►
From YouTube: 2019 05 23 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
Yeah,
let's
start
our
meeting
today,
so
agenda
is
there.
Everybody
can
grab.
A
First
item
is
from
me:
the
planning
board,
so
I
proposed
it
to
send
how
about
a
new
board
for
the
planning
and
then
use
the
current
board
for
work
because
out
the
label.
How
the
labels
are
handled
in
one
board,
so
come
here
said,
is
good
to
go
and
I.
Think
I'm
going
to
do
it.
Everybody!
Okay,
with
that
the
issue
is
so
when
you
have
the
p1
p2
piece
and
as
severe,
it
is
in
a
safe
on
same
board.
A
When
you
move
the
issues
across
the
columns,
the
priority
and
the
severity
levels
will
be
removed.
Then
we
lose
information,
that's
the
important
information,
so
the
workaround
is
just
to
have
another
board
to
list
the
P
priorities
and
severe
case
then
use
this
current
board
to
do
the
works
tracking
for
the
compound
for.
Can
we
use
the
bug
scrub
board
for
the
priority
and
severity,
the
one
that
should
I
link
it
here?
The
one
that's
already
set
up
yeah,
but
but
that's
a
box
cross
scrub
right,
it's
a
we
want
to.
A
B
Thing
you
want
to
call
out
I
think
it
may
be
a
buck.
So
if
it's
not
working
for
us,
let's
push
for
the
plan
team
to
fix
it.
I
think
there's
a
bug
already
file
and
like
any
move
things
and
they
will
get
labels
get
removed.
I'll
try
to
find
it,
but
let's
put
more
pressure
for
our
people
to
fix
it,
because
we're
billing
deal
at
all
and
use
it.
It's
not
working
then.
A
Good
point:
I
was
thinking
it's
a
bug,
but
I'm
I
don't
know
if
it's
a
feature
yeah
but
supposed
to
be
a
bug,
I
think
because
the
work
for
all
levels
they
should
be
in
that
should
behave
you,
my
mother,
while
you
move
across
columns,
the
old
one
gets
removed
at
the
new
one
gets
added,
yeah
cool
Mac.
If
you
know
a
bot
number,
can
you
just
post
here
or
maybe
it's
just
a
register,
a
new
bug?
Yes,.
B
C
A
A
That
seems
so
standard
away.
Could
do
it
eating
do
that?
Okay,
silence
means
I,
think
everybody
is
fine.
Continue
to
use
that
and
Camille
suggests
stated
different
questions.
Definitely
I
can
change
the
questions
there,
so
I
agree.
I
will
make
the
change
the
Odin
question.
I
had
is
aa
rotating
questions.
A
D
It's
if
those
proposal
I
think
that,
like
when
we
have
you
people
joining,
maybe
they
we
have
some
ideas.
How
I
can
improve
that?
So
maybe
right
we
could
just
keep
it
as
nice
and
like
run
the
iteration
on
the
book
question
like
when,
like
one
week
or
two
weeks
from
the
new
people
join.
Maybe
this
is
the
right
time.
Okay,
wait.
I
need
to
London,
okay,.
D
Do
whatever
you
prefer,
we
like
I,
think
it's
fine
to
keep
it
lassies.
I
also
think
that,
like
climbing,
like
two
questions,
he's
probably
like
like
something
that's,
it's
the
easiest
to
answer,
because
when
I
look
at
the
our
stand-ups
so
far
like
we
mostly
answered
like
two
questions,
maybe
and
looks
like
possibly
to
five
questions,
probably.
D
A
A
Yes,
it's
okay.
We
just
erased.
You
know
real
time
when
the
amendments
come
up.
We
just
erasing
in
the
channel
and
the
ping.
Are
this
smoker
people?
So
everybody
is
aware
we
have
some
impediments,
then
we
work
actively
to
address
that.
Maybe
we
can
thread
off
of
any
response
that
has
an
impediment
given
it.
D
A
C
Yeah
just
for
inspiration
in
plan
team,
we
use
similar
set
of
questions
if
you
are
interested,
how
it
works,
feel
free
to
check
G
underscore
plan
underscore
stand
up
channel
and
you
can
see
history
of
our
looks.
It's
it's
quite
similar
to
what
we
use
in
memory
team.
But
if
you
wonder
how
to
include
impediments,
question
inspiration
is
to
make
it
part
of
your
today
plan.
Yes,
so
if
someone
is
booked,
you
can
mention
it
in
the
upcoming
plan.
They
yeah.
A
Good
yeah
good
suggestion
I
will
do
that
so
just
a
roll
either
into
the
second
question
into
the.
What
I
will
do
today,
good
I'll
do
that
cool
I
think
we
have
a
consensus
here,
and
the
next
item
is
really
just
a
reminder:
Celeste
remember
that
I
did
jus.
First
at
your
service,
we
want
to
do
the
planning
just
a
groom,
the
backlog
and
the
plan
for
next
to
release.
So
please
remember
that
activity
and
it's
just
a
reminder
of
course.
A
C
Yeah
I
created
a
proposal,
optimization
issue
which
is
linked
from
the
documents
which
is
related
to
well.
The
point
of
the
issue
is
proposal
of
using
a
technician
for
match:
request.
Widget
I
came
across
this
when
debugging
the
customer
issue
or
performance
issues
and
I
think
that
this
might
be
a
nice
improvement
also
on
github.com
and
rather
for
other
customers
too,
and
it's
something
I
would
like
to
look
at.
C
The
question
is:
if
there
is
overall
consensus
that
this
has
this
can
be
scheduled
soon
or
if
you
think
that
there
are
some
other
more
important
tasks.
For
example,
we
should
look.
We
should
finish
Puma
Puma
setup
first
and
perhaps
then
look
at
this
or
if
you
think
that
this
has
lower
priority
and
it
should
be
scheduled
later
so
I
wanted
to
bring
this
up.
So
it's
not
forgotten
because
it's
something
I
would
like
to
look
at
soon,
if
possible.
A
C
D
I
think
that,
like
this,
for
the
it's
rather
simple,
optimization
with
a
big
impact,
so
maybe
I
mean
we
would
still
have
to
like
checked
it,
but
like
on
your
proposal,
I
think
make
sense
to
do
it.
First,
III
kind
of
like
I
say
that,
like
most
of
the
work
we
finish
for
Puma
this
release
and
we
basically
testing
and
not
remaking
a
lot
of
changes,
things
B.
So
doing
that
it
seems
to
me
to
me
right
not
make
a
lot
of
sense.
A.
C
Some
cool,
so
in
that
case,
can
we
schedule
this
for
12.1
and
we
will
see
them.
I
suspect
it
might
be
really
tricky
task,
so
I
cannot
promise
that
there
will
be.
The
problem
is
that
there
are
many
loose
ends
which
we
have
to
take
care
of
when
they
say
when
invalidating
in
Turkish.
I
think
that
the
impact
is
will
be
too
big,
so
it's
very
attractive
to
at
least
try
to
fix
this.
D
A
A
A
A
B
Thanks
for
that,
so
this
was
a
request
from
said
I,
think,
but
I
think
this
team
should
be
aware.
We
started
to
find
defining
on-prem
use
case,
I
shied
away
from
calling
it
on
print
quality
because
I,
don't
think
all
the
engineering
to
be
responsible
quality
as
everybody
know,
but
we're
defining
on-prem
use
case
expertise
or
code
engineering
counterparts
that
have
knowledge
or
have
worked
on
customer
performance
issues
and
the
to
person
on
our
team.
Now
our
mock,
Lapeer
and
cenastone
be
on
loan.
B
No,
that's
why
we
try
to
document
it
and
make
it
really
clear.
The
the
failure
mode
is
obviously
everybody's.
Well,
the
past
quarter.
A
lot
of
focus
has
been
on
Kela
but
calm.
We
need
to
start
pulling
it
back
and
make
sure
our
self-managed
customers
are
happy
as
well,
and
this
is
kind
of
like
a
mechanism
to
to
highlight
who
is
who
has
more
knowledge
in
this
area
that
we
can
work
with
a
pull
em
in
that's
the
context.
Thank.
A
A
D
D
We
like
getting
these
things
mentioned
starting
validating
that
as
soon
as
possible,
because
most
of
these
items
are
like
we
had
them
done,
exercise
King
Keeler,
because
we
didn't
start
working
on
that
yet
and
that
kind
of,
like
figuring
out
the
defaults,
validating
that
on
the
prayer
environment,
ensuring
that
we
have
all
the
metrics
and
like
it's
kind
of
really
a
lot
of
housekeeping
right
now
to
be
done
tonight
to
finish
Puma.
So
there
is
a
lot
of
items
because
there
is
like
this.
A
D
A
B
So
yeah,
it's
not
new,
just
wanna
close
out
the
loop
that
this
team
is
equipped
for,
measuring
any
performance
improvements
or
degradation
with
their
code
changes
so
I'm,
not
tracking
anything
else.
If
there
is
please
let
us
know
and
I
will
fan
out
to
Italians
or
not
to
to
work
on
getting
this
implemented.
But
prometheus
is
their
high
speed.
Their
CPU
memory
measurement
is
already
there.
It's
in
G
underscore
memory.
B
A
I
actually
have
a
question
about
the
test
frame.
Oh
so
last
time
one
lamia
was
still
attending
the
meeting.
I
I
think
there
was
one
thing
that
one
metrics
coroner
board
cannot
be
set
up
was
or
was
not
set
up,
and
let
me
I
ran
into
a
roadblock
her
there.
She
was
working
with
other
folks.
I
am
not
sure,
was
a
progress
there.
I
thing
is.
B
Already
there
from
what
I'm
seeing
in
G
memory,
if
not,
let's
this
asks
there
one.
Second,
let
me
scroll
up
this
was
this
before
contribute
chin
yeah,
it's
just
before
the
contribute.
Yeah
well,
would
assume
it's
done
by
now,
as
a
the
last
I
talked
with
Romney
she's,
essentially
done
with
Peapod.
Now
we're
setting
up
the
other
con
trim
test
bed,
the
second
test
bed
up
or
first
an
so
cool.
D
It
seems
like,
for
my
perspective,
that,
like
we
have
everything
it's
just,
our
dashboard
is
not
really
like
it
configured
to
show
everything,
because
it's
really
like
a
test
of
the
dashboard,
like
we
I,
think
that
the
next
thing
to
happen
and
I'm
waiting
to
OMA
sampler
to
be
like
finish.
It's
like
pushing
that
to
the
tripod
and
like
seeing
and
running
load
balancer
using
our
TV.
D
That
I
may
need
to
like
to
follow
that
story
of
like
misery.
Guts
I
mean
SSH
access
to
the
servers,
editing
access
to
the
Gotama.
There's
probably
like
the
question
to
yarn.
You
can
need
some
of
these
successors
to
measure
that,
but
I
think
that's
like
we
have
to
figure
out
that
we
have
to
go
through
these
by
ourselves.
C
Yeah
yeah,
a
related
question
to
this
I
was
wondering
about,
is
what
will
be
our
strategy
to
actually
measure
the
performance
in,
for
example,
unicorn
versus
Puma,
because
Eskimo
try
to
let
me
paste
the
link
here.
So
everybody
can
hear
so
this
issue.
So
we
now
have
two
servers
right,
pre-production
and
player.
C
Puma
and
I
was
wondering
if
we
want
to
compare
performance,
unicron
versus
Puma,
how
we
would
proceed
if
he
would,
he
would
run
some
automated
tests
on
Puma
server,
then
switch
it
to
unicorn
and
run
this
test
again
and
then
compare
graphs
or
if
we
have
two
identical
environments
which
we
can,
which
we
can
compare
at
the
same
time.
So
I'm
not
exactly
sure
how
I
assume
developers
should
proceed
with
measuring
differences
between
two
episodes.
C
D
One
thing
that
we
may
have
to
figure
out
to
have
like
the
consistent
matrix
if
we
like
a
way
to
cycle
the
worker
processes
to
force
them,
we
start
so
he
always
starts
from
the
same
point.
This
is
one
thing
ii
think
is
really
like.
I
don't
think
that
we
can
run
these
things
concurrently,
because
you
should
just
like
storages
databases
and
Ready's,
but
these
I
think
that
we
can
run
this
segment
shiny
and
we
should
get
pretty
much.
C
B
I
would
recommend
looking
through
the
existing
test
time
we
have
for
rugged.
I
would
assume
we
can
replicate
the
same
format
so
have
it
have
an
issue
ongoing
issue
once
or
not
understand
have
been
doing
is
when
they
run
a
test.
They
just
put
a
table
in
in
an
update
with
like
a
screenshot
of
Ravana.
Then
you
can
compare
point
A
to
point
B
really
boring
solution.
We
don't
have
any
two
dashboards
and
like
to
saving
time
right.
This
is
probably
the
best
that
we
have
and
yeah
pasting.
B
Richard
had
itself
yes,
so
maybe
we
need
to
ping
Lupin
Sanada
Tonya
mark
lucky
I
can
help
out
as
well,
but
I
think
is
busy
with
Gilly
and
plus
1.
So
we
have
CI
jobs.
That
hits
the
end
point
and
then,
depending
on
which
end
point
you
want
to
hit,
you
can
get
hit,
commit
history
or
whatever,
and
you
can
like
generate
low
it.