►
Description
Memory Chronicles: Manage, Trace and Fix Memory Leaks - Giovanny Gongora, NodeSource
Speakers: Giovanny Gongora
Let’s start talking about taking the time to set up a proper test environment to repeat the exact same conditions as in production. Performance metrics? Response time higher? Memory leak? We would walk through different memory leak situations to identify and analyze patterns that produce an increase of memory, CPU usage, and the load average over time, without any apparent reason. Learn about memory and how to access Node.js memory using a V8 Inspector & Chrome Dev Tools, create a dump of the heap memory for the inspected application, with a lot of details about the memory usage, watching memory allocation in real time and finally a compilation of best practices about how we could end fixing the memory leak. In the end, we'll be able to use the right tools to monitor, understand, and debug the memory consumption of a Node.js application on time.
A
A
A
What
I'm
going
to
try
to
do
today
is
show
you
or
explain
you
a
few
cases
where
you
can
have
memory
issues
I'll
try
to
go
through
most
of
the
cases.
It's
really
hard
to
actually
have
a
definite
guy
or
something
like
you
will
see
in
a
YouTube
video.
This
one
words:
one
link
mega
4k,
full
HD,
fixes
all
the
issues.
It's
really
hard
to
get
that,
especially
for
memory
issues.
Why?
Because
they
can
happen
for
many
reasons
and
that
you
can
fix
them
on
me
how
many
ways.
A
So
what
we
are
going
to
do
is
just
try
to
talk
about
some
of
the
cases
I
have
seen.
Some
of
them
are
worse
than
others,
but
that's
it
yeah,
my
name.
It
I
think
you
all
know
now
my
name
and
my
birthday,
that's
that's
my
email,
that's
my
Twitter!
If
you
wanna
reach
out
so
sometimes
I
don't
treat
too
much,
but
if
you're
into
that,
that's
my
tutor.
A
What
was
this
talk
is,
you
know,
just
has
a
safe.
Obviously
this
is
more
about
talking
and
if
you
really
want
to
or
if
you
have
a
memory
leak
or
you
have
a
something
wrong
with
really
your
application
and
you
want
to
probably
discuss
a
bit
a
little
bit
more
in
deep.
We
can
just
talk
later.
The
talk
or
I'll
be
giving
a
workshop
at
2
p.m.
just
around
the
corner,
about
memory
leaks,
and
we
can
just
run
some
examples
and
just
hold
let
you
know
how
to
fix
them
and
how
to
identify
them.
A
A
You
can't
remember
the
number
of
the
times
or
you
can
remember
the
last
time.
Okay,
okay,
yeah
I
mean
that
this
is
this
pretty
interesting,
because
that
there's
something
great
about
JavaScript
and
node.js
in
general
is
that
you
really
don't
need
to
learn
from
the
beginning
about
how
to
deal
with
this
type
of
issues
and
I.
Think
that's
what
makes
node
and
JavaScript
really
a
language
that
everybody
can
get.
It
start
and
actually
start
to
build
stuff,
because
you
really
don't
need
to
learn
this.
A
The
issue
is
that
when
you
have
to
learn
this,
it's
on
Friday
night
when
production
is
down-
and
someone
is
calling
you
hey,
why
did
the
server
is
what
application
is
not
working?
Why
wait
Sam
or
have
you
seen
this
application?
Well,
why
this
server?
What
this
service
is
not
running,
so
it
does
it
don't
worry,
it
happens
in
real
life,
it
happens
in
T.
It
happens
in
all
the
companies
do
not
worry
you're,
not
alone.
A
It
happens
to
only
including
the
best
one
and
I
can't
make
sure.
Be
sure
that
that's
true,
we
have
my
add
node
source.
We
have
been
working
with
different
companies.
Some
of
them
have
products
that
you
are
using
right
now
and
with
real
critical
process
that
they
need
to
handle
in
node.js
applications
and
they're
really
and
complicated
applications.
A
A
What
the
timeline
is
or
what
the
hippo
prefer
we
hit.
Profiler
is.
The
timeline
is
just
a
tool
that
I
think
how
many
of
you
have
used
a
chrome,
dev
tools,
yeah.
Okay,
you
can
see
that
the
timeline
determine
is
really
not
something
that
I'm
just
going
to
tell
you
exactly
what
the
issue
is.
That
is
pretty
much
for
you
to
understand
when
or
at
what
time
it
is
started
to
slow
down,
because
one
of
the
things
with
with
node.js
application
is
that
you
not
only
can
do
servers.
A
A
So
it
doesn't
mean
that
that
the
leaks
or
memory
issues
only
happen
on
applications
that
you
get
little
or
something
like
that
or
customers
just
joining
the
your
application,
or
something
that's
probably
one
of
the
biggest
issues
to
to
try
to
understand
that,
because
you
know
something,
and
you
know
that
this
could
probably
be
an
issue
it
doesn't
matter.
It
doesn't
mean
that
it's
going
to
apply
every
sound.
So
you
have
to
be
really
sure
that
you
know
your
application.
You
know
what
are
the
bottlenecks
and
that's
something
we
are
going
to
discuss
later.
A
A
A
No,
it's
four,
so
ft
SS
t
is
for
it,
snapchatting
need
or
five
it's
not
yet
again.
Why?
Because
it
really
depends
on
the
application,
but
if
you're
just
going
to
start
digging
and
just
start
to
understand
what
application
is
doing,
you
really
need
to
embrace
that
work.
That
way
to
figure
out
things
and
I
didn't
invent
this
I.
You
know
what
up
it
was
not
surely
someone
at
the
Gmail
team,
it's
cool,
Lauren
alley.
So
thanks
for
thanks
to
her
for
that.
A
What
this
means
is
that
you
take
heat
profile
at
the
beginning
of
your
application
as
a
cold
star.
Why?
Because
you
need
to
make
sure
how
your
application
is
in
memory
when
it
starts,
then
you
start
to
increase
the
load,
so
you
just
start
sending
more
and
more
requests
or
whatever
your
application
does
and
after
certain
point
could
be
one
minute.
Two
minutes
three
minutes
you
take
an
order.
He
was
a.
He
is
an
option
and
after
more
time
it
could
be
another
two
or
three
minutes.
You
take
a
different
one.
A
So
why
would
you
do
that?
Or
what
would
you
take
three
snaps
just
at
different
times?
Well,
it's
because
you
really
want
to
make
sure
which
objects
are
the
ones
that
are
in
you
know
at
the
time
and
the
only
way
to
know
that
is
knowing
how
your
application
started
and
after
a
period
time
which
objects
are
newer
or
didn't
disappear
from
the
first
snapshot
compared
with
the
second
a
snapshot,
and
that,
if
you
want
to
make
complete,
be
completely
sure
you'd
compare.
The
second
is
option
with
the
third
a
snapshot.
A
A
So
there's
something
in
the
chrome,
dev
tools
and
I
think
how
many
of
you
have
take
HIPAA
snapshot
before
yeah
and
there's
a
bunch
of
columns
that
you
see
a
percentage
of
for
each
column,
and
usually
you
don't
know
what
what
any
of
them
mean
and
is
fine
I
mean
it
takes
some
time
to
understand.
All
of
that
in
what
is
important,
but
was
something
I'd
really
highly
recommend
is
to
check
our
for
the
distance.
A
There's
a
column
called
distance
in
the
in
the
chrome
developer
tool
when
you
take
a
HIPAA
snapshot
and
something
I
have
seen
for
for
it
for
a
while.
Is
that
the
closest
and
the
AR
day,
the
constructor
that
has
the
the
closest
or
the
minimum,
the
closest
it
will
be
like
two
or
three
number
in
the
distance
column,
it's
usually
the
one
that
is
responsible
for
the
memory
leak.
So
that's
something
it's
not
completely
true
for
all
the
applications,
but
I
have
seen
many
cases
and
that's
what
I
can
usually
recommend.
A
Yeah
so
check,
for
contrast,
constructors
are
usually
check
some
of
the
cold
stock.
That's
something
I'm
going
to
recommend
at
the
end
that
how
many
of
them,
how
many
of
you
name
your
functions
or
you
just
make
function
the
arguments
and
all
the
stuff?
Well,
when
you
don't
name
your
fractures
and
I,
think
there
was
kind
of
fix
by
the
time
and
and
I
think
in
Note
8.
A
Something
was
implemented
on
VA
that
it
didn't
matter
that
you
did.
You
didn't
name
your
applications,
because
something
in
b8
was
more
intelligent
than
us,
so
it
was
kind
of
generally
names
for
unnamed
functions.
However,
when
you
start
to
look
at
the
HIPAA
snapshot,
you
start
to
see
in
the
call
stack
that
there's
a
lot
of
functions
and
they
don't
have
name
so
there's
actually
no
way
for
you
to
track
where
this
call
is
being
done
or
or
how
you
can
actually
fit.
A
They
fix
the
issue,
so
highly
recommend
just
put
a
name,
don't
put
an
X
or
like
Y
or
something
like
that
on
your
functions.
It's
not
really
helpful
seriously
when
you
are
trying
to
debug
stuff
in
especially
for
HIPAA.
Sorry
in
memory,
you
really
want
to
make
sure
that
you
have
all
the
compute
information,
and
this
is
only
talking
if
your
application
is
not
unified
or
transpile.
A
This
is
something
I.
This
is
something
that
people
actually
don't
know
is
that
you
can
generate
Riddick
redic,
it's
not
shirts.
What
this
means
is
when
you
take
a
HIPAA
snapshot,
it
takes.
You
are
going
to
see
all
the
strings.
Most
of
those
strings
have
sensitive
data.
So
if
you
are
going
to
send
these
HIPAA
snapshot
to
someone,
just
please
make
it
protected.
So
any
of
the
strings
I
mean
strings
are
at
the
content
of
this
race.
A
That's
not
important
for
the
person
or
the
company
or
whatever
or
the
service
that
is
going
to
face.
They
they
issue
or
actually
do
consulting
for
you,
but
do
it
because
it's
probably
going
to
have
all
your
secrets
of
your
application,
including
credit
cards,
information
and
all
that
stuff,
so
really
take
care
of
that.
A
But
but
you
can
naturally
do
core
dumps
on
note'
and
and
in
this
way,
because
when
they
talk
about
this,
people
do
not
do
not
know
that
you
can
actually
do
that
and
actually
have
more
interest
in
information
because
we
card
them.
We
can
in
real
time
actually
it's
like
a
travel
time
when
you
can
actually
know
the
state
or
an
application
or
something
at
certain
point
or
a
certain
period
of
time,
and
that
doesn't
happen
with
the
HIPAA
snapshot.
A
You
are
the
one
that
is
trying
to
collect
all
the
information
in
the
memory
at
certain
time,
so
curtains
are
really
interesting.
They
take
a
while
to
understand,
especially
if
you
are
not
familiar
with
any
longer
on
suppose,
plus
a
or
C
to
create
a
card
occurred
on.
Usually
what
you
have
to
do
is
most
of
the
time
just
send
us
a
sign
now
to
award
or
you
can
put
on
your
card
the
process
that
award.
A
But
what
that
means
is
that
you
have
to
modify
your
code
to
actually
trigger
the
cardem
and
the
curtain
in
the
court.
Don't
is
actually
is
expected
to
be
triggered
when
there's
a
new
handle,
as
as
exception.
However,
you
can
I
mean
you
can
do
that
or
you
can
just
use
this
option
when
you
run
your
application,
so
you
don't
have
to
modify
anything
and
it's
going
to
avoid
every
time
you
have
a
new
handle
exception.
A
It's
just
way
better
do
not
modify
your
code
and
especially
if
you're
on
production,
you
cannot
modify
your
code
to
actually
get
all
the
in
handle
expressions
or
every
time
you
throw
something
to
do
something.
Weird
and-
and
something
is
when
you
get
the
core
file
with
the
previous
one,
with
the
process
that
award
you.
Can
you
get
more
stack
traces
that
are
not
really
relevant
today
to
the
to
the
problem?
Why?
Because
you
are
modifying
the
code
to
actually
make
it
break
at
some
point
to
actually
abort.
A
So
don't
actually
do
it
if
you
really
have
to
make
sure
that
you
are
awarding
others
at
some
point
and
you
are
not
able
to
trigger
I
mean
because
it's
really
it's
pretty
normal
I
mean
how
many
of
you
have
had
an
issue
that
is
on
production,
but
it
never
happens.
It
never
happens
on
development
yeah.
It's
like
everyday
right.
It's
it!
That
happens
all
the
time
and
usually
there's
a
few
things
that
you
can
do
for
that,
especially
with
the
stress
testing.
A
A
You
can
use
that
to
actually
inspect
and
get
all
the
information
that
I'd
like
you,
actually
need
from
the
core
file
there
this
week
we
can
go
more
in
deep
and
that,
but
that's
that's
trying
to
explain
how
a
debugger
works
for
C++
and
I
think
that's
just
too
much.
There's
something
called
L
know
that
it's
kind
of
plug-in
for
adult
Eevee
yeah
a
little
to
be.
What
was
was
the
thing
with
this
is
that
you
can
actually
get
a
stack
trace,
JavaScript
s
actress,
so
it's
really
useful,
but
I
think
is
I'm,
not
sure.
A
If
it's
highly
level
at
this
moment,
something
it's
something
that
is
not
good
about
core
part
is
that
you
have
to
make
sure
that
the
debugger
that
you
are
using
and
the
core
file
are
on
seeing
conversions.
Why?
Because
if
you
are
using
at
the
work
or
part
that
has,
for
example,
at
something
in
v8
or
a
different
version
of
you
than
you,
the
worker
has
then
you're
going
to
have
a
problem,
because
if
that
change,
if
there's
something
that
changed
between
versions,
then
the
de
Vitoria
is
not
going
to
be
able
to
catch
that.
A
So
you
are
missing
information.
You
are
making
core
files
out
of
nothing
and-
and
it's
not
not
not
going
to
be
worth
it
because
you're
not
getting
the
whole
story
or
the
whole
picture
for
the
application,
and
that's
one
of
the
things
I
would
say
just
make
sure
that
you
are
running
at
the
same
variance
between
to
the
blogger
and
the
moment,
you're
building
the
core
file.
A
If
not,
then
you
are
going
to
have
a
hard
time
trying
to
figure
out
why
your
core
file
doesn't
have
certain
things
because
is
probably
named
different
in
some
times
he
ate
something
in
the
structure.
Just
changed
the
name
and
because
of
that,
there's
no
way
to
know
that.
There's
no
way
that
the
debugger
and
the
core
files
are
going
to
link
that
because
they
don't
know
what
happened
from
Isis.
A
A
You
can
accent
as
you
get
the
from
ice
finish,
but
it
doesn't
mean
at
the
state
of
the
promise
can
change
during
that
time
until
it
awards.
So
that's
one
of
the
unknown
things
about
promises.
That
means
this
is
one
of
many,
and
I
think
you
on
the
previous
talk
just
so
that
so
make
sure
that
you
understand
from
isis
really
well,
if
not
you're,
going
to
have
a
really
hard
time.
I
think
we
all
have
we
promises
already
a
hard
time,
but
I
actually
recommend
I
think
away.
A
So
please,
if
you
still
not
using
a
secure
way
just
do
it.
It's
really
important
that
you
try
to
embrace.
This
is
why
cleaner
is
much
easier
to
the
block
and
it's
going
to
be
less
painful
I.
Actually
it
really
highly
recommended
so
I
have
some
tips
for
different
situations.
I
know
that
people
usually
look
for
something
to
not
do
so.
They
can
avoid
the
sort
of
things.
A
One
of
this
is
on
server-side,
rendering
I'm
not
saying
you
should
not
do
server-side.
Rendering
is
that
you
should
do
it
properly.
Why
I'm
saying
this,
because
we
have
seen
companies
the
doing
server-side,
rendering
that
they
have
are
really
big
our
views
on
how
the
I
do
the
transportation
or
the
tour
or
all
the
process
of
the
service
are
rendering?
And
what,
when
you
do
that
you
have
to,
you
have
to
know
that
there's
a
cost,
there's
a
trade-off
and
what
that
means
is.
A
They
are
simple
and
no
jeaious
application
can
start
on
20
megabytes
or
30
megabytes.
If
it's
too
complex
of
a
memory
on
a
cold
star,
however,
for
a
service
on
Ranieri,
it
can
start
on
200
megabytes
of
memory,
300
megawatts
hot
memory,
and
that's
just
starting
it's
doing
nothing.
Why?
Because
most
of
the
server-side,
rendering
all
the
components
or
everything
that
is
being
processed,
it's
a
storage
memory.
A
So
when
you
to
keep
a
snapshot
of
services
rendered
applications,
you
are
going
to
see
many
strings
or
many
template
in
the
strings
on
your
memory
and
that's
how
pretty
much
a
server-side
rendering
works.
So
it's
going
to
be
white
harder
to
actually
strip
all
those
objects
or
strings
or
whatever
service
our
engine
is
created.
So
if
you're
using
server-side
rendering
make
sure
that
you
know
what
is
the
trailer
trade-off
and
you
know
the
cost
and
that's
something
people
do
not
know
the
same.
They
usually
and
I
understand
that.
A
There's
reasons
for
some
architects
to
use
something,
but
usually
memory
is
not
that
because
we
are
so
used
to
when
something
doesn't
work.
It
just
just
put
more
more
memory
more
cores
and
let's
make
because
it's
pretty
easy
in
AWS,
just
to
increase
the
it
doesn't
matter
if
he,
if
he's
not
working,
just
put
more
memory
and
more
cores
and
we
were
fine.
A
Ok
still,
it's
monitoring
tools,
make
sure
that
you
are
choosing
the
right
monitoring
tools.
Why?
How
many
of
you
have
been
using
dynaTrace
at
some
point,
probably
of
you?
Well,
if
you
do
see
paper
fire
or
the
HIPAA
snapshot
you
are
going
to
or
if
you
see
the
stock
a
the
coal
stock.
What
you're
going
to
see
is
that
dynaTrace
is
crapping
almost
every
function,
every
call.
Why,
because
is
dynaTrace,
needs
to
know
what
is
happening
with
your
application,
but
the
way
they
do
it,
they
grab.
A
A
That's
agents
or
monitoring
tools
that
they
just
grab
the
clothes
because
they
need
to
get
information,
but
it
doesn't
mean
that
they
are
very
responsible
and
that's
something
that
you
were
with
CPU
profiles
in
keep
CPU
snapshots,
so
just
make
sure
a
deltas,
a
few
tools
that
cause
an
overhead
on
memory.
But
you
really
have
to
understand
how
the
monitoring
tool
that
you
are
implementing
actually
works.
If
not,
then
you're
going
to
have
a
hard
time
and
your
start
calling
dynaTrace,
hey,
hey,
you
just
brought
my
application
and
that's
that's
not
good
cause.
A
A
Okay,
oh
I'm,
not
sure
yeah.
If
you
can
see
it
well,
this
is
just
how
a
application
you
saw
press
API,
it's
just
pretty
simple.
Nothing
special
is
it
has.
It
has
a
database
that
has
been
loaded
with
random
data,
and
this
is
it
probably
the
closest
exercise
that
I
can
find
to
explain
people
why
you
should
close
your
connections,
because
if
we,
this
Way's.
A
A
Yeah
just
ignore
the
fact
of
the
car
that
just
appeared
on
my
screen
and
8,000
tabs
on
my
okay.
This
is
this.
Is
sin
solid?
If
you
never
have
used
that
or
seeing
that
it's
monitoring
tool,
you
can
do
other
stuff?
One
of
this
this
one
has
something
interesting
is
that
you
can
see
this
graph
is
CPU
and
for
the
x-axis
and
memory
on
the
y-axis.
A
You
see
that's
and
it
just
got
increase
in
memory
and
that's
fine
I
mean
that's
how
it's
supposed
to
work
if
it
recovers,
that
will
be
the
natural
way,
but
it
doesn't
recover.
He
never
garbage
collect
the
rest
of
the
objects
that
he
just
created
and
doesn't
need
anymore.
So,
if
you
see
any,
if
I
keep
increasing
the
load,
I
mean
another
100
users.
A
A
You
see
it's
going
to
increase,
increase,
increase,
increase
in
memory
that
is
never
going
to
recover,
and
this
is
a
memory
leak.
This
is
actually
memory
leak.
Why?
Because
the
application
is
not
recovering.
That's
that's
how
you
identify
that
you
are
having
memory
issues.
So
how
do
we
fix
this?
It's
actually
pretty
simple.
You
just
do
kind
close
after
you
said,
send
answer
and
you
the
issue
was
that
you
not
closing
the
connection
after
you
answer
now
you
fix
it.
A
The
memory
leak,
so
this
is
one
of
the
things
that
I
really
make
people
understand
is
that
it's
just
follow
back
best
practices.
Don't
think
that
if
you're
doing
something
missing
something
and
that's
the
application
it
doesn't
break,
it
doesn't
mean
that
it's
not
doing.
We
run
things
on
the
inside
because
he
it
actually
is
and
he's
going
to
break
later
and
really
hard,
and
it's
going
to
be
way
harder
to
to
recover
and
try
to
debug
and
understand
what
the
application
is
doing.
A
A
A
Yeah
so
close
your
connections,
please
do
it,
you
don't
want
a
lot
of
things
being
kept,
especially
if
you
use
a
poll
how
many
of
you
suppose.
Well,
you
probably
use
it
or
grab
fuel.
When
you
keep
your
connections
open
the
object
created
for
a
polar
or
graph
field,
they
are
a
store
in
memory,
so
you
are
keeping
all
of
that
and
not
getting
garbage
collected.
A
A
Because
some
of
its
own
a
few
times
is
that
what
you
serve
things
and
you
can,
if
you
can
actually
try
different
ways,
it's
really
helpful
for
you
to
lease
this
car
than
what
you
are
serving
things
and
something
interesting,
and
we
have
seen
a
lot
is
review
your
package
darf
that
our
CS
CSS
or
in
JavaScript
or
JavaScript,
or
something
like
that.
Just
please
review
that.
Make
sure
that
you
know
what
their
package
is?
How
are
you
doing?
That's
a
lot
of
translation.
That's
a
lot
of
compilation!
A
That's
a
lot
of
things
are
happening
there
in
those
package.
If
you
are
really
not
sure
or
if
you
think
that's
a
cool
package
just
make
sure
that
it
doesn't
impact
you
for
your
memory
and
a
special
performance.
A
lot
of
these
package
store
many
things
on
memory
because
they
have
to
do
templating
of
the
things
that
you
do
and
what
that
translates
to,
for
example,
JavaScript.
So
please
check
that
if
you
have
a
dependency,
just
check
it
and
keeping
them
in
naming
your
functions,
that's
something
nice
point
before!
Please
do
it!
A
So
you
don't
get
unforgettable
cold
here
in
snapshots
that
you
cannot
understand
or
fix
yeah,
and
this
one
is
the
most
interesting
one,
because
I
have
seen
combinations
of
nodejs
and
Java
back-end
servers
in
many
other
combinations,
and
one
of
the
thing
is
I
have
seen
Java
servers
or
database
or
other
service
that
same
information
to
know.
Gs
servers
it
services.
A
Sometimes,
okay,
crazy
right,
I
mean
and
I
and
I
could
just
keep
going,
keep
going
and
keep
going
on
all
the
reasons
and
possibilities
that
you
can
have
actually
an
issue
with
memory.
But
this
is
pretty
much
like
a
hobby
or
a
general
view.
How
do
you
look
for
and
what
probably
you
can
you
don't
want
to
do,
or
at
least
try
to
avoid
as
much
as
you
can
and
try
to
fix
it
later?