►
From YouTube: CHAOSS Webinars: GrimoireLab
Description
Overview of the GrimoireLab system, one of the software projects produced by CHAOSS. Webinar in the series of CHAOSS Webinars. April 17th, 2018.
Slides: https://speakerdeck.com/jgbarah/chaoss-webinars-grimoirelab
A
A
Let's
go
to
the
next
slide.
Cows
is
working
group
hosted
by
the
Linux
Foundation
devoted
to
produce
and
integrated
open
source
software
for
analysis
of
a
development
and
to
produce
a
set
of
metrics
which
are
useful
for
analyzing,
the
health
and
the
situation
of
open
source
software
projects.
Today
we
are
going
to
talk
about
good,
more
lab,
which
is
one
of
the
tools
that
we
have
developed
in
in
respect
to
this
goal
of
producing
open
source
software
for
analyzing
software.
A
This
is
a
can
of
staff
that
you
can
produce
with
doing
more
lab
on
the
right.
You
have
a
dashboard.
This
is
a
screenshot
of
open,
a
feed
or
bitter
spot.
I
owe
you
want,
you
can
go
and
check
the
real
thing
and
there
you
have
an
actionable
dashboard
based
on
Cubana
what
you
can
drill
down
and
have
a
look
at
what's
happening
in
the
period
from
many
different
points
of
view.
A
Kimura
is
capable
of
analyzing
a
lot
of
different
data
sources
from
git
repositories
from
turret
or
pot,
Scylla
or
slack
or
IRC,
or
mailing
list
up
to
like
20
or
25,
and
produce
this
kind
of
words,
and
you
can
also
produce
reports
which
are
PDF
PDF
files,
where
you
have
a
description
of
what's
happening
in
Priya
from
the
software
development
point
of
view,
Indonesian
groom,
our
lab
is
also
a
set
of
Python
tools,
which
means
that
you
can
also
use
them
from
Python.
You
can
load
your
own
code
to
do
anything
you
may
want.
A
A
The
idea
is
to
have
as
much
information
as
possible,
ideally
all
the
information
available
in
the
original
data
source
and
store
it
into
a
database
which
is
right
here
and
that's
because
once
you
have
all
the
information
into
the
database,
you
don't
need
to
go
again
to
the
primary
sources
so
that
you
need
to
prepare
staff
again
from
the
original
API
or
from
the
original
cuyps,
which
usually
is
much
more
efficient.
So
in
the
end,
the
idea
is
having
a
copy
of
everything
here
in
the
database
as
you
compute
the
row
index.
A
So
the
first
path
for
the
data
that
we
have
is
extraction
extracting
is
done
mainly
by
Percival.
Percival
is
a
set
of
libraries
which
try
to
provide
a
single
API
for
retrieving
information
for
any
of
the
data
sources.
Then
we
have
Artful,
which
is
a
way
of
orchestrating
percival,
to
retrieve
information
from
a
large
number
of
repositories
in
parallel
continuously,
and
we
have
a
processor
of
all
that
information
which
is
from
our
elk
America
elk
is
basically
driving
out
on
Percival
and
storing
the
raw
information
in
the
database.
A
We
are
using
elastic
search
database
on
this
part.
You
have
something
which
is
quite
interesting
when
you
need
to
analyze
a
software
development
which
is
identifying
the
many
entities,
a
developer
solution
for
even
very
simple
staff,
like
counting
how
many
people
are
participating
in
the
project,
you
need
to
find
out
the
different
identities
at
the
blawker
are
using
and
merge
them.
In
addition,
you
can
also
profile
them
a
bit
so
like
for
which
company
they
are
working
in
somewhere
else.
A
They
want
to
track
that
kind
of
information
from
having
statistics
visiting
companies
in
addition
of
statistical
base
it
on
developers,
that's
the
stuff
done
by
a
salty
hat,
which
is
a
tool
that
we
have
for
managing
identities
into
a
Marriott,
V
or
MySQL
database.
That
information
is
combined
with
the
raw
indexes
to
produce
the
in
which
indexes
and
rich
indexes
are
indexes
designed
to
be
more
simple
to
visualize
and
to
be
more
simple
to
produce
reports
with
them.
A
We
have
moderate,
which
is
a
tool
the
body
to
the
configuration
of
the
software
so
moderately,
usually
how
you
run
everything
together
so
that
you
produce
a
simple
configuration
file
and
from
that
configuration
file
you
define
sorry
in
that
configuration
file.
You
define
the
data
sources,
you
define.
How
do
you
want
them
to
be
deal
with
and
Indian
moderate
is
capable
of
producing
both
the
documents
and
and
the
dashboards.
A
We
have
a
couple
of
tools
more
that
are
not
in
this
diagram
and
we
are
still
integrating
them
right
now,
one
for
managing
identities
in
our
web,
browsers
that
you
can
do
merging
of
identities
or
identifying
of
identities
in
the
browser
and
another
one
for
dealing
with
the
configuration
of
moderate,
mainly
the
list
of
stories
and
mr.
credit,
and
we
also
have
a
new
library
for
doing
enrichment.
Those
tools
have
been
out
right
now
in
the
process
of
being
integrated
with
everything.
A
The
next
slides
will,
in
a
bit
more
of
detail
with
some
of
the
projects
I'm
going
to
go
very
quickly
through
them,
so
I'm,
just
repeating
myself.
First
of
all,
first
step
is
retrieving
information
from
the
data
sources.
You
have
the
complete
list
of
data
sources.
If
you
go
to
the
purse
or
depository
and
you
have
there
a
list
of
what
is
supported
by
Kumar
lab
and
after
which
is
orchestrating
the
data
retrieval
paper.
A
So,
while
Arthur
is
basically
dealing
with
with
jobs
and
its
job
is
usually
the
retrieval
of
information
from
our
repository
and
young
jobs
can
be
done
in
the
loop
incrementally
so
that
you
can
go
and
visit
the
story
once
again
and
get
all
the
incremental
information
that
you
need.
I
mean
what
happened
in
repository
since
the
last
time
I
visited,
then
we
have
enrichment
and,
as
I
said
in
Richmond,
is
basically
combined
a
combination
of
the
information
in
identities
and
the
information
in
data
structure.
A
So
you
get
the
raw
indexes
right
here
to
combine
them
with
the
information
in
identities,
and
you
do
some
messaging
of
the
data,
and
you
do
things
like,
for
instance,
for
tickets.
It's
important
to
know
how
long
the
tickets
were
opened.
So
here's
where
you
do
the
competing
for
that
simple
metrics
like
how
long
a
ticket
has
been
open
or
how
long
does
it
take
to
answer
a
message?
A
This
is
the
way
of
running
it.
You
just
keep
installing
model
of
remark
model
which
is
the
main
packets,
the
mint
they
mean
Python,
packets,
driving
everything,
and
then
you
run
moderate
with
this
configuration
file
that
you
need
to
produce.
You
have
the
tailored
information
in
a
gumar
lab
tutorial,
where
it
explains
you
how
to
produce
these
files,
which
are
the
files
needed
for
configuring
moderate
and
that's
it.
So
today
the
idea
was
just
to
introduce
a
gumar
lab.
You
can
try
it
with
a
single
line.
A
If
you
have
a
Tokyo
driver
installed,
so
you
can
just
docker
run
this
container
here
with
this
configuration,
and
this
is
going
to
analyze,
Bluemont
lab
itself.
The
only
thing
that
you
need
is
a
github
token
that
that
you
can
obtain
from
the
github
website
graciously,
and
then
you
just
run
this
again.
You
have
details
on
the
a
gumar
lab
tutorial
by
the
way,
as
of
today,
we
are
moving
to
more
lab
tutorial
to
meet.