►
From YouTube: Ops Section Product Walk Through - Error Tracking
A
A
Ok,
so
this
issue
just
an
overview
air
tracking
is
important
because
it
allows
you
to
link
the
kind
of
real
user
experience
that
might
be
generating
the
errors
to
your
development
process
as
well
as
your
operations.
So
a
big
primary
use
case
for
this
is
while
a
developer
is
working
on
an
application
or
an
update
to
an
application.
A
So
my
expectations
in
going
to
just
want
to
is
that
I
will
be
able
to
start
with
a
fresh
application.
I'm
going
to
use
one
of
our
auto
dev
ops,
enabled
templates
plummet,
force
entry,
configure
it
configure
century
in
the
github
application
and
then
start
seeing
errors.
You
can
get
lab
itself
so
with
that
I
will
get
started.
I
keep
my
notes
over
here.
This
was
just
the
tracking
issue.
A
I'm
gonna
start
with
this
is
a
group
that
I
have
an
attached
cluster
to
actually
just
attached
it.
So
hopefully,
I
won't
run
into
too
many
issues
with
the
with
using
a
kubernetes
cluster
to
start
a
new
project,
but
I
am
going
to
go
ahead
and
create
a
new
project.
I
had
one
previously
for
this
Opie
tests
that
I'd
done
for
previous
walkthrough,
but
I've
already
plumbed
it
for
sentry
separately.
So
I
want
to
go
through
the
process
of
doing
it
from
scratch.
A
A
Now
one
thing
I
think
I
need
to
do
is
to
go
back
to
the
I.
Don't
think
I
set
up
this
end
point
I
just
pushed
provision
ingress.
Let
me
make
sure
I
do
this
so
I
give
it
a
base
to
mean
no
IP
needed.
I,
oh
okay,
now
I
think
when
I
run
Auto
dev
ops
on
this
new
project,
it
says:
hey,
I
could
add
a
cluster
or
I
could
enable
auto
dev
ops,
I
have
a
kubernetes
cluster
just
make
sure
it
shows
up
here:
operations
kubernetes,
it
says:
I
got
the
group
cluster
available.
A
B
A
Next
thing
I
want
to
do
is
adjust
the
application
for
century
and
before
I
do
that
I
need
to
create
a
sentry
project,
but
doing
a
little
bit
of
this
blind.
So
let's
look
for
some
docs
sentry
and
get
that
error
tracking.
That's
probably
what
I
be
okay
error
tracking
allows
focused,
easily
discover
and
view
errors
in
the
application
may
be
generating,
helps
them
develop
or
efficiently
we're
gonna
use.
B
A
That's
cool,
there's
a
way
to
deploy
sentry
I
can
use
the
hosted
version
I'm
going
to
do
that.
Okay,
so
enabling
sentry,
Here
I
am
sign
up
for
sentry,
so
I
will
sign
up,
but
I
already
have
an
account.
Let's
go
to
sentry,
ok,
so
this
is
a
project
that
I
already
have,
but
let's
have
some
other
projects.
This
one
is
one
I
mentioned
I'd,
already
plums,
separate
personal
project.
A
So,
let's
add
one
a
new
project
here:
okay,
come
on
sentry
that
one
does
I
me
to
do
navigate
to
your
projects,
yeah,
okay,
so
it
doesn't
really
tell
me
that
I
need
to
create
a
new
century
project
for
this.
For
you
for
my
individual
get
lab
projects,
I
can
probably
just
also
reuse
the
existing
one,
but
I'm
gonna
make
a
little
note.
A
A
A
B
A
B
A
A
Okay
and
I'm
just
committing
the
master
here.
I'm,
not
gonna,
do
much
requests
so
I'll
push
that
the
next
thing
it
says
I
need
to
inform
Centrino
desk
a
about
my
DSN
Candace
DSN
is
a
unique
code
for
sending
specific
information
to
my
sentry
project,
so
I'll
do
that
it's
here
around
so
I'll
use
this
user
1.
A
A
I
want
to
say:
if
you
go
to
slash
users,
it's
going
to
have
generate
a
sentry
error
and
then
I
know
from
a
previous
example
that
the
language
to
sentry
says
I
can
just
call
it
an
undefined
function,
but
I've
realized
that
there's
an
issue
in
the
sentry
dogs,
at
least
I-
couldn't
figure
out
how
to
actually
make
this
execute,
and
what
I
really
needed
was
something
like
this
line.
That
I
will
show
you.
B
A
A
A
The
production
one
is
generating
errors
to
century
as
well.
Okay,
cool,
so
two
commits
later
one
got
me:
the
sentry
packages
and
the
second
one
added
this
line
to
capture
something
it's
capturing
locally
and
I,
validated
that
in
century,
so
it's
sending
out
appropriately
now
the
question
is:
I
would
like
to
see
if
this
is,
if
I
can
start
capturing
these
from
a
deployed
application
in
get
lab
okay,
so
this
is
where
I
was.
Let's
look
at
this
I've
got
some
pipelines
running
I
got
three
pipelines
running.
A
Look
at
this
one,
it's
the
latest
one
that
is
also
about
supplying
that
depression.
I
was
just
gonna
cheat
a
little
bit
for
timewise
and
cancel
out.
I
know
these
jobs
in
the
Test
section.
I
I
can
run
without
website.
They
can
they'll
still
forward,
even
if
these
jobs
fail
so
they're
not
required
so
I'm,
just
gonna
cancel
them
for
Speed
purposes.
A
Well,
that
is
happening,
look
at
some
of
those
previous
pipelines,
so
this
one
has
deployed
to
production,
so
I
should
be
able
to
see
in
my
environments
a
new
environment
and
I
wanna.
Look
at
what
this
looks
like
live.
Okay,
so
there's
my
welcome
to
Express
live
public
published
by
Otto
DevOps,
so
I'll
wait
for
the
other
two
pipelines
to
finish
running
then.
Hopefully,
if
I
end
up
going
to
you
know,
it's
just
confirm
this.
When
I
go
here,.
B
A
Gonna
give
me
that
response,
but
I
should
not
be
seeing
any
new
errors
generated
in
century,
which
I
don't
only
one
event
has
gone
into
that
one,
but
when
these
other,
when
this
pipeline
finishes
I
should
when
I
hit
that
same
URL
generate
a
new
error
in
century.
Okay,
let's
go
back
so
this
was
about
getting
century
set
up
said
you
know,
deploy
century
plummet
to
your
application.
That's
what
I've
just
done
now.
Here's
the
instructions
for
connecting
sentry
to
get
lab
so.
A
A
Let
me
read
the
instructions
carefully,
find
a
generate
a
sentry
auth
token
and
then
navigate
to
your
settings
in
operations.
Ok,
so
here
I'm
in
sentry,
let's
figure
out
how
I
get
an
auth
token.
The
link
here
gave
me
something
to
go
to
find
it
or
generate
authentication,
because
you
can
find
or
create
authentication
tokens
within
sentry-go.
Okay,
you're
am
I,
have
a
token
I'm,
just
gonna
use
that
one.
A
Error
tracking,
so
it's
active.
This
is.
A
A
A
A
Okay,
there's
my
one
with
one
event
having
five
minutes
ago,
so
cool
sentry
I
can
see
my
sentry
errors
right
here
and
get
that
that's
pretty
cool
and
there's
a
little
pump
out
to
go
see
it
in
sentry
itself.
Awesome,
let's
check
in
on
that
pipeline
and
make
sure
it's
sending
ones
to
production.
Okay,
so
this
is
now
deployed
production
past.
So
this
should
be
our
newest
or
newer
instance,
and
so
now.
A
B
A
A
A
B
B
A
B
B
B
A
A
A
A
A
B
A
A
A
A
A
B
B
B
B
A
B
A
A
A
B
A
B
B
B
A
A
A
A
B
B
A
A
A
A
A
So
it
is
generating
events
31
events,
not
sure
what
took
that
time.
But
let's
go
back
to
these
errors,
can
now
see
unique
errors
in
century
really
good
for
demo
purposes,
sure
total
different
types
of
errors
in
different
environments,
both
in
century
and
get
lab,
probably
good
for
next
iteration
on
the
walkthrough,
but
alright,
despite
the
trouble
to
get
that
specific
thing
deployed
and
generating
errors
from
our
production,
environment,
centric
connection
easy
to
plumb.
Your
application
with
century,
generally
speaking,
even
easier
to
have
this
single
pane
of
glass
view
into
your
century
errors
from
gitlab.
A
How
much
I
know
we
will
be
working
on
improving
documentation
was
fairly
straightforward
and
easy
to
follow
so
yeah.
Oh,
no
I
think
a
success,
some
follow-up
for
what
might
have
been
happening
that
took
it
so
long
to
start
pushing
the
errors
from
the
production
application.
But
with
that
I
will
close
this
out
thanks.
Everybody.