►
From YouTube: How to find K8s logs in GCP
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
All
right,
well,
yeah.
I
was
just
planning
on
like
just
sort
of
showing
you
like
what
I
do
in
certain
cases
and
like
specifically
just
giving
you
like
a
walkthrough
of
like
where
everything
lives
in
the
console,
yeah
and
maybe
just
like
tips
and
tricks.
I
guess
there's
not
a
whole
lot
to
be
honest
but
yeah
anyway,
so
console.cloud.google.com,
it's
the
url,
and
this
all
of
this
sort
of
applies
to
either
demo
prod,
but
for
reference
for
the
people
who
have
products.
A
A
First,
you
probably
want
to
pin
stuff
here
because
it's
like
impossible
to
find
things
quickly
if
you're
like
scrolling
through
this
list,
I
think
you
can
also
search
for
them
at
the
top,
but
you
kind
of
need
to
know
what
they're
called.
So
I
just
pin
everything
here.
A
I
find
it's
easier
to
find
what
I
want
so
yeah,
let's
start
with
the
cluster,
so
it's
just
the
one
cluster
in
this
project
at
the
moment,
nexus
dev,
gitlab
ci
cluster,
and
you
can
generally
just
go
straight
to
services
and
ingress
because
that's
if
you're
looking
for
like
your
particular
review,
app
and
like
why
it's
failing
this
is
kind
of
like
the
easiest
way
to
find
those
logs.
A
A
So
you
want
to
look
for,
like
whatever
the
closest
one
is
to
yours,
like
you
can
see
here,
there's
obviously
a
couple
of
branches,
three
branches
that
were
named
the
same
or
all
very
similar,
so
they
had
like
review
feed
commit
at
the
start.
So
it
might
be
hard
to
find
exactly
the
one
that
you're
looking
for
in
this
case.
A
A
So
so
we've
got
this
view
app
here
yeah.
So
if
you
look
at
this
one
right,
that's
got
the
exact
same
like
prepended
service
id
there,
so
that
could
be
the
the
way
to
like
work
out
exactly
which
service
pertains
to
your.
Mr.
A
That
the
start
yeah,
we
might-
I
mean-
that's
probably
a
good
idea
right
like
I
I
sort
of
use,
I
mean
review
review.
Is
we
can't
remove
that?
But
everything
after
that
like
fit,
for
example
like?
Maybe
we
just
want
to
change
that
and
just
start
putting
in
exactly
what
it
is
there
right
anyway.
The
one
that
you
want
to
look
at
usually
is
the
auto
deploy
one
there's
generally
three
within
every
service,
so
with
every
deployment,
so
there's
a
postgresql
one
and
then
a
headless
postgres.
A
I
don't
know
what
the
difference
is,
but
the
auto
deploy
is
the
actual
app
pod
so
app
service.
A
So
if
we
look
at
that
one
so
so
yeah
like
if
if
it
fails
to
deploy
properly
you'll,
see
like
that
you'll
see
here
like
something
other
than
okay,
it'll,
probably
say
like
you
know
it's
in
a
crash
loop
back
off
or
something
like
that.
A
If
it
didn't
deploy
properly.
That's
generally,
what
happens
if
the
app
doesn't
launch
properly?
The
first
time
the
pod
or
the
service
gets
deployed
because,
if
it,
if
it
launches
properly
the
first
time
and
then
like
a
subsequent,
commit,
makes
it
break
when
the
deploy
happens,
it'll
generally
roll
back
to
the
last
working
version,
so
it'll
be
still
okay,
so
sort
of
you
know,
I
guess,
take
this
status
with
a
grain
of
salt
because
it
might
not
be
the
status
of
the
version
you're
trying
to
deploy.
A
A
Or
at
the
ci
cd
for
that
branch,
maybe
then
you
could
see
like
how
many
deploys
you've
tried
to
do
so.
C
A
So
so,
what's
this
one
called
commit
hood
hooks,
three?
Okay.
So
if
we
filter
the
pie
plans
by
commit
hooks
three
yeah,
so
you
can
see
like
the
first
time-
oh
sorry,
which
one's
the
first
one
yeah.
This
is
the
first
one,
so
the
first
time
that
camille
deployed
this
it
failed
and
then
the
second
time
it
passed.
So
I
mean
that
might
be
one
way
of
seeing
whether
it's
the
first
time
or
not,
there's
probably
a
way
in
kubernetes
to
see
as
well.
I'm
just
not
that
good
at
it.
So
yeah.
C
B
A
A
A
So
if
you
look
at
the
filters
here,
this
is
like
pre-done
for
you,
based
on
what
you
clicked
on
the
service
that
you
were
looking
at.
So
it
has
a
bunch
of
stuff
because
it's
like
pointless
because.
C
There's
some
errors.
I
guess
seven
looks
like.
A
Yeah,
so
if
so
on
this
thing,
it's
kind
of
easy
to
find
the
errors
and
when
they
happened,
or
you
can
just
click
on
the
error
filter
and
that
will
show
you
those
there.
A
So
this
these
areas
here
are
actually
just
normal.
This
happens
when
the
pod
gets
when
it's
upgrading
to
a
new
version.
So
you
see
these
in
prod
as
well.
I
don't
know
a
nice
way
to
get
rid
of
them,
but
sometimes
you
know
if
the
app
fails
to
start
or
whatever
you'll
get
all
the
logs
about.
Why
didn't
you
know
what
exception
happened
or.
A
It
so
so
yeah,
that's
kind
of
how
you
get
to
the
logs.
A
D
A
Yeah,
so
this
this
runs
on
the
like
get
lab
ci
runners,
so
you'll
just
have
to
go
directly
to
the
job
to
find
the
logs
but
yeah.
This
is.
C
A
A
B
D
This
is
for
your
whitelist
validation,
right
yep.
I
bet
there's
a
something
being
passed
in
that's
being
stripped
out.
D
All
right
yeah!
No,
I
think
this
lets
us
know
how
to
get
there
and
how.
B
D
The
regular
logs
that
the
json
is
being
split
up
over
multiple
lines
instead
of
just
yeah.
A
D
A
So
it
looks
like
it's
it's
taking
the
like
raw
string
like
it's
stringifying
that
json
first
and
then
it's
like
doing
line
by
line
locking
on
this,
which
is
just
totally
annoying,
because
what
it
should
be
doing
is
a
json
payload
with
the
entire
of
this,
in
that
one
blog
line.
So
if
anyone
can,
if
anyone
knows
how
to
fix
that,
I
think
it
only
applies
to
the
request,
like
the
request.
Logging
right,
I
think
our
custom
logs
are
working
fine.
D
Oh
one
moment
go
hold
there
go
back
up
just
a
little
bit.
That
is
a
this
looks
like
it
might
be
from
the
test.
Doesn't
it
because
it's
got
that
undefined
request
complete
the
tests,
one
wouldn't.
A
C
C
A
It's
super
hard
to
read:
okay,
cool
yeah,
and
then
that
should
make
like,
if
we
get
that
right,
that
should
make
filtering
a
lot
easier
as
well,
so
like
you
should
be
able
to
do
things
like
json
payload
dot
and
then
like,
if
you
actually
had
a
json
payload
it'll,
actually
like
allow
you
to
filter
by
specific
fields
on
that
object,
so
you'd
be
able
to
do
like
give
me
all
the
ones
from
this
ip
address.
Something
like
that.
A
If
you
wanted
to
look
for
a
specific
problem
right
for
a
specific
person
and
there's
like
a
whole
bunch
of
docs
on
how
to
use
this,
I
don't
oh
here
we
go
learn.
So
if
you
want
to
learn
more
about
it,
go
and
have
a
look
at
this
stuff,
all
right,
all
good
with
that
for
now,
yep
yep,
all
right,
so
the
next
one.
A
Where
is
it
memory
store?
Okay,
so
we
do
have
a
redis
store,
which
is
shared
among
all
of
the
review
apps.
What
is
going
on.
B
A
So
this
is
our
redis
cluster.
There's
nothing
really
to
see
here
honestly,
but
just
so
you
know
that
it's
there
and
then
the
the
main
difference
between
like
prod
and
dev
is
that
in
dev
there
is
no
like
managed
sequel.
It's
just
every
single
instance.
Every
single,
like
service
gets
its
own.
A
A
Yeah
so
yeah,
so
there's
no
managed
sql
in
in
the
dev
environment,
but
there
isn't
prod.
So
I
mean
this
this.
This
applies
less
to
to
you
both,
but
this
applies
to
dominic,
so
just
like,
so
you
know
dominic
how
things
are
set
up
there.
Is
this
separate
sql
instance
now
connecting
to
this
is
not
great
at
the
moment.
Basically,
you
have
to
add
your
like
home
network
here
in
the
allowed
list
of
networks
before
you
can
access
it.
A
Basically,
ideally,
we
want
to
have
like
a
proper
way
to
do
this.
Through
I
mean
we,
don't
originally
wanted
to
use
a
bastion
host,
but
apparently
there's
a
proper
way
to
do
it
through
the
cloud
sql
proxy
features
which
we
haven't
properly
set
up.
A
Okay,
so
I'm
kind
of
waiting
on
it
to
be
more
free
to
be
able
to
do
that
for
us,
but
yeah.
That's
that's
this
database
here
and
and
for
staging
staging
actually
has
its
own
database
within
the
kubernetes
cluster.
So
that's
more
like
the
dev
environment,
probably
we
should
make
staging
closer
to
prod's
setup,
but
I
just
haven't
gotten
around
to
doing
it
yet.
A
And
then,
in
pride,
all
of
this
is
in
a
a
vpc.
So
it's
not
like
super
complicated
compared
to
dev.
It's
just
a
tiny
bit
more
complicated.
A
A
I
don't
think
they've
yeah,
I
don't
think
I've
actually
done
it
yet
I'll
follow
up
on
that.
A
I
will
probably
record
a
separate
video
for
like
how
to
connect
to
this,
because
I
don't
want
to
like
share
it
too,
broadly
so,
but
I'll
show
you
how
to
like
set
up
the
connection
for
okay,
the
sql,
the
single
instance
and
how
to
like
run,
queries
and
stuff
against
it,
but
yeah
that's
pretty
much
it
I
mean
we
have
like
a
little
bit
of
time.
So
is
there
anything
else
that
you
wanna,
that
you
all
wanna
go
and
look
at
or
learn
more
about.
C
I
don't
think
so.
I
think
the
build
stuff
was
mainly
the
big
one
for
now.
I'm
sure
there's
lots
of
other
cool
stuff
in
here.
Unless
there's
something
else
you
use,
you
know
pretty
often
that
kind
of
helps
you.
A
A
Yeah,
so
this
is
an
app
called
lens,
which
is
like
kind
of
like
a
front
end
for
k9s,
which
is
like
the
more
feature
fully
featured
okay,
it's
cli
tool.
A
So
this
is
pretty
like
useful
as
well.
If
you
don't
want
to
like
log
into
the
gcp
console
every
single
time,
and
this
you
just
have
to
set
up
like
a
service
account
connection
in
gcp.
A
B
Actually,
let's
pick
one
right:
let
me
do
a
better
job
of
this.
Do
they
have
services
yeah
citizens?
A
A
Yeah,
it's
just
called
lens,
I
don't
know
if
it's
got
like.
B
C
For
other,
is
it
google
answer?
No,
they
have
one,
of
course.
Oh
that's
something
else.
Remember
that.
B
A
A
B
B
D
D
Our
pretty
print
is
just
based
on
the
app
env
variable
yeah,
so
we
can
do
some
local
testing
on
that.
C
D
Yeah
or
just
wait
for
the
unleash
logs,
because
they'll
show
up
in
30
seconds
or
something
yeah.
A
Constantly
I
actually
kind
of
want
to
filter
them
out.
I
don't
find
them
that
useful,
I'm
considering
just
turning
them
off.
To
be
honest,
if
we
can,
we
don't
really
use
feature
flags
that
much
yet
or
at
all.
Yet
I
I
see
us
using
it
in
the
future,
but
yeah
we
don't
use
them
that
much
now,
so
probably
not
super
useful
to.
C
C
A
Is
yeah?
Let
me
know
if
there's
anything
else
that
you
like
run
into,
we
can
do
like
some
short
videos
as
well
on
those
I'll
I'll.
Take
this
recording
and
I'll
put
it
in
now.
I
actually
might
upload
it
to
youtube
because
we
have
like
a
public
youtube
network,
so
I'll
check
it
on
there
and
then
I'll
put
it
in
our
handbook
as
well,
so
that
future
people
can
look
at
it.