►
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
Alright,
if
we
had
music,
it
would
be
playing
right
now,
but
we
don't
so
I'm
just
going
to
talk
about
stuff
I.
Welcome
to
this
week
in
Cassandra
planet
Cassandra,
we
are
pretty
excited
for
this
week.
I
am
back.
I
was
missing
last
week.
Luke
did
a
good
job,
probably
better
than
I
have
in
any
of
the
things
I've
done
and
there's
a
good
chance.
I'm
gonna
be
voted
out
of
this
week,
Cassandra
sometime
in
the
near
future,
but
that's
okay.
A
We've
got
Luke
Tillman,
we
got
Patrick
McFadden,
we've
got
three
high
and
we
have
our
new
wonderful
guest.
Carlos
rollo
he's
going
to
be
talking
to
us
about
some
of
the
consulting
I
experiences
that
he's
had
the
past
some
of
the
weird
problems
that
he's
fixed
in
some
of
the
old
versions
of
Cassandra
that
he's
unfortunately
had
to
encounter,
but
the
real.
A
A
A
So
what
do
we
got
first
here
before
we
get
going
with
you,
Carlos
we're
gonna
talk
about
the
news,
because
that's
important
right
this
week,
we've
got
a
couple.
New
bug
fix
releases
for
Cassandra.
We
got
two
to
one
saw
a
bug
fix
released
into
to
these
guys.
Keep
going.
This
is
good
stuff
to
one
is
definitely
super
stable
release
and
now
there's
a
handful
of
boats
that
got
fixed
in
it,
but
pretty
solid
overall
tutus,
probably
the
place
that
most
people
are
putting
their
new
clusters
in
production.
C
A
It
just
hit
two
to
six,
so
that's
pretty
cool
and
then
of
course
you
know,
we've
got
a
few
blog
posts
and
what
do
you
know
lucky
us?
We've
got
Luke
Tillman
here,
he's
gonna,
give
us
a
little
love.
You
know,
he's
got
11
token
ring
to
rule
them
all
it.
Doesn't
he
or
a
tree
versus
so
Luke.
Once
your
once.
You
give
us
a
little
background
on
this.
D
He
made
this
kind
of
offhand
comment
that
turned
out
to
be
and
and
the
people
that
we
were.
You
know,
kind
of
doing
this
boot
camp
with
were
pretty
experienced.
Cassandra
users,
you
know
they've
been
at
it
for
a
while
new
know,
Cassandra
pretty
well,
and
he
made
some
comment
when
he
was
explaining
replication,
where
he
basically
said
well,
there's
only
one
token
ring,
there's
only
one
range
and
and
then
he
said
well,
you
know
it's
I
understand.
D
D
It
looks
like,
but
really
like
you
know,
everything
is
built
on
top
of
this
idea
that
there's
only
one
token
range,
you
know:
that's
that's
divided
across
all
the
nodes
in
in
all
your
data,
centers
and
replication,
and
data
centers,
and
all
these
other
things
are
kind
of
built
on
top
of
it,
and
that
was
kind
of
like
an
aha
moment.
I
guess
for
me
where
I
was
like,
like
I
guess:
I'd,
never
really
thought
about
how
kind
of
deceiving
it
can
be.
D
Sometimes
when,
when
we,
you
know,
have
these
visual
depictions
of
Cassandra
a
lot
of
times
we're
showing
them
group
we're
showing
nodes
group
by
data
center
and
thing
like
things
like
that,
and
and
really
you
know
it's
it's
kind
of
hard,
sometimes
for
beginners
to
get
their
mind
around
some
of
these.
You
know
kind
of
terms
like
cluster
and
and
rang
and
data
center,
and
and
that
kind
of
thing
so
and
that
basically
the
the
thrust
of
the
blog
post
is
hey.
Take
a
look
at
these
pictures.
D
A
Yeah
well,
this
is
that
I
think
this
is
one
of
those
things
that's
really
important.
Maybe
this
is
a
fundamental
thing
and
I
love.
I
love
the
fact
that
you
were
able
to
use
the
content
in
datastax
academy,
so
you
didn't
have
to
rehash
everything
that
everybody
needs
to
know
about
how
a
snitch
works.
You
know,
because
otherwise
it
turns
into
like
okay
I
have
to
cover
35
different
topics
in
my
blog
post
and
then
you
end
up
with
like
the
longest
blog
post
ever
that
perhaps
needs
a
table
of
contents
and
yeah.
A
So
it
was
cool
to
see
the
you
know
being
able
to
link
to
Academy.
Where
now
we've
actually
have
you
know
these
small
segments,
the
one
that
you
did,
patrick
I
thought
was-
was
really
good
on
the
snitch.
It's
just
a
few
minutes.
It's
like.
If
you
just
want
to
learn
about
snitch
works,
you
can
watch
this
five
minute.
Video
yeah.
B
Got
me
got
to
be
a
snitch
yeah.
Well,
I
think.
Actually
I
did
a
talk
called
night.
I
have
99
problems
in
its
niche
81
and
thank
you
jay
z,
but
that's
the.
I
think.
That's
that's
probably
way
more
in
the
weeds
and
some
people
want
to
get,
but
it
is
pretty
critical.
I
mean
topology.
Awareness
is
in
the
snitch.
B
The
snitch
is
what
deals
with
what
data
center
you're
in
in
the
data
placement
and
if
you're,
in
rock
awareness
like
making
sure
your
date
is
not
unlike
one
rack,
so
the
power
goes
up
that
and
then,
if
you
think
about
what
we,
how
we
manage
data,
you
know
with
the
different
stitches
like
easy
to
snitch
or
Google
snitch
or
property
file
snitch,
it's
all
about
where's.
My
data
in
this
big
ring
around
the
world,
so
yeah
yeah.
A
I
just
want
to
back
up
just
a
brief
moment
with
the
table
of
contents
thing,
because
I
think
I
think
it's
important
to
discuss
the
epic
tome
of
knowledge
that
should
be
high,
has
posted
this
week
about
sazzy
that
actually
did
need
a
table
yeah
this
dude,
I
love
this
post
like
I
sat
down
and
I
was
like
probably
half
an
hour
of
just
going
through
the
details
of
this.
It
was
really
really
good.
Well,.
E
Want
to
keep
the
knowledge
in
the
same
place.
You
know,
because
you
want
to
know
about
something
something
about
sassy.
You
have
just
one
place
to
go:
you're
and
probably
in
the
future,
we
may
and
reach
our
official
documentation
with
more
content,
but
at
least
right
now
for
people
want
to
try
out
sassy.
They
have
something
to
to
read.
A
Yeah,
well,
it's
really
nice
to
see
that,
like,
if
you're
interested
in
working
on
it,
it's
really
nice
to
be
able
to
read
like
technical
details
without
before
you
dive
into
the
code,
so
you
have
at
least
a
frame
of
reference.
Ok,
this
is
the
purpose
of
it
right.
You
talk
about
how
the
tree
is
built
that
b+
tree.
E
Code,
if
you're
in
via
developer,
try
to
work
on
so
and
if
you
are,
and
if
you
don't
have
enough
time
to
read
through
all
this
stuff,
just
go
to
the
last
chapter,
where
I
give
some
recommendation
when
you
should
use
sassy
and
when
you
should
avoid,
because
it's
not
magical,
it
has
some
like
drawbacks.
So
it's
all
about
trade-off.
Yep.
A
C
C
C
A
But
yeah
the
other
thing
I
did
like
about
that
posters.
It
also
pointed
out
some
things
that
need
to
go
with
road
map
and
the
architectural
changes
that
need
to
happen
with,
probably,
like
you
know,
increased
performance
even
more
like
it's
just
really
cool
to
like
see
how
someone
can
debug
something
like
you
know.
Okay,
this
is
how
you
can
get
into
the
Cassandra
Kota
base
and
like
learn
how
this
functionality
works.
Any
there's
a
bike,
a
patch
there,
which
I
think
is
amazing
on.
A
B
Be
fair
and
I
think
a
lot
of
it.
So
you
know
we
do.
We
have
been
talking
to
Pablo
and
that
team
quite
a
bit
about
the
roadmap
and
I
would
just
say
stay
tuned.
This
is
just
the
beginning.
There's
a
lot
of
really
cool
things.
Coming
with
sassyxx,
it's
been
all
bad
yeah.
Actually,
it's
not
bad
at
all.
You're
scanning
it
yeah.
C
A
No,
no,
no,
like
advanced,
query,
planner
stuff,
you
know,
there's
there's
other
features
that
this
is.
This
is
why
I
really
really
like
tick
tock,
where
they
schedule,
because
it's
like
every
month,
you're
like
what's
coming
out
and
you
get
to
like
look
at
some
cool
stuff
and
hammer
on
it
and
test
it
Mike,
you
know.
Should
we
hide,
you
know
you
did
all
that
work
like
hammering
stuff
on
like
that
cluster,
that
you
got
a
hold
of
you.
We
didn't
have
to
wait
nine
months
to
test
that.
I
really
like
that.
A
Roughly
your
beautifully
way
bigger
like
super
push,
the
other.
The
other
thing
I
wanted
to
talk
about
before
we
go
the
week
it
caught
this
as
ething
is
another
sri
hi
thing:
we
have
the
Achilles
release
with
support
for
you
da
that's
right,
yeah.
E
So
just
the
the
idea
is
you
define
you
declare
your
UDF?
Are
you
Leah
in
a
Cadiz,
and
at
least
we
build
at
compile-time
a
query
like
a
dsl
so
that
you
can
apply
your
function
in
a
type
safe
manner?
For
example,
if
you
have
like
a
text
to
integer
function,
you
can
only
apply
this
function
on
a
text
column
and
not,
of
course,
on
uuid
column,
because
it's
just
not
possible
from
the
type
safety
point
of
view.
Yeah.
A
Very
cool,
yes,
well
right,
so
I
think
that
covers
all
are
a
lot
of
blog
posts
that
we
were
going
to
go
over
it
Carlos
guess
what
word
so.
C
A
You
there
you
go
wait.
You.
E
B
And
I
were
talking
earlier,
you
and
I
have
been
probably
doing
consulting
on
Cassandra
well,
probably
about
the
same
amount
of
time
at
different
time
periods,
but
I
think
it's
really
useful
for
people
to
understand
the
same
mistakes
to
get
over
and
over
and
over.
Maybe.
C
So
what
I
see
awful
lot
and
that
it's
so
you
have
a
relational
database
and
at
some
point
you
decided
that
it
doesn't
scale
you
want
Cassandra,
but
the
problem
is
you
decide
to
jump
in
Cassandra
and
you
just
make
exactly
the
same
data
model.
You
know
when
then
you
figure
out
well,
why
can
I
query
by
this
or
that?
And
you
start
throwing
secondary
indexes
on
all
that
and
I
see
tables
with
all
fields
with
all
secondary
indexes.
I.
Every
single
thing
is
on
the
secondary
index
and
then
it
it
will
work.
C
It
will
work
until
we
have
three
nodes
because
you're
not
scaling
on
your
dev
environment
to
work
for
sure.
Then
you
get
to
that
five
six
nodes
and
bum
there
you
go
it's
not
performing
anymore.
It's
you
need
to
rewrite
your
little
model
at
this
point
it
probably
at
some
amount
of
data,
which
is
then
you
need
to
change
applications
and
all
that
it
gets
messy.
So
that's
the
day.
C
A
C
Versioning
yeah
versioning
people
yeah-
if
it
works,
don't
touch
it
until
they
figure
out
their
weird
barking
apps
and
then
need
soap
right
all
the
way.
Then
it's
not
easy
to
operate
anymore
because
you
need
well
it's
not
that
difficult.
But
it's
a
lot
of
trouble
might
involve
yeah
some
months
Oh
foot
to
push
stuff
to
a
decent
level
re.
Can
you
you've
had.
A
C
So
normally
the
the
most
difficult
process
is
actually
convincing
people
that
they
need
to
move
so
like.
Well,
you
know,
I
have
traits
and
it
still
works.
Why
should
I
cql
will
actually
prefer
to
upgrade
version,
but
don't
change
the
driver
and
they
go
all
the
way
to
21
and
they
are
sell
one
actor
using
trip
yeah.
C
So
this
is,
but
so
that's
the
most
bit
difficult
point
at
then
it's
then
the
next
point
is
actually
make
expected
test
systems
that
you
can
actually
show
that
stuff
works.
You
know
it
don't
be
afraid.
Yeah
we
put
to
the
test
system.
Next
to
yours,
we
load
some
data,
you
can
test
your
apps
and
things
work
and
then
eventually
they
will
make
the
move.
It's
never
a
fast
process.
It
takes
it's
a
month.
Yeah
I
never
seen
anything
go
faster
than
23
mar
yeah.
Even
if
it's
just
big
balls
database.
B
So
we
should
probably
talk
about
hardware
too,
because
I
think
you're.
You
said
this
earlier
when
you
get
these
people
that
are
so
in
a
bad
place
and
then
they're
like
I'll
just
add
more
hardware,
sometimes
there's
some
really
bad
choices
in
hardware
as
well.
C
Yeah,
it's
that's.
The
sink,
people
think
well,
I
read
it's
adding
water
nodes
and
it
puts
well.
But
then
everything
is
on
an
ass
of
a
son
and
there
you
go.
It
doesn't
matter
how
many
nodes
you
are.
You
are
eating
the
same
storage
all
over
again.
So
that
is
a
really
really
bad
thing
and
it's
seen
quite
often
every.
C
A
C
Makes
that
difficult
to
handle,
for
example,
I?
Don't
remember
the
name
of
one,
but
it
was,
was
really
weird
at
four
and
then
also
problems
with
cpu.
That
was
at
some
point
like
people
don't
understand
the
cpu
matrix
and
I
get
like
CPU
steals
of
sixty
percent,
and
then
people
complain
that
the
center
is
not
performing
yeah.
C
B
Think
you
and
even
I
were
spending
some
time
on
slack
one
time,
and
you
were
talking
about
this
cluster.
That
was
just
on
the
edge
disaster
and
it
was
still
holding
up
even
though
they
were
doing
horrible
horrible
things
to
it,
and
it
was
still
serving
data,
but
yeah
it
goes
yeah.
It
goes
that
whole
thing
that
Cassandra's,
like
the
honey
badger
of
databases,.
C
B
A
C
C
B
A
And
we
do
have
a
public
job
opportunities
right
now,
there's
like
almost
ten
job
opportunities
sitting
around
Netflix
last
pickle
weather
company
walmart
we're
hiring
for
an
evangelist
just
there's
a
bunch
of
stuff
in
there
like
if
you're
looking
for
a
job
check
that
out
the.