►
From YouTube: CI WG demo: Syndicate with iRODS
Description
Date: 02/03/17
Presenter: Chun-kun "Amos" Wang
Institution: UNC-Chapel Hill
South Big Data Hub
A
On
to
Chun
Kuen
mo
swang
will
be
talking
to
us
today
about
syndicate
dr.
Wang
received
a
BS
and
master's
degree
in
national
Chun,
Qiang
University
in
Taiwan
in
2009
and
2011.
He
is
currently
working
towards
the
PhD
degree
in
computer
science,
University
of
North
Carolina
at
Chapel
Hill
this
research
interests
through
the
parallel
programming
and
system
verification
that
will
hand
it
over.
B
B
So,
first
of
all,
the
syndicate
is
the
story
surface.
That's
build
a
coherent
storage
abstraction
from
the
already
deployed
component,
so
it
can
be
include
all
storage,
l,
edge,
caches
and
deposit
providers,
and
one
of
the
nice
features
that
it's
fully
decoupled
education
from
providers
and
the
transparently
leverage
them
without
modification.
So
it's
not
only
offer
consistent
semantics
across
the
multiple
providers,
but
also
over
a
flexible,
a
programming
model
to
application,
so
the
user
can
just
define
their
own
provider
as
a
non
stick
functionality.
B
So
we
don't
have
to
take
care
about
what
is
the
underlying
are
base?
You
are
going
to
use
so
the
next
slide.
It
shows
the
Syndicates
architecture
so
in
according
to
I
think
it's
indicated
architecture.
We
have
three
major
components
and
first
one
is
the
user
syndicates
gateway,
so
X
Q
stands
both
in
this
indicates
gateway.
The
users
Sinek
has
gateway,
is
interface
with
H
edge
cases
for
an
advertising
endpoint,
so
its
response
to
the
application,
read/write
request
and
six.
B
B
So
right
now,
I
have
already
setup.
My
virtual
machines
am
I
running
all
the
experiments
on
the
docker
container
and
the
god.
Docker
container
is
the
virtual
machine,
so
we
can
use
the
docker
container
to
simulate
the
behavior
of
a
distributed
file
system.
So
here
right
now,
I'm
already
boost
my
docker
container.
So
you
can
see
here
there
is
a
5
container
right,
one.
It's
a
replica
gateway
and
users
gateway,
an
acquisition
gateway
and
Iraq
itself
and
MX
represent
Makati
pass
surface,
so
it's
manager,
metadata
server
and
then
control
control
that
actually
data
location.
C
A
B
Now
there
is
nothing
exists
in
The
Syndicate
file
system,
because
I
didn't
upload
anything,
so
this
window
just
show
you
I
can
upload
something
wrong:
I
rot
so
right
now,
I
love
into
I,
rot,
container
and
then
I
think
Savannah
I
already
upload.
A
one
I
rot
file
is
called
I,
rot
files
that
50
and
I
already
uploaded,
but
I
have
to
activate
my
acquisition
gateway
to
death.
The
Syndicate
is
nowhere
what
kind
of
file
you
already
upload
so.
B
So
so
right
now,
I
are
going
to
activate
acquisition
gateway.
So
it's
already
accept
and
then
provided
on
the
data
set
to
The
Syndicate
file
system,
and
then
we
we
execute
the
LS
again.
So
we
can
see
the
file
so
right
now
we
got
okay.
I
see
the
Iraq
file
that
we
see.
I
already
provide
upload,
one
arm,
I
rot,
they
have
database,
and
then
we
can
look
at
take
a
look
within
use
page
on
to
take
a
root.
Allah
I.
B
Ok,
so
this
is
the
automatic
upload
file
for
Iraq
from
the
script.
So
this
my
are
just
just
just
a
way.
I
I
can
get
a
date
happen
to
make
that
file
system
and
then
the
next
one
I'm
going
to
show
a
flat
bar.
If
we
wanted
to
upload
some
files
on
the
user
gateway,
which
you
can
do
this
the
first
of
all
week,
we
asked
a
local
file,
subject
e
and
then
we
just
time
this
is
a
sample.
B
B
B
C
B
C
C
B
Currently,
we
don't
have
the
we
don't.
We
don't
provide
the
performance
comparison
because
we
develop
some
instruction
and
put
some
shocking.
There
is
still
I
mean
there's
a
potential
box,
so
we
verify
the
correctness
is
our
first
priority
and
after
we
finish
that
we
will
to
go
in
to
look
at
the
performance,
II
issue.
A
Okay,
well
hearing
no
other
questions.
I
want
to
thank
amis
for
a
great
presentation,
and
thank
you
all.
Does
anybody
have
before
we
close
for
the
weekend?
Does
anybody
have
any
other
announcements
or
things
they
want
to
share
with
the
group?
I
see
Jennifer
hammocks
on
the
phone
and
her
Ser
spoke
project
I
think
is
pretty
interesting
that
they
were
one
of
the
spoke
T
eyes
that
were
awarded.
Does
anyone
want
to
jump
in
and
add
anything
today?
Oh.
C
This
is
Jen
can
I
just
a
quick
word
about
next
meeting
yeah.
Please
do
I'm
really
excited,
although
barely
following
this
discussion,
I
will
come
to
representing
a
community
that
is
mostly
biologists
with
very
few
data
scientists,
so
I'm
hoping
to
have
a
bunch
of
data
scientists
next
week.
What
I
plan
to
do
is
describe
to
you
our
data
and
hope
that
people
can
advise
me
about
solutions
for
our
data.
Just
so,
you
all
know
where
I'm
going.
A
So
look
for
that.
We
will
be
beginning
to
wrap
up
our
demo
and
begin
to
move
towards
more
real
working
meetings.
But
if
folks
have
any
recommendations
for
additional
demos
that
you
think
we
really
should
have
it's
not
too
late.
You
certainly
could
make
some
recommendation
as
we'd
like
to
hear
it
so,
and
you
have
Carl's
email,
I
believe
as
well
as
mine,
and
so
please
send
us
your
ideas
any
last
thoughts
before
we
close
for
the
weekend.
A
Okay,
I
know
it's
Friday,
people
are
probably
there
so
well.
Thank
you
all
speakers
I
really
appreciate
and
thank
the
rest
of
you
for
for
joining
us,
and
and
thank
you
all
so
Regan
and
Jason
for
asking
great
questions.
We
really
appreciate
it
and
with
that
we
will
close
have
a
good
weekend.
All
thank.