►
From YouTube: Centaurus TSC Meeting 1/26/2021
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
Information,
okay,
hello,
everyone-
this
is
centaurus
tsc
meeting
and
welcome
everyone,
and
this
is
the
meeting
logistic
as
mentioned.
I
will
be
posting
the
meeting
minutes
and
meeting
agenda
in
this
link.
I
will
also
be
sending
this
information
again
in
the
calendar
after
this
meeting
is
done,
and
this
is
the
agenda
for
today,
doctor
shone
will
be
leading
the
meeting
and
before
we
start
do
we
want
to
go
around
and
introduce
each
other.
B
Yeah,
maybe
I
can
start
with
me,
okay,
introduction
hold
on
everybody.
See
me
yes,
yes,
okay,
great
well,
hello.
Everyone
good
evening
or
good
morning
depends
on
where
you
are
again.
My
name
is
leading
the
cloud
lab
here
at
virtual
way.
I'm
being
visual
aware
with
I
don't
know
six
years
from
now.
I
feel
like
a
very
short
time,
but
I've
been
six
years
or
now,
but
since
last
march
we
basically
work
from
home
almost
a
year
now,
so
we
start
with
almost
a
year
now.
B
I
think
everybody
is
like
that
and
before
I
work
before
I
come
to
the
official
way
I
work
at
microsoft
for
around
seven
eight
years
and
then
before
that
is
at
t
for
about
10
years.
I
guess
so
that's
based
on
my
experience.
My
my
my
interest,
of
course,
is
a
cloud
related
to
technology.
B
So
personally
I
I
like
sports,
so
I
especially
the
basketball
I
actually
played
basketball.
I
joined.
I'm
actually
was
a
college
team,
it's
not
a
good
team,
but
I
was
in
the
college
team
play
so
I
like
to
watch
mba
a
lot
and
los
angeles
is
still
my
favorite
teams
now,
I
think
other
than
that.
Well,
I'm
just
talking
need
to
be
the
cloud
lab.
We
do
we're
basically
doing
open
source
project
and
open
collaboration,
so
we
hopefully
this
year.
B
We
continue
to
collaborate
with
some
of
you.
You
you
guys
and
we
continue
to
do
the
open
source
project.
Centaurus
is
a
key
project
started
with
a
virtual,
but
now
it's
part
of
nina's
foundation.
I
really
appreciate
the
executive
for
all
your
support
for
this
project,
I'm
hoping
that
the
community
will
grow
the
communities
strong
that
I
can
continue
to
support
this
con
centaurus
project.
B
Hopefully
we
have
customers
started
using
this
and
doesn't
matter
who
we
are,
and
so
thank
you
for
supporting
this
ecosystem
and
we
have
other
projects
from
like
a
storage
project.
We
have,
I
think
we
have
ai
research
sum
project
and
then
we're
also
looking
for
some
of
our
continuous
collaboration
with
this
cr2
on
the
80
side
on
the
ai
side.
As
soon
as
she
only
can
talk
about
that
a
little
bit
during
the
centaurus
project
update
okay,
I
know
I
assume
I
know
almost
everyone's
here
so
welcome.
C
I
can
go
next
all
right,
hey
hello,
everybody
nice
to
meet
you
guys,
I'm
really
excited
about
this
centaurus
project,
and
so
I've
been
in
open
source
for
quite
a
while.
I
was
on
the
board
of
openstack
foundation
for
about
four
years,
and
then,
after
that
I
was
on
the
board
of
cncf
cncf,
which
you
know
cloud
native
foundation
under
linux
foundation.
C
Right
now,
I'm
a
board
alternate,
I'm
still
on
the
board,
and
I
think
it's
really
good
time
for
me
to
transition
to
centaurus,
because
I'm
very
excited
about
the
centaurus
project,
because
I
think
that
the
cloud
infrastructure
is
over
10
years
old.
The
the
open
source
version
of
cloud
infrastructure
is
over
10
years
old.
It
is
time
that
we
need
to
introduce
new
technologies
to
in
to
this
area,
and
I've
talked
to
a
lot
of
users,
especially
the
users
who
are
currently
using.
You
know
other
solutions
right
now.
They
are
all.
C
They
all
think
that
they
need
more
scalability.
They
need
more
performance,
they
need
more
interoperability
to
meet
the
needs
of
5g.
You
know
ai
era,
so
I'm
very
excited
to
be
part
of
this
project.
My
day
job
is,
I
work
for
a
future
way,
I'm
responsible
for
future
ways,
open
source
operations,
ensuring
that
you
know
all
of
our
open
source
projects
are
operated
according
to
compliance
and
also
with
open
source
best
practice,
I'm
very
happy
to
meet
everybody
here.
Thank
you.
D
Okay,
I
I
can
probably
I
can
do
next
hi
everyone.
My
name
is
juanin,
I'm
based
in
seattle,
together
with
delta
jung.
I
also
from
the
future
cloud
lab
in
cloud
lab.
We
have
several
teams,
we
have
computing
networking,
storage,
ai,
I'm
leading
the
computing
and
networking
team
they're,
basically
working
on
cloud
infrastructure.
D
D
E
Well,
I
can
briefly
go
like
so
yeah.
This
is
a
deeper
crazy
part
of
the
cloud
lab.
You
know
I've
worked
with
shawnee,
you
know
young
and
annie
and
everybody
here,
so
I've
been
here
a
future
way
for
a
while
now,
actually
for
seven,
eight
years
now
so
yeah,
so
I've
been
involved
with
the
cloud
computing
for
a
long
time
actually
so
way
back
in
2010.
E
When
you
know
I
was
part
of
that,
I
was
I
triple
e
chair
actually
for
the
one
of
the
cloud
initiative
and
prior
to
that,
you
know,
I
worked
on
a
lot
of
enterprise
software
companies.
You
know
oracle,
cbo
and
bunch
of
startups.
You
know
and
pretty
much
in
silicon
valley,
all
this
time
and
yeah,
so
we
we're
really
excited
about
this
project.
Actually,
so
I
think
we
have
a
great
opportunity
and
there's
the
fact
that
you
know
the
whole
landscape
application
landscape
is
changing
in
front
of
our
eyes.
E
E
I
was
receivable
for
a
long
time.
Actually,
so
that's
one
of
the
things
yeah.
I
wanted
to
mention
that
even
tom
siebel
just
started
a
startup.
Actually
a
company
called
c3
and
they
are
positioning
their
crm.
They
call
it
ai
crm
actually
and
their
valuation
is
skyrocketed.
It's
like
27
billion
dollars,
valuation,
some
ridiculous
amount.
So
what
I'm
saying
is
the.
E
Bring
this
up
is
because
everything
is
all
ai
now,
basically,
so
I
think
so
the
work,
though
that's
the
one,
that's
one
of
the
challenge
the
cloud
infrastructure
has
and
which
we,
I
think
shawnee
is
gonna,
talk
about
that.
Actually.
B
Yeah
back
and
forth
in
the
background
and
180
I
was,
there
will
implement
the
stable
for
the
18t
crm
and
journey.
That's
the
early
stage
of
I
think,
but
quite
a
journey
there,
yeah
yeah
yeah
yeah
yep.
A
C
G
Hi
everybody,
my
name
is
shahram
dustar,
I'm
a
professor
of
computer
science
at
technical
university
in
vienna,
in
austria
in
europe
and
well,
I'm
working
in
distributed
systems
pretty
much
in
cloud
computing
service,
oriented
systems
and,
more
recently,
meaning
like
five
six
or
something
years
in
the
area
of
edge
computing
internet
of
things,
and
also
kind
of
obsessed
with
the
principle
of
elasticity.
This
is
also
how
I
I
got
to
work
with
ing
and
deepak
and
xiaoning.
G
We
have
a
project
with
future
way,
which
is
super
exciting
and
now,
more
recently,
I'm
looking
into
bringing
this
principle
of
building
service
level
objectives
and
agreements
to
the
world
of
elasticity.
G
F
I
can
go
next
now
just
to
con
to
reuse
the
context
of
professor
dostar,
because
I'm
also
from
vienna
the
same
europe
and
also
working
on
the
same
project
with
with
the
future
way
for
how
long
now
a
year-
and
it's
been
really
a
lot
of
fun,
and
now
it's
it's
really
kind
of
enjoyable-
that
to
see
that
this
is
growing
to
a
whole
different
level
and
you
know
centaurus
becoming
part
of
linux
foundation.
F
It's
really
a
privilege
for
me
to
to
be
able
to
to
be
on
the
on
the
committee
and
to
be
able
to
you
know,
cooperate
and
collaborate
with
you
guys.
So
briefly,
my
background
been
dealing
with
cloud
computing
for
the
past.
F
I
think
something
like
eight
ten
years
did
my
did
a
phd
in
that
area,
also
overlapping
with
the
edge
computing
and
iot
so
basically
try
to
combine
cloud
computing
and
internet
of
things
and
edge
computing,
and
lately
I've
started
my
own
company,
so
I
co-founded
a
company
with
actually
a
colleague
of
mine
from
from
from
vienna,
where
I
used
to
work
before
and
yeah.
That's,
I
think,
a
little
bit
about
myself
and
again
really
super
excited
to
be
part
of
this
team
and
looking
forward
to
to
to
the
future
here.
H
Mute
all
right,
hi,
I'm
prashant.
I
hope
everybody
can
see
me.
I
don't
know
how
my
background
is
right
now,
so
today
is
a
republican
day
in
india
and
just
to
give
my
background,
is
I'm
a
founder
of
founder
and
ceo
of
click
to
cloud
inc,
and
I
have
two
successful
exit
before
this.
One
was
a
product
that
everybody
knows
hp
highly
and
that
we
built
with
eric
ruder
and
few
other
people
in
back
in
the
earlier
days,
and
before
that
I
had
a
company
called
argent
solutions
which
got
acquired
by
persistent
system.
H
So
this
is
my
third
venture
and
we,
when
we
met
with
dr
xiao
and
annie
about
this
project,
I
got
super
excited
because
I
believe
this
technology
is
forward
looking
very
exciting
in
terms
of
the
ability
to
do
like
cloud
native
computing
and
5g
cloud
kind
of
concept.
H
So
I
feel,
like
click
to
cloud,
is
already
contributing
towards
it
from
the
beginning
and
the
team
is
very
engaged
and
we
see
a
lot
of
future
in
coming
days
to
do
it
and
to
add
the
last
sentence.
I
also
been
worked
in
microsoft,
almost
for
eight
to
ten
years
from
microsoft,
licensing
to
different
windows,
genuine
advantage
team,
where
I
handled
many
projects
for
microsoft.
I
Great
thanks
hi
this
is
sunil.
You
are
able
to
hear
me
and
hope
you
can
see
me
yeah
yeah,
yeah,
I'm
also
from
india
joining
from
bangalore.
I'm
part
of
soda
foundation,
part
of
the
toc
and
leading
the
architectural
work
group
and
heading
the
engineering
part
of
the
soda
foundation
projects.
As
of
now
last
two
years,
and
it's
very
happy
to
see
anthony
most
of
you.
I
I
think
dr
ingen
prashanth
deepak
to
some
extent
and
eli
also,
and
I
think
we
work
together
one
or
the
other
way
we
connected
in
the
past.
I
think
it's
really
nice
to
meet
you
again
and
catch
up
so
before
the
soda
foundation.
Soda
foundation
is
basically
working
on
an
open
data
framework
from
linux
foundation,
something
similar
to
centaurus,
and
probably
I
will
be
the
knowest
in
this
group
about
this
project.
I
So
before
that
I
have
been
working
with
open
source.
Quite
some
time
be
part
of
the
advisory
board
of
your
project
when
it
was
incubated,
and
then
we
incubated
qbh
in
edge
computing
in
cncf,
and
now
it
is
in
incubating
state
and
also
worked
on
distributed
computing
projects
like
amino
and
cloud
projects.
That's
where
some
of
you
work
with
us,
so
it's
very
nice
and
exciting
to
be
part
of
this
project
and
always
open
source
and
technology
is
close
to
our
heart
yeah.
J
All
right
just
to
know
exactly
how
it
goes
so
good
morning,
everybody,
my
name
is
nikita.
I'm
enough!
I'm
a
cto
and
founder
of
company
called
great
gain
systems.
The
companies
behind
apache
ignite
one
of
the
top
projects
in
apache
software
foundation.
I'm
also
involved
with
a
few
other
projects,
an
apache
software
foundation.
J
So
hopefully
I
can
bring
some
of
the
some
of
the
experience
of
many
many
years
in
in
yet
another
open
source.
Community
apache
is
a
quite
large
community,
as
you
probably
most
of
you
know.
This
is
the
home
for
most
data
processing
open
source
software,
as
we
know
it
today.
J
Actually
thinking
about
deepak-
and
I
we've
met-
you
know
when
huawei
actually
wasn't
invested
in
my
company
in
brigade
systems
and
deepak
was
on
the
board
and
that's
the
kind
we
met
many
years
ago.
So
I'm
really
excited
about
this
project.
J
I
think
it's
long
overdue
of
having
some
deluxe
and
taurus
going
to
be
an
open
source
arena
and
really
challenging
the
current
status
quo.
With
you
know,
with
the
kubernetes
and
whatnot
technology
has
been
developed,
you
know
decades
ago,
quite
literally
and
kind
of
overcome,
all
these
limitations
and
and
whatnot
so
really
really
excited,
and
we
already,
you
know,
had
some
experience
using
apache
ignite
in
placement
hcd
and
in
the
kubernetes
and
with
some
decent
results.
J
So,
overall
again,
just
like
everybody
else,
I'm
excited
to
be
in
this
project
excited
to
see
what's
going
to
happen
in
the
next,
you
know
12
18
months,
so
let's
go
and
we'll
see
located
here
in
silicon
valley,
so
welcome
everybody.
A
Okay,
hi
everyone,
my
name
is
moani
and
I'm
the
pm
for
collab,
and
I
work
with
doctor
zhong
and
xiaoning
deepak
and
any
and
everyone
here
so
I'll,
be
helping
with
the
logistic
of
the
tsc
meeting
going
forward
so
feel
free
to
send
me
any
questions
and
I'm
very
excited
to
be
a
part
of
this
project
and
like
everyone
else,
and
welcome
everyone.
A
B
Yeah
all
right,
I
think
we're
gonna
go
ahead
and
start
it
again.
Welcome
everyone
looking
forward
to
working
with
all
of
you
closely,
hopefully
making
the
centaurus
realize
the
vision
of
the
centaurus
in
the
in
the
in
their
growing
their
ecosystems.
So
here
is
the
agenda
for
today.
I
know
we
have
only
one
hour
so
time
and
need
to
be
tight,
so
we're
going
to
go
through
some
of
the
project.
B
Overviews
basically
discuss
the
logistics
here
and
they
are
governing
how
we
are
governing
the
project
itself
and
then-
and
the
mission
today
is
confirm
the
the
members
nominations
and
the
chair
illumination
for
the
tac,
the
technical
student
committee
today.
This
meeting
is
the
first
gsa
meetings
and
then
we'll
talk
about
xiaomi
will
give
some
centaurus
project
updates,
and
currently
we
have
two
projects
on
this
entire
umbrella
project,
which
is
octo's
compute
and
the
misa,
which
is
the
networking
project.
B
But
I'm
hoping
that
we
are
hoping
that
we
have
more
projects
during
the
centaurus
like
storage
like
age,
maybe
ai.
So
we
can
talk
about
that
and
then
maybe,
if
we
have
time,
we
can
discuss
a
little
bit,
how
the
centaurus,
how
we
do
our
outreach
in
the
marketing
and
then
any
ideas
so
we
can
collect.
I
don't
think
we
may
not
have
time
more
in
more
detail
on
this,
but
we
can
give
some
of
the
volunteers
that
plan
the
goal
that
will
be.
That
would
be
great.
B
So,
let's
move
on
next
page,
so
centaurus
project.
I
can
just
learn
quickly,
as
you
know,
in
the
red
before
I
think
yes
right
before
the
christmas
last
year,
they
have
become
a
official
linux
foundation
project
and
we
we
have,
they
have
a
blob.
Everyone
has
know
that
so
it's
officially
a
some
menus
foundation
project.
So
the
meeting
we
have
here,
we
were
it's
a
public
meeting.
B
We
will
record
it,
we
have
many
minutes
and
then
we're
published
and
then,
of
course,
linux
foundations
have
access
to
it
and
we
do
have
a
snack
channel
created
for
the
centaurus
project
and
this
this
is
a.
This
is
a
stack.
This
is
a
snack
channel
that
created
me
specifically
for
centaurus
on
non-technical
communication.
B
Like
meetings
logistics,
any
information
sharing
and
our
notification
that
can
use
just
a
snack
channel
technically,
we
have
everybody,
know
there.
We
have
a
snack
channel
setback
for
the
actors
that
will
have
a
snack
channel
for
the
visa,
so
in
technique.
I
can
go
still
using
that
snack
channel
governing
board
discussions.
If
I
understand
it
correct,
because
this
is
a
linux
news
foundation
project,
but
not
a
board,
not
a
foundation.
B
So
we
we
don't
have
a
board
for
this
project,
but
we
still
need
a
group
that
to
discuss
non-technical
issues
like
strategy
directing
users,
group
and
outreach
and
marketing
and
all
logistics,
there's
all
non-technical
issues.
So
we
are
actually
maybe
stop
here
to
see
anybody.
That
has
a
suggestion
for
that
any
volunteer.
I
want
to
be
a
a
leading.
This
group
will
call,
I
don't
know
what
we
call
a
governing
group
whatever.
It
is
because
we
do
have
a
go
forward.
We
do
have
a
non-technical
discussion
going
on
here.
C
I
think
what
we
can
do
is
we
definitely
we're
going
to
need
an
outreach
committee
to
help
us
build
up
the
ecosystem
and
we
need
a
marketing
committee
to
talk
about
marketing
issues.
But
as
far
as
other
stuff
strategy,
we
can
just
have
project
things.
We
can't
use
the
word
board
because
we
don't
have
a
board
and
if
we
want
to
have
a
board,
then
we
there's
some
legal.
You
know
documents
we
have
to
sign
under
linux
foundation's
guidance.
C
So
at
this
point
we
can
just
have
you
know,
project,
discussion
or
overall
project
discussion,
something
like
that
and
but
but
the
committees
that
will
really
need
to
do.
C
The
work
is
the
outreach
committee
and
marketing
committee
and-
and
I
think
outreach
committee
will
cover
the
users
right
and
marketing
committee
will
cover
the
marketing
issues
and
overall
strategy
and
direction
and
technical
strategies
should
come
from
a
tsc,
this
forum
and
that's
a
technical
strategy
and
direction
so
that
I
think
we
can
cover
by
these
three
committees
right
now
and
unless
you
want
to
have
another
overall,
you
know
we
can
have
an
overall
project
group
for
anybody
who's
interested
to
find
out
about.
C
B
Okay,
any
other
feedback
add
suggestions.
I
know
that
any.
I
know
that
we
cannot
use
on
board,
but
we
still
need
a
group
or
some
some
form
of
communication
that
yeah.
F
B
Discuss
some
of
our
rating
and
marketing
group
origin
group
user
group-
maybe
some
of
the
non-technical
discussions.
C
We
can
just
call
it
the
project
discussion
yeah,
something
like
that.
Overall
project
discussion,
yeah.
H
B
Yeah
I
actually
put
in
there
with
all
much
thinking
there
thanks
for
that
water,
hey.
What
here
means
non-technical
strategy,
nutanix
directions-
I
don't
know,
there's
any
non-technical
direction,
but
strategy
for
for
how
to
grow
the
community
strategy,
how
to
get
marketing
there,
the
other
things
but
all
non-technical
stuff.
B
D
E
Accurate
yeah,
I
mean
so
I
think
that
for
prashant,
then
nikita
and
soda
foundation
for
factory
so
feel
free
to
have
your
you
know.
Customers
or
your
ecosystem
get
involved
in
that
actually
that'd
be
really
good,
because
once
the
end
user
and
customer
and-
and
you
know
the
companies
get
involved,
a
lot
of
use
cases
will
come
up
and
that's
how
the
this
you
know
this
project
believe
all
that.
So
you
can,
you
know,
feel
free
to
involve
your
customers
or
your
user
base
ecosystem.
H
H
C
Yeah
so
doctor
shows
I
was
just
thinking
you
know
so
for
overall
project
governing
strategy
meeting
or
group
we
can
do
that
and
then
and
then
the
people
who
are
you
know,
participating
in
the
outreach
committee
and
marketing
committee
can
participate
in
that
group
yeah.
The
thing
is,
you
know,
want
to
avoid
having
too
many
meetings,
but
at
the
same
time
we
need
to
make
sure
we
cover
our
bases.
C
So
I
think
yeah,
it's
good
to
have
overall
strategy
group
and
then,
under
that
group
we
have
two
committees
right:
one's
user
committee.
One
one
is
marketing
committee,
one's
outreach
committee
and
in
the
future,
if
we
want,
we
can
even
have
a
user
committee.
If
you
want
to
focus
on
users,
like
user
cases,
use
cases,
user
requirements
yeah,
that's
also
possible.
B
Okay,
okay!
Well,
actually,
what
I
mean
here
is
we're
probably
not
able
to
discuss
all
this.
What
the
this
group
is,
what
the
system
will
focus
on,
but
hey,
let's
just
mention
that
and
we
want.
We
need
a
way
to
have
a
communication
for
governor
not
for
governor
forgot
the
project
governor.
I
think
again,
it's
not
for
non-technical
stuff,
so
our
and
then
this
group
can
go
off.
You
know,
can
go
off
setting
the
strategy
and
the
groups
how
many
groups
within
that
project.
B
So
I
think
the
group
can
do
that.
So
I
think
here
we're
just
acting
looking
for
volunteers
to
see
who
want
to
be
the
group
discussions
who
want
to
contribute.
I
know
a
lot
of
you
guys,
as
companies
has
has
a
business
groups.
Can
you
involve
that?
I
think
christian
just
mentioned
there
need
to
be
that.
So
I
will
recommend
this
if
I,
if
everyone
agree
that
from
virtual
wayside,
annie
will
be.
B
Maybe
can
I
wanted
here
to
lead
to
that
this
project
governing
project,
whatever
the
governing
project
governing
group
or
for
the
first
heroes,
and
then
we
can
and
has
everybody
you
can
just
get
together
with
everybody
they're
interested
in
this
and
discuss
and
and
see
how
we're
going
forward
with
that
is
a
plan.
You
know
we
don't.
B
I
don't
think
we'll
have
time
to
discuss
what
exactly
how
what's
the
group
it
is,
but
let's
have
people
and
they
go
off
to
discuss
this
and
then,
if
you're,
okay,
annie
and
I
will
recommend
the
group
yeah.
I
I
think
that's
a
good
good
start.
I
think
any
also
mentioned
about
this
governing
part.
I
think,
as
you
mentioned,
maybe
this
this
is
independent
project,
so
governing
board
is
not
there
that
a
project
group
can
be
there
and-
and
he
can
then
come
up
with
just
a
governance
structure
where
user
group
marketing
and
tse
and
then
the
common
meeting
can
be
representatives
from
each
of
this
group
so
that
we
can
go
ahead
with
the
strategies
and
the
non-technical
stuff.
I
B
From
what
I
heard,
we
have
at
least
has
two
people,
not
here
persian.
One
has
from
kidney
cloud
project
just
mentioned
that
rapport.
Can.
She
can
also
help
on
this,
so
rep
and
annie
will
work
together
to
form
this
kind
of
group
and
then
start
to
form
some
of
the
subgroup,
exec
and
and
the
plan
for
that.
C
Yeah
so
the
way
I
I
like
to
do,
this
is
because
right
now
the
hierarchy
is
tsc
is
because
we
don't
have
a
board.
I
mean
we
can
explore
having
a
board
later,
but
having
a
boy
is
more
complicated
because
then
there's
also
budget
and
all
that
kind
of
things
that's
involved
right.
So
I
think.
C
B
C
There
yet
so,
since
the
project
is
very
small,
that's
why
we
don't
have
a
board.
I
just
want
to
explain
to
the
people
of
the
group
here
so,
but
we
do
still
need
to
have
you
know
a
group
to
discuss
non-technical
strategy
direction
and
then
outreach
and
marketing
that
kind
of
stuff
non-technical
stuff.
So
it's
still
important
to
have
this
group,
so
I
would
still
like
to
urge
all
the
member
companies
senate
representative,
because
you
know
like,
for
example,
it
would
be
great
to
do
some
events
in
europe.
C
Right,
I
mean-
and
you
know,
kukan
europe.
We
can
do
something
in
kukan
europe
and
then
centaurus
can
be
a
sponsor,
and
then
people
in
europe
can
give
us
feedback
can
give
this.
You
know
group
people
feedback
or
we
can
do
something.
Asia
pack
in
india,
so
it'd
be
good
to
have
people
who
are
in
india
to
give
us
feedback.
C
So,
even
though
it's
not
a
technical
forum,
but
I
would
still
like
to
encourage
you
know
this
member
company
to
send
representative
and
to
participate
and
we'll
try
to
make
this
forum
as
infrequent
as
possible.
I
know
we
all
have
a
lot
of
meetings
and
I'll
also
make
sure
that
the
meeting
minutes
go
out
to
everybody,
so
you
know
at
least
even
if
you
don't
have
time
to
participate.
You
know
you
know.
C
What's
coming
up
what
kind
of
marketing
events
we're
doing,
what
kind
of
outreach
you
know,
initiatives
we're
doing
and
and
still
we
we
still
need
your
every
member
company's
support
to
to
grow
our
ecosystem.
B
Okay,
great,
it
sounds
like
you
have
a
lot
of
ideas
which
is
good,
so
so
you
can
go
offline
with
your
contact
all
the
member
companies
here
and
with
rapport
together
we
can.
We
can
come
up
with
a
plan
for
for
at
least
for
this
year,
for
the
centaurus
project.
Yeah
that'd
be
good.
B
Okay,
if
everybody,
okay
with
that,
no,
we
will
move
on
to
the
next
page,
so
next
widget
next.
This
is
our
main
sense.
Today,
this
meeting,
since
it's
a
tsa
meeting,
I'm
win
the
project
of
launch
officially
from
english
foundation,
I'm
the
acting
chair
of
tsc.
B
So
I
play
my
role
here
today,
but
going
forward
the
this
is
a
tsa
road
responsibility.
This
is
very
actually
in
fact,
standard.
One
for
many
of
the
open
source
project
is
some
projects
called
toc
technique,
oversight,
committees,
some
good
tac
steel
committees,
and
so
I'm
not
going
to
go
through
this.
Everybody
probably
already
knows
understand
this.
B
B
So
we
have
this-
is
our
tsc
members
nomination
nominees.
We
have
a
two
from
virtually.
I
want
the
one
we're
thinking
about
getting
a
huss
number,
so
we
come
with
a
seven
two
from
virtually,
which
is
t
pack
and
charlene,
and
then
one
from
green
gain
and
nikita
here,
one
from
click.
B
The
cloud
person
is
here
and
then
well
from
soda
soda
are
foundations
for
sun
near
is
here
and
then
one
from
teo
wayne
and
professor
drew
starts
here
and
then
for
the
event
as
a
star
company
here
in
in
europe.
Stefan
is
here
so
that's
our
current
nominees
and
for
the
chair.
B
I
sent
out
the
email
last
week
to
see
with
any
other
nominee
for
the
tsh
here
right
now
we
have
one
xiaomi
sean
has
already
have
talked
to
you
individually,
and
he
is
the
called
the
creator
of
october's
project
misa
project.
So
he's
he's
a
creator
that
you
know
from
technical
perspective,
he's
the
founder
of
this
two
projects.
So
that's
why
we
nominate
him
to
beat
the
chair
and
and
again
money
will
be
the
ttsc
secretary.
B
One
nominee
will
actually
work
on
other
nominees
if
he,
if
you
want
to
so
this
is
the
tsa
members
nominees
included,
chairs
technical
members
and
the
secretary.
So
we
come
down
to
work
any
any
suggestions.
H
H
D
C
So
the
further
sorry,
I'm
more
of
a
you,
know
compliance
person.
So
I
want
to
make
sure
you
know
we're
doing
this
official
way.
So
it
looks
like
right.
Three,
four
five
we'll
have
seven
tsc
representatives
right
and
then
now
we
need
to
elect
the
tsc
chair.
So
only
these
seven
people
have
voting
right
to
elect
the
tsc
chair.
I
just
want
to
make
that
clear.
C
So
so
we
need
to
ask
this
seven
people,
anybody
who
obj,
since
we
only
have
one
nominee,
and
so
we
there's
no
point
to
really
vote,
but
we
would
like
to
ask
if
there's
any
objection.
B
As
soon
we
are
confirmed,
all
the
members
confirmed
chair
confirmed,
I
think,
from
my
perspective,
if
you
are
not
no
no
objectives
here,
but
if
you
do
have
concern
the
questions,
do
not
the
whole
team
knows,
and
I
think
we
as
anything
we,
since
it's
now
official
this
condition.
We
want
to
operate
on
the
on
the
open
source,
spiritual
and
open
source
principles
in
open
discussions,
so
any
concerns
which
is
that
the
whole
team
knows
so
right.
B
Yeah,
the
d-pack
and
the
challenge
have
been
involved
a
lot
with
technique
with
centaurus,
and
that
also
has
other
kind
of
responsibilities
in
other
projects
this
year.
I
would
probably
also
focus
on
other
practices
too,
so
my
time
is
maybe
they
name
it,
but
whatever
that
helps
need
me
from
needing
me
help
from
centaurus.
B
I
would
definitely
help
that's
no
no
easier
at
all,
and
then
I
also
probably
be
part
of
the
project
governing
group
on
the
overwatch
koi
to
help
to
help
how
we
grow
the
economy,
the
ecosystems,
the
communities
and
they
use
the
marketing
group
outreach.
So
I
can't
help
on
that.
I
thought
submission
on
the
conference
side,
what
I
come
from
with
sponsors
and
which
companies
will
want
to
want
to
give
a
talk.
So
all
this
the
non-technical
I
can
help
on
that.
C
Yeah-
and
I
also
want
to
explain
why
you
know
another
reason
that
dr
schoen
is
not
on
the
tsc,
because
we
want
to
make
sure
we
have
a
you
know,
pretty
good
balance
of
diversity.
We
don't
want
to
have
one
company
has
more
a
lot
more
people
than
other
companies.
You
know,
so
you
know
that's
why
we're
futurists
already
got
two.
That's
all
you
know.
I
mean
anybody
like
push
any
other
organization.
If
you
want
to
contribute
more
tsc
you're
welcome,
but
we
want
to
make
sure
that
there's
a
good
balance
balance
there
yeah.
F
H
B
Absolutely
makes
sense:
yeah
yeah,
I
can
definitely
in
the
group-
can
help
this
project
again.
We
want
to
make
sure
this
is
real.
You
know
open
source
project,
open
communities
in
you
know
of
everybody,
and
then
I
think
the
chair
also
the
tsa
can
discuss
themselves.
I
think
the
tsa
members
you
guys
can
actually
go
off
to
decide
the
terms
for
tier
could
be
year
to
year.
It
doesn't
matter
it's
up
all
up
to
you
guys
now
for
for
how
to
technically
set
the
directions
for
the
centaurus.
F
H
C
Yeah
we
want
to
make
sure
that
we're
doing
this
project
you
know,
because
now
it
is
linux
foundation
project
and
we
need
to
operate
this
project
as
a
really
a
pure.
You
know
open
source
project,
so
everything
we
do
has
to
follow
the
you
know
open
principle,
right,
open
design,
open
collaboration,
open
communication.
That
means
everything
we
do
has
to
be
done
in
an
open
way
and
everything
we
do
has
to
be
documented
and
a
collaboration
and
only
any
discussion.
You
know
whether
it's
very
controversial
discussion
or
you
know
bad.
You
know
or
constructive
discussion.
C
I
Yeah
yeah
great
any
yeah,
I
agree
to
I
mean
what
you
discuss
now.
I
have
just
one
thought:
maybe
you
can
check
the
feasibility
that
usually
us
such
kind
of
projects.
We
also
could
have
something
like
a
project,
advisors,
yeah,.
F
I
C
B
That's
a
good
idea.
Actually,
what's
the
thing
about
this
is
you
I
have
people
from
anita's
foundation
right,
join
the
advanced
group
or
whatever
the
group
will
have
perfect.
C
Yeah
for
project-
that's
great
yeah,
okay,
I'll,
take
that
action
item
I'll
look
into
that
yeah
and
then
also
so
for
the
overall
project
discussion
group.
We
can
always
update
this
tsc.
You
know
body
right
to
make
sure
that
all
the
tcc
members
are
aware
of.
What's
going
on
in
their
on
the
non-technical
side,.
B
Okay,
I
think
if
I
have
no
objection
and
saying
that's
it,
that's
for
the
tsc
for
centaurus,
thanks
again
for
everyone
try
to
to
be
part
of
the
ltsc
and
looking
forward
for
your
contributions
technically
in
any
any
technical
aspect
aspect.
So
let's
move
on
thanks
again
for
everybody.
So
this
is
a.
I
think
this
is
for
the
tsc
meeting.
For
my
time,
frequency.
I
think
I
think
we'll
probably
want
to
leave
this
to
tsc.
B
H
Will
that
be
okay
to
bring
it
to
the
next
meeting
like
list
of
people
who
could
be
potentially
there
as
any
was
suggesting,
like
sunil
being
part
of
soda
and
huawei?
Both
is
one
option
here,
but
I
think
other
customers
can
also
come
from
different
angle
like
otc
or
maybe
from
telefonica
or
other
angles.
C
I
think
what
would
be
good
is
the
tsc
okay,
I'm
just
you
know
using
the
other
projects.
I
you
know
I
was
involved
in
how
they
do
the
best
practice,
so
the
tse
needs
to
have
a
charter
right.
You
need
to
decide,
you
know
the
process
to
welcome
or
to
qualify
a
new
tsc
member
right.
So
but
I
mean
I'm
sure
at
this
point,
the
more
users
companies
we
have
the
better.
C
I
think
because
that's
how
we
can
make
this
project
successful,
but
I
do
think
there
should
be
some
official
process
for
a
tsc
member
to
vote
to
say:
okay
yeah,
for
example,
orange
is
interested
in
participating
in
this
project.
Are
we
gonna?
You
know
you
know,
what's
the
qualification
process,
and
sometimes
it's
just
it's
just
a
process,
it
doesn't
have
to
be
very
difficult.
It
should
be
like
oh
yeah,
okay,
they
have
to
show
their
intent.
Why
they're
interested
and
then
okay?
Now
they
are
part
of
this
project.
C
H
C
What
I
was
thinking
we
can
have
a,
I
think,
what
I
suggest
is,
you
know
we
can
look
at
other
projects
and
then
I
can
go
get
the
charter
verbiage
like
and
then
we
can,
you
know,
modify
a
coin
to
you
know
what
this
tsc
body
wants
right
and
then
you
you
vote
for
it
saying
this.
Is
the
process
to
welcome
a
new
member
company
and
and
then
we
just
create
that
process
and
have
this
body
vote
for
it
and
then
once
that's
done
then
yeah.
We
should
you
know
and
that's
part
of
our
reach.
C
We
go
out
and
recruit
all
these
other
member
companies
and
then
they
all
have
to
go
through
this.
You
know
pro
enrollment
process.
D
Yeah,
I
think
the
first
thing
that
tlc
probably
need
to
do
is
we
define
our
chart
right
and
the
chart
should
include
this
process
like
how
we
accept
the
new
members,
how
we,
how
we
operate,
how
we
vote,
basically
all
the
the
basic
rules
then
and
process.
Then
the
tlc
can
follow
the
charter
to
operate
right,
yeah,
something
like
a
topic
for
your
first
yeah.
We
can
prepare
yeah,
we
can
prepare.
D
E
C
Yeah,
but
we
want
to
our
goal,
is
to
grow
our
ecosystem
right
I
mean
anybody
can
join,
can
die
and
anybody
you
know
everywhere
in
the
world
can
die
in
right,
but
we
do
want
to
recruit
all
these.
You
know
user
users
to
come
be
part
of
the
tsc,
because
if
they
are
part
of
tsc,
then
they
can
help
us
grow
the
ecosystem
because
they
have
more.
D
Yeah,
but
I
think
depends
on
their
intention
if
they
are
only
one
part
of
the
ecosystem.
Maybe
they
can
join
this
outreach
group
or
other
user
groups
only
if
they
really
want
to
contribute
on
the
technology
side,
and
they
want
to
correct
here
in
technology
right.
We
don't
want
to
make
everybody
part
of
the
tlc.
Then.
C
D
And
what
tlc
should
do,
how
they
should
do
it?
What's
the
process?
All
this
thing,
we
don't
want
to
make
the
process
heavy,
but
we
we
do
want
to
have
a
clear
framework.
So
we
can.
The
committee
can
operate
yeah.
J
One
piece
of
advice
I
can
give
it
to
you
guys,
you
know
coming
from
this
nikita
from
from
apache
world
is
to
you
know.
I
understand
that
you
know
this
is
very
typical
kind
of
typical
mindset
when
the
you
know
when
the
commercial
companies
set
not
open
source
project.
There
is
this.
You
know
default
drive
to
put
some
processes,
committees
and
kind
of
a
lot
of
necessary,
not
in
deregulatory
way
but
necessary
bureaucracies.
J
Try
to
try
to
you
know
you
know
limit
this
urge
to
you
know,
put
processes
around
it.
I
would
advise
to
make
it
as
as
open
and
as
free
initially
as
possible.
You
you'll
have
time
when
it
becomes
messy
and
unmanageable
to
kind
of
correct
this
later,
but
you'll
be
surprised
how
many
people
are
really
fickle
about
this.
These
details,
I
mean
you
know.
Cnc
generally
has
a
lot
more
heavyweight
process
than
apache.
Apache
is
pretty
much.
J
You
know
kind
of
wild
west
in
many
ways,
but
even
on
the
purchase
of
the
foundation,
people
frown
upon
processes
and
bureaucracies
and
and
whatnot
you'll
be
surprised
how
sometimes
small
things
deter
people
from
joining,
because,
obviously
it's
it's
everybody's
free
time.
Nobody
get
paid,
it's
just.
Basically,
it's
really
in
the
volunteering
work
and
the
last
thing
people
need
when
they
donate
the
time
is
some
kind
of
you
know,
framework
or
whatnot.
You
have
to
have
a
minimal.
J
Of
course
you
have
to
have
you
know
technical
committee
and
some
basic
voting
rights
so
that
you
know
you
can
control
the
project.
Maybe
even
internally
can
have
it.
But
again
I
would
advise
you
know,
outboard
externally,
to
try
and
minimize
that
yeah
and
that's
you.
H
I
think
that's
a
very
good
point
that
you
have
bought
it
up,
because
what
I
have
seen
in
some
other
things,
like
some
employees
of
the
companies,
become
part
of
many
group
and
they
try
to
direct
like
every
single
thing
and
then
it
doesn't
make
any
sense
of
like
being
an
open
source
community,
and
I
hate
that,
because
ultimately,
open
source
is
all
about
ensuring
that
people
have
contributions.
People
voices
have
been
heard,
otherwise
they
become
chair
in
all
the
30
communities
and
then
they
start
basically
doing
the
things
as
they
wish
right.
J
H
J
It'll
come
to
this
by
the
way,
if
you
look
at
successful
projects,
whether
it's
hadoop
and
spark
or
kafka
in
an
apache
software
foundation,
there
is
they
reach
the
point
where
it
becomes
extremely
messy.
There's
hundreds
of
people,
it's
so
you
know,
and
then
you
need
to
have
some
structure,
but
that.
E
Yeah
you
need
to
have
connected
critical
mass
first,
you
know,
and
then
can
I
have
all
this.
Exactly
yeah
process
is
around
again
yeah.
D
B
You're
very
good
nikita
very
good,
suggested
very
good
idea
comments,
because
I
I
was
involved
in
the
same
safe
toc
at
the
beginning,
back
to
2015,
and
then
I
see
the
growth.
I
think
I
agree.
I
was
very
good
at
comments
that
we
agree
will
make
the
at
the
beginning
at
least
right
and
at
this
stage
make
the
as
a
free
as
simple
as
open
as
possible
yeah.
So
we
can
can
can,
can
join
and
contribute
and
understand
this
project.
Definitely
yeah.
J
F
J
Real
part
of
the
community,
a
real
membership
community
will
be
the
developers
and
sometimes
it
could
be
individual
developers
you
know
could
be
yes.
You
know,
you
know,
members
of
the
companies
live
in
the
tcs
and
whatnot.
So
just
a
piece
of
advice
that
initially
I
would
concentrate
our
efforts.
You
know
in
outreach
and
marketing,
which
is
initially
probably
going
to
be.
Almost
the
same
thing
is
to
really
outreach,
and
you
know
you
know,
attract
as
many
developer
types
into
the
project
and
that's
a
little
bit
different.
J
So
that's
kind
of
my
piece
of
advice,
because
again
it's
a
just
a
matter
of
kind
of
sequencing.
You
know
right
now,
there's
nothing
to
use,
so
we
have
to
build
something
to
be
used
and
we
need
to
basically
get
as
much.
You
know.
Developer
feedback
develop
participation
into
it.
J
So
that's
not
gonna,
be
no.
You
know
thought.
H
So
what
I
can
do
in
the
next
meeting,
I
can
share
some
experts
that
I
know
who
can
be
really
using
this
project,
because
when
I
met
with
dr
xiao
and
annie
about
this
thing,
I
was
like
mind
blown
with
this
concept
and
I'm
very
passionate
about
this
whole
cloud
native
5g
kind
of
technology.
That
is
coming
up
front
and
we
believe
that
there
are
true
users
who
can
be
part
and
who
can
really
contribute
to
it.
H
And,
as
you
said,
nikitake
in
the
beginning,
we
should
be
very
open
and
we
should
like
great
get
the
right
set
of
expert
so
that
it's
more
about
contribution
rather
than
basically
giving
too
much
power
to
certain
individuals,
because
this
is
an
open
source.
Community
and
open
source
community
thrives
when
everybody
feels
that
they
are
part
of
this
ecosystem
and
it's
not
governed
by
one
or
two
individuals.
H
J
Yeah,
it's
a
good
point,
but
you'll
be
surprised
that
you
know
there's
a
very
delicate
balance:
yeah,
nobody
likes
actually
wild
west,
and
so
you
know,
if
you
look
at
the
history
of
hadoop-
and
you
know
some
of
the
big
projects
in
apache,
I
mean
people
decry
the
situation
where
there's
like
you
know,
half
a
thousand
developers
and
nobody
knows
what
everybody
else
is
doing.
That's
not
the
right
approach,
so
you
have
to
have
leaders
and
leaders
have
to
be
vocal
and
they
have
to
be
known.
J
Keep
it
as
flexible
as
possible.
Another
basically
tidbit
over
you
know,
experience
is
that
probably
the
best
way
to
attract
developers
is
to
do
you
know
you
know,
is
to
do
a
bunch
of
meetups
and
bunch
of
small.
You
know
local
presentations
everywhere,
hopefully
we're
now
in
the
cupboard
area.
You
know
error,
we
don't
have
to
travel
anywhere.
We
can
do
all
of
that.
You
know
just
literally
just
with
the
right
now
for
zoom,
and
you
know
in
two
projects
that
I
was
involved.
Very
you
know
heavily
in
apache.
J
That
was
the
boon
you
know
like.
In
apache
night,
we
were
able
to
grow
by
leaps
and
bounds
by
literally
just
doing
a
bunch
of
meetups
and
bunch
of
presentation
about,
because
that's
the
way
people
learn
about
new
project
right,
there's
literally
it's
a
very
economical
way.
It's
a
you
know
it's
a
many
of
us
can
do
it,
especially
those
that
are
actually
involved
in
a
from
future
way.
Side
of
things
you
know,
can
do
very
interesting
technical
presentations
and
that's
a
very
good
coverage.
J
You
know
presentations
get
recorded,
you
know
at
least
we
should
get
recorded.
We
should
have
a
youtube
channel,
which
is
the
primary
channel.
You
know
marketing
channel
today
for
learning
anything,
and
that
has
you
know
hello
effect.
You
know
we
do
the
presentations,
you
know
we
caught
them
on
youtube.
We
put
them
on
youtube,
they're,
searchable,
they're,
available
and
over
a
short
period
of
time.
In
the
weekend
you
know
six
months.
We
can
develop
the
library
of
these
presentations,
you
know
and
have
a
much
better
coverage.
F
J
J
So
everybody's
competing
for
the
same.
You
know
brilliant
people
who
are
passionate
and
you
know
and
participate
in
communities,
so
we
have
to
make
an
effort.
This
is
not
you
know.
This
is
not
one
of
those
situations
where
we
just
hey,
saying:
hey,
centaurus
and
and
wait
everybody
to
join
you'll
be
surprised.
You
have.
G
J
An
effort
you
know
you
have
to
get
people
excited,
I
believe,
there's
plenty
of
excitement
in
this
project.
I
think
it's
you
know,
as
I
mentioned
in
the
beginning,
it's
a
it's
massively
overdue
for
something
like
this
to
to
appear
in
an
open
source
arena.
Not
google,
not
amazon,
driven,
not
vm3
when
we're
driven.
So
there's
plenty
of
being
excited
here,
and
we
just
need
to
make
this
in
no
concerted
effort
to
get
this
message
out.
B
Yeah
yeah,
we
just
need
to
talk,
we
just
did
go
out
and
talk
and
the
the
project
itself
yeah.
J
I
agree,
I
would
say
basically
almost
the
first
order
of
you
know
of
action
is
to
really
you
know,
probably
for
you
guys
out
of
future,
where
you're
most
closely
related
to
the
technical
side
of
this
right
now
is
to
really
start
talking
about
it,
and
obviously
you
know
many
of
us
can
help,
but
you
guys
are
closest
at
this
point
and
has
best
you
know
the
most
technical
arguments
projects
concerned.
So.
E
You
know
one
of
the
opportunities
I
see
is
actually
not
just
us
doing
the
meetups.
You
know,
but
presentation,
obviously
we'll
do
that,
but
it'll
be
much
more
powerful.
If
we
collaborate
with
you,
for
example,
you
know
you
have
your
own
ecosystem,
for
example,
grid
gain
or
you
know.
J
E
I
E
See
a
lot
of
opportunities
in
data
management.
You
know
the
scenario
soda
group
right
there,
a
lot
of
companies,
I'm
pretty
sure
they
must
be
looking
for
this
kind
of
platform.
You
know
private
cloud,
amazon,
scale,
private
cloud
and
you
know
grid
gain.
You
know
financial
services
and
telecom
and
all
that-
and
this
is
a
great
opportunity.
E
B
Well,
I
have
to
control
from
time
perspective
reacting
over
three
minutes
and
we
still
have
some
items
but
just
quickly
sucks
for
the
advice.
I
seem
very
good.
I
think,
there's
a
lot
of
work
away.
We
are
waiting
for
us
to
do,
and
especially
with
for
the
tsc
for
all
uk's
and
offices,
in
a
lot
of
work
to
do
to
to
grow
their
ecosystems
and
also
to
grow
not
only
the
ecosystem,
but
also
the
the
the
robustness
of
the
the
project
itself.
B
Okay,
listen
again
at
the
time.
I
I
I
think
next
item
is
project
update,
but
since
this
is
the
time
is
where
our
need
to
be
time,
if
everybody
has
a
couple
three
more
minutes,
a
second
channel,
you
can
talk
about
quickly,
the
planning
and
but
again
this
is
a
draft.
I
remember
this
is
the
draft
and
this
you
guys
can
go
to
discussion
about
adding
what
is
missing.
B
What
is
what
needs
to
be
added,
and
so
everybody
has
a
couple
for
a
few
more
minutes
that
shannon
can
go
through
this
hi
money.
B
Yeah,
I
just
want
to
make
sure,
because
this
is
for
this
first
meeting,
make
sure
the
time,
if
you,
if
anyone
that
has
other
obligation
need
to
go,
I
think
I
think
that's
fine.
I
think
and
then
now
that
this,
whatever
you're
going
to
talk
about
is,
I
will
be
here
anyway.
Right
will
be
open
or
be
shared.
D
Exactly
yeah
all
the
information
here
in
the
public
and
will
be
publicly
accessible
because
we
are
already
around
over
time,
so
I
will
try
to
be
quick.
This
is
just
a
brief
update
of
the
projects
where
we
are
and
what's
the
plan
for
the
new
year,
so
I
will
have
three
parts.
First,
just
give
a
brief
review
of
this
project.
D
It
mainly
includes
now
includes
three
projects,
one
in
the
contactors,
that's
for
compute,
the
other
one
is
misa
for
the
cloud
networking.
These
two
are
the
major
component
component
projects.
We
also
have
a
third
one
called
dashboard.
That's
the
portal.
The
portal
is
for
the
for
both
actors
and
visa,
so
user
can
manage
all
these
computer
and
the
network
elements
resources
in
an
intuitive
way.
These
are
the
three
major
components
in
addition
to
that,
we
have
some
several
collaboration
projects
on
site
projects.
D
They
are
still
not
officially
part
of
the
centaurus
project,
but
can
potentially
become
the
aware
cloud
which
we
are
among
feature
weight
until
when
we
went
and
there's
a
scalable
strong
consultant
that
I
stole
that's
among
feature,
we
are
in
a
great
game
and
feature.
We
also
have
a
global
schedule
inside
the
project.
So
this
overview
of
all
the
projects
related
to
centaurus
project.
D
For
the
status
last
year,
we
have
four
to
five
releases
for
actors
and
misa.
We
basically
follow
our
cadence,
our
seasonalities
every
three
months
on
release,
but
depending
on
the
different
features,
some
features
probably
will
take
a
longer
time.
Some
features
will
have
a
short
time,
so
it's
just
roughly
every
three
months,
but
it's
not
exactly
every
three
months.
D
D
D
D
Also,
we
finished
the
multi-tenancy
improvements
because
kubernetes
doesn't
have
multi-tenancy,
so
we
added
a
blt
monitoring
model,
a
strong
intensive
model
implementation
and
we
greatly
improved
the
scale
bridge.
We
implemented
a
partition-based
architecture
and
it
already
passed
the
20k
test
for
of
the
benchmark
test.
The
original
kubernetes
can
only
pass
5k
or
10k,
but
now
we
already
passed
the
20k
and
we
are
trying
more
a
larger
scale
with
with
this
scale-out
architecture.
D
For
me
that
we
implemented
the
basic,
is
basically
architecture
and
supported
basic
virtual
network
model,
including
vpc
subnet,
and
points
we
also
implement
some
advanced
features
like
a
direct
path
and
that's
a
performance
optimization
and
also
we
have
scale
endpoint
and
the
covariance
network
policy
support
for
dashboard.
As
I
just
mentioned,
we
just
implemented
some
features
about
actors,
but
not
for
me
right.
D
The
the
most
interesting
part,
probably
is
planning
the
technical
direction
for
this
year.
This
is
also
we
want
to
get
all
the
get
the
feedbacks
from
our
tsc
members
and
other
men
other
other
participants
in
the
new
year.
So
far,
we're
thinking
about
the
three
things
one
is
that
we
continue
to
enhance
the
base
platform
by
basic
platform.
We
mean
that
this
basic
fundamental
feature
that
a
large
cloud
infrastructure
platform
should
have.
D
D
So
I
will
spend
several
minutes
here
for
the
best
platform.
We
will
continue
to
increase
scalability
to
support
the
public
cloud
scale
and
we
also
add
missing
network
features
in
visa
and
do
some
hardware
offloading
performance
research
and
we
increa.
We
enhance
the
wrong
time
in
the
photo.
So
for
this
part
it's
basically
continued
our
efforts
in
last
year,
but
we
have
started
two
new
areas.
One
is
for
support
for
ai
and
a
the
other
one
is
edge
for
ai.
D
When
we
talk
about
system
ai,
that
two
different,
there
are
usually
two
directions:
one
in
the
system
for
ai
the
other
one
is
ai
for
system
this
year
from
engineering
perspective,
we
are
mostly
focusing
on
system
for
ai.
We
are
trying
to
solve
two
problems
that
we
heard
from,
not
our
customers
and
the
partners.
D
One
is
that
the
gpu
using
gpu
utilization
in
in
some
research
people
mentioned
that
about
one
third
of
gpus
in
production
they
have
an
average
optimization
less
than
20.
D
The
other
part
is
that
we
want
to
optimize
the
distributed
training
or
just
training
performance,
so
that
the
machine
learning
training
can
have
less
waiting
time
and
a
shorter
duration
time.
D
We
already
did
some
work
one
or
two
years
ago
to
support
to
support
some
machine
jobs.
This
year
we
plan
to
do
a
more
comprehensive
job
cover
all
these
training
jobs
and
for
the
training
we
will
research,
some
new
scheduling,
algorithm
to
reduce
the
average
job
waiting
time
and
also
we
try
to
implement
the
elastic
training
that
will
support
a
trinity
job
to
dynamically
scaling
us
out.
D
So
when
there
are
expert
gpus
available
that
can
leverage
this
extra
available
gpu
to
make
the
training
and
time
shot
in
terms
of
ai
system
for
system,
we
will
continue
the
project
that
we'll
work
with
the
feature
we
welcome
to
you
in
on
the
machine
learning
dream,
auto
scaling
and
sal.
D
D
For
the
edge
we
are
planning
to
implement
the
actual
agent
and
misa
agent
on
edge
sites.
So
these
h
engines,
where
the
lightweight
agents
and
also
support
some
like
offline
situations
that
are
common
in
the
edge
of
computing
and
also
we
will
optimize
our
scheduling
algorithm.
So
if
our
class
is
across
cloud
data,
centers
and
edge
sites,
how
we
can
do
this
edge
aware
schedule
to
dynamically
offloading
some
services
to
edge
sites
based
on
the
latency
and
the
customer
requirements,
and
how
to
optimize
scheduling
algorithm,
to
support
this
affinity
and
affinity
requirements.
D
If,
because
for
some
exercise,
their
sites
are
very
small,
it's
not
easy
for
them
to
meet
these
requirements,
and
also
we
have
other
information.
We
have
a
side
project
for
called
global
scheduler,
that's
also
edge
related
and
we
have
our
open
source.
Another
open
source
project
called
qh,
where
we
are
thinking
how
we
can
integrate
with
them.
D
D
For
example,
we
probably
have
scalability
6
second
unit
6,
networking
6
h,
probably
ui
depends
on
depending
on
which,
depending
on
how
we
can
organize
people
more
efficiently
to
to
move
faster
and
to
align
with
the
tech
directions.
These
are
just
some
ideas.
D
We
so
far
we
don't
have
a
concrete
plan
on
this,
but
if
we
have
some
use
cases
or
some
ideas
for
this
part,
I
think
it
will
be
great
that
we
have
some
some
features
or
planning
for
this,
at
least
so
we
can
be
prepared
for
that.
H
D
J
Yeah,
just
the
one
comment
from
my
side:
we
also
it's
just
nikita-
I'm
enough
we're
on
igniting
against
site
we're
also
working
with
we're
kind
of
thinking
we're
developing
a
kind
of
a
next-generation
database
engine.
If
you
will
what
we
call
multi-planar
database
engine,
where
you
know
one
of
the
planes
that
we're
exploring
is
the
execution
planes
so
right
now
we're
java
base,
obviously
executing
just
a
standard.
J
You
know
x86
architecture,
but
we
also
look
not
looking
but
we're
developing
any
ways
to
like
upload
sql
processing,
for
example
the
gpu
and
cmd,
and
what
we're
finding
is
that
you
know
a
lot
of
people
just
ignored
the
cmd
processing
and
that's
in
many
ways
much
easier
concept
than
the
gpu,
because
for
gpu
processing
it's
a
very
limited
infrastructure.
J
We
should,
we
shall
say,
see
what
in
nvidia
clustering
comes
around
in
the
networking
kilometers
comes
around.
But
fundamentally,
when
you
deal
with
the
gpus,
it's
a
it's
a
narrower
case,
because
you
have
to
upload
that
to
gpu
for
gpus
to
be.
You
know
useful
in
many
ways.
Well,
the
cmd
operations
can
often
not
always
but
often
get
the
same
performance
for
the
end
users,
because
you
basically
get
the
same
parallelization
in
the
cmds
right
now
prevalent
in
right,
you're,
starting
with
ssc
to
avx
instructions
available
everywhere
in
every
cloud.
D
I
think
we
are
we're
talking.
Probably
I
shouldn't
make
them
more
clear.
There
are
two
different
things
for
this.
For
this
hardware
of
loading,
I
mean
to
accelerate
the
networking
process,
it's
probably
overloaded
to
the
smart
techniques.
It's
not
related
to
the
to
this
topic,
the
gpu.
This
gpu
is
okay,
it's
mostly
for
the
machine
learning
training,
because
we
want
to
improve
our
platform
to
give
better
experience
for
machine
learning
applications
yeah.
These
are
two
different
two
different
types.
J
All
right
so
yeah
so
say
for
folks
like
me
and
coming
from
this
background,
you
know
we
need
this
clear
kind
of
you
know
differentiation,
because
if
you
know
if
I
would
be
looking
from
kind
of
my
vintage
point,
the
minute
I
see
you
know
the
gpu
word
mind
is
thinking.
This
is
the
research
second
schedule
I
can
use
you
know
and
when
I
see
the
gpu
immediately
trying
to
look.
Where
is
my
cmd
support?
J
No,
because
it's
you
know
from
an
application
standpoint,
it's
just
the
different
planes
for
execution.
My
code
I
can
execute
on
in
the
classic
x86.
I
can
execute
on
a
different
instruction
set
and
I
can
execute
a
different
instruction
set
on
different
hardware
cards
with
gpu
cards,
for
example.
So
I
don't
you
know
that's
kind
of
my
view
of
the
world.
E
E
D
Those
things
in
mind
there
are
different
devices
like
tpu
the
council,
accelerator
machine
learning,
training,
yeah.
K
J
F
J
F
J
J
J
So
just
again,
I
don't
think
we
we're
saying
the
same
thing.
I'm
just
saying
that
you
know
just
to
get
a
little
bit
more
clarity
to
the
end
user
community.
You
know.
B
But
it's
more
broad,
abstract
layer
of
manage
the
heterogeneous,
the
resources,
what
blood
execution
you.
D
Know
hydrogenation,
but
for
the
projection
plan
I
don't
want
to
here
is
the
plan
that
we
will
start
with
gpu,
because
gpu
so
far
is
still
the
main
streaming
devices
that
people
are
using
and
the
tpus
like
they
are,
for
example,
google,
tv
users.
They
are
still
available
in
their
cloud.
So
for
us
it's
not
easy
to
like
develop
a
feature
on
top
of
them
and
use
it
for
users.
D
J
B
Thank
you.
You
cannot
see
me
but
anyway,
I
think
it
is
a
good
time
actually,
when
this
meeting
now,
I
think
we
all
24
minutes,
which
is
good
everybody
passionate
about
it.
Everybody
has
got
very
good
ideas,
so
it's
great
in
fact,
so.
B
I
think
I
think
that's
that's
all
for
us
this
meeting
today.
H
H
Shouting,
it's
very
very
elaborate
and
very
nicely
done.
So
thank.
H
B
No,
no,
no!
No!
No!
No!
No!
I
would
recommend,
I
would
say
I
would
recommend
the
maybe
set
up
a
next
meeting
me
instead
of
a
next
meeting,
maybe
and
then
the
meeting
itself
discuss
the
frequency,
how
you
guys
want
to
change
to
whitney
monthly
and
is
up
to
the
tc
now
yeah,
probably
we.