►
From YouTube: 2022-05-09 Database Scalability WG
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).
B
All
right,
let's
get
started,
welcome
to
the
database
scalability
working
group
monday
may
9th.
It's
like.
I
have
the
first
updates
just
a
couple
of
housekeeping
things.
B
Okay,
I
can
move
on,
looks
like
nothing
in
items
two
through
five,
so
I
think
we
can
get
to
the
the
fun
part
of
this
fabian's
item
about
the
timeline.
Take
it
from
here.
A
C
A
A
So
I
think
that
this
is
a
good
time
to
put
a
more
firm
plan
together
when
certain
things
need
to
happen,
because
once
we
start
rolling
out
things
in
staging
and
production
right,
we
can't
just
do
that
on
the
whim.
We
need
to
communicate
it.
We
need
to
get
signs
of
sign
offs
and
so
on
and
so
forth.
So
I've
I'm
proposing
a
timeline.
A
We
can
talk
about
that,
but
the
the
overall
plan
for
failing
for
bringing
bringing
about
decomposition
is
also,
I
think,
the
most
boring
plan
that
we
could
come
up
with.
Essentially
we're
going
to
up
adopt
most
of
the
run
books
that
already
exist
within
infrastructure
to
upgrade
the
database.
A
There
are
a
few
ideas,
let's
say
simply
put,
we
could
add
another
ci
cluster
and
then
we
upgrade
one
of
them.
If
that
goes
wrong,
we'll
fall
back
to
the
the
second
one
as
a
one
standby.
There
are
some
other
ideas
and
then,
when
that
is
all
let's
say,
successful
anything
that
comes
afterwards
that
would
require
rolling
forward.
A
A
That's
that
in
terms
of
timeline,
there
are
essentially
four
major
milestones.
There
are
a
few
others
that
are
in
the
issue
that
I
actually
didn't
put
in
here,
but
we're
going
to
do
three
or
we
intend
to
do
three
runs
on
staging
the
first
one
is
going
to
be
a
dry
run,
so
we're
not
going
to
make
any
lasting
changes.
We
just
need
to
validate
our
our
run
books
and
we're
ambitiously
targeting
that
for
the
beginning
of
next
week,
camille
and
tong
are
going
to
work
on
this.
A
Then
we're
going
to
aim
to
get
a
run
in
with
rollback
on
the
23rd.
These
are
you
know
my
my
dates
right
now.
They
may
shift
by
a
couple
of
days
depending
on
on
other
needs,
but
sort
of
the
the
general
issues
should
align.
So
this
the
second
run
is
essentially
executing
everything,
including
the
failover,
but
then
also
testing
the
rollback
after
that,
so
we'll
reset
to
phase
four.
After
that
we're
doing
this,
because
when
we
do
it,
we
can
essentially
validate
that
the
failover
was
successful
and
that
our
rollback
scenarios
were
also
successful.
A
A
We
need
an
sre
to
actually
help
us
go
into
into
staging
and
validate
that
all
of
the
the
bits
work.
The
hosts
are
there
and
so
on
and
so
forth,
and
I
think
camille
can
speak
to
more
details
if
required.
But
this
is
not
something
that
we
can
do
by
ourselves.
We
want
to
collaborate
with
an
sae,
so
that's
the
that's
the
plan
and
yeah.
It
gives
us
six
ish
weeks
ambitiously
to
get
there
and
we
still
have
a
few
things
to
do,
but
I
think
it
is.
It
is
doable.
A
C
Yeah
there
is
another
move
yeah
there
is
another
meeting
with
all
stephen
infrastructure
managers
on
wednesdays,
so
naked
probably
can
bring
this
over
and
fabian.
If
you
want
to
join
there,
I
can
invite
you
for
this.
For
this
week,.
A
Sure
I've
also
pinged
everyone
on
on
this
plan
already
for
transparency.
B
Yeah
and
I
think
even
ahead
of
that
meeting,
we
do
have
the
slack
channel
with
of
those
same
people
the
decomposition
stakeholders
channel
so
make
sure
we
share
this
plan
today
and
ping
ping,
those
stakeholders
for
confirmation,
I
think,
yeah.
In
the
last
meeting.
I
think
there
was
general
agreement
that
raphael
could
be
available
for
supporting
this,
but
we'll
just
make
sure
we
can
get
additional
confirmation
and
see
if
there
are
any
concerns
with
the
plan.
A
My
main
concern
is
that
there
may
be
other
priorities
that
I'm
unaware
of
that
actually
take
people's
time,
and
I
think
that
is
something
that
we
need
to
align
on,
where
the
decomposition
effort
sort
of
clocks
in
just
so
that
we
ensure
that
everybody
understands
what
what
can
and
can
be
done,
because
otherwise,
we'll
run
into
challenges.
D
D
But
let's
make
sure
that
I
know
there
are
conflicts
and
some,
as
you
can
see,
the
managers
and
freaking
not
attend
this
meeting
and
I'm
not
empowered
to
make
people
decisions
so
I'll
pass
along
the
info.
But
I
was
made
aware
that
there
was
this
other
meeting.
I
didn't,
I'm
not
invited
to
it
and
I
didn't
know
what
it
was,
but
I
think
it's
you
run
that
meeting
right.
C
Nick
now
runs
through
the
meetings
so
nick,
please
see
what
will
be
interested
in
that
meeting,
because
now
we
are
into
a
critical
planning
stage
now.
A
D
Really
want
to
make
sure
that
we
don't
drop
it
on
the
floor
because,
oh
you
know
jerry
was
enough
and-
and
I
know
that
at
least
can
and
they've
attended
the
other
one.
So
they'll
be
the
authoritative
sources
to
say
yes,
so-and-so
is
committed
and
these
states
look
fine.
We
did
have
a
conversation.
I
think
last
week,
where
we
said
functionally,
compensation
will
take
priority
over
the
operating
systems.
Last
postwars
13
upgrade.
So
I
think
we're
committed
to
push
those
back
if
we
need
to
in
order
to
get
this
done.
First.
D
But
again,
infra
managers
have
the
authoritative
word
on
that.
B
D
B
Else
who
who
wasn't
aware
of
the
meeting
the
weekly
decomposition
stand
up
on
wednesdays,
wanted
invite?
Let
me
know.
D
B
B
Of
my
attempt
with
this
with
the
item,
I
had
here
the
mr
to
clarify
the
exit
criteria,
just
saying
all
right,
it's
decomposition
and
then
cleaning
up,
and
if
anybody
wants
to
propose
something
else,
we
can
do
it,
but
otherwise
that's
that's
what
bmr
says
is
remaining
for
this
group.
A
A
All
right,
that's
all
I.
B
A
B
I
think
camille's
reaction
was
confirmation
enough
that
we're
heading
on
the
right
path,
so
yeah.
Thank
you.