►
From YouTube: Grafana Community Call 2022-01-20
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
So
welcome
to
this
month,
community
call
happy
to
have
you
all,
and
you
know
it's
it's
it's
been
a
while
since
since
last
time
I
think
the
last
call
was
in
november.
A
We
skipped
the
one
in
december
because
of
the
holidays,
but
now
we're
back
and
I'm
happy
to
to
be
joined
by
the
cloud
data
sources.
Team
who's
gonna
share
a
little
bit
about
what
they've
been
working
on
lately
and
maybe
some
of
the
what's
going
on
in
the
future.
A
And
if
you
have
any
questions
on,
you
know
any
of
the
data
sources.
The
cloud
data
sources,
please,
you
know,
ask
questions
directly
or
again.
If
you
don't
want
to
join
in
with
audio,
you
can
ask
in
the
chat
as
well,
so
I'm
going
to
head
it
over
to
the
team
and
they
can
introduce
themselves.
B
Thanks
marcus,
is
it
fine
if
I
share
my
screen
totally.
B
I'm
assuming
silence
means
yes,
yes,
cool!
Well,
thank
you,
everyone
and
similar
to
what
marcus
says:
yeah.
It's
our
first
grafana
community
call
of
the
year
so
great
for
everyone
to
join
us
and
very
slow.
As
a
reminder,
we
meet
every
third
thursday
of
every
month
where
the
next
one
will
be
the
17th
of
february
yeah,
and
each
meeting
will
have
a
different
theme.
So
today
is
focused
on
the
cloud
data
sources,
theme
cool.
B
So
moving
on
to
the
agenda
then,
seeing
as
this
is
the
first
time
our
squad
has
actually
hosted
this
community
call,
we
thought
we
would
give
a
short
introduction
to
what
we
do
and
who
is
in
the
team.
B
We
will
talk
a
bit
about
what
we
have
been
up
to
with
our
azure
google
and
aws
plugins,
and
then
we've
got
some
short
demos
prepared
for
the
team
by
the
team
and
finally,
we
will
have
time
then
for
q,
a
or
any
feedback
that
you'd
like
to
give
to
us
directly
if
you
have
worked
with
any
of
our
plugins.
B
So
what
is
cloud
data
sources
then?
And
what
do
we
actually
do?
Our
mission
statement
is
working
to
make
grifan
an
indispensable
tool
to
manage
observability
in
public
clouds
and
nurturing
the
relationships
of
them
and
our
community
to
make
grafana
better.
But
what
does
this
actually
mean
simply
puts
you
know
we
work
together
with
cloud
providers
and
we
build
plugins
to
integrate
with
their
services
and
apis
into
grafana
and
this
collaboration,
and
it
enables
us
to
not
only
have
feedback
from
the
grifana
side,
but
also
from
their
side,
because
they
know
their
users.
B
You
know
much
better
than
we
could
so
the
current
plugins
that
we
currently
own
and
ones
that
you
may
have
used
before
azure
monitor
azure
data,
explorer
cloud
monitoring,
formerly
known
as
stackdriver
cloudwatch,
but
only
the
metrics
side
at
the
moment,
and
quite
recently,
I'm
not
sure
if
you've
seen
but
we've
released,
redshift
and
athena
as
well.
B
So
now
that
you
know
who
we,
what
we
do
like,
who
are
we
so
to
introduce
myself?
I'm
vicky,
I'm
the
engineering
manager
for
the
squad
eric?
Would
you
like
to
go
next.
C
Sure,
hello,
everyone,
my
name
is
eric
and
I'm
an
engineer
on
the
team,
so
I'm
basically
in
stockholm,
sweden
and
I've
been
at
the
company,
for
I
think
it's
in
total
about
two
years
and
for
this
period
I've
been
focusing
on
the
plugins
and
mainly
plugins,
that
are
related
to
cloud
providers.
So,
like
cloud
watch
cloud
monitoring
and
azure
monitor
yeah.
Sarah,
do
you
want
to
go
next.
D
Sure
hi,
I'm
sarah,
I'm
based
in
new
york,
I've
been
with
griffana
since
april
and
yeah.
That's
it
andre
syrup.
G
Fine
I'll
go
next
hi,
I'm
yael,
I'm
based
in
france
and
I've
joined
grafana
almost
three
weeks
ago.
So
I'm
the
newbie
in
the
team.
B
Cool
so
now
yeah
now
that
you
know
who
we
are,
let's
talk
about
what
we
have
been
doing
now
for
past,
I
guess
few
months
or
weeks,
starting
with
azure,
as
you
may
or
may
not
know,
most
of
our
plugins
have
end-to-end
tests,
but
for
adx
these
aren't
there.
So
one
of
the
recent
focuses
that
we
would
like
to
work
on
is
improving
the
adx
plug-in,
and
to
do
that,
we
would
like
to
add
in
these
end-to-end
tests,
which
are
missing
and
similar
to
our
other
end-to-end
tests
for
other
plug-ins.
B
D
D
F
Yeah,
so
what
I've
been
working
on
recently
is
separating
out
the
google
authentication
code
into
its
own
packages
for
front
end
and
back
end,
so
to
make
it
easier
basically
to
develop
future
google
plugins,
because
they
all
use
the
same
auth.
B
And
for
aws
there's
a
few
things
we've
been
working
on
to
start
with,
andres
can
talk
a
bit
about
the
aws
sdk.
A
E
So
the
story
here
is
that
at
the
beginning
we
started
with
just
one
one
plugin
for
aws.
It
was
cloud
watch,
it's
mostly
the
most
used
one,
but
we
are
starting
to
to
work
on
more
and
more
and
more
plugins
and
we
are
seeing
a
pattern,
a
common
pattern.
So,
for
example,
you
use
you
use
the
same
authentication
methods
to
authenticate,
regardless
of
the
plugin,
because
you're
using
aws
authentication
or
now
we
are
seeing
more
and
more
sql
based
plugins.
So
because
of
that,
we
need
to
refactor
it
a
bit.
E
So
it's
easier
for
for
us
to
create
new,
plugins
or,
and
also
is,
is
faster,
so
this
graphene
aws
sdk.
At
the
same
time,
this
is
the
upstream
aws
sdk,
which
is
obviously
maintained
by
aws,
but
it's
also
using
something
that
we
call
sqlds
library
that
again
this
is
a
library
that
we
are
trying
to
develop.
E
So
it's
easier
to
to
develop
generically
databases,
plugins
related
to
sql
databases,
but
now
bigquery,
that
is
from
from
google,
also
snowflake,
which
is
kind
of
its
own
thing,
but
all
of
them
are
trying
to
use
this
library
and
and
more
packages
in
the
future,
so
yeah.
We
are
really
trying
to
scale
up
this.
This
approach.
B
Cool
thanks,
andres,
and
also
I
listed
the
current
plugins
that
we
own
currently,
but
we
are
in
the
process
of
taking
over
time
stream
and
the
logs
parts
of
the
cloud
watch
plugin
as
well
as
x-ray,
and
the
reason
why
we're
in
the
process
of
taking
this
over
is
because
they
are
returned
by
other
teams
in
grafana.
But
we
would
really
like
to
consolidate
the
ownership
of
all
of
these
plugins,
since
we
communicate
with
the
cloud
providers
a
lot.
So
it
would
be
easier
for
us
and
yeah
sarah.
D
Yeah,
this
is
a
another
small
improvement
that
we're
working
on,
there's
an
open
pr
right
now,
but
we
hope
to
merge
it
for
8.4
with
cloudwatch
metric
search
options.
So
if
you
write
math
expressions
for
cloudwatch,
look
out
for
some
some
nicer
syntax,
highlighting
and
autocomplete
suggestions
that
hopefully
can
help
make
writing
these
queries
a
little
bit
easier
yeah
and
if
you're
interested,
you
can
check
out
the
pr
and.
B
B
You'll
be
able
to
check
out
the
pr
by
the
code
because
I
don't
think
we're
sharing
these
slides
with
the
links
cool
and
then
we
can
move
on
to
the
demos
then
so,
firstly,
eric
do
you
want
me
to
stop
presenting.
C
So
hi
again,
as
you
know,
aws
has
this
huge
I.t
conference
every
december
and
in
the
in
last
year's
event,
I
think
the
first
of
december
they
announced
something
that
they
called
metrics
insights.
Matrix
insights
is
a
sql
query.
Engine
that
you
can
use
to
query
metrics
and
we've
been.
We
in
the
team
have
been
working
closely
with
the
cloudwatch
team
during
the
whole
fall
to
add
support
for
metrics
insights
also
to
the
cloudwatch,
plugin
and
graphone.
C
So
for
those
of
you
who
have
used
who
have
used
the
cloud
watch,
query
editor
before
you'll
notice
that
this
is
a
completely
new
ui
here.
So
not
only
are
we
adding
the
metrics
insights
support
we're
also
making
like
a
total
total
rewrite
of
this
front
here.
C
So,
if
you've
used
the
cloud
watch
plugin
before
you
have
probably
used
metric
search,
this
is
we've
sort
of
decided
to
split
up
the
query
of
32
different
metric
nodes
and
that
is
metric
search,
sort
of
the
asset
mode
and
then
metric
query,
which
is
basically
matrix
insights.
C
So
not
only
I've
done
that
we've
also
sort
of
separated
these
metric
search
metric
step
queries
from
the
metric,
the
math
expression
quiz.
C
The
way
to
go
into
the
math
expression
mode
would
be
to
click
the
code
toggle
here
and
I'm
not
going
to
go
into
depth
about
this
because
sarah
has
already
touched
the
subject.
So
instead
I'll
go
into
the
metrics
insights
mode
and
what
you
see
here
is
a
code
editor
that
allows
you
to
enter
raw
sql
queries.
E
C
C
The
way
that
metrics
insights
works
is
that
it's
going
to
scan,
I
think
up
to
10
000
metrics,
and
then
it
will
return
up
to
500
of
them.
So
since
this
is
a
sequel-
and
it
has
basically
all
the
standard
sql
support
that
is
out
there,
you're
now
allowed
to
do
otherwise.
So
for
this
I
can
order
by
maximum
and
sort
them
in
descending
order,
and
that
will
return
me
the
like
the
the
the
instances
that
had
the
highest
average
of
cpu
utilization
for
the
past
three
hours.
C
So
for
for
this
type
of
query,
it's
probably
better
to
switch
over
to
some
other
visualization.
So
I
can
use
bar
gauge
and
I
can
also
limit
the
result
to
say
five
instances.
C
So
this
is
one
of
the
advantages
with
metrics
insights.
It
does
allow
you
to
do
the
top-end
type
of
queries,
which
is
pretty
cool.
The
cloudwatch
or
the
sort
of
metrics
insights
editor
in
the
graphola
plugin
also
has
a
builder.
C
So
if
I
switch
over
to
the
builder
mode,
it
will
allow
me
to
create
sql
queries
using
these
fields.
So
if
I
want
to
do
the
exact
same
query,
I
would
buy
I'll
equip
by
instance,
id
followed
by
max
in
in
descending
order,
and
then
we
can
limit
the
result
to
five.
So
that's
basically
the
same
query
as
I
showed
you
in
the
raw
code
editor.
C
So
I
can
also
mention
that
metrics
insights
is
enabled
through
the
same
api
as
we
used
for
metric
search.
So
if
you
have
the
data
source,
if
you
have
the
data
source
configured
in
grifon
already,
this
will
just
work.
There's
no
extra
configuration
necessary
and
it
also
it
does
work
with
template
variables,
so
you
can
use
them
in
the
builder
or
in
the
code
mode
and
of
course
this
works
with
alerting
so
yeah.
It's
really
excited
about
this.
H
Because
it's
kind
of
for
me,
it's
a
little
bit
more
complicated
before
that
you
kind
of
have
everything
concentrated
on
your
left
side,
right
all
the
modules.
So
when
the
somebody
is
trying
to
build
your
query,
you
kind
of
start
with
your
sql
and
then,
if
you
want
write
sql
itself
right,
there
is
was
kind
of
more
intuitive
that
now
you
have
to
look.
Oh
there
is
a
small
builder
called
thing
which
you
have
to
change.
H
Yes,
but
I
mean
this
builder
code,
it's
kind
of
somewhere
on
the
side,
because
right
now
you
already
have
a
lot
of
buttons
right.
When
you
have
a
query,
you
have
this
copy
delete.
Then
you
have
this
left
side
which
you
have
to
fill
to
actually
get
the
data
you
need,
and
now
you
have
is
another
switch
builder
code
which
actually
have
to
be
the
first
thing
you
are
going
for
when
you
create
the
query.
H
C
Yeah
yeah,
I
think
that
that
the
the
point
of
adding
this
switch
is
because
they
are
two
completely
different
type
of
queries.
I
mean,
if
you're,
entering
a
math
expression
here.
That
is
something
completely
different
than
building
your
query
using
the
builder
so.
E
C
H
C
Because
you
have
this
kind.
H
Of
list
boxes
right
on
your
left
side,
and
now
you
have
this
like
a
radio
on
your
right
side.
Why
not
to
make
it
the
same?
And
it's
maybe
it's
a
if
it's
a
different
functionality,
because
I've
never
used
it.
Maybe
it
have
to
be
like
the
first
thing
on
your
left
side,
which
you
have
to
choose.
You
want
to
do
build
your
code.
Oh
I'm,
going
with
the
builder.
Then
you've
got
this
region
and
cloud
watch
metrics
right.
If
it's
a
code,
you
have
your
edit
box.
H
C
H
Okay,
thank
you
because
every
time
when
I
try
to
explain
or
teach
someone
how
to
use
this
query
editor,
it
takes
a
while
right.
You
have
to
use
a
mixed
data
source.
If
you
have
different
data
source
like
sql
and
api
right,
then
you
have
to
configure
put
the
this
one
like
sql.
Here
you
have
to
do
the
payload
that
and
that
and
now
oh
there
is
one
more
thing
you
have
to
do
the
code,
so
the
less
we
can
teach
and
more
kind
of
streamline
is
going
to
be
it's
easier
for
the
customers.
C
Yeah
now,
but
I
I
I
agree,
and
it's
a
bit
confusing
with
the
cloud
watch
data
source,
because
it
is
the
a
the
api-
gives
you
access
to
a
whole
bunch
of
different
types
of
queries.
So
it's
kind
of
hard
to
to
replicate
that
in
a
ui
but
yeah.
That's
really
good
feedback.
Well
I'll.
Take
that
bring
that
to
the
team
and
and
the
ux
designers
here.
A
So
another
question
on
that:
are
we
looking
to
to
kind
of
you
reuse,
that
same
pattern
of
the
the
builder
code
switch
for
the
other
cloud
data
sources
as
well,
and.
A
It
like
built
into
the
actual
platform
so
that
it
looks
if
I,
if
I'm
hearing
mikhail
er
right,
you
know
it
would
be
nice
if
all
data
sources
use
a
similar.
You
know
pattern
and
switching
between
builder
and
code,
because
I
I
think
it's
really
powerful
being
able
to
switch
between
them.
But
you
know
if
all
data
sources
have
the
same
way
of
doing
that.
That
would
benefit
all
users.
I
think.
C
Yeah,
no,
I
that's
my
intention
here
at
least
that
will
have
the
same
sort
of
design
for
all
our
cloud
data
sources.
I
think
that's
the
like
the
initial
idea
that
patrick
had
when
he
made
a
design
for
this
creator
too,
and
you
know
if
I
switch
over
to
the
postcards
data
source,
if
that's
sort
of.
C
But
obviously
this
is
an
angular
and
we
wanted
to
have
something
similar
in
react.
So
I
think
that
the
intention
here
is
that
we're
going
to
use
something
similar
for
for
all
our
react
based
code,
editors
or.
C
If
we
look
at
the
cloud
monitoring
editor,
it
has
something
it's
not
using
the
of
the
new
the
new
ui,
but
it
has
also
has
two
different
modes:
a
code
editor
and
a
visual
query
builder,
and
this
toggle
is
is
here.
So
I
would
like
to
use
the
same
pattern
for
cloud
monitoring
as
we
use
for
for
for
cloud
virtual.
Actually.
C
Definitely
I
can
just
mention
that
this
new
ui
that
you're,
seeing
here
with
this
type
of
header
here
and
also
fields
grouped
logically
grouped
into
different
rows
that
will
most
likely
become
like
this-
will
form
the
new
baseline
for
how
queer
editors
will
look
in
the
future.
There's
work
going
on
moving
this.
It
has
already
been
moved
into
a
separate
package.
It's
it's
sort
of
experimental
right
now,
but
I
think
this
is
already
being
added
to
the
bigquery
data
source
and
a
few
others.
C
H
E
C
Yeah,
like
that's
a
really
good
question
and
for
the
moment
there's
no
style
guide.
We
do
have
a
sky
library
with
components,
and
I
think
that
these
all
these
components
are
not
added
to
to
that
storybook
design
library.
Yet
it
will
be
eventually-
and
I
hope
that
this
will
be
paired
with
some
guidelines
on
how
to
write
query
editors
and
how
they
should
look,
but
yeah
we're
not
there
yet.
But
okay,
we
will
be
some
time.
H
Because,
every
time
when
I
look
how
other
plugins
developed
from
grafana,
also
from
community
from
what
marcus
does
in
his
own,
it's
like.
Oh,
this
is
the
new
elements
you
can
use
like
great.
This
looks
more
interesting
right
because
there
is
a
history
with
the
years
of
history
from
where
it
started
right:
the
old
plugins,
which
used
one
way
and
a
new
one.
So
I
like
this
design
more
so
maybe
I
mean
there
is
something
we
can
build.
Some
kind
of
style
guide
for
everyone.
C
Yeah,
no
definitely
yeah
that's
what
we're
aiming
for
at
least,
but
it
will.
It
will
take
some
time.
I
think
the
the
ui
components
are
still
sort
of
in
a
in
a
beta
mode,
so
we
need
to
sort
of
fine-tune
them
and
make
sure
that
they
are
ready
for,
but
publicly
ready
for
for
using
an
external
plywood
as
well.
H
C
H
H
C
Yeah
I
mean
you
can
definitely
use
it
just
be
aware
that
it's
in
like
a
very
early
state,
so
things
apis
might
change
over
the
next
couple
of
months,
but
I
mean
feel
free
to
use
it.
I.
A
C
H
C
That
is
not
completely
decided
yet.
I
think
that's
one
of
the
reasons
why
this
was
put
in
an
external
package,
but
potentially
we
could.
It
could
still
be
being
placed
in
like
an
external
npm
package,
the
way
it
is
for
today
in
this
experimental
package,
but
the
future-
I
don't
know
marcus,
haven't
heard
anything.
A
We
we
do
recognize,
you
know
the
issues
that
you're
highlighting
that
we
have
very
different
ways
of
building
query
editors.
We
are
aware
of
this,
and
I
think
that
is
one
of
the
reasons
why
we're
looking
into
creating
a
new
set
of
components
that
are
specifically
made
for
query
editors
that
specifically
take
into
account
all
the
different
ways.
You
know
creating
logical
operators,
for
example,
filters
and
and
reducing
the
aggregations
and
all
that
and
making
it
streamlined
like
you're
saying
we
probably
will
need
some
form
of
style.
A
On
top
of
that,
but
first
I
think
that
we're
focusing
on
finding
the
right
balance
for
the
components
themselves,
and
hopefully,
just
by
using
these
components.
It
will
become
more
streamlined
in
in
in
a
way.
But
I
know
that
we
are
constantly
working
on
these
internally
and
you
know,
if
you
start
using
them,
feel
free
to
forward
any
feedback
you
have,
you
can
look
for
it
in
any
channel
of
your
choosing,
and
you
know
if
you
run
into
any
issues
or
you
feel
like
this
is
not
the
what
I
would
expect.
A
The
query
editor
to
look
like
in
2022.
That's
also
very
valid
feedback,
and
once
we
feel
like
we
have
something
that
we're
happy
with.
I
would
be
surprised
if
it
didn't
make
it
to
main
branch,
but
I
don't
think
that
we
can
make
any
promises
at
this
point.
E
So,
as
we
said,
we
have
a
couple
of
brand
new
plugins,
so
I'm
going
to
present
you
dredshift.
This
was
released
on
november
this
year,
the
same
around
the
same
time
than
the
global
project.
It's
obviously
open
source.
It's
in
this
in
this
github
repository
you
can
visit
it
and
you
know
typical,
create
new
issues,
give
a
star
whatever
you
feel
comfortable
with,
and
then
give
you
a
quick
demo.
So
what's
spreadshift,
if
you
have
never
seen
it,
it's
an
aws
service
which
is
a
cloud
data
warehouse.
E
E
So
in
the
plugin
page
you
can
see
all
the
readme
and
all
the
information
and
let's
see
how
you
can
create
a
new,
a
new
data
source.
So
let's
go
to
the
data
source,
let's
create
a
new
one
and
let's
look
for
it.
It's
here
the
bottom
dress
shift
and
when
you
create
it,
you
may
recognize
this.
If
you
have
ever
used
an
aws
data
source
in
which
you
have
to
introduce
some
common
details.
This
is
what
I
said
at
the
beginning
for
the
sdk
that
we
are
working
on.
E
All
of
this
is
common,
so
it
makes
sense
to
have
it
in
a
common
sdk.
So
here
I'm
not
going
to
enter
into
too
many
details,
but
you
have
several
ways
of
authenticating
against
ews.
Since
this
is
a
an
ec2
instance,
I
could
use
the
the
iron
role
associated
with
the
machine
or
I
could
use.
I
prepared
this
instance
also
be
compatible
with
the
default
way
of
connecting
with
the
sdk
or
you
can
also
use
an
access
and
secret
key
or
a
potential
file,
and
then
you
can
fill
this
this.
E
All
these
fields
are
optional,
I'm
going
to
enter
into
third
of
what
they
mean,
but
the
one
thing
that
is
necessary
is
to
select
a
default
region
in
this
case
us2
and
then
to
authenticate
against.
This
is
common
to
aws,
but
this
this
part
is
specific
to
redshift
and
to
authenticate
on
redshift.
You
can
either
use
something
called
temple
temporary
credentials
or
you
can
use
an
aws
secret,
managed
temporary
credentials.
They
give
you
you
have
to
enable
some
feature.
E
What
basically
allows
you
to
have
a
temporary
password
that
will
that
will
expire
or
use
a
secret
manager.
In
this
case,
I'm
going
to
use
a
cpu
manager
just
need
to
pick
a
secret
manager,
and
it
will
pick
the
details.
For
me,
this
is
the
information
that
is
encoded
in
the
secret
and
I
need
to
select
a
database.
E
E
So
that's
for
the
configuration
of
the
data
source.
Let's
see
an
example,
for
you
can
query
data
here.
So
if
we
go
to
this
demo
dashboard
that
I
have
prepared,
we
can
see.
Let's
start
this
basic
cable
example
as
any
other
sql
data
source
in
grafana.
You
can
write
sql
code
here.
So,
for
example,
here
I'm
just
you
know,
selecting
from
this
table
this
volume-
and
I
will
enter
into
details
about
this
mean
later.
E
But
as
you
can
see
here,
you
have
syntax
highlighting
you
have
some
very
quite
basic,
it's
not
as
smart
as
the
auto
completion
features
that
eric
showed
you
before,
but
you
got
also
some
kind
of
calculation.
E
If
you,
if
you
want
to
so
for
example-
and
then
I
I
will
go
into
that
later,
because
here
I
have
some
macros
that
I
can
also
use-
and
these
are
very
useful
to
explore
your
database.
So
you
can
list
here
the
schemas
in
your
database.
You
can
browse
tables
like,
for
example,
the
becoming
one.
Don't
forget.
Example:
sorry,
this
is
the
one
that
we're
using
here,
or
you
can
also
explore
the
different
columns
that
you
have
in
your
in
your
in
your
table.
E
E
This
is
a
little
bit
more
advanced,
but
if
you
want
to
modify
the
way
the
frames
are
are
formatted
you
can
or
you're
using
time
series
you
can
select
it
from
here
and
also
the
values
are
empty.
You
can
modify
those,
but
that's
not
really
important.
We
can
see
here
that
we
are
requesting
the
the
information
and
the
result
is
the
same.
E
So
that's
for
the
for
the
sqliter.
We
can
see,
for
example,
an
example
of
a
time
series
in
this
case.
I
need
to
go
a
little
bit
older
data,
because
this
is
some
demo
data
they
have
here.
Let's
go
a
little
bit
into
the
past,
and
here
we
can
see
a
time
series
again
quite
similar,
but
you
can
see
here
that
I'm
formatting
the
query
as
a
time
series
and
then
I'm
using
a
very
useful
macro
in
order
to
filter
the
the
query
by
the
time
that
we
select
here
in
the
in
the
graffana
filter.
E
E
E
So
if
we
go
to
the
variables-
and
I
can
show
you
an
example
of
a
redshift
template
variable-
you
can
see
that
you
can
also
use
sql
to
query
your
database
on
being
able
to
return
variables
in
this
case
I'm
selecting
the
different
environments
in
this
in
this
table.
So
these
two
values
are
can
later
be
used
here.
So
if
I
select
a
different
environment,
we
can
see
here.
For
example,
this
table
is
using
that
environment
as
a
as
a
template
variable.
So
the
data
is
changing.
E
E
So
I
have
here
a
notation
that
say:
okay,
if,
in
my
environment
the
humidity
goes
over,
90
95
just
create
an
annotation,
and
this
is
the
kind
of
things
that
you
can
do,
and
here
there
is
the
the
toggle
for
the
for
the
annotation,
and
you
can
see
here
that
we
are
starting
to
see
some
annotations
and
the
event
is
is
happening
at
some
point
where,
where,
when
it
happened,
that's
one
way
of
using
this.
E
But
usually
when
things
like
that
happens,
what
you
want
to
do
is
to
alert,
and
obviously
this
is
compatible
with
the
graph
analytics
system
and
I'm
going
to
enter
into
the
details.
But
here
you
can
see
that
I
have
created
an
alert
saying
that
whenever
the
temperature
here
I'm
monitoring
the
temperature,
so
whatever
the
temperature
that
the
purchase
goes
over
at
30,
we
are
going
to
to
fire
an
alert,
and
I
think
that's
all
that's
all.
I
wanted
to
show
to
show
you.
Obviously,
it's
also
compatible
with
the
login.
A
I'm
kind
of
curious
from
a
plug-in
development
perspective.
Again,
you
mentioned
the
the
sql
ds
package
earlier,
so
it
feels
like
all
of
these
are
using
the
the
from
the
demo
that
eric
showed
and
this
one
as
well.
All
of
them
are
using
the
same
sql
ds
package
right.
E
Yeah
in
the
case
of
eric
cloudwatch,
it's
not
really
an
sql
data
source.
It's
only
using
sql
language,
so
that's
kind
of
different.
It
could
use
part
of
that
functionality
right,
but
it's
not
using
that
in
this
case
it's
more
specific
sql
databases
or
similar
kind
of
databases
like,
for
example,
athena
that
sar
is
going
to
show
you
now.
It's
also
using
sql
kind
of
data
on
a
language,
so
that
one
is
is
using
this
qlds.
A
So
so
my
follow-up
question
then,
would
be
like
if
I'm
a
plug-in
author,
that
I'm
building
a
plug-in
that
has
a
sequence
sql
dialect.
Could
I
use
this
package
to
get
all
the
nice?
You
know
auto
completion
and
syntax
highlighting
today
is
that
something
that
I
could
bring
in
in
my
plugin,
or
is
that
only
for
grafana
built-in,
exactly.
E
No,
the
goal
the
goal
is
to
have
the
autocompletion
and
everything
there.
The
completion
is
still
not
part
of
it.
We
are
still
working
on
this
experimental
feature
but,
for
example,
to
change
connections
and
to
connect
to
a
different
database.
All
that
logic
is
already
implemented
in
in
the
in
the
library.
So
if
you
want
to
have
that
in
your
plugin
yeah
for
sure
that's
that's
available,
and
it's
something
that,
for
example,
right
now,
the
bigquery
data
source
that
is
being
developed
is
using
is
using.
C
Yeah
actually
just
out
there
that
this
is
still
work
in
progress
there's,
but
there
there's
like
work
going
on
to
add
support
for
for
specifying
your
custom
dialect
for
sql
and
then
you'll
get
auto
completion
out
of
the
box
yeah
if
you
integrate
with
monaco.
So
this
is
work
that
we
started
and
then
sort
of
the
bi
squad
took
over
we're
gonna
try
to
to
consume
that
package
for
for
all
our
sql
data
sources
in
the
near
future.
C
So
I
mean
that
could
be
a
really
good
feature,
but
when,
when
that
day
comes
that
it's
we're
gonna
announce
it
in
some
way.
So
people
will
know.
H
E
G
D
D
D
Athena
lets,
you
write,
sql
queries
to
analyze
and
visualize
your
s3
data,
which
is
kind
of
cool
so
again
to
configure
it's
pretty
similar
you're,
going
to
see
the
same
aws
login
things
that
you
would
see
anywhere
else.
D
We
just
have
a
little
section
at
the
bottom
here
that
lets
you
sort
of
browse
what
databases
you
have
available
and
specify
which
defaults
you
want
to
use
once
you've
sort
of
got
it
set
up.
You
can
click
over
to
dashboards
and
you
can
see
we
have
a
few
curated
dashboards
for
you
that
are
hopefully
not
too
difficult
to
set
up
with
data
you
might
already
have
in
s3
I'll
go
through
this
athena
cost.
You
said
cost
and
usage
report,
one
which
is
pretty
fun.
D
You
can
see
here
we
sort
of
can
look
at
monthly
costs
and
cost
per
service,
and
things
like
that
again,
we
see
the
template
variables
that
we
were
talking
about
earlier.
D
If
you
wanted
to
sort
of
like
you
know,
only
look
at
cost
per
a
certain
region
or
cost
for
a
specific
type
of
service.
You
can
do
that
if
you
wanted
to
like
dig
into
how
we
sort
of
set
up
those
variables.
D
You
know
again,
you
can
write
sql
queries
as
you
would
elsewhere.
This
is
a
very
simple
one.
Just
show
all
the
tables
and
filter
it
up
for
just
the
ones
that
we
care
about.
We
also
have
annotations
here.
You
can
see
one
that's
for
if
a
budget
has
exceeded
so
again
we
have
a
sql
query
that
when
we
turn
it
on
it
can
highlight
yeah.
Here
we
go.
We
have
these
little
lines
here
that,
like
say,
oh
hey,
the
budget
exceeded
at
this
point
in
time.
D
If
we
dig
into
how
do
these
panels
work,
you're
gonna
see
they
look
very
very
similar
to
the
ones
that
we
have
for
redshift.
You
can
we're
set
up
with
default
regions.
Data
sources
and
databases,
but
you
can
also
make
queries
for
different
regions,
data
sources
and
database
as
you'd
like
we
have
these
macros
here
for
tables
and
columns
that
we
just
looked
at
in
redshift.
D
We
have
the
same
ones
here
we
have
the
same
macros
for
things
like
time,
filter
where
you
can
take
take
a
time
and
and
make
sure
that
it
maps
to
whatever
you
have
selected
yeah,
I'm
trying
to
think.
If
there's
anything
else,
we
still
have
the
alerting
you
can
set
up
in
the
same
way.
D
Yeah
it's
pretty
much
pretty
much
it.
I
guess.
H
D
H
Right
because
now
more
and
more,
I
have
more
and
more
requests
when
you
can
actually
configure
the
data-
and
I
mean
it
makes
sense-
actually
create
like
a
form
which
can
deselect,
and
I
mean
search
in
updating
the
data,
because
so
some
customers
actually
wants
to
go
beyond
just
regions
of
data.
This
is
what
I'm
working
on
right
now,
so
it
will
be
greater
actually,
instead
of
just
showing
oh,
we
can
do
select
or
show
we
can
actually
do
update
and
inserts.
E
H
C
Yeah,
I'm
not
sure
about
that,
but
if
yeah,
if
grafana
is
supposed
to
be
like
a
management
tool
for
sql
and
other
services,
that
without
like
a
will,
have
to
keep
like
like
a
completely
new
story
in
mind-
and
I
don't
think
we're
there
at
this
point.
So
it
is
an
interesting
thought,
but
for
now
this
is
something
that
we
want
to
avoid
our
users
to
do
so
when,
in
the
data
source
documentation,
we
specify
only
a
set
of
permissions
for
this
data
source.
E
H
A
So
if
we
want
to
add
another
interface
for
rights
or
or
something
like
it,
then
you
know
we
would
have
to
figure
out
how
to
do
that
in
groupon,
because
a
dashboard
is
typically
like,
like
eric
mentioned
and
or
I
think
it
was
andres
that
you
know
it's
made
to
be
read
only.
A
But
with
that
being
said,
we
are
seeing
a
lot
of
you
know:
companies
that
are
building
control
panels
and
inside
their
dashboards,
for,
for
you
know,
controlling
sensors
and
devices
and
like
that,
so
it's
definitely
possible,
but
yeah
grafana
is
the
the
user
experience
for
writing.
Data
back
is
not
quite
there
yet,
because
it's
not
what
grafana
has
traditionally
been
designed
for
it's
possible,
but
it
might
not
be
the
greatest
experience.
A
Right
and
I
mean
you
can
write
data
back
from
your
query-
editor,
if
you
choose
to
you
know,
if
you
have
a
button
that
writes
data
back,
you
can
totally
do
that.
It
might
not
be
what
all
users
expect,
but
for
for
your
data
source,
it
might
make
sense.
H
B
Cool,
well,
that's
it
for
the
demos
that
we
had
and
thank
you
for
all
the
feedback.
Do
you
have
any?
Does
anyone
have
any
more
feedback
for
any
of
the
plugins
that
we've
discussed
or
any
plugins
that
you
know
we?
I
guess
we've
owned
we
own,
but
we
haven't
spoken
about
or
do
you
have
any
suggestions
for
any
plugins
that
you
would
like
to
see
as
well.
H
So
I
mean
coming
back
to
my
story,
so
now
I
mean
you
can
read
the
data
right
about
the
cloud
watch
different
providers,
but
you
are
not
interested
to
manipulating
data
right.
I
think
no
point
because
right
now
I
mean
I
have
like
this
client
who
said:
oh,
we
create
this
grafana
dashboard.
Now
they
hosted
on
with
devices
everywhere
on.
H
H
Should
I
create
grafana
data
source
to
like
have
a
user
hv
http
api
inside
the
grahada
or
for
aws?
How
I
can
actually
I
mean
there
is
aws
api,
how
I
can
stop
instance
from
grafana
and
same
time
monitors
the
services
which
run
with
us
everything
in
grafana
and
don't
go
to
aws,
and
because
I
mean
it
makes
sense
if
some
services
surrounding
aws
some
of
the
gcp
and
azure-
and
you
have
a
mix
of
that
right-
you
don't
go
to
want
to
go
to
different
panels
for
all
these
providers.
A
So
what
I'm
hearing
you
say
is
that
you
would
like
to
have
a
way
to
to,
for
example,
provision
you
know,
instances
in
aws
create
new
machines.
E
A
Yeah
provision
like-
and
you
know
I
I
think,
application
plugins-
is
a
really
good
place
to
do
this,
because
you
can
you
can
have
you
know
you
have
total
freedom
and
how
you
build
that
you.
A
I
know
you
have
first-hand
experience
with
your
your
writer's
application,
so
you,
you
know
already
know
how
to
do
this,
but
I
I
think
yeah
dashboards
are
very
much
designed
to
be
read
only
so
I
don't
know
the
cloud
data
sources
team
if
there's
any
plans
on
creating
like
application
plugins
for
these
I
know
you're
the
cloud
data
sources
team
and
not
the
cloud
app
plugins
scene,
but
but
still
you
might
have
some
insights.
J
Hi,
so
my
name
is
talina
and
I'm
working
with
our
cloud
partnerships.
What
you've
what
you've
mentioned.
I
have
has
actually
come
up
in
discussions
before.
How
could
we
actually
see
what
grafana
instances
are
out
there
across
different
cloud
platforms
or
even
on-prem,
and
how
could
we
make
that
more
manageable?
That's
not
something
that's
currently
available,
but
it
is
something
we've
talked
about
now
and
there's
some
underlying
work
that
would
have
to
be
done
before.
That
would
be
possible.
So
it's
in
discussions
not
there
yet.
J
It
would
be
a
long
term
thing,
there's
like
in
some
of
the
underlying
work
we're
talking
about.
It
would
enable
that
as
well
as
several
other
things,
but
we
don't
have
all
those
defined
yet
so.
H
A
We
we
have
started
the
the
process
of
discussing
this
and
and
trying
to
identify
common
components
for
data
sources.
I
think
what
we've
had
in
the
in
the
past
is
that
you
know,
because
all
of
the
all
data
sources
are
well.
A
lot
of
them
are
very
different
from
each
other,
but
again
some
are
very
same
as
well.
A
So
I
think
it's
tricky
sometimes
to
create
like
a
unified
interface,
and
I
think
that's
the
the
previous
components
have
led
you
to
let
you
basically
do
whatever
you
want,
and
I
think
that
instead
now
for
the
the
stuff
that
eric
has
been
working
on
to
to
find
those
abstractions
that
make
sense
for
a
lot
of
data
sources.
A
A
I
think
it's
going
to
make
life
a
lot
easier,
but
I
I
will
make
a
point
to
you
know
we
probably
need
some
form
of
style
guide
to
at
least
some
recommendations
for
external
plug-in
developers,
in
what
components
you
choose
and
how
to
think
about
where
they
should
go
and
how
to
build
your
query.
I
think
it
makes
whole
sense
to
have
something
like
that:
we're
not
there
yet
obviously
very
good
feedback.
A
H
You're
good,
so
I
mean
I
wanted
to.
I
talked
to
you
briefly
about
that:
the
difference
between
back-end
data
source
and
the
front-end
only
data
source-
and
you
said
that
you
do
you-
go
for
the
back-end
data
source
when
you
need
alarms
right
alarming
and
that's
the
only
difference,
because
I
mean
most
of
the
things.
If
you,
if
some,
if
your
application
api
has
http,
I
mean
you,
don't
need
to
do
a
backend
right,
but
actually
I
mean
for
http.
H
H
C
I
think
marcus
correct
me
if
I'm
wrong,
but
we're
always
recommending
to
develop
back-end
plugins
these
days,
because
that
will
allow
you
to
it
will
give
you
a
few
things
that
are
pretty
nice,
like,
for
example,
alerting
will
come
out
of
the
box
if
you
just
implement
the
backend
data
source,
plugin
api,
so
yeah.
H
A
No,
it
won't
if
you're,
using
a
front
and
only
data
source
you
can
to
get
alerting,
and
the
reason
is
that
the
requests
are
going
coming
from
going
from
your
browser
to
that
service.
So
if
you
shut
down
your
browser
tab,
that
means
that
there
is
no
query
being
run
so
that
that's
the
that's
the
only
reason.
That's
the
reason
why
alerting
only
works
for
the
backend
data
sources
because
you
they
can
keep
running
in
the
background
on
the
server,
whereas
front-end
data
sources.
A
They
only
run
in
your
browser
tab
right-
and
you
know
you
don't
want
to
keep
your
your
browser
tab
up
to
to
get
those
alerts
right.
H
Now
I
understand,
but
still
I'm
kind
of
using
some
kind
of
backend
right
http
is
the
standard
one,
but
I
mean
I
probably
have
a
problem.
Okay,
I
have
to
experiment
with
that.
But
alerting
is
the
only
only
difference
right
because
I
mean
when
you
do
back-end,
you
have
to
spend
your
time
writing
the
go
code
and
then.
E
H
Write
the
same
logic
on
the
front
end,
so
I
just
prefer
to
for
some
third
party
which
I'm
not
supposed
to
have
alerting
right,
not
all
of
them
supposed
to
have
it.
You
can
only
do
front
end.
Is
there
any
other
difference.
A
A
Solve
it
with
those
you
kind
of
need
to
go
to
back
in
data
sources
and
the
second
one
is:
if
you
have
non-http
requests
like
if
you're
connecting
over
another
protocol,
for
example,
then
you
would
also
need
to
use
the
back-end.
But
your
if
you
have
a
rest
api
in
the
background
and
you
don't
need
alerting.
A
We
highly
recommend
backing
data
sources,
because
that's
all
where
all
the
development
is
going,
but
you
can
definitely
do
it
with
a
front-end.
Only
one
but
like
eric
says,
if
we
we
are,
are
strongly
pushing
for
back-end
data
sources.
E
A
So
query
caching
would
have
to
be
implemented
by
the
back-end
right,
so
that
is
kind
of
up
to
you.
But
but
let's
say
that
you
have
you
have
a
hundred
users
of
your
your
dashboard.
They
will
all
send
requests
to
your
your
database
right.
So
you
would
open
potentially
a
hundred
connections
rather
than
have
one
connection
from
your
final
server
and
then
having
them
all
go
through
that
and
then
you
can
do
more
aggressive,
caching
and
stuff
like
that.
A
But
again
it
depends
on.
You
know
it
sounds
like
you
have
a
rest
api
in
the
background
and
you're.
Not
you
don't
really
care
about
caching
either
so
yeah.
I
I
can't
strongly
recommend
like
I.
I
can't
tell
you
not
to
use
it,
but
we
we
recommend
backend
data
sources
right
because.
E
A
Yeah
I
did
that
for
the
json
api
plugin
and
now
I'm
kind
of
stuck
because
I
don't
have
the
some
dependencies
aren't
available
on
the
go
side.
So
I
can't
actually
do
that
because
I
would
have
to
re-implement
those
dependencies.
So,
yes,
you
can't
prototype
but
be
aware
that
if
you
want
to
move
it
at
some
point,
you
kind
of
need
to
make
sure
that
those
dependencies
are
available
on
the
go
side
as
well
just
from
personal
experiments.
H
I
thought
there
is
a
back-end
what
started
to
be
implemented
on
typescript,
not
on
the
go.
Is
there
any
timeline
when
it's
going
to
be
like
we
can
use
it.
A
Yeah,
I
think
there
are
several.
I
don't
think
anyone
are
official
ones.
I
know
there
is
one
in
bust
as
well
there's
one
in
in
c
sharp
and
there's
one
in
java.
I
think
in
python
we
we
have
a
lot
of
passionate
developers
who
are
working
on
these
separately,
but
I
don't
I'm
not
aware
of
any
official
commitment
from
the
grafana
team.
A
A
A
It
is
a
rough
reference
implementation,
so
it's
it's
probably
going
to
stay
stabilized
at
some
point
and
maybe
we'll
see
more
implementation.
A
A
So
I
think,
unless
there
is,
I
we're
we're
five
minutes
over
and
I
just
want
to
thank
you
all.
You
know,
especially
the
cloud
data
sources
team
for
joining
in
and
sharing
what
you're,
working
on
and
and
really
cool
stuff.
I'm
looking
forward
to
have
you
in
the
future
and
so
we're
gonna
have
another
community
called
next
month.
A
We
haven't
decided
on
a
theme
yet
so,
if
you're
really
interested
in
a
certain
topic-
and
you
want
the
graffana
team
to
to
share
their
thoughts
of
what
they're
working
on,
please
let
us
know
and
we'll
make
it
happen.
So
until
next
time
see
y'all.