►
From YouTube: CNCF Research End User Group: HPC:HTC End User Landscape
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).
C
A
It
all
right,
so
we
can
start.
I
guess
so.
Yeah
welcome
everyone.
Today's
session
is
the
first
one
with
this
new
platform,
so
we
got
a
few
people
here,
so
maybe
we
lost
some
on
the
way,
but
hopefully
not
many.
A
The
topic
for
today
is
this
hpc
and
htc
end
user
landscape
review.
We
sent
a
form
two
weeks
ago
with
a
couple
questions,
and
this
should
be
hopefully
triggering
some
of
the
discussion.
I
see
some
people
don't
have
microphones
available
so
hi
alex.
D
I
think
we're
fine
other
people
were
just
muted
or
if
they're,
okay,
just
me
too,
okay.
D
A
So
yeah,
so
the
topic
today
was
this
hpc
htc
and
user
landscape
jamie.
We
put
some
questions,
it's
kind
of
more
to
trigger
discussion.
Today
we
can
go
through
the
replies
and
maybe
stop
on
every
topic
and
discuss
a
bit
and
hopefully
like
one
thing
that
would
be.
Nice
is
to
kind
of
come
up
with
some
next
steps
of
what
is
needed
in
this
area
for
the
cloud
native
tooling
and
what
would
be
really
useful
for
people.
D
Yeah,
I
guess,
make
sure
people
put
their
names
on
the
agenda
as
well
like
normal
on
the.
Actually.
We
don't
need
that
anymore,
because
we'll
get.
A
D
F
A
A
G
A
All
right,
so
maybe
we
just
go
through
yeah
everyone,
if
you
can
add
your
names
there
and
we
can
start
by
going
through
the
through
the
questionnaire.
I
think
we
can
stop
at
each
one
and
if
anyone
has
anything
to
highlight,
we
can
discuss
in
detail.
A
So
the
first
question
was
what
kind
of
solutions
people
are
using
for
high
performance
computing
or
high
throughput
computing
and
other
batch
like
workloads.
So
in
total
we
actually
got
eight
responses,
which
is
not
too
bad.
I
would
say
because
it
was
kind
of
a
long
question,
so
the
top
two
were
slurm
and
pure
kubernetes,
which
is
a
pretty.
A
I
was
a
bit
surprised.
Actually,
then
we
had
two
for
hd
condor
and
then
we
had
one
for
armada,
none
for
volcano.
I
had
added
it
just
because
it's
like
a
native
kubernetes,
scheduler
or
cloud
native
scheduler.
There
was
one
for
code
flow
which
is
interesting
and
then
ancient
torque
system.
D
On
this
well
one
thing
that
strikes
me,
so
people
are
clearly
using
more
than
one
thing
as
well,
so
that's
not
yeah,
we've
only
you
know
got
five
five
people
in
stone
and
vanilla
they're
also
not
unique,
so
yeah.
That's
that's
kind
of
interesting.
A
Maybe
maybe
a
question
here
is
for
the
ones
that,
if
anyone
here
has
answered
with
more
than
one,
is
that
like
from
a
transition
to
something
new
or
is
it
a
plan
to
maintain
both
in
parallel.
D
We're
an
example
of
people
doing
both
for
a
transition.
H
The
kidnaps
actually
but
yeah,
so
I
was
wondering
too,
you
know,
are
we
not?
I
would
be
surprised
if
there's
really
nobody
using
volcano
and
armada
and
other
things
like
that
right
so,
like
you
know,
maybe
maybe
it's
also
kind
of
a
call
like
like.
Maybe
we
need
to
put
out
feelers
into
those
user
groups
and
try
to
get
them
more
involved
with
this
sig
right,
because
people
who
are
doing
who
are
actually
using
that
would
certainly
have
overlap
in
what
we're
doing.
H
I
think
I
wonder
if
there's
ways
we
could
reach
out
to
those
groups
or
even
coop
flow
too.
I
mean
you
know.
Those
are
I
feel
like
those
are.
I
don't
know
too
much
about
armada
and
volcano,
but
I
feel
like
those
are.
Those
are
non-zero
user
communities
right
now,
torque,
I'm
not
really
interested
in
reaching
out
to
that
community.
I'm
just
kidding,
but.
A
All
right
but
yeah,
I
think
that's
that's
a
good
point
actually
and
for
volcano.
They.
They
have
quite
a
good
structure
of
like
weekly
meetings.
I
think
they're,
mostly
targeting
or
at
least
they're
mostly
have
end
users
in
asia.
For
now
and
those
are
weekly
meetings
and
then
they
have
like
every
two
weeks.
They
have
a
meeting
that
is
kind
of
europe
and
north
america
friendly
one.
E
Other
note
on
the
volcano
thing,
depending
on
the
group,
they
will
not
be
able
to
access
google
docs
or
submit
to
google
forms
so
like
if
trying
to
like
that,
might
segment
off
a
you
know
other
community
there,
but
they.
E
Yes
right,
but
if,
if
creating
a
form
for
a
global
audience,
surveymonkey
will
work
in
in
china.
A
Okay,
that's
good!
Oh
okay!
I
see
what
yeah,
but
in
terms
of
reaching
out,
maybe
maybe
that's
a
good
point.
We
we
can
take
as
an
action
just
to
to
advertise
this
group
in
these
communities
to
see
if
they
are
interested
in
joining
it.
It's
not
like.
We
cover
every
time
this
kind
of
work
topic
either
like
there's
a
lot
of
things
that
they
will
probably
not
be
so
interested
in.
A
All
right,
the
other
thing
I
don't
know
if,
like
slurn,
has
a
pretty
strong
presence
here.
A
E
I
can
I
can
speak
for
my
old
job
yeah.
They
were
using
both
a
mix
of
slurm
and
kubernetes.
There
wasn't
any
real
transition
plan
to
go
from
one
to
the
other,
but
that's
also
was
the
university
of
michigan
and
a
good
chunk
of
their
users
were
very
familiar
with
you
know:
slurm.
They
didn't
really
want
to
like
interrupt
their
workflow.
E
H
Yeah
I
mean
or
nl.
Obviously
you
heavily
using
slurm
lsf,
I
put
them
all
kind
of
together
right
tour.
Slam
lsf.
I
guess
you
know
twerks
kind
of
on
the
outs.
I
guess
these
days,
but
you
know
pbs
all
those
right.
You
know
a
lot
of
times.
You
know,
certainly
for
us
right,
the
the
vendor
that
we
buy
the
super
computer
from
at
the
scale
that
they're
building
right.
You
know
like
they're
gonna,
it's
gonna
come
it
came
with
lsf
right.
We
bought
an
ibm
machine.
H
It
came
with
lsf,
so
you
know
those
sorts
of
things
I
don't
think
are
going
away
for
that
traditional
hpc
community
kind
of
like
what
bob's
talking
about
right
and
so
for
us,
it
was.
The
name
of
the
game
is
like
how
do
we?
How
do
we
bridge
that
gap
as
much
as
possible?
How
can
people
use
the
slurm
commands
s
batch
from
inside
of
a
container
that
sort
of
thing?
So
that's
the
direction
that
we've
gone
but
yeah?
I
don't.
I
don't
see
those
getting
supplanted.
I
mean
you
know.
H
E
I
guess
for
what
it's
worth,
I
do
see
more
people
looking
to
transition,
or
at
least
support
running
both
largely
just
because
it's
a
lot
honestly
easier,
especially
these
days
to
get
up
and
going
in
kubernetes
to
potentially
burst
out
to
some
place
and
a
lot
of
the
at
least
at
my
old
job.
A
lot
of
the
researchers
were
more
interested
in
using
things
like
keep
flow
and,
and
it
just
made
it
a
lot
easier
to
get
going.
There.
C
Yeah
and
see
what
one
sorry
go
ahead,
no
go
ahead.
Go
ahead.
I
was
going
to
say,
like
one
question
there
would
be.
Is
there
anything
that
is
prohibiting
people
from
moving
towards
using
vanilla
kubernetes
as
the
scheduler
of
choice,
or
is
it
mostly
just
familiarity
with
the
old
stuff?
So,
let's
continue
using
what
is
not
broken.
I
guess.
A
Yeah,
I
think
the
answer
there
is
there
are
things
missing
and
at
least
for
us
at
least
for
us,
the
things
that
are
missing
is
like
priority
queueing,
the
notion
of
a
queue
on
top
of
just
the
workloads
on
kubernetes,
and
then
the
notion
of
fair
share
to
optimize
the
cluster
usage.
A
A
There
was
a
very
nice
talk
at
the
last
coupon
from
I
forget
his
name,
you,
one
from
apple
apple.
A
C
Yeah
so
then,
and
then,
therefore
a
follow-up
would
be
kubernetes
allows
for
custom
schedulers.
So
I
mean,
in
fact
I
think
volcano
is
an
example
of
that
are
people
building
their
own
custom
schedulers,
because
they're
supposedly
have
not
done
myself
but
supposedly
fairly
straightforward
to
build.
So
is
it
something
that
people
consider
and
people
say
we'll
just
build
our
own
scheduler?
Is
that
an
option.
E
E
Gets
considered,
however,
that
has
largely
changed,
I
think,
as
of
the
1
21
release,
so
like
this
this
past
year,
there
is
a
significantly
more
hooks
added
to
potentially
you
know
to
make
x,
writing
or
extending
schedulers
easier.
A
I
think
okay
yeah,
it
might
be
just
it-
might
be
more
than
just
hooking
the
scheduler
as
well
like.
If
you
want
to
introduce
queues,
you
actually
need
to
to
handle
the
persistency
of
those
queues,
and
if
you
want
to
have
multiple
queues
and
a
priority,
there's
quite
a
bit
of
logic
there
that
all
these
systems
are
very
good
at,
because
they've
been
developed
for
ages
now,
a
lot
of
them.
So
so
it's
not
like
a
obvious
transition.
B
We
looked
at
that
briefly,
and
the
issue
for
us
was
that
we
wanted
to
be
able
to
schedule
across
multiple
clusters,
and
so
then
you
look
at
cube,
fed
and
that
it
didn't
all
work
to
do
multiple
classes,
which
is
why
we
ended
up
writing
armada.
B
I
mean
it
was
easy
enough
to
do
the
custom
scheduler
part
of
it,
but
not
the
multi-cluster
part.
So.
A
I
think
it's
good
points,
that's
a
good
point.
Actually,
like
the
experiments
we've
been
doing
with
managing
things
like
hd
condor
with
kubernetes,
even
if
we
are
still
submitting
like
under,
we
could
have
multiple
clusters
managing
the
condor
daemons
and
then
have
central
schedulers
somewhere
else.
So
you
could
kind
of
benefit
from
the
kubernetes
like
operations,
simplification,
but
then
still
use
condor.
D
I
always
thought
your
reason
for
not
moving
away
from
condors
less
about
the
lack
of
features
and
kubernetes
more
just
the
sort
of
inertia
of
being
able
to
change
user
behavior
and
the
fact
that
they
will
know
how
to
use
condo
and
thousands
of
them
yeah
but
fair
share.
Like.
D
B
I've
been
joining
a
bunch
of
hpc
meetups
and
it's
amazing
how
focused
that
group
of
people
is
on
hardware
like
they're
so
into
the
late
breaking
hardware,
and
how
much
we're
going
to
be
able
to
pump
over
this
pci
pipe
and
the
dram
and
this,
and
that
and
and
they're
just
fascinated
at
throwing
more
hardware
at
the
problem
as
opposed
to
what
we're
talking
about,
which
is
how
to
use
that
hardware
more
efficiently.
B
C
One
last
question:
sorry,
so
do
you
guys
anticipate
that
the
existing
kubernetes
or
the
existing
kubernetes
scheduler,
the
default
category
that
comes
with
kubernetes,
will
have
options
for
a
bunch
of
these
going
forward?
Or
is
this
always
going
to
be,
like
you
know,
default
scheduler
can
only
do
this
if
you
want
something
more
specialized,
either
build
your
own
scheduler
or
like
use
this
other
open
source,
scheduler
and
whatnot.
Is
that
where
do
you
see
that
going.
A
A
F
C
A
I
think
so,
actually
the
this
was
pretty
overwhelming
on
premises.
I
think
from
all
the
responses
we
got,
one
that
mentioned
hybrid,
so
I
think
this
is.
I
think
the
main
question
is:
is
this
staying
like
this
or
are
people
looking
at
hybrid
deployments
as
well.
H
We're
certainly
evaluating
and
exploring
hybrid
for
us.
The
bigger
issues
were
things
like
the
united
states
government
data
protection
stuff
around
fed
ramp
authorizations
and
things
like
that.
That
being
a
government
entity,
that's
that's
the
biggest
barrier,
but
we
are
starting
to
explore
that
that
hybrid
thing,
but
not
not
really
for
hpc,
it
would
be
more
for
for
workloads
that
could
that
I
don't
know
we
haven't,
we
don't
we
don't
certainly
don't
have
any
clear
workloads
that
are
like.
Oh,
this
would
be
perfect.
D
I
would
imagine
a
lot
of
this
group
have
got
a
relatively
established.
Infrastructure
have
already
have
on-prem,
so
it
would
start
there.
Probably
all
various
different
degrees
of
security
concerns
as
well,
and
you
sort
of
know
what
you
have
and
how
to
trust
it,
and
also
probably
just
large
data
sets
as
well,
which
is
probably
a
factor
which
might
keep
you
on
prem,
because
you
know
transferring
large
amounts
of
data
around
the
cloud
could
be
prohibitively
expensive
and
also
needing
that
the
equivalent
amount
of
compute
to
be
able
to
make
good
use
of
it.
E
One
of
the
reasons
the
university
of
michigan
was
looking
at.
It
was
because
a
lot
of
the
grants
were
coming
with
like
cloud
credits,
so
it'd
be
a
lot
easier
to
give
people
one
interface
that
they're
familiar
with
and
just
sort
of
abstract
it
all
away.
So
the
cloud
credits
could
go
to
you
know
it
could
be
gcp,
it
could
be
amazon,
it
could
be
wherever
but
they're
still
getting
an
interface
they're
familiar
with
and
know
how
to
work
with.
D
It'll
be
interesting
to
see
what
a
new
org
would
do
that
would
fit
into
this
group.
So
if
there
was
a
new
company
or
institution
invented
tomorrow,
where
would
they
go?
B
I
suppose
the
counter
argument
jamie
to
like
that
the
big
data
sets
we
have
on
prem
is
that
companies
that
use
cloud
data
sets
like
data
providers
who
are
cloud-based
initially,
then,
if
you're
in
the
cloud,
you
don't
have
to
move
them
as
far
to
your
on-prem
location.
So
you
know
we
might
even
be
in
that
state
for
some.
A
All
right,
I
can,
I
cannot
hear
we
I
think
the
answer
for
hybrid
is
is
ours,
so
we
are
already
deploying
some
workloads
in
this
hybrid
mode
and
the
ones
we
do
are
the
ones
that
are
from
this
embarrassing
parallel
type
of
workload.
But
we
also
have
a
couple
where
we
actually
established
links
network
links
between
our
on-premises
data
center
and
some
regions
in
different
clouds.
A
It's
much
easier
to
do
what
was
describing,
which
is
you
will
you
depend
on
the
kubernetes
api
and
you
you,
you
just
use
it
for
workloads
that
can
be
loosely
coupled
and
don't
have
like
interdependencies
that
would
require
a
low
latency
or
some
sort
of
special
network
connectivity
and
the
motivation
is
really
bursting
and
especially
for
accelerators,
which
we
don't
have
many
on
premises.
Right
now,.
C
Mind
sharing
like
how
much
do
you
guys
go
into
public
cloud
like
when
do
you
when
there
is
this
thing?
How
many
number
of
nodes
do
you
spin
up
in
public
cloud
for
these
kinds
of
when
that
happens,.
A
Well,
it
depends
like
which
unit
for
for
for
the
batch
systems,
we
can
really
tune
the
amount
of
resources
that
are
there
for
for
things
like
the
ml
workloads
using
things
like
kubeflow,
for
example,
we
actually
auto
scale
the
clusters,
so
they
will.
They
will
only
scale
up
when
workloads
go
there
and
we
try
to
define
policies
on
what
can
go
there.
A
No
it's
possible
because
we,
for,
if
you
choose
a
region
within
a
cloud
you
can
set
up
this,
this
extensions
of
the
network,
and
we
do
that.
I
can
expand
our
own
premises.
Data
center
to
this
specific
region,
but
this
is.
C
A
A
Okay,
so
then
we
move
to
a
question
which
was,
if
not
already,
do
you
plan
to
move
these
workloads
to
kubernetes,
please
expand
and
yeah.
The
couple
of
questions
we
had
was
no,
but
the
ones
with
more
details.
It
said
we
have
workloads
in
kubernetes.
That's
for
hpc
some
use
governance
to
launch
jobs
on
supercomputers.
I
guess
that
kind
of
makes
sense.
A
Then,
for
some
workloads,
I
guess
it's
the
answer,
portability
being
the
reason
and
trying
to
burst
this
is
in
line
with
what
was
describing
earlier.
I
guess
mostly
already
on
kubernetes
planning
interested
or
another.
So
I
guess
the
the
next
question
is:
what's
stopping
us,
we
already
covered
up
it.
I
don't
know
if
anyone
wants
to
add
something.
E
A
D
D
Presumably,
I
suppose
really
is
probably
what
I
expected
to
see
in
a
way
we
can't
really
tell
within
the
both
how
much
is
one
or
the
other
and
we've
got
different.
I
mean
in
our
case
anyway.
We've
got
different
groups
of
users
where
some
people
are
a
bit
more
sort
of
power
users
and
do
access
kubernetes,
directly
and
obviously
the
administrators
thereof,
but
most
of
the
our
researchers
anyway
go
through
tools
which
we
build
for
them
to
help
them
do
what
they
need
to
do
rather
than
using
kubernetes
directly.
C
C
So
cube
c
user
cube
ctl
is
direct.
Anything
outside
of
cube
ctl
is
indirect,
essential
right
yeah.
So
some
kind
of
python.
D
A
It's
important
also
for
the
all
the
role-based
access
control
that
we've
discussed
in
the
past
and
the
past
and
credential
management
of
this.
E
We
had
people
that
like
wanted
access
for
like
troubleshooting
purposes
or
just
to
diagnose
problems,
but
yeah
we
were
able
to
like,
and
then
people
wanted
direct
access
to
the
api,
and
we
got
really
good
at
having
our
back
profiles
to
allow
that
sort
of
thing
and
make
sure
people
you
know
couldn't
you
know,
couldn't
get
out
of
there
basically
name
space.
C
I'm
also
learning
a
little
bit
about
some
of
these
modern,
newer
projects,
or
at
least
modern
and
newer.
For
me,
which
is
ray,
be
a
v-a-e-x
anything
one
more.
I
think.
That's,
no,
not
that
stacks
is
different,
dusk,
yeah
and
and
some
of
those
things
I
believe,
the
way
they
expect
you
to
run
with
kubernetes.
C
Is
you
have
your
local
cube
config,
so
they,
the
the
das
scheduler,
will
run
pods
inside
kubernetes,
so
the
user
who's
using
it
doesn't
know
that
you
know
I
mean
they
know
that
there
is
kubernetes.
They
have
to
set
up
some
things,
but
they
are
not.
The
user
is
not
the
one
who
runs
the
cube,
ctl
created
or
whatnot.
So
it
is
again.
I
don't
know
how
many
people
are
using
these
modern
services.
Yet
again
I
don't
know
about
modern
sorry.
C
I
keep
saying
modern
as
if
it's
modern
for
me,
but
but
but
it's
possible
that
some
people
may
not
they,
since
they
themselves
don't
use
kubernetes,
it's
some
abstraction
layer
there
it's,
those
might
be
the
cases
yeah.
A
D
Let's
move
on
so
scale,
so
we
just
asked
compute
resources
in
terms
of
order
of
magnitude
of
cpu
cores
from
100
to
over
ten
thousand
the
bigger
well,
not
so
majority,
but
yeah.
The
the
biggest
response
is
large,
which
maybe
isn't
too
surprising,
because
that's
the
kind
of
thing
we're
all
doing.
D
I
don't
know
who's
got
less
than
100.
Cpu
calls
interesting
what
they're
up
to
they've
got
one
cluster,
I
suppose
and
they're
playing
with
it.
Well,
that
was
20.
That
was
two
out
of
the
eight
responses.
Actually.
C
A
D
A
A
A
So
the
replies
are
pretty
much
integrating
them,
although
only
one
in
one
case
or
no
like
yeah,
still
quite
relevant
with
a
thousand
or
more,
we
still
have
quite
a
bit
there.
So
one
one
question
I
had.
I
don't
know
if
people
want
to
say
other
things
about
this,
but
one
question
that
I
had
was
what
types
of
gpus
are
this:
is
it
all,
nvidia
or,
and
also
is
there
any
sort
of
virtualization
or
is
it
all
like
pci
password
like
and
dedicated
cards
for,
the
jobs.
G
Or
shout
for
help
on
the
chat,
if
you
can't
communicate,
I
just
want
to
say
that
the.
H
We
we
added
some
gpus
it
that
the
hardware
took
like
three
months
to
come
in
and
we
using
the
gpu
operator
got
the
nodes
up
and
running
allocatable
in
the
cluster.
In
like
two
days
you
know
the
gpu
operator
was
awesome,
and
I
really
can't
say
enough
about
that.
I
think
it's
really
cool.
How
that's
how
nvidia
is
able
to
kind
of
do
that
and
just
kind
of
throw
that
over
the
fence,
and
I
don't
even
know
how
much
they
support
it.
H
H
So
actually
we
just
here
gpus
in
to
start
doing
some
of
that
stuff
with,
but
we
haven't
haven't
played
with
those.
Yet
those
are
sitting
on
the
floor
getting
getting
installed,
hopefully
in
the
next
week,
so
but
yeah
I
know
we
haven't
it
was
they
were
voltas,
I
believe,
was
the
ones
we
have
today
so
but
yeah
so
so
then
you
know
we
get
like
a
jupiter
notebook
that
allocates
a
full
volta
and
they
use
it,
like
maybe
less
than
10
of
the
time.
H
A
Yeah
we
we
offer
also
the
possibility
to
do
this
virtual
gpu,
that
nvidia
already
supported
with
t4s
and
v100s,
but
it
was
kind
of
time
sharing.
Oh
okay,
we
realized
that,
in
addition
to
being
very
unstable
in
terms
of
performance,
there
were
limitations
in
doing
things
like
that
there
were.
There
were
some
bits
of
functionality
that
that
were
not
available
for
for
for
this
sort
of
driver.
It
also
needs
an
additional
license,
but
that
we
we
managed.
A
B
D
You
mean
other
other
vendors
other
than
nvidia
for
gpu.
So
it
was
a
question
I
think.
D
I
don't
think
we've
got
on
to
that.
Yeah.
D
Yeah,
no,
I
think
we're
just
nvidia
at
the
moment.
A
It's
easier
for
now,
but
yeah.
We
would
like
to
get
something.
In
addition,
there
are.
There
are
sites
because
we
collaborate
with
a
bunch
of
sites
around
the
world
and
there
are
sites
that
have
amd
cards
as
well,
so
we
started
looking
at
integrating
them,
but
they
run
properly
code
but
yeah
for
now
it's
it's
all
anything
yeah.
I
think
they've
got
pretty
mad
markets
there
or
we
get,
but
we
also
have
issues
with
the
delivery
times.
A
All
right
I'll
move
to
the
next
one,
because
we're
actually
going
fast
on
time
as
well.
So
the
next
one
is
other
types
of
accelerators
I
put
here
fpgas,
but
actually
one
another
reason
we
burst
into
the
cloud
is
to
use
things
like
tpus
as
well.
So
I
don't.
C
D
B
B
You
know
we're
looking
at
all
the
graph
cores
and
sub
novas
and
what
are
some
of
the
other
ones
takians
or
what
are
those
other
ones
ascension,
or
something
like
that?
There's
a
there's,
a
bunch
of
those
things
that
are
being
tested
and
played
around
with,
but
nothing
that's
gone
near
to
production
or
kubernetes
status.
So.
E
Honestly,
I
experimented
with
it
back
when
I
was
at
the
university
but
like
outside,
of
mounting
the
device
into
into
the
container
like
beyond
that,
not
really
it
never
got
beyond.
Essentially
me
messing
with
it.
E
Okay,
I
haven't
really
looked
at.
I
haven't
looked
at
it
really
since
then,.
A
Okay,
now,
maybe
maybe
we
take
it
as
an
action
act,
I'm
also
to
to
investigate
a
bit
where
we
are
with
this
maximum
peaking
inspiration.
A
A
For
those
that
replied
like
this
is
just
like
seen
as
a
like,
an
extra
pci
device
that
is
given
to
the
job
or
how
does
that
work.
E
Well,
oh
sorry,
as
I
know,
there
is
a
way
of
mounting
the
device
directly
in
there.
Intel
actually
has
like
an
operator
that
that
does
it
too,
if
I
recall.
E
It's
all
through
device,
plugins.
A
I
think
I
don't
know
if
anyone
wants
to
add
anything
to
what
is
already
here.
I
think
we
see
yeah
x
509
in
kerberos
so
off
and
the
main
thing
would
be:
how
are
these
credentials
being
maintained
and
like
refreshed
for
long-lived
jobs
and
things
like
this?
I
guess
everyone
has
this
sorted
out
or
any
problems
there.
D
A
There
you
go
bye,
okay,
so
maybe
we
jump
to
storage,
then
jamie.
Do
you
want
to
take
this
one.
D
Yeah
sure
so
yeah
question
around
how
we
handle
data
in
our
clusters,
what
kind
of
file
systems
people
use
or
other
quite
split
lots
of
ceph
cfs,
that
is,
people
choosing
multiple
as
well,
but
yeah
lustre,
gps
hdfs
as
well.
Quite
I
mean
there's
yeah
lots
of
different
various
responses.
I
don't
know
is
anyone
interested
to
know
if
the
hdfs
people
are
on
the
call?
Actually,
I
haven't
talked
much
about
that
previously
in
our
group.
B
Yeah
to
be
interested
group,
whether
any
hdfs
users
are
looking
at
ozone
apache
ozone
is
a
replacement
for
hdfs.
If
any
hdfs
users
are
on
the
call
be
interested.
A
All
right,
I
just
saw
here
also
in
the
chat
nathan.
I
don't
know
if
you
cannot
turn
the
microphone,
because
I
just
saw
a
couple
of
comments
that
you
had.
That
would
be
quite
interesting,
which
would
be
how
many
sites,
how
many
of
these
sites
are
using
containers
in
sloan.
H
Sort
of
I
don't
know
I
mean
so,
did
you
see
that
apptaner
is
the
new
singularity
they
just?
They
just
announced
that
the
other
day
the
I
feel
like
I
feel,
like
you,
know,
hpc
containers
people
want
more
than
than
what
they
think
they
want
kind
of
thing
right.
I
feel
like
the
name
of
the
game.
You
know
we
were
working
for
a
while
on
trying
to
replicate
singularity,
contain
or
hpc
containers
with
podman
and
really
the
amount
of
holes
that
you
kind
of
poke
in
the
container.
H
It
turns
into
more
of
a
sieve
than
it
does
like
a
container
right,
because
you
really,
you
really
want
to
bind
mount
up
all
all
of
your,
your
your
blast,
libraries.
You
know
the
gpu
line.
You
know
you
want
to
pull
all
that
stuff
in
off
the
host
right.
You
know
it
kind
of
kind
of
necessarily
breaks
that
isolation.
H
You
know
I
mean
I
still
think,
there's
really
good
stuff
about
it
and
and
even
like,
like
nurse
nurse,
showed
that
with
what
are
they,
it's
not
singularity,
they've
got
their
they've
got
another
one
based
on
docker,
but
that
python
applications
actually
perform
faster
across
a
cluster
in
a
container
than
outside
of
a
container
right.
It
has
to
do
with
the
python
looks
up
paths
for
linking
for
dynamic,
libraries
and
stuff.
H
It
doesn't
have
as
many
paths
to
look
up
in
a
container,
because
the
way
that
the
way
that
you
link
in
a
container,
I
guess,
is
compared
to
like
a
normal
hpc
host.
So
it's
kind
of
funny,
but
but
I
don't
know
I
mean
I
don't
know
we
get.
We
get
tons
of
requests
for
people
to
to
to
support
hpc
containers,
but
and
people
do
use
them,
but
I
don't
know
I
feel
like
I
feel
like
we
always
have
to
have
this
like
hard
conversation
of
like
okay.
H
I
Well,
that's
there's
been
a
lot
of
research
like
a
good
number
of
sites
on
how
to
get
the
performance
out
of
it,
like
the
common
trick
now
is
to
bind
mount
the
mpi
layer
in
so
that
you
use
you
know,
especially
on
the
craze
and
then,
of
course,
breaks
a
whole
bunch
of
other
stuff.
I
None
of
these
limits
are
new.
Actually,
let
me
go
look
up
the
paper
or
the
presentation
I
have
on
it
these
these
lip
these.
These
issues
have
been
around
a
long
time.
Yep.
I
There
was
a
group
when
you
want
to
use.
When
you
want
to
go
fast,
you
will
lose
compatibility
yeah.
I
mean
necessarily.
I
H
There's
a
good
good
quote
from
another
guy
in
a
different
lab
who
said
that
that
hpc
containers
is
teaching
a
whole
new
generation,
the
of
of
linking
for
errors
right
library
linking
errors
right.
You
know-
and
it's
it's
so
true,
because
you're
right,
that's
what
you're
doing
you're
mounting
it
off.
If
you
want
to
get
the
performance
so
yeah
here.
I
I
put
the
the
link
in
there,
I
mean
we
did
this
back
in
2017
and
these
aren't
solvable
by
containers
or
anything
else.
I
mean
and
the
whole
world
of
issues
come
in
when
you
want
to
swap
architectures
or
compile
against
ssc,
4
versus
scc3
or
whatever.
I
There's
a
lot
of
problems
with
that,
especially
with
the
move
to
the
the
single
floats,
the
gpus.
I
mean
you
get
the
fancier
nvidia
ones
with
the
double
floats,
it's
not
as
much
of
an
issue,
but
it
still
matters,
and
then
the
lack
of
the
ieee
float
standard
being
consistently
implemented
completely
makes
it
entertaining
yeah.
I
posted
the
link
of
a
lot
of
the
limits
that
you
know
been
around
for
a
while.
I
I
I
mean
right
now,
there's
a
lot
of
glue
work
that
goes
in
for
like
getting
jupiter
books
to
work
on
hbc
or
some
of
them
run
them
on
kubernetes
and
then
burst
out
to
hbc
and
stuff
like
that
really
nice
to
know
about,
you
know
what
the
sites
really
need.
What
they're
doing
I
mean
I
understand
the
use
case
of
you
know
you
want
to
use
coop
flow,
you
use
argo
or
something
like
that
or
hell
you
don't
care
how
it
runs.
You
just
wanted
to
run.
I
But
yeah
you
hit
a
lot
of
the
complications.
I
mean
in
a
lot
of
cases.
You're
gonna
have
to
recompile
absolutely
everything
to
get
it
to
the
full
performance.
You
know
when
you're
jumping
from
your
laptop,
which
may
be
like
an
armed
chromebook
to
you,
know
a
zeon
box
or
something
like
that,
or
even
a
power,
eight
or
power
power,
one
power
where
we
power
ten.
Now
we
should
probably.
A
Yep,
let's
browse
with
the
rest,
and
then
we
can
come
back.
I
think
we
can.
D
A
Till
we
started
late
as
well
thanks
a
lot
yeah.
Should
we
browse
real
quickly,
then
monitoring
monitoring,
we
see
prometheus
okay,
fluently.
A
A
All
right,
so
this
is
coming
a
bit
to
what
nathan
was
just
referring,
which
is
how
our
container
image
is
built.
I
think
this
is
a
one
of
the
replies
I
had
for
him,
which
is
in
most
cases
we
don't
have
people
building
locally,
they
just
push
somewhere
and
there's
some
sort
of
ci
cd
that
will
build
for
multiple
architectures,
so
those
systems
here
we
get
gitlab,
jenkins,
tecton
and
then
manually.
A
But
then
gitlab
tech
manual
again
very
likely
manual,
okay,
there's
quite
a
lot
of
manual
jenkins.
A
A
I
A
To
to
a
branch,
and
then
the
the
runner
will
will
get
the
web
hook
and
will
just
pull
clone
the
code
and
build
locally
on
where
the
hardware
the
run
is
running
and
we
basically
replicate
them
on
all
the
architectures.
A
All
right,
then,
we
go
through
registries,
so
we
have
all
it's
the
answer.
D
Or
any
any
hard
issue
to
raise,
we
use
artifactory
we've
run
into
some
scaling
problems
with
it,
but
we've
recently
started
looking
at
dragonfly,
it's
like
a
sort
of
caching
yeah
and
it's
very
early
days,
but
it
looks
pretty
good.
Actually,
we
started
originally
looking
at
something
called
kraken,
which
I
think
was
out
of
uber,
but
it
seems
to
have
died
in
a
ditch.
So
then
we
sort
of
moved
sideways
onto
dragonfly
and
it
looks
pretty
good
and
that's
sort
of
taking
some
of
the
pain
away
from
mars
factory.
A
Anyone
else
all
right,
let's
go
through.
I
think
we
only
have
two
more
so
languages.
It's
pretty
much
like
half
is
bison
and
then
the
other
half
split
some
four
turn.
So
that's
pretty
good.
A
A
Yeah
at
home
sounds
pretty
reasonable.
I
think
I
think
that's
it
I
don't
know
do
we
want
to
highlight
anything
particular
already
three
minutes
over.
A
A
So
maybe
maybe
we
take
those
as
as
topics
for
you
for
our
next
session
yeah,
otherwise
yeah.
Thank
you
very
much
everyone
and
we
meet
in
two
weeks
for
jamie's
jen's
session.