►
From YouTube: 2019-04-18 Memory 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
A
Okay,
the
meeting
time
we
talked
about
alternate
the
meeting
time
every
other
week,
so
I,
that's
Ramya,
is
basically
yeah.
Thank
you
for
your
flexibility,
but
let's
see
if
we
can
just
pick
another
time
that
is
better
for
you
and
we
can
alternate
every
other
week.
If
you
have
suggested
just
to
cut
here,
then
I
will
you
know
everybody?
Where
is
everyone
so
figure
time
that
most
of
people
most
people
can
joy?
A
Not
maybe
not
everyone,
but
we
can
alternate
that
every
week
and
he'll
you
know,
Tonya
mentioned
that
the
new
hair
maybe
made
the
offer
maybe
made
so
once
the
new
hair
comes.
We
can
adjust
the
meeting
time
again.
Yeah
so
yeah
just
to
let
us
know,
was
what
time
works
better
for
you
will
I
will
reschedule
the
meeting
to
alternate
every
other
week.
C
A
C
A
D
Talked
briefly
yesterday
with
Dalia
about
that,
and
we
concluded
that,
like
we
gonna
spent
our
next
meeting
before
me,
taking
off
on
figuring
out
the
exact
date
idea,
and
it
should
be
basically
right
after
the
contribute,
so
we're
trying
to
figure
out
like
if
this
is
like
the
the
right
time
or
if
there
is
something
else,
I
preliminary
targeting
at
the
right
after
contribute
right
now.
Okay,.
A
So
our
target
is
right
after
the
target.
A
E
B
A
Speaking
of
photon,
part-time
yeah,
you're,
officially
the
first
member
of
the
team,
okay
cool,
we
know
that
and
so
I.
My
next
topic
is
a
suggestion
for
the
first
iteration.
So
basically,
I
want
to
show
this
really
quick.
I,
put
all
the
known
issues
here
and
used
this
things
like
the
functional
areas
and
the
task
hive
to
categorize.
So
you
can
see
the
category
category
and
type
here
and
it
need
a
pivot
table
analysis.
So
I,
the
the
concentration
is
the
tumor
Ruby
language
itself
or
Ruby
versions
and
psychic.
Those
are
in
functional
areas.
A
We
see
most
tickets
were
issues
here
and
then
the
ticket
type
is
a
feature
request
and
investigation.
So,
looking
at
that
list,
table
I
suggested
that
first
layer
iteration,
we
start
from
something
within
this
box.
This
house,
so
I
went
through
this
idea
with
Camille
and
he
was
also
sinking.
The
the
pure
Puma
probably
is
the
first
target
we
can
focus
on.
So
that's
well
defining
this
in
this
square
and
of
course
there
are
more
more
things.
We
need
to
nail
down
more
detail.
We
need
to
nail
down.
So
that's
my
sauce.
B
Yes,
so
to
me
it
it
makes
all
the
sense
related
to
this
is
yeah.
I
agree
that
puma
sounds
like
a
good
starting
point.
I
related
to
this
is
a
little
bit
I
added
later
to
the
document.
Is
that
I
wonder
if
we
already
have
a
to
link
for
measuring
our
improvements,
whichever
we
do,
because,
supposing
that
we
start
with
Puma,
and
we
do
some
improvements,
do
we
have
any
tooling
by
which
we
are
going
to
measure
this
I
think
it's
related
to
the
memory
test
bed,
but
I
could
be
wrong.
Yeah,
I,
I.
E
B
A
Qe
team
has
been
working
on
that
and
we
have
the
testbed
or
pilot
or
proof
of
concept
of
the
test
back
up
and
the
some
tests
are
running
there
already.
So
probably
I
don't
have
more
detail
here
but
may
be
wrong.
It
has
tell
us
what
measurement
award
metrics
we
can
collect
from
that
test
bed
already
and
that's
that
how
us
to
measure
the
Puma
improvements.
C
So,
as
of
now,
we
just
have
our
desks,
which
would
okay
the
desks
that
we
have.
It
actually
collects
some
front
end
metrics,
like
the
using
site,
speed
and
also
we
have
you.
We
have
the
integration
in
Prometheus,
which
has
bunch
of
metrics
later
from
web
and
extract
the
CPU
and
memory
usage
and
a
lot
more
metrics
as
well
so
yeah.
So
those
are
the
things
available
and
we
have
the
inter
patience
really,
so
we
have
yeah.
So
this
is
what
we
have
as
cool.
A
Sounds
like
we
have
the
basic
numbers
ready
to
be
poured.
So,
let's
work
with
that.
I
mean
let's
use
that
platform
and
as
we
discover
that
with
maybe
we
need
more
data,
then
we
can
always
enhance
the
platform
to
gather
that
data
or
collect
that
data.
Sometimes
we
have
something
working
already
and
that
will
be
good
for
our
first
later
iteration.
E
A
E
A
E
A
A
Cool
okay
and
also
test
of
coming
back
to
this
topic,
so
the
table
I
searched
I,
shared
and
I
bet
the
the
square
that
is
those
are
cells
is
here
so
full
marks
that
Kuechly
ruby
is
a
target
all
the
functional
areas
and
then
the
investigation,
the
feature
requests
are.
The
feature.
Requests
actually
actually
includes
that
we
need
to
add
the
metrics
Okuma
metrics
to
pro
seers
from
sears
and
those
kind
of
things.
Our
feature
requests,
so
those
those
things
are
well
included
in
that
cell.
So
we
can
work
on
that.
A
A
Also
prioritize
our
focus
so
I.
Think,
although
there
are
our
tickets,
were
issues
ironic,
Italy
and
vanilla,
omnibuses
dogs,
blah
blah
blah,
but
I
think,
based
on
the
analysis,
I
suggest
the
party
will
be
these
three
products
and
the
ticket
types
here.
So
if
you
have
concern,
please
let
me
know,
but
that's
my
sauce
I'm,
just
sharing
my
shot.
My
sauce
here
and
I
will
work
with
Camille
to
articulate.
We
already
had
some
ideas:
how
to
articulate
the
issues
yeah
and
acceptance
criteria.
Of
course,
this
is
my
suggestion.
We
have
a
backlog
to
work
on.
A
We
have
a
backlog
of
engineer,
Erickson
stories
that
we
can
actually
start
the
hands-on
work
in
for
first
iteration
and
the
future
iterations
and
a
stretch
goal
is
to
even
further
break
down
the
stories
actual
tasks.
So
it's
more
granular
breakdown
of
the
stories.
So
that's
my
my
thinking
of
acceptance,
criteria
for
first
iteration
and
Camille
has
are
coming
here
as
a
subside.
The
best
oh
right
for
now
is
to
work
on
the
pool
and
I
receive
an
ad
for
that.
Yes,.
E
E
D
D
There
was
also
the
common
that
we
needs
a
feature
like
the
ability
to
kill
you
in
West
when
they
are
like
running
for
longer
periods
and
then
some
some,
some
of
them
some
of
those
but
I
would
say
everything
is
related
to
latency
and
time
of
the
processing
and
and
the
footprint
of
the
of
the
processing
single
requests.
Yeah.
A
E
Can
tell
because
I'm
I
think
to
my
is
a
new
thing
right.
I,
don't
think
two
months
supports
from
you
get,
so
we
need
to
measure
things
before
it
goes
into
Puma
and
measure
response
time,
but
you
we
until
we
instrument
whatever
new
a
new
component
but
yeah
I.
Think
that's
a
good
point,
but
so
let
me
know
on
that.
We
need
you
to
watch
for
this
thanks
for,
like.
D
Historically,
we
we
had
very
short
period
of
history
like
four
years
ago
when
we
use
Puma
for
very
short
periods.
We
chose
in
the
end
unicorn
because
it
had
like
some
very
nasty
multi-threaded
marks
that
we
did
not
have
people
a
unicorn
and
technically
like
we
are
a
different
infrastructure,
for
example,
for
measuring
the
the
queuing
time
like
we
do
that
already
for
unicorn.
So
it's
rather
like
replicating
the
behavior
that
we
use
for
the
unicorn
or
some
uma.
D
How
long
the
request
leaves
in
the
queue
and
it
attacks
funny
like
better
if
you
have
the
higher
load,
so
it's
really
like
Puma.
Does
it's
a
key?
What
would
be
interesting
to
know
if
OMA
does
provide
you
more
fruit
than
Unicode
in
the
given
time
slice?
Basically,
because
it
would
be
like
the
measure
of
the
of
these
server
being
more
efficient
in
processing
requests
on
one
of
the
trees,
at
least
I
mean
more
throughput
in
terms
of
the
amount
of
the
request
process
like
their
given
timeframe,.
A
B
We
already
discussed.
The
first
item
was
about
the
memory
test.
Bed
I
have
one
more
item
there.
I
was
going
through
your
document,
I,
don't
notice
any
knowledge
about
nodejs,
which
seems
like
another
nice
thing
to
look
at
because
it
consumes
a
lot
of
memory.
So
I
was
wondering
if
it's
in
scope
of
this
team
or
not
everything.
A
About
memory
and
performance
is
in
scope,
so
would
you
mind
to
you
know,
make
a
issue
together
with
some
memory,
so
we
can
track
it
and
then
further
in
the
future,
we
shall
break
down
to
more
actionable
items
and
yeah
I
will
use
the
same
pivot
table
analysis
to
see
when
we
can
prioritize
that
yeah
I
think
everything
about
memory
and
performance
is
excope
of
this
team.
Okay,.
A
E
A
Okay,
my
next
item
is
I:
haven't
updated,
really
quick,
a
hiring
update,
so
our
pipeline-
this
is
a
dev
site.
Deficit
pipeline
is
kind
of
dry.
We
almost
around
out
of
all
our
candidates
and
it
just
pulled
two
more
candidates
from
other
jobs
this
morning.
So
now
we
have
nine
active
interviews
and
we
have
a
lot
of
the
of
those.
A
Are
we
haven't
heard
back
from
the
candidates,
so
I'm
working
with
recruiter
to
be
more
proactive
to
recharge
so
right
now
when
candidate
is
in
last
round
of
interviews
with
her
manager
and
executive,
so
that
that
carried
array,
this
is
in
this
stage,
all
others
are
still
at
the
screening
or
the
tech
interview
stage.
So
that's
a
quick
update
after
hiring
and
not
too
bad,
but
we
need
to
spend
more
time
on
that
I'm
working
on
that,
so
just
update
yeah
yeah.
A
How
much
should
we
focus
on
the
real
yeah
yeah
I
will
give
me
I
will
pass
you
a
couple
links,
so
we
are
Camille
and
I.
We
already
defined
our
criteria
for
hiring
and
also
for
both
the
manager
and
the
engineer
and
also
the
global
global,
optimization,
optimization
plan,
the
exit
plan.
You
know,
engagement
and
exit
I
will
share
the
link
with
you,
so
yeah
thanks.
C
C
So
we
have
Bob
site,
speed,
indicated
and
there's
a
channel
called
key.
We
have
performance
and
what
we
do
is
like
every
time
the
test
is
run,
the
site,
speed
metrics,
are
actually
posted
to
this
slack
channel.
So
if
I
don't
have
it
scheduled
periodically,
maybe
like
twice
a
day
to
start
with
and
yeah,
so
that's
about
the
site,
speed
integration.
This
is
similar
to
the
one
that
we
have
in
production.
C
Right
like
we
have
a
site,
speed,
integration
of
production
and
we
get
static
URLs
and
we
just
keep
checking
the
front-end
metrics
like
the
first
time
first,
which
will
pain
or
yeah.
There
are
a
bunch
of
metrics
like.
If
you
actually
take
a
look
at
QA
performance,
you
should
be
able
to
see
the
latest
runs
results
and
now
we'll
also
be
adding
loud.
C
Few
more
URL
looks
like
which,
which
would
usually
take
very
long
time,
and
this
would
be
static
and
it
will
be
always
domain
in
the
our
Testament
like
say
a
man
with
a
lot
of
discussions
and
very
huge
files
and
how
this?
How
long
does
it
take
to
actually
load?
All
of
those
things
like
that?
Will
be
there
and
the
things
to
do
is
like
a
bunch
of
meaningful.
Urls
should
be
added.
So
for
that
we
are
actually
planning
to
input
few
ripples
like
from
github.
C
So
that
is
one
thing
to
be
done,
and
second
thing
is
to
create
at
the
corner
dashboard,
wherein
it
would
actually
show
the
trend
that
story,
trend
and
comparisons
between
the
previous
week
and
previously
and
previous
day,
and
so
on.
So
that's
it
could
be
in
this
line
and
prometheus
computation
is
also
set
up
and
again
here
we
need
to
set
up
at
the
fauna
dashboard,
which
will
now
show
the
metrics
that
we
actually
are
interested
in.
A
Great
cool
very
cool
looks
like
the
the
pro
process
is
ready,
it's
up
and
running,
so
we
probably
should
be
ready
to
work
on
the
pull
my
pushing
Pullman
metrics
into
the
pro
chrome,
since
there
is
a
there
is
a
issue
about
that
and
we
come
here
and
I
just
talked
about
it
this
morning.
So
that's
probably
the
first
thing
we
want
to
do
before.
We
actually
do
something
to
Puma,
because
we
need
a
data.
We
need
to
see
the
data
before
we
do
anything.
E
Before
we
move
on
I
want
to
chime
in
here,
so
our
team
is
getting
into
a
strange
situation
because
now
testing-
although
this
has
its
chicken
and
egg
problem,
we
can
set
up
big
repo
to
test
because
github
import,
it's
fairly
I'm,
either
fix
opponents
there
as
well.
Right
so
I
mean
to
no
performance
need
to
import
and
imports,
not
working,
because
import
can't
do
performance
import.
So.
E
To
let
you
know
instead
tableau
said
here
with
everybody
that
that
performance
is
a
key
challenge
right.
It's
big!
It's
due
to
sidekick,
being
killed
every
post
day
on
them.
The
link
there
some
complexity
here,
because
you
get
lab
import,
is
being
done
by
managed
by
the
managed
team,
so
they
may
be
favors
from
for
us
asking
the
depicts
on
this
there's.
Also
one
fix
going
in
that's
the
feature
flag
is
disable.
A
Mac,
can
you
schedule
a
working
meeting
to
discuss
what
we
can
do
there?
Maybe
we
can
chop
down
chop
the
big
repo,
multiple
state,
import,
I,
don't
know
if
that's
possible,
but
let's
schedule
a
working
meeting
or
or
even
you
or
offline
or
basic
discussion
what
we
can
do
probably.
E
E
C
E
The
one
okay
one
second
I'll
just
link
Stan's
discussion
and
then
that's
the
issue
and
then.
E
E
C
E
So
I
have
both
my
managers
here
and
then
I
think
it's
it's
beneficial
for
us
to
discuss
it
here
as
well,
but
my
plan
will
premier
is
now
the
resonance.
An
expert
with
I
think
we
should
define
a
time
line
for
for
the
people
that
are
translating
in
and
trash
transitioning
out
and
I'm,
not
sure
if
you
just
do
it
in
a
table
in
the
stock
tree
and
I'm.
Okay
with
that
as
well
or
come
up
with
a
new
issue.
E
But
but
my
plan
is
have
Ramya
and
tonight
help
train
Tonya
and
the
new
question
and
then
Rama
needs
to
move
back
to
there
because
there's
a
there's,
also
other
fires.
We
need
to
put
out
in
the
dev
side
of
things
as
well,
but
so
now
I
will
stay
on
for
a
bit
longer
because
he
configured
the
Gatling
load
testing.