►
From YouTube: Buyer Experience GitLab Lighthouse CI Server
A
Hi
everybody:
this
is
Javier
Javi
on
the
digital
experience
team
marketing
site,
so
I'm
going
to
be
talking
about
the
lighthouse
CI
server
project
that
I
set
up
pretty
much
it's
a
web
app
that
cracks
historical
Lighthouse
data
across
our
main
branch.
This
is
the
URL
to
that
project
and
it's
under
the
digital
experience
subgroup.
And
so,
if
I
look
here,
what
I
have
going
on
pretty
much?
A
A
Let's
see
what's
going
on
under
the
hood,
see
what
else
I
want
to
talk
about,
so
you
might
notice
this
recent
dip
in
performance-
and
this
is
intentional
so
before
we
were
tracking
Lighthouse
performance
scores
in
desktop,
and
so
then
this
was
an
MR
to
change
that
to
traffic
to
mobile.
This
is
to
align
with
their
Q3
okrs
of
improving
mobile
site
performance.
A
One
note
I
want
to
talk
about
quickly
is
how
the
results
can
vary
for
Lighthouse,
depending
on
the
computer,
that
you're
running
on
whether
you're
running
it
on
the
cloud
different
Hardware,
a
lot
of
Internet
different
things
that
can
affect
Lighthouse.
So
one
thing
to
note
is
that
this
Lighthouse
server
will
give
us
a
sort
of
Apples
to
Apples
comparison
for
how
to
how
our
site
is
doing.
I.
A
Think
that's
pretty
important
because,
as
we're
doing
a
ton
of
work
to
improve
our
mobile
site
performance
it'd
be
nice
to
see
what
is
affecting
the
scores
and
how
how
much?
If
I
go
into
Heroku
over
here,
we're
still
looking
at
my
notes,
go
into
Heroku.
This
is
where
I'm
hosting
it
I'm
just
using
their
lowest
tier
print
plan
to
look
at
how
the
app
is
doing.
This
is
I
mean
dashboard.
Looking
thing,
I
wanted
to
Showcase
that
you
could
run
SQL
queries.
A
This
data
is
being
held
in
a
postgres
database,
so
some
pretty
interesting
things,
Laura
you're,
killing
it
with
48
Mr
since
July,
so
good
job
with
that
next
steps
for
this
app
it'd
be
nice
to
I,
haven't
opened
an
issue
this,
but
it'd
be
nice
to
concurrently
measure
both
desktop
and
mobile,
set
performance.
That's
something
that,
though
I
house
CI
server
team
has
addressed
as
being
important
for
this,
but
they're
not
really
supporting
it.
A
So
we're
gonna
have
to
find
a
way
to
do
that
ourselves
and
in
that
issue,
where
they
talk
about
how
to
measure
both
of
those
concurrently,
they
talk.
There's
some
people
in
the
comments
that
are
talking
about
ways
to
do
that.
So
that's
one
thing:
perhaps
another
thing
that
could
look
to
do
is
moving
all
of
this
to
gitlab
infra
right
now.
This
is
essentially
segmented
from
any
of
the
other
gitlab
data.
That's
being
collected
it'd
be
great
to
unify
all
of
this
together.