►
From YouTube: Data Engineering Internship - Session 2
Description
Radovan and Chris discuss dbt DAG sharding
A
A
C
A
B
A
So
so
I
would
say,
for
us
is
to
analyze
models
in
that
lineage
I
don't
know.
Did
we
have
a
chance
to
take
a
look
into
that,
but
that's
first
step
for
us
and
then
we
can
drill
it
down
like
okay.
This
is
a
full
referred.
This
is
incremental.
This
is
you
etc,
etc,
because
after
that
we
will
get
the
facts
in
dimensions
and
as
per
that
also,
we
need
to
consider
this
kind
of
stock
in
Salesforce,
but
we
need
a
list
of
models.
First
right.
B
A
Can
we
can
be
definitely
I
mean,
maybe
if
you're
able
to
put
in
the
scripture
like
to
put
a
let's
say
all
models
right
now,
very
sufficient
for
you
to
have
list
of
models,
so
you
can
track
like
okay
and
create
a
table
later
on,
like
okay,
we
have
this
model,
it's
incremental,
it's
fully
fresh
and
at
the
end
it
will
be
dimensional
effect.
Something
like
that.
If
you
know
what
I
mean,
because
with
that
guideline
for
us,
we
can
move
forward.
A
So
we
need
to
analyze
the
models
list,
all
of
them
tuck
them
and
check
the
extraction
frequency
for
Salesforce
settings.
Okay,
this
this
again,
this
can
be
done
very
quickly,
but
this
is
the
let's
say,
number
one
action
for
us
and
after
that,
what
this
book
with
that
is
it's
about
to
try
to
optimize
the
code,
keep
them
low
and
convert
full
refresh
to
incremental,
where
it's
possible.
So
we
need
to
go
one
by
one
and
analyze
all
models
together
like
okay.
C
A
A
C
B
B
C
A
B
C
C
A
C
A
B
A
B
C
B
A
B
C
A
A
Now
I
think
so
far
so
good,
because
this
is
full
referred,
don't
know
the
reason
for
that.
Maybe
we
can
do
a
curriculum
and
convert
it
to
incremental,
but
before
that
we
need
to
find
a
lineage
actually
what
we
want
to
to
wrap
up
and
what
it
will
want
to
fix
and
check
right,
yeah,
okay,
let's
go
back
to
the
documentation.
So
for
now
we
have
this
opportunity
job
between.
C
B
A
Models
related
to
cell
supports
I
create
one
additional
heading
in
this
description
of
the
issue
and
we
need
to
list
all
of
them
once
when
you're
done
with
that
we
can
attack
them
also
also
under
on
the
Fly.
You
can
do
that,
so
it's
not
a
problem
and
then
we
need
to
evaluate
each
model
and
try
to
optimize
and
see
packs
and
dimensions
what
we
can
do
there,
but
still
we
need
to
First
Define
the
scope
of
the
project.
Otherwise,
we'll
never
finish
because
you
see
it's
it's
many
of
them.
A
Definitely
I'm
for
that,
because
you
know
better
the
domain
knowledge.
You
have
the
domain
order,
the
main
scope,
what
needs
to
be
done
there
and
how
and
also
it's
a
good,
let's
say,
sweet
spot
between
our
cooperation
between
native
engineering
analytics
are
yes,
like
you
know,
the
more
knowledge
I
will
support
you
with
this
execution
when
it
comes
to
python
for
the
data
stuff,
so
I
would
say
for
us.
Let's
pick
this
agenda
here,
we
need
to
find
the
scope
first
and
complete.
The
scope
then
tag
them,
which
is
not
a
problem.
A
Just
put
the
tag
there
and
then
check
the
extraction
frequencies
for
sales
points
importance.
Teacher
can
do
that
immediately
and
then
evaluate
the
models
workflow
and
what
actually
we
can
optimize
so
for
now
for
sure
we
can
switch
99.
This
opportunity
Source
says
set
DC
from
Full
to
incremental
somehow
right
and
it
will
be
really
quick
win
for
us.
So
I
would
say.
If
you
go
to
the
issue
now
and.
B
A
A
Mean
to
do
a
logical,
let's
see,
shape
of
the
project
like
okay,
we
will
analyze
these
10
models
or
something
like
that.
Do
you
know
better
than
me,
I'm
lost
here,
because
I
know
okay.
This
is
the
first
step.
It
makes
sense
after
that
we
have
a
couple
of
branches
so
to
see
what
is
the
best
option
for
us
so.
B
A
A
Quick
catch
up
because
then
I
would
say
we
can
Define
or
finish
that
part.
Then
we
can
move
on
and
split.
The
work
like
okay
I
will
bring.
The
Crickets
is
just
a
second
to
my
attention.
Then
we
can
move
on
and
unless
the
data
and
then
see,
okay
applied
action
for
each
of
these
models
and
after
that
it
will
be
much
easier.
So
we
can
balance
the
work
we
can
move
forward
and
also
when
it
comes
to
Dimension
effects.
That
will
be
the
toughest
point.
So
I
will
extend
the
duration
of
that
thing.
A
B
A
A
Yep
I
think
we
are
done
for
now,
just
a
short
recap.
What
we
have
here
is
Amazon
models
related
to
Salesforce
recruiter
requirements
here,
five
models:
you
will
do
that,
along
with
your
instigation
check
extraction
there.
I
can
do
that.
You
can
sync
next
well,
then
avoid
the
models.
So
these
four
or
five
points
under
the
next
ability
model
workflow
and
what
we
need
to
optimize
is
on
the
plate
for
Wednesday,
so
I
would
say
we
are
still
on
good
track
and
then
we
needed
to
to
do
a
deep
dive
on
that
right.
Yeah.