►
From YouTube: UX Showcase Monitor APM Monitoring with GitLab
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
All
right,
hi,
everyone,
I'm
Nadia
from
monitor,
APM
and
today,
I-
will
give
you
a
bit
of
an
overview
of
what
we're
currently
working
on.
I
will
tell
you
about
how
we
are
evolving,
our
metrics
dashboard.
So
what
are
metrics
dashboards
metrics
dashboards
are
basically
a
tool
for
getting
insight
into
your
application
performance.
So
without
such
tools,
it
would
be
very
difficult
for
you
to
foresee
and
prevent
performance
problems,
because
using
metrics
dashboards
you
can
track
such
metrics
as
memory
usage,
latency
error
rate
things
like
that.
So
then
you
can
proactively
act
on
those
problems.
A
Essentially,
metrics
dashboards
are
visualization
tools.
So
I'll
tell
you
a
bit
more
about
about
how
metrics
work.
So,
basically
we
use
our
dashboards
to
visualize
Prometheus
metrics
Prometheus
is
a
very
popular
and
well-loved
tool
for
instrumenting
your
application
to
track
different
types
of
performance
metrics.
So
we
work
with
Prometheus.
Our
users
can
integrate
Prometheus
into
github.
A
In
fact,
they
can
also
install
Prometheus
on
their
cluster
very
easily
with
one
click
in
the
cluster
application
settings,
and
by
doing
so
they
will
enable
the
tracking
of
metrics
on
their
application,
and
they
can
see
the
metrics
in
their
dashboards
nicely
visualized.
So
they
can
see
any
weird
spikes
in
the
metrics
right
away
and
act
on
them.
So
who
are
the
people
who
use
metrics
are
me?
A
Persona
is
a
DevOps
engineer,
so
it's
someone
who
deals
all
day
with
appliquéd
infrastructure
and
they
are
usually
the
ones
who
are
monitoring,
duplication
performance
and
they
would
be
going
to
the
metrics
dashboards
and
looking
at
the
charts
and
so
on.
However,
we
noticed
the
trend
that
nowadays,
because
monitoring's
become
more
accessible
developers
are
starting
to
also
use
metrics
and
developers
are
starting
to
monitor
their
applications.
So
now,
suddenly,
there's
a
need
for
us
to,
on
the
one
hand,
build
a
tool
that
is
very
robust.
That
is
very
powerful.
A
That
would
cover
the
needs
of
DevOps
engineers,
but
also
it
has
to
have
a
very
small
learning
curve.
It
should
be
very
intuitive
and
easy
for
someone
to
just
try
it
out.
So
for
any
gifts,
lab
user
to
go
to
metrics,
to
hook
it
up
to
their
application
and
to
see
if
they
can
track
any
interesting
metrics
that
would
help
them
keep
their
application
running
smoothly.
So
currently,
metrics
dashboards
is
our
huge
priority
on
the
APM
team.
A
So
we
have
a
lot
of
ongoing
efforts
to
evolve
github
metrics,
so
some
of
you
have
probably
seen
this
comment.
Sid
commented
on
it
and
it's
also
in
our
handbook.
If
we
can
get
our
user
to
adopt
one
more
stage,
they
will
be
three
times
more
likely
to
convert
to
a
paying
customer.
So
knowing
that
it
becomes
very
important
for
us
to
increase
the
adoption
of
the
monitor
stage
and
specifically
right
now
we're
focusing
on
metrics
dashboards.
So
how
can
we
get
get
lab
users
to
adopt
metrics?
A
What
what
are
the
things
that
we
can
be
doing
to
make
it
easier?
So
one
of
the
things
we're
doing
is
dog
fooding.
We
have
an
internal
customer
within
github,
our
own
infrastructure
team.
That
can
be
a
really
great
source
of
feedback
for
us.
So
if
we
were
to
dog
food
with
the
infrastructure
team,
we
would
be
able
to
get
feedback
much
faster
and
we
would
have
direct
access
to
a
customer.
A
Who
will
you
know
who
helped
us
build
a
good
product
so
we're
seeing
it
as
a
very
powerful
way
to
build
a
product
together
with
like
a
real
engaged
customer,
so
we're
partnering
with
our
infrastructure
team?
Who
is
helping
us
define
what
are
those
gaps
in
the
experience,
and
we
have
an
epic
that
we're
currently
working
on
to
fill
those
gaps
to
make
sure
that
we're
coming
closer
to
bridging
those
future
gaps.
A
A
Who
are
those
people,
and
actually
an
interesting
thing
that
we
noticed
in
our
research
is
that
there
are
many
developers
who
don't
need
such
robust
functionality
as
our
infrastructure
team,
but
whose
main
point
pain
point
is
the
steep
learning
curve
for
enabling
monitoring
so
instrumenting
your
application
to
start
tracking
metrics,
it's
very
very
difficult.
So
we
want
to
learn
more
and
we
want
to
dig
deeper.
We
want
to
better
understand
what
are
those
pain
points
of
our
average
user
and
who
are
all
of
those
different
users
that
we're
serving
so
to
do
that.
A
A
For
example,
an
incident
some
kind
of
problem
within
the
application
reviewing
metrics
is
a
very
important
part
of
the
that
workload.
So
in
this
research
we
want
to
better
understand
how
all
the
different
parts
of
triage
workflow,
like
review,
metrics
and
looking
at
Ayres
and
escalating
the
problem
to
other
team
members.
How
do
metrics
fit
in
that
workflow?
Because
if
we
understand
it,
then
we
can
connect
metrics
to
other
part
of
gitlab
in
a
way
that
would
increase
the
adoption
of
metrics.
A
So
we're
hoping
that
we
will
uncover
some
interesting
insights
through
this
research.
Another
research
we're
running
is
how
our
users
use
graph
Anna,
so
we're
interviewing
a
mix
of
internal
and
external
participants
to
see
what
is
the
difference
between
how
how
our
own
infra,
how
other
external
users
how
they
use
graph
Anna?
What
are
their
preferences?
A
What
are
the
top
features,
etc,
etc,
because
because
our
own
team
uses
graph
Anna
for
their
dashboarding,
it's
important
for
us
to
understand
what
is
it
that
they
need
and
what
is
it
that
they
love
so
much
and
what
are
also
the
gaps
in
graph
Ana's
user
experience
that
maybe
we
can
fix
in
G
lab.
We
can
solve
those
problems
so
and
of
course,
we
also
want
to
focus
just
on
the
overall
user
experience
and
the
UI
metrics
are
currently
at
viable
level
of
maturity.
A
There
are
plenty
of
other
things
to
fix
plenty
room
for
improvement,
so
I'll
just
go
through
some
of
the
main
things
that
we
are
currently
implementing.
This
is
the
work
that
is
happening
in
this
milestone
and
in
the
next
milestone,
so
we
are
going
to
allow
the
user
to
view
chars
in
fullscreen,
which
is
very
helpful
for
someone
who's
reviewing
a
lot
of
data,
it's
good
to
have
all
of
that
screen
real
estate,
the
where
you're
also
able
to
view
more
of
the
legend.
A
A
We're
also
adding
chart
annotations,
which
are
essentially
chart
comments
that
developers
use
to
add
extra
context
to
the
events
in
charts.
So
there's
a
spike
in
symmetric.
You
can
add
a
comment
directly
to
the
spike
and
now
everyone
else
on
your
team
knows
what
happened.
That
makes
it
easier
to
investigate
problems,
because
now
there's
this
extra
context
attached
to
it
and
it's
even
possible
to
automatically
create
those
comments.
So
a
lot
of
the
infrastructure
teams
would
be
automatically
created
comments
for
deployments,
for
example.
A
A
Another
important
problem
that
we're
solving
is
navigating
the
dashboards,
so
any
given
team
can
be
dealing
with
like
hundreds
of
dashboards,
so
it
can
be
very
difficult
to
navigate
through
them.
So
we
have
a
search
function
that
you
can
use
to
navigate
the
dashboard
or
navigate
between
different
dashboards
and
we're
now
introducing
a
way
to
star
the
dashboards.
So
you
have
quick
access
to
your
favorite
dashboards
and
another
very
exciting
issue
that
we're
working
on
actually
Nick
post
is
running.
A
It's
very
big
effort
to
bring
in
more
consistency
into
how
we
do
data
visualization,
inget
lab.
Currently,
there's
a
lot
of
inconsistency.
I
think
many
of
the
teams
that
deal
with
data
visualization,
you
know,
we've
been
at
a
low,
surely
level
and
we're
working
hard
on
introducing
new
functionality.
In
all
honesty,
I
know
it's
not
just
a
problem
of
APM,
but
everyone
is
struggling
with
working
on
pyjamas
components.
It's
just
something
that
gets
pushed
back
and
we
prioritize
stage
work.
A
So
this
effort,
I,
think,
is
going
to
be
very
helpful
to
get
all
hands
on
deck.
To
finally
make
you
know,
create
all
other
components
make
it
consistent,
reduce
the
cost
of
implementation,
which
is
in
the
end.
What
were
you
know
what
we
want
to
achieve
as
well,
so
yeah,
it's
all
very
exciting,
and
thanks
Nick
for
running
this
yeah,
and
this
is
it.
This
is
the
overview
of
what
we're
working
on
APM.
Let
me
know
if
you
have
any
questions.