
as
far
as
I
understand
for
licenses,
we
used
to
select
star
and
consume
the
entire
thing.
So
that
was
that's
where
I
was
going
with
that.
F
Okay
and
the
expectation
for
the
blueprint
is
to
basically
summarize
that
problem,
give
an
example
with
the
license
table
summarize
solutions
that
we
know
that
we
want
to
pursue
right.
Okay,.
C
But
it's
also
like
the
the
one
solution,
you're,
confident
and
like
it
should
you
know
it
doesn't
have
to
be
correct.
It
can
be
challenged,
but
it
should
be
like.
I
think
this
is
the
direction
the
organization
needs
to
go
and
even
if
there's
three
policy
possibilities
like
definitely
have
a
favorite.
D
So
andre,
can
you
andreas?
Do
you
think
you
can
give
us
based
on
that
scope?
Can
you
give
us
an
estimate
on
either
today
or
on
wednesday
or
sure?
E
So
yeah
we
are
wrapping
up
all
the
actions
from
the
previous
weeks.
So
hopefully
I
will
start
working
on
on
my
blueprint
at
the
end
of
this
week
and
I
will
have
a
deadline
set
by
next
meeting.
I
will
have
an
estimate
dennis.
Can
we
collaborate
together
because
it
feels
like
the
ci
builds
table
is
similar
to
some
extent,
of
course,
of
course,
and
I
can
create
a
first
draft
and
then
we
can
discuss
cool
things
or
we
can
think.
A
I
put
an
april
7
once
on
the
on
the
starting
one,
which
I
think
is
overly
optimistic.
It's
not
going
to
happen
because
we
don't
have
a
dri
for
it
yet
so
we
will
update.
But
for
me
and
looking
at
the
table,
it
just
shows
me
where
we
are
so.
Thank
you.
B
Soon,
you're
next
yeah.
Actually
I
want
to
circle
back
to
what's
happening
next
on
that
execute
the
heck
on
the
reset
so
question
two
eric.
B
C
Sure,
the
in
general,
we
should
this-
this
section
was
empty.
We
should
always
have
something
here
like
this
is
the
most
important
thing
out
of
this
working
group
is:
is
maintaining
momentum
and
driving
completion
here?
So
even
if
we
start
the
meeting
with
nothing,
we
should
be
populating
this
with
like
yeah.
This
is
this:
is
the
stuff
people
are
committing
to
delivering,
so
the
things
I
picked
up
but
already
are
we
talked
about
the
dris
below
driving
the
blueprints
and
we
added
dates
for
those
chun.
C
Has
the
head
count
reset
item,
and
then
I
heard
it
mentioned,
but
it
was
something
about,
like
folks,
are
kind
of
consumed
with
the
rapid
action
currently,
but
eventually
they're
going
to
close
that
work
and
then
start
executing
on.
What's
here,
I
don't
know
if
I
wrote
that
correctly.
So
I
put
a
question
mark
after
it.
C
Does
that
seem?
Does
that
seem
right?
Should
we
do
this?
Should
we
have
this
at
the
end
after
blockers
and
discussion,
or
how
do
we?
How
do
we
make
sure
that
we're
always
exiting
the
working
group
with
a
we're
all
on
the
same
page
about
this,
because,
like
already,
we
got
to
the
point
where
we
covered
the
head
count
reset,
but
now
china's
asking
do
I
have
approval
or
not
yet
so
we
must
have
missed
something
up
above.
D
I
think
it
should
be
the
last
thing
in
the
meeting,
because
discussion
may
also
have
some
things
and
expectations
around
it.
So
yeah,
it
seems
like
moving
to
the
the
last
thing
on
the
agenda
makes
the
most
sense.
Okay,.
C
I'm
gonna
so
I'll
move
it
in
the
template
of
above
for
next
time
and
then
I'll
just
get
rid
of
follow-up
items
because
it
feels
like
that's
the
same
thing.
B
Yeah
is
it
fully
approved
the
whole
package,
or
only
the
one
portion
of
it?
There
are
two
portions
one
is
the
db
team
icon
reset.
One
is
staffing
for
charlie.
B
Let
me
enlarge
it,
so,
let's
just
the
total
head
countries
request
is
here
at
the
top
of
the
table.
You
know
11.5
accounts
and
actually
it's
separated
into
two
requests.
One
is
the
database
team
headcount
reset
that's
for
database
team,
temporary
staffing
time
box,
the
first
three
months
that
this
part
requires
three
foreign
engineers
and
then
the
sharding,
the
sharding
team,
actually
is
another
part
of
the
request.
So
I
just
I'm
just
wondering
is
the
whole
package
approved
or
only
one
part
of
it
at
this.
B
D
So,
are
you
looking
for
like
e-group
approval?
Are
you
looking
for
more
like
product
approval?
I
think
products
has
proved
it
from
the
perspective
they're
supportive
of
investigating,
seeing
if
we
can
find
appropriate
head
count
where
we
can
make
other
prioritization
decisions,
so
basically
we're
at
the
next
step
where
we
can
start
identifying
people
from
that
perspective.
If
that's,
what.
B
D
Yeah
so
we've
gotten
an
oops
enough
sign
off
at
this
point
so
and
I
think
scott's.
D
C
Let's
just
not
use
the
word
approved
because
it's
not
approved
until
it's
a
specific
proposal
where
this
this
person
is
being
drawn
away
from
this
team.
There's
going
to
be
an
impact
to
that
backlog
and
they're
going
to
be
moved
here,
so
you
have
kind
of
like
directionally.
It
seems,
okay,
go
figure
out
the
heads
and
then
that's
the
plan
that
can
either
be
approved
or
unapproved
or
changed
or
counter-proposed,
or
something
like
that,
and
so
I
think
that's
so.
I
think
you're
halfway
to
request
one
and
then
for
request
two.
C
D
Yeah,
so
I
specifically
did
not
want
to
do
that
eric
and
the
reason.
Why
is
because
I
agree
with
your
sentiment,
but
the
problem
is:
is
that
going
to
product
and
then
two
weeks
later,
coming
back
with
okay,
now
here
we're
gonna,
do
another
head
count
reset
we're
better
off
just
going
after
it
in
the
same
swoop,
and
if
we
decide
that
we
need
to
ask
for
more,
then
we
can
make
an
adjustments
in
process
associated
with.
That
would
be
my
recommendation.
C
I
mean
I'm
more
worried
about
individuals
and
and
manager
and
managing
relationships
being
rendered
apart,
the
name
about
product
management,
I
mean
we're
iterative,
they
should
pm,
I
think,
is
totally
fine
consuming
a
high
rate
of
change
on
these
things.
But
if,
if
we're
at
the
point
where
we
can
say,
we
can
name
these
three
people
we're
going
to
transition
to
the
starting
team
and
heads
up,
we
may
do
this
again
next
week,
then
yeah.
We
can
do
that
if
even
without
the
subject
matter.
D
Yeah,
I
think
I
think
we
have
to
start
the
conversation
now
because,
like
even
if
we
untangle
those
folks,
it
may
be
april
19th
before
we
see
them
right.
So,
like
my
view
is,
is
this
is
kind
of
working
in
parallel?
So
the
only
question
is:
is
you
know
what
urgency
do
you
want
to
have?
If
we
don't
have
this
urgency,
then
we
may
not
see
those
books
until
may
19th.
D
As
an
example
short
of
us
saying,
drop
everything,
and
this
is
the
most
important
thing
which
then
puts
a
different,
much
different
development
experience
for
them.
Yeah.
C
A
Because
if
there
is
like
a
more
general
thing
but
charlie,
we
know
specifically
and
the
blueprint
is
something
that
we
need,
someone
with
math
skills
to
produce.
C
D
Yeah,
so
so
my
staffing
strategy
is
this
is
working
on
specifically
what
needs
to
be
done
as
a
general
mechanism,
and
then
we
push
it
to
the
teams,
the
stage
teams
potentially
to
actually
do
some
of
this
additional
charting
work.
So,
depending
on
how
that
plays
out,
the
question
would
be
is
how
many
you
need
for
general
mechanism,
implementation
in
the
short
term
to
basically
accelerate
the
program,
and
then,
if
we
need
to
come
back
later
on
and
say,
okay
now
we
need
like
stage
specific
work
done.
D
D
C
Yeah,
I
like
the
urgency,
I
would
separate
them
into
two
proposals,
but
I
would
work
on
both
of
them
kind
of
concurrently.
C
Let's
say,
and
we
can
do
another
headcount
reset
later
and
just
be
transparent
about
there
may
be
another
one
coming
yeah,
I
mean
okay,
okay,
I
think
from
my
standpoint,
I've
tried
to
provide
visibility
that
the
that
these
ass
are
coming
and
that
it
seems
like
it's
better
to
come
up
with
a
list
of
names,
as
opposed
to
like
the
usual
process
of
finding
one
from
this
group,
one
from
that
group.
So
I
think
you
know
we're
proceeding
on
that
front
and
that
way
product
has
a
list
to
react
to.
C
We
can
see
if
it
works
and
go
from
there
but
happy
to
change
course.
If
that
doesn't
seem
like
the
right
approach
here,
I
agree.
I
think
this
is
less
about
drawing
like
relative
investments
in
areas
of
the
product,
the
roadmap.
It's
more
about
individual
sort
of
skills
and
experience
in
terms
of
fitness
for
this
type
of
project.
B
I'll
separate
the
issues
into
two
two
proposals,
and
I
think
next
week
my
outcome
is
the
names
for
the
common
set
and
the
three
names
for
charlie.
Is
that
expected
from
me
just
to
double
check?
Does
it
take
a
week?
B
No,
I
I
think
we
can
strive
for
towards
like
enough
this
week.
Okay,.
C
Yeah
I
mean
in
general,
I
would
treat
this
this
working
group.
This
meeting
is
like
we're
checking
at
least
this
often,
but
don't
block
anything
for
the
next
working
year,
try
to
work
on
everything
async.
Otherwise
we
become
interrupt
driven
and
things
that
could
take
a
month
and
a
half
end
up
taking
six
months,
because
we're
always
waiting
for
the
next,
the
next.
The
next.
B
F
B
Yeah
just
a
quick
question:
who
will
be
the
pmdri
on
this
working
group.
C
Yep,
so
I
think
you
use
me
for
now,
I'm
just
discussing
please
do
with
lobbying
this
morning,
but
it'll,
I
think,
be
one
of
us
but
put
me
down
for
now
and
we
can.
A
C
A
Evolves
the
item
I
had
is
related
to
the
comment
earlier,
so
I
I
do
think
that
it's
important
for
this
to
have
awareness
about
some
of
the
midterm
stuff
like
the
primary
keys,
work
and
ci
builds.
I
don't
want
to
reborn
it
because
it's
being
discussed
elsewhere,
but
one
of
the
things
I
was
trying
to
do
today
is
sort
of
navigate
the
pk
issues
and
I
just
got
lost.
A
So
if
there
is,
I
was
going
to
ask
that
there
is
one
issue
per
pk
or
whatever
a
set
of
issues
that
I
can
track
specifically
versus
trying
to
like
there's
a
lot
of
work
involved
in
there,
and
there
is
like
research,
work
and
figuring
out
this
and
figuring
out
that.
But
if
there
is
one
place
where
I
can
track
a
more
overall
view
of
what
what's
going
on
here,
I
would
love
to
hear
about
it,
and
maybe
I
should
just
reach
out
to
craig
directly.
E
We
have
added
a
summary
in
the
description
of
the
epic
and
we
are
going
to
try
and
update
it.
Can
you
check
it?
Can
you
check
if
it
is
sufficient?
Okay,
yeah?
I
can
do
that
yeah
so
because
we
know
that
there
are
so
many
issues
there.
So
if
you
can
check
the
top
epic
and.
D
A
Okay,
I
will
I
will.
I
will
try
that
then
thank
you
and
then
tune
since
we're
starting
to
run
low
on
time.
I,
yes,
I
will
create
a
separate
call
so
that
the
recordings
come
to
me.
I
thought
there
was
a
way
to
transfer
call
ownership
but
I'll
create
a
new
one
and
not
take
the
calendar.
Invite
okay,
thank
you,
and
that
is
it
for
today,
cool.
Thank
you.
Everyone
see
you
once.