2. Market
Segment
Core
Banking
Business
Solutions
e-‐Health
e-‐Administration
Brokering
customers’
IT
needs
&
Services
MS
Azzure
Salesforce
Gigaspaces
Amazon
EC2
OVH
Sky
Compu*ng
Appistry
Spring
3. Historical
Background
&
Uniqueness
gnúbila
technological
assets
leverage
on
state-‐of-‐the-‐art
in
High
Performance
Computing
(HPC),
in
particular:
CERN
LHC
Grid
technology
} 10
years
of
Research
and
Development,
} Technology,
knowledge
and
experience
transfer
since
2005,
» Involved
in
key
European/international
initiatives
for
more
than
5
years,
} Developed
and
intensively
exercized
interoperability,
» Technology
applied
to
and
validated
for
different
sectors
and
customers
gnúbila
leverages
on
the
Grid
to
offer
a
PaaS
forward…
to
a
Sky
of
Computing
facilities
} Multitenant
PaaS
that
can
cope
with
different
IaaS
providers
» Interoperable/integration
middleware,
» Built-‐in
Private
Cloud
capabilities,
» Native
data
distribution
capabilities,
» Hybrid
Cloud
Brokering
} Multitenant
DBMS
that
can
cope
with
different
Data
Infrastructures
» Interoperable/integration
middleware,
» Built-‐in
graph-‐oriented
G
DBMS
} Easy
and
rapid-‐cycle
time
to
develop
and
customize
to
new
Corporate
needs
4. Technology
-‐
Value
Proposition
(1) To
facilitate
IT
legacy
assets
migration
to
the
Cloud
(with
non-‐invasive
integration,
nor
costly
reengineering),
» Cost-‐effective
migration,
» Sustainable
Business
Intelligence
(2) To
offer
compatibility
with
/
abstraction
from
mostly
utilized
technologies
in
databasing,
application
server,
» No
vendors
lock-‐in,
» Potentially
significant
license
costs
savings
(3) To
provide
a
powerfull
and
innovative
new
multi-‐tenant
PaaS
towards
Private
and
Hybrid
Clouds
» From
inhouse
facilities
to
online
dedicated
resources
to
keep
control
over
sensitive
Entreprise
data,
» Efficient
brokering
to
outside
elasticity
5. Service
-‐
Value
Proposition
(1) To
accompany
ISVs
and
SIs
migrating
to
the
Cloud
» Technical
support
in
designing
and
developing
service-‐oriented
architectural
solutions,
» Expert
consulting
in
standards
and
interoperable
architectures,
» Expert
consulting
in
evaluating
and
selecting
appropriate
Clouds
(2) To
broker
Customers’
needs
towards
Cloud
offerings
» Expert
consulting
in
evaluating
and
selecting
Cloud
solutions,
» Expert
consulting
in
Cloud
pricing
and
business
model
strategies
(3) To
manage
cost-‐effectively
and
technically
migrations
to
the
Cloud
» Experienced
project
managers
in
HPC,
Grid
and
Cloud
computing
developments,
» High
quality
developments
and
processes
6. France
Service
-‐
Value
Proposition
(1) Setup
of
research
and
development
projects
» Project
setup
expertize
in
various
funding
tools
§ EU
FP7,
IMI,
EUREKA,
IDEA2
§ UN
ITU,
US
AA
§ FR-‐ANR,
IT-‐PON
» Network
of
professionals,
consultation
and
lobbying
activities
(2) R’n
D
project
management/pilotage
» Technical
coordination,
» Subcontracting,
» Specialized
technical
expertize
§ Cloud/Grid/HPC
computing
§ Biomedical
Informatics,
e-‐Health
(3) Software
architecture
design
consultancy
Research
» Model
Driven
Engineering
» Service
Oriented
Architectures
design
and
developments
» HPC/Grid/Cloud
integration/transition
7. OVERVIEW
PRESENTATION
LAYER
G
Developer
-‐
SaaS
MIDDLEWARE
LAYER
G
Pandora
–
PaaS
G Technology
RESOURCE
LAYER
G
Core
-‐
IaaS
8. PRODUCT
G
-‐
PANDORA
§ Secure
• Encrypted storage, secure and anonymous data transfers
• Innovative identification means, identity delegation
• Role-based access controls, customized security model
§ Elastic
• Distributed, extensible and federated storage
• Service Oriented Architecture (SOA), modular design
• Standards compliant, interoperable solutions
§ Powerful
• Unlimited application and data scale-up,advanced scheduling techniques
• Multi-tenant Cloud IaaS brokering
• Dedicated access to private Cloud resources
9. PRODUCT
G
-‐
Pandora
§ Distributed platform providing access to heterogeneous Computing and Storage Resources:
• G is Secure:
· Flexible authentication mechanisms: SAML2.0, X.509 Certificates
· SSO Enabled:
– Shibboleth http://www.internet2.edu/shibboleth
– JASIG CAS http://www.jasig.org/cas
• G is Powerful:
· Based on the JSAGA http://grid.in2p3.fr/jsaga/
- Implementation of the Simple API for Grid Applications (SAGA) http://saga-project.github.com/
- Specification from the Open Grid Forum (OGF) http://www.gridforum.org/
– Unlimited application and data scale-up with advanced scheduling techniques
• Elastic/Fault tolerant:
· Can be installed in different servers to resist to failure and
to increase capacities in case of high load
• Extensible:
· Create your own service with your own business logic:
– Framework which provides simple APIs and services to
interact with heterogeneous middleware and data sources.
» Multiple available services frontend:
• Based on Apache CXF http://cxf.apache.org/
• JAX-WS
• JAX-RS
10. G
–
Pandora
G
–
FedEHR
Use
networked
IaaS
resources
Start
with
one
PaaS
node
1
Add
other
3
PaaS
nodes
2
Add
nodes
duplicates
4
Query
from
anywhere
in
the
network
Add
SaaS
gateways
11. Federated
Electronic
Health
Records
G
-‐
FedEHR
• Patient-‐centered
EHR
management
• Extensible
repositories
and
data
Integration
possibilities
Anonymizing
Adding
Concept
Anonymizing
Adding
Concept
Anonymizing
Querying
Adding
Concept
Anonymizing
Adding
Concept
• Ontology
concepts
ready
• HL7,
OpenEHR
and
SNOMED
CT
Ready
• Build-‐up
high-‐performance
data
distribution
strategies
12. PRODUCT
Federated Electronic Health Records
G
-‐
FedEHR
§ Patient-centered EHR management
§ Extensible repositories • Build-‐up
high-‐performance
data
distribution
strategies
• Automated
replication
of
chosen
nodes
to
slave
nodes
• Distribution
of
data
between
master
nodes
Extraction
13. Data
Mart
(1)
Select
Medical
(2)
Select
Data
to
(3)
Generate
(4)
Execute
Procedure
Query
Query
Query
Downl.
Visualisation
Charts
Maps
Cloud
Data
Warehouse
(ETL)
OPBG
1
OPBG
2
OPBG
3
OPBG
n
14. " Data model completely hidden to users
" Database can be extended and distributed,
" Query results can be displayed using various tools
" Google Chart Tools
" Google Earth/Maps
" Gephi
15. G
–
Pandora
G
–
FedEHR
Hardware
Software
• We
can
manage
the
full
process
from
design
to
on
site
deployment
• Your
hardware
is
delivered
pre-‐configured
on
site
• Your
data
structure
is
defined
from
User
requirements
• Data
concepts
are
added
independently
of
your
data
structure(s)
• Hardware
and
software
processes
run
in
parrallel
Importer
data
structure
user
rDelivery
equirements
Node
development
definition
Pre-‐installation
Re-‐Packaging
deployment
On
site
Ordering
installation
16. PCDR
Paediatric Cardiology Digital Repository
• Distributed (FedEHR)
· Data stored in each node
· All data accessible from any
Connected Node
• “Out of node” anonymised (configured)
· Anonymised in the repository
· Can retrieve Patient in his original
node
• Grid Calculation Ability (Pandora)
• Secured Data (Pandora)
Facts & Figures:
§ 2 major acquisition sites
• 6 hospitals
§ 2.547 million patients
§
• 17.665 million
EHRs
117.769 medical
+
procedures
17. GINSENG
ANR-‐10-‐TECSAN-‐008-‐01
Health Watch Network
• Distributed (FedEHR)
· Data stored in each node
· All data accessible from any
Connected Node
• Grid Calculation Ability (Pandora)
• Secured Data (Pandora)
• Data deduplication service
Facts & Figures:
§ 1 hospital
§ 2 sanitary surveillance centers
§ 11 private clinics
§ 10 years of patient data in
Auvergne region +
18. G4Life
gnúbila
Cloud
Computing
offer
is
based
on
Grid
Computing
evolving
towards
a
PaaS
model
gnúbila France backbones
" Data
and
Processing
Middleware,
major international initiatives
" Mainframe
at
a
fraction
of
the
cost,
" Highly
scalable
to
expand
as
needed,
" Interoperability
to
create
homogenous
data
Sim-e-Child repository(ies),
" Abstraction
technology,
" Integrate
standards,
" Easy
and
rapid-‐cycle
time
to
customize
to
Corporate
needs,
" Security
based
on
state-‐of-‐the-‐art
encryption
technologies,
" From
1FA
to
3FA,
Sentinel " Single
Signon
SALTY
19. Biomedical
Applications
Decision
Health – Case Study &
Support
Knowledge
Discovery
Complex
Modelling &
Simulation
Medical Image
&
Data Processing
Mammogrid+
Health-‐e-‐Child
Digital
Repository
neuGRID
GRID
20. Distributed
compu/ng
infrastructures
for
Biomedical
Research
(1)
TranslaKonal
medicine
EU
FP6
Health-‐e-‐Child,
www.health-‐e-‐child.org
IST-‐2006-‐027749
EU
FP7
Sim-‐e-‐Child,
www.sim-‐e-‐child.org
IST-‐2009-‐248421
FR
ANR
GINSENG,
www.e-‐ginseng.org
ANR-‐10-‐TECS-‐008-‐04
EU
FP7
Model-‐Driven
Paedigree
na
• 10
EU
projects,
FR
RSCA
SenKnelle,
www.e-‐senKnelle.org
na
• 1
EU
with
Pharma
industry,
IT
OPBG
PCDR,
www.gnubila/pcdr
na
NL
MAASTRO
TRAiT
na
• 3
FR
projects,
NL
VUA
NG
na
• 4
industrial
contracts
• PCDR
(2)
NeurodegeneraKve
disease
biomarkers
• MAASTRO
EU
FP7
neuGRID,
www.neugrid.eu
RI-‐211714
EU
FP7
DECIDE,
www.eu-‐decide.eu
RI-‐261593
• VUMC
EU
FP7
N4U,
www.neugrid4you.eu
RI-‐283562
• GAAIN
US
AA
GAAIN,
www.gaain.org
(3)
Medical
imaging
FR
ANR
VIP,
h_p://www.creaKs.insa-‐lyon.fr/vip/
ANR-‐09-‐COSI-‐013-‐02
(4)
Clinical
Trials
EU
IMI
EMIF
na
Distributed
compu/ng
infrastructures
for
Biodiversity
EU
FP7
CReATIVE-‐B,
www.creaKve-‐b.eu
RI-‐284441
Research
Distributed
compu/ng
infrastructures
and
interoperability
EU
FP7
SHIWA,
www.shiwa-‐workflow.eu
RI-‐261585
EU
FP7
outGRID,
www.outgrid.eu
RI-‐246690
EU
FP7
EUDAT,
www.eudat.eu
RI-‐283304
FR
ANR
SALTY,
h_ps://salty.unice.fr
ANR-‐09-‐SEGI-‐012
22. Our
References
in
Research
&
Industry
David
Manset
CEO
gnúbila
France
23. • 10
EU
projects,
• 1
EU
with
Pharma
industry,
• 3
FR
projects,
• 4
industrial
contracts
Awards
/
Prizes:
2011
-‐
Best
Live
Demo
Award
@
EGI'11
Technical
Forum,
2011
-‐
Magic
Quadrant,
Entreprise
ApplicaKon
Server
"Visionaries"
@
GARTNER,
2010
-‐
Top
25
Startups,
Cloud
and
ICT
Summit
2010
@
European
TechTour,
2010
-‐
Technology
Transfer
Award
@
IN2P3
/
CNRS,
SenKnelle
Network,
2010
-‐
Best
IT
Development
Award
@
Health-‐e-‐Child
Final
Conference,
Awards
&
Prizes
2010
-‐
Best
Live
Demo
Award
@
EGEE'10
User
Forum,
2009
-‐
Magic
Quadrant,
Entreprise
ApplicaKon
Server
"Visionaries"
@
GARTNER,
2009
-‐
Best
Live
Demo
Award
@
EGEE'09
Conference,
2008
-‐
Best
Exhibit
Award,
1st
Prize
@
ICT'08
Conference,
2008
-‐
Best
Poster
and
Demo
Award
@
HealthGrid'08
Conference,
2008
-‐
Best
Demo
Award
@
EGEE'08
User
Forum,
2007
-‐
Best
Demo,
Runners
Up
Award
@
EGEE'07
Conference,
2006
-‐
Gold
Medal
@
35th
Salon
InternaKonal
des
InvenKons
de
Genève.
24. Magic
Quadrant
Enterprise
Application
Servers,
2009
Gartner
locates
G
technology
in
the
visionaries
quadrant
and
defines
G
as
an
enterprise
application
server
based
on
a
grid
computing
and
advanced
dynamically
extensible
proprietary
DBMS
foundations.
Strengths
• G
is
a
very
sophisticated
and
advanced
platform
providing
a
wealth
of
functionality
(multimedia
content
management,
data
federation,
application
integration,
adapters,
natural
language
support,
metadata
management,
rule
management,
G Technology
orchestration
and
several
others)
on
top
of
the
Specialized
Press
core
enterprise
application
server
functionality,
in
turn
providing
elastic
scalability
and
shared-‐
everything
multi-‐tenancy
support.
• G
has
a
significant
installed
base
of
approximately
500
customers,
mostly
mid-‐size
organizations,
using
the
product
to
run
a
variety
of
custom
and
packaged
applications
in
health
care,
local
government,
financial
services
and
other
industry
sectors
both
in
Europe
and
Latin
America.
• G
is
natively
"cloud
capable,"
and
is
available
on
top
of
Amazon
Web
Services
EC2.
MaatG
is
developing
technology
to
support
a
go-‐to-‐market
strategy
aimed
at
providing
various
forms
of
"G
as
a
service"
in
partnership
with
telecom
operators
and
other
service
providers.
25. Who's
Who
in
Application
Platforms
for
Cloud
Computing:
The
Enterprise
Generalists,
2009
The
description
of
G
technology
given
by
Gartner
is:
• The
G
platform
has
its
technology
roots
in
academic
research
on
grid
computing
(gLite)
and
in
advanced
DBMS
technology.
But,
starting
from
the
seeding
technology
prototyped
in
2000,
G
has
evolved
to
become
a
full-‐fledged
application
platform
offering
enterprise
application
server,
portal,
orchestration,
data
and
application
integration;
application
development
tools;
content
management;
and
multimedia
capabilities.
• The
G
platform
is
a
cloud
platform/SEAP,
as
it
natively
provides
a
variety
of
cloud-‐enabling
features,
including
elastic
scalability,
shared
processing
and
shared-‐everything
multitenancy
support
and
advanced
extreme
transaction
processing
capabilities,
including
a
proprietary,
integrated
distributed
caching
technology.
• The
core
of
the
G
platform
is
a
very
sophisticated,
proprietary,
metadata-‐driven
DBMS
supporting
a
dynamically
extensible
data
model,
and
is
also
capable
of
federating
traditional
relational
DBMS
(RDBMS).
• The
G
platform
is
a
combination
of
Java,
C
and
Perl
code,
but
programmers
can
use
any
combination
of
C,
C#,
Java,
Perl
and
PHP
for
application
development,
in
addition
to
the
platform's
native
metadata-‐driven
tools.
Specialized
Press
• The
G
platform
has
a
vast
installed
base,
but,
for
the
most
part,
it
supports
SMBs
and
on-‐premises,
single-‐
tenant
vertical
applications.
The
advanced
design
of
G
is
theoretically
capable
of
supporting
a
very
large
number
of
tenants
and
users
on
a
single
logical
instance.
However,
these
capabilities
still
need
to
be
proven
in
large-‐scale
deployments,
although
the
company
is
working
on
several
high-‐profile,
business-‐critical
projects
in
the
banking,
healthcare,
transportation
and
electronic
fund
transfer
industries.
Finally
Gartner
gives
the
following
recommendation:
Consider
the
G
platform
when
looking
for
a
multitenant,
elastically
scalable
and
transaction
processing-‐oriented
technology
platform
either
to
support
aPaaS
solutions
(whether
in
the
context
of
private
or
public
cloud
scenarios)
or
to
implement
SaaS
offerings.
26. Cool
Vendors
in
Platform
and
Integration
Middleware,
2008
• Gartner
describes
G
is
cool
because
"G"
combines
data
management,
data
and
application
integration,
application
server,
and
development
tools
in
an
intrinsically
multitenant
platform.
Distinctive
of
"G"
is
its
innovative
database
management
system,
based
on
a
hybrid
relational/network
model.
Data
in
"G"
is
stored
in
objects
that,
in
turn,
are
grouped
into
types.
Although
objects
of
a
certain
type
are
semantically
homogeneous
(for
example,
they
are
all
about
"customers"),
the
structure
of
each
object
can
be
different
(that
is,
not
all
the
customer
objects
hold
data
according
to
the
same
structure).
This
enables
flexibility
in
data
manipulation,
provides
support
for
a
variety
of
query
languages
(XQuery,
SQL
and
XPath),
and
makes
it
possible
to
federate
different
data
sources.
"G"
enables
distributed
architectures
and
also
comprises
application
development
tools,
an
application
container,
an
OWL-‐based
dynamic
schema
builder,
a
job
manager,
Ajax
support,
adapters
for
various
data
sources
and
integration
with
the
gLite,
Open
Grid
Services
Architecture-‐compliant
grid
platform.
• Gartner
identified
the
following
challenges
of
G:
"G"
has
been
extensively
proven
in
hundreds
of
projects,
Specialized
Press
but
its
ability
to
support
large-‐scale
(that
is,
thousands
of
concurrent
accesses
and
hundreds
of
transactions
per
second),
business-‐critical
deployments
with
high
performance
and
scalability
needs
further
validation.
Given
the
originality
and
sophistication
of
"G,"
it
is
challenging
to
articulate
a
crisp,
compelling
and
distinguishing
value
proposition.
• Finally
Gartner
shows
who
should
care
about
G
platform
are:
IT
architects,
development
managers
and
application/data
integration
managers,
especially
those
working
for
European
companies,
struggling
with
heterogeneous
and
highly
distributed
data
sources,
and
looking
for
a
data-‐centric,
rapid
application
development
platform
with
powerful
data
federation
capabilities,
should
look
at
"G,"
including
for
SaaS-‐
style
deployments.