►
From YouTube: Quality (Mek) Group Conversation (Public Livestream)
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
A
B
I
make
thanks
for
the
great
presentation
I
really
enjoyed
reading
it
and
also
before
I
actually
dug
to
my
question.
Thank
you
very
much
for
all
the
hard
work
in
the
Quality
department.
Jenny
has
helped
about
at
geo
and
I
really
appreciate
everything
that
you
are
doing.
So
thanks
for
that.
So
I
had
one
question
on
slide.
8.
B
But
most
of
the
items
in
there
seem
to
be
dealing
with
performance
of
upgrades,
I'm,
also
interested
in
learning
and
understanding
what
your
thoughts
are
on
actually
testing
for
upgrade
robustness,
as
in
customers,
may
attempt
to
upgrade
and
may
fail
for
various
reasons,
and
that
obviously
may
have
some
impact
on
their
their
production
systems.
So
that's
my
question.
Cool.
A
I
can
give
an
intro
and
I'll
pass
it
on
to
Tonya
who's,
leading
this
effort.
So
we
are
constrained
with
effort.
We
are
forced
to
work
smart,
so
we
are
killing
two
birds
in
one
stone
with
upgrading,
implicitly
and
also
testing
performance.
Now
we
that
is
only
one
facet.
The
other
facet
is
the
combination
of
installation
architectures
that
our
self-managed
customers
have.
We
can't
test
all
of
them,
which
is
why
the
reference
architecture
comes
into
play.
A
So
there's
two
streams
of
work,
testing
and
also
design
the
architecture
having
them
be
on
the
architecture
and
those
are
the
ones
that
we
fully
tests
and
tell
our
customers
hey.
We
have
vetted
this
this
architecture,
so
those
are
the
two
facets
in
play.
Tonya,
if
you
want
to
chime
in
into
for
the
details,
what
your
thoughts
are
yeah.
C
Yes,
so
that
was
a
great
overview.
We
we
are
doing
this
a
kind
of
a
multi-prong
effort
and
we
can
add
more
information
into
that
epoch.
We
are
testing
like
Mike
said.
The
the
performance
is
definitely
a
viewpoint
of
ours,
but
I
know
that
Jenny
is
also
addressing
no
downtime
upgrades
and
being
able
to
test
those
making
sure
that
that's
robust,
that's
our
main
focus
right
now.
As
far
as
the
no
downtime
upgrade
part
Jenny.
Is
there
anything
you
want
to
add
there
about
additional
areas
that
you're
looking
at.
C
D
D
If
we're
grooming
are
we
grooming
for
Bingley
problems
that
are
consistent
or
getting
started
problems
or
both
and
whether
we
should
be
considering
some
type
of
either
Q&A
session
associated
with
that
to
surface
these
kinds
of
bugs,
because
I
think
a
number
of
folks
at
least
you
know,
Stan
Stan
basically
ran
a
Q&A
session
and
he
he
had
ten
folks
show
up
with
ten
different
problems
which
points
to
we
nefeli
need
a
guru
but
also
kind
of
points
to.
We
should
make
sure
we're
surfacing
me
somehow.
Yes,.
A
So
it's
the
circle
of
life,
thanks
Tonya.
For
that
term,
this
effort
actually
came
up
from
the
discussion
with
me
and
Stan,
and
the
issues
that
I
identified
here
are
the
ones
that
are
blocking
people
from
onboarding
or
it's
a
persistent
bug,
that's
affecting
productivity.
The
grooming
is
set
up
between
me,
Stan
and
enlisting
other
folks.
We
are
going
to
start
out
this
week
and
to
start
and
then
for
next
week
we
may
move
the
time
around,
so
more
people
can
join.
A
But
to
answer
your
question,
it
is
designed
to
immediately
fix
the
the
hardship
and
getting
set
up,
and
this
is
hopefully
a
temporary
thing.
We
have
one
new
hire
in
engineering,
particularly
Albert,
who
may
be
taking
this
on
how
his
time
that
we
work
on
stopping
later
on
to
to
have
full-time
engineers.
Looking
at
the
GDK
we
did-
that's
that's
kind
of
where
we
are
at
stan
has
been
setting
up
in
Q&A
sessions.
Monthly
I
expect
that
to
still
continue
and
the
feedback
from
that
will
feed
into
this
grooming,
yeah.
B
Want
to
I
didn't
want
to
come
in
with
the
second
question
right
after
so
that's
in
space.
So
one
question
that
I
had
is
I
know
me
working
on
reference,
architectures
and
I
think
this
is
a
really
cool
idea,
also
for
testability
and
I
really
like
that,
but
I
was
wondering:
do
we
actually
have
data
available
from
our
customers?
You
know
so
we
we
know
you
know
like
at
this
point
with,
let's
say
a
thousand
uses.
B
You
know
forty
percent
of
our
customers
deploy
in
that
specific
way,
or
they
follow
a
specific
pattern
so
that,
if
we,
but
when
we
design
these
reference
architectures,
we
are
not
designing
something
that
we
think
is
great,
but
something
that's
actually
relatively
close
to
what
our
customers
are
already
doing.
Great.
A
Question
so
we
are
in
fact
actually
doing
this
and
we
are
in
the
process
of
adding
it
and
it
marks
it
handbook
on
the
playbook
or
say
the
information
that
we
asked
customers.
This
is
an
output
of
the
self
managed
together
the
working
group
where
we've
come
up
with
a
few
logs
on
usage
patterns,
I
think
Italy,
workhorse
production
logs
and
there's
not
a
lot
I
forgot,
but
essentially
we're
looking
at
endpoints
and
response
times
of
their
environments
and
also
the
type
of
data
that
we
ask.
A
Is
it
mostly
binary
digit,
binary
data
or
text
data,
so
they're
checking
in
JPEGs,
which
is
a
really
cognitive
load
on
the
storage
and
then
maybe
talk
about
either
in
fsor
as
I
see
later
afterwards.
But,
yes,
we
are
having
those
discussions
and
if
you
look
at
the
the
link
in
the
slides
and
I
will
not
vocalize
our
customer
names,
we
have
outreach
issues
to
those
two
big-name
customers
to
have
this
information
in
and
I
believe
that
that
should
be
more
clear.
Given
this
week,
I've.
C
Linked
the
merge
request
that
Mike
mentioned
into
the
document,
if
there
is
any,
if
anyone
has
any
input
on
things
that
we
can
add
to
the
playbook
right
now,
it's
just
just
an
initial
set
of
steps,
but
they're
not
a
sub
two
step,
three
additional
things
that
we
can
do.
We'd
love
you
to
contribute.
So
if
you
have
any
ideas
there,
please
please
help
all.
B
E
You,
sir
Thanks,
alright,
thanks
for
again
particular
presentation
and
for
adding
a
feature
abuse
as
well.
I
know
that
we've
been
talking
about
how
to
improve
the
or
the
way
we
tackle
the
backlog.
I
know
that
your
team
has
been
doing
quite
a
lot
of
working
in
either
planning
or
implementing
improvements
to
the
tolling.
E
A
A
This
is
a
current,
mr
coach
slack
channel
that
I
believe
the
plan
was
to
link
together
and
have
it
be
an
open
slack
channel
with
Kitter,
so
people
can
just
find
and
click
join
and
then
and
then
interact
with
the
team
and
them
our
coaches.
That's
one
facet
that
that
we
were
looking
at
as
part
of
contribute,
actually
I,
think
Ben
and
I,
and
an
r/a
were
which
were
huddling
together.
We
did
not
have
the
chance
to
investigate
this,
yet
it
goes
in
line
with
like
mocha
OGS,
where
there's
a
button
on
the
project
page.
A
You
click
one
button
really
easy
one
quick
sign
in
and
then
you
get
interact
with
the
engineers
right
away
at
main
trainers.
That's
one
facet,
I
think
automation
still
needs
to
be
improved.
There's
no
clear
line
of
I,
don't
want
to
say
ownership,
but
like
the
standby
effect
and
intentionally
is
there
because
we
have
so
many
Amar's
in
so
many
facets
of
the
application,
and
when
you,
when
you
randomly
ask
one
maintainer
to
review
he
or
she
might
not
have
the
best
knowledge
like
hey.
A
This
is
not
my
area
of
expertise
so
that
that
link
isn't
there
yet
so,
there's
still
some
short
hair
can
review
this.
Can
someone
else
take
it
on
I
think
adding
a
flavor
of
maintaining
of
what
specific
area
or
which,
which
group
do
you
belong
to?
Might
help
here
to
review
that
type
of
em
are
right.
Now
it's
a
it's
an
innovation
of
the
river
but
I
do
think
the
single
repo
will
help
I
think
that's
one
less
project
to
manage
going
forward
and
we
do
need
to
actually
make
more
attention
to
the
satellite
projects.
A
E
A
F
A
E
Don't
really
have
any
data
on
this,
but
I
mean
I.
Do
have
the
feeling
that
essentially,
since
it's
so
so
new
contributors
will
get
confused
at
least
in
the
first
weeks
and
I
know
that
I
know
Kyle.
You
and
Ray
have
been
talking
about
this
analysis.
There's
an
ongoing
conversation,
but
yeah
I
think
I
think
it's.
It
would
be
it
going
to
start
to
be
entities
and
weight
from
our
side
from
the
Tony
delicious,
a
more
than
happy
to
help
I
see.
A
Let
me
let
me
link
you
to
the
issue
and
there's
an
issue
called
final
changes
for
single
repo,
and
there
are
some
community
contributions
being
asked
if
they're
already,
of
course,
cross-linked
using
that
and
then,
if
their
improvement
items,
let's
work
with
the
delivery
team
to
make
that
make
sure
it's
improved
and
I
may
have
cut
off
somebody.
So
please
vocalize.
If
I
did.
A
A
F
You
so
on
what
items
in
the
backlog
like
which
of
the
three
items
in
the
backlog,
Decius
and
I
like
most
important,
to
help
continue
kind
of
the
growth
in
our
department.
F
A
Sorry
slide
18
you're,
saying
the
growth
in
our
department
or
the
most
important
thing.
How
about
most
important
thing?
Okay,
I
would
actually
say
the
middle
one,
which
is
guarantee
roll
back
path
for
self-managed
I.
Think
we're
testing
upgrades
right
now.
We
also
need
to
test
roll
back
because
unknown
unknowns
if
they
upgrade
and
they
run
into
something
we
don't
want
them
to
be
dead
in
the
water,
and
this
came
up
when
the
self-managed
working
group
as
well
change
management.
We
can't
expect
the
customer
to
run
as
fast
as
we
are.
A
Sometimes
they
have
two
or
three
months
three
times
just
to
provision
something
and
a
fallback
plan
to
that
is
to
roll
back
so
they're,
not
sitting
in
the
water,
so
downgrade
testing,
no
food
for
thought
books
on
the
call.
It's
also
important
and
I'll
move
that
to
the
top
of
the
list.
Next
time
you
have,
the
GC
goes
clear.
Thank
you
for
the
question.