►
From YouTube: The Apache Cassandra® Corner podcast w/ Mary Grygleski
Description
For this episode, I sat down with Mary Grygleski of DataStax. Mary is a Developer Advocate on the Astra Streaming team, evangelizing for and helping folks with Apache Pulsar®. She has a long history as both an advocate and an engineer in tech, organizes the Chicago Java Users' Group (CJUG), and has been recognized as a Java Champion. Listen in and hear how we discussed Java, Apache Cassandra®, Pulsar, and Mary's talk for the upcoming Cassandra Summit.
A
B
B
B
B
So
yes,
I
am
Mary
quickleski!
Thank
you!
Aaron
and
I'm,
a
developer
Advocate
with
the
streaming
team
at
data
stacks
and
so
I
the
area
that
I
primarily
advocating
for
it
would
be
the
event
streaming
platform
Astra
streaming
data
stacks
and
then
it's
powered
by
Apache
Pulsar.
So
a
lot
of
you
know
all
of
our
things
that
data
Stacks
are
really
open,
source
powered
so
so
essentially
yeah
Apache,
Pulsar,
so
event,
streaming
kind
of
more
catered,
For,
The,
Cloud
native
platform.
B
So
that's
what
I'm
advocating
for
and
then
in
my
spare
time
or
you
know
off
hours,
I
would
be
running
community,
so
I'm
also
the
Chicago
Java
users
group
president
yeah
and
then
I'm
also
Java,
Champion
too
so
yeah
and
I'm
interested
yeah
and
I'm
interested,
not
just
in
Java,
but
everything
else
and
now
nowadays,
even
like
getting
into
AI
like
our
company,
is
doing
and
things
like
that,
and
not
just
that.
Just
a
lot
of
variety
of
different
things.
B
I
think
I'm,
just
very
interested
in
So
yeah,
so
and-
and
it's
just
some
background
2-
is
that
I
was
an
engineer
before
becoming
an
advocate
at
IBM.
That's
where
I
started
being
you
know
started
off
my
efficacy
career
in
2018,
so
it's
been
five
years,
I've
been
doing
efficacy
back
then
I
was
doing
more
Java,
focused
reactive
technology
and
now
here
doing
event
streaming
yeah
and
so
I
was
an
engineer.
B
Doing
you
know
different
products
too,
and
in
the
90s
I
was
actually
doing
Unix
and
c
and
then
in
2000,
switching
doing
Java,
Java,
open
source
and
and
then
you
know,
from
products
to
engineering
applications
and
then
now
back
to
doing
products
and
efficacy
so
yeah.
So
that's
kind
of
a
nutshell
about
me
so
yeah,
oh.
A
A
A
Certainly
are
a
Lacross
over
there.
I
guess
is,
is
where
I'm
going
with
that
yeah
and
actually
that
was
kind
of
where
you
know
I
was
a
Java
developer.
You
know
it
really
wasn't
my
strong
suit
15
years
ago,
when
I,
when
I
first
started
doing
it,
but
it's
like
that's.
That
was
the
same
time
that
I
got
into
Cassandra
and
you
know
just
again
learning
so
much
about
the
the
Java
ecosystem.
A
B
A
B
About
oh,
go
away,
but
no
it's
it's
a
very
you
know
a
long
proven
system.
It's
it's
you
know
is
all
these
things.
That
is
so
much
like
open
source
options
to
it
too,
and
the
community
too,
is
amazing
right,
so
big
yeah
yeah,
even
the
whole
Apache
software
Foundation,
you
know
most
of
the
software
is
written
in
Java
as
well,
and
the
world
is
pretty
much
all
the
back.
End
systems
are
still
powered
by
Java,
there's
just
very
robust
yeah.
It's
proven
so
yeah
yeah,
yeah,
yeah,
yeah,
I.
A
B
B
A
B
B
A
B
A
community
is
nice,
but
I
think
the
key
again.
You
know,
of
course
you
know.
Ideally
if
you're
programming
you're
a
technical
person
would
be
even
better,
but
the
thing
is
I'm
finding.
Also,
interestingly,
there
are
some
folks
that,
if
you're
more
like
in
other
aspects
of
you
know
working
in
the
computer
industry,
but
you
kind
of
take
an
interest
or
you
know
very
active
in
doing
promotion
of
java
you'll
be
recognized,
you're,
helping
to
run
Java
users,
group
meetups
and
things
like
that,
and
maybe
you
are
less
of
a
Hands-On
technical.
B
There's
a
variety
of
different
ways,
but
it
is
true.
You
definitely
have
to
make
yourself
known
to
the
community
and
and
be
active
too
I.
Think
that's
that's.
What
I'm
finding
yeah
is
it's
kind
of
very,
very,
very
crucial
in
this
case.
Yeah
I
found
it
so
and
also
everything
is
internally
nominated.
So
if
you
know
someone
you've
done
enough
work,
the
person
will
be
confident
they
can
nominate
you
to
kind
of
like
that.
So
yeah,
okay,
okay,
oh,
it
works,
yeah,
yeah,
good.
A
A
B
B
B
B
A
A
Actually,
you
know
I
I,
think
I
have
something
that
that
would
be
good
to
talk
about.
That
is
both
Cassandra
and
Java,
that
yeah
yeah.
No,
it's
it's
an
example.
I
built
out
I.
Think
I
sent
this
to
you
at
the
get
repo
I
did
on
building
out
the
new
Vector
search
feature
and
like,
like
a
spring
boot
service
that
that
runs
and
kind
of
serves
that
out.
You
know
and
and
helps,
helps
you
do
like
an
a
n.
A
B
A
B
A
B
A
B
A
B
B
B
Around
the
hillside,
so
you
can
kind
of
think
of
you
know
kind
of
using
a
data
event
streaming
ingestion
platform
such
as
that,
because
your
iot
devices
are
collecting
all
the
data
they
keep
sending
to
the
server
and
then,
basically,
you
have
some
point
of
sending
it
back
into
your
app
on
the
server
and
processing
this
data,
maybe
that
that
would
be
a
good
good
use
case
scenario.
Thinking.
A
That's
a
cool
idea,
though
yeah
yeah
and
I
never
would
have,
would
have
thought
about
yeah
using
it,
for
you
know,
for
you
using
an
iot
device.
You
know
spread
out
across
a
pasture
like
that,
but
yeah,
that's
that's
cool
yeah,
so
yeah
and
then,
of
course
you
have
all
that
data
from
the
iot
streams
coming
in
and
you're.
You
know
that's
where
you
want
the
yeah,
the
guaranteed
delivery
and
the
the
exact
order
that.
B
B
Monitoring
any
kind
of
suspicions
activities
you
know,
based
on
the
data
you
need
to
analyze,
so
the
ingestion
part
too,
is
very
important
because
there
are
any
given
point
in
time.
You
have
so
many
different
activities
going
on
so
yeah.
You
collect
all
those
inputs,
yeah
and
through
an
event
streaming
platform
that
that
yeah.
That
would
be
also
good
use
cases
for,
for
that.
B
Yeah
and
of
course,
also
recommendation
system
right:
there,
people
doing
shopping
and
you're
collecting
all
their
clicks.
You
know
keystrokes
and,
and
then
you're
analyzing
and
sending
them
down
and
machine
learning.
You
know,
algorithms
to
analyze
their,
you
know
their
patterns
and
then
figure
figuring
out.
You
know
giving
out
recommendations
as
they
are
shopping
too
so
yeah,
so
those
you
can
think
of
event
streaming
to
me.
B
I
feel
almost
seems,
like
you
know,
if
you
kind
of
explain
this,
you
know
to
my
grandma
who
doesn't
know
you
know,
does
isn't
technology,
you
know
focus
and
you
can
say
well,
it's
sort
of
like
infrastructure
in
your
house.
You
walk
into
a
house,
you
don't
really
see
it.
These
are
pipes
that
are
running
underneath.
You
know
that
Behind,
the
Walls
and
your.
B
They
are
so
essential
too.
They
are
like
you
know,
that's
where
the
pipes
are
kind
of.
You
know
when
you
turn
on
faucet.
It
sends
out
the
water
to
you
and
it
kind
of
drains
back
to
the
output.
You
know
whatever
exhaust,
not
exhausted
waste,
you
know
pipes
and
things
like
that.
So
I
think
in
some
ways
event
streaming.
It's
kind
of
it's
like
that.
You
know
in
the
if
I
have
to
put
in
analogy
that's
what
event
streaming
the
type
of
platform.
B
A
B
Sure
so,
yes,
so
again,
right
I'm
doing
streaming
for
data
Tech,
so
we
have
Apache
Pulsar,
so
Apache
Pulsar
too
again,
you
know
event
streaming
for
ingestion,
so
here
it
is
I'm
thinking
right.
Data
Stacks
is
not
only
about
event
streaming
I
mean,
of
course,
Cassandra
is
the
main
thing
and-
and
you
know
so,
I'm
kind
of
been
thinking
of
you
know.
There's
got
to
be
ways
of
integrating
the
two
together
and
for
sure.
Yeah
again,
you
know
so
with
event
streaming.
B
My
talk
actually
is
really
to
explore
right
about
building
a
data
Pipeline
and
then
using
Apache
pulsar
and
of
course
too.
You
know
just
like
water
too,
when
you
flow
through
you
kind
of
water
goes
through
different
treatment
plans.
So
it's
like
data
coming
through.
You
want
to
do
some
transformation.
For
example,
you
you
can
have
data
that
is
flowing
through.
That
only
sends
the
message.
Each
one
of
them
carries
some
kind
of
IDs
right
along,
but.
B
Only
ID,
so
you
need
to
do
some
sort
of
lookup
too
to
your
tables.
So
that's
where
I
think,
let's
say
Cassandra
can
come
into
play
because
that's
where
your
your
Cassandra
is
storing.
Let's
say
it's
coming
in
user
ID.
You
need
to
use
look
up
more
information
about
the
user.
So
you
then
you
know
your
data
comes
in.
You
want
to
do.
Look
up
to
your
Cassandra
database,
based
on
the
user.
B
Id,
look
up
the
information
you
know
in
your
in
a
data
store,
so
that's
actually
what
I'm
working
on
is
is
actually
building
this
pipeline,
using
both
Apache,
pulsar
and
Cassandra,
and
basically
and
from
Poster's
perspective
I'm
using
Pulsar
functions,
and
these
are
lightweight
functions
that
you
can
plug
into
your
pipeline
and
that
will
do
message,
transformation
or
selection
filtering
whatever
you
need
to
do
so.
Insight
is
just
bite.
B
Size
function
that
you
can
I
can
do
query
against
a
Cassandra
and
and
search
for
the
information
I
need,
and
then
you
transfer
form
your
data
essentially,
and
then
you
send
them
off
to
your
Downstream
to
to
your
destination.
To
a
sync
like
that.
So
essentially,
my
talk
is
about
that
and
just
coming
up
with
this
demo
scenario,
to
show
to
demonstrate
how
we
can
integrate
the
technology
together
and
essentially
provide
solution
to
solve
some
business
problems
so
yeah.
So
that's
what
it
is
in
a
nutshell:
that's
what
my
talk
is
about
so
yeah.
Oh.
A
B
Oh
yes,
yes,
yes,
you
can
you.
Can
you
can
do
that
too?
So?
Yes,
so
essentially
data
gets
sent
to
the
topic.
So
from
the
topic
too,
that's
where
you
know
you're,
you,
you
get
your.
You
know
your
subscriber
you're,
getting
the
messages
so
you're
consuming
it.
So
from
the
consuming
side
to
you
can
basically
have
you
know
your
your
functions
can
be
called
to
in
in
there
right
and
oh.
A
B
Right
so
so
you
should
be
able
to
then
yeah
do
whatever
you
need
and
then
basically
from
there,
then
it
gets
sent
Downstream
to
whoever
destination.
You
need
to
go
to,
let's
say
you're,
building
a
set
of
pipelines
that
are
all
connected
together
and
that
that's
possible
too
right
and
or
your
output
can
be
sent
to
another
topic
as
well.
So
so
that
will
be
another
topic
for
somebody
else
to
pick
up,
so
you
can
think
of
chaining
all
your
data
pipelines
together
so
I
think
in
in
this
case.
B
You're,
like
you,
know,
we're
kind
of
treating
data
like
they
can
kind
of
flow
through
the
system
quite
easily.
In
this
sense
you
know,
rather
than
you
know,
the
more
traditional
way
is.
Let's
say
you
make
restful
calls
and
things
like
that
and
all
that
stuff,
which
is
different
kind
of
approach
to
doing
things,
but
in
this
case
you're
right.
A
B
A
B
A
B
B
B
To
be
able
to
do
some
lookups,
for
example,
things
like
that
so
yeah
yeah,
you
should
be
able
to
to
do
that
as
well.
So
yeah
yeah
yeah.
It's
because
the
function
is
think
of
it,
yeah
they're
bite-sized
they
they
can
be
essentially
another
kind
of
interface
or
a
place
where
you
can.
You
know
invoke
another
procedure
right.