Transcript ESnet

Cyberinfrastructure and Networks:
The Advanced Networks and Services
Underpinning the Large-Scale Science of
DOE’s Office of Science
William E. Johnston
ESnet Manager and Senior Scientist
Lawrence Berkeley National Laboratory
1
ESnet Provides Global High-Speed Internet Connectivity for
DOE Facilities and Collaborators (ca. Summer, 2005)
ESnet Science Data Network
(SDN) core
Japan (SINet)
Australia (AARNet)
Canada (CA*net4
Taiwan (TANet2)
Singaren
CA*net4
France
GLORIAD
(Russia, China)
Korea (Kreonet2
MREN
Netherlands
StarTap
Taiwan (TANet2,
ASCC)
SINet (Japan)
Russia (BINP)
CERN
(USLHCnet
CERN+DOE funded)
GEANT
- France, Germany,
Italy, UK, etc
LIGO
PNNL
ESnet IP core
MIT
BNL
JGI
LBNL
NERSC
SLAC
TWC
QWEST
ATM
LLNL
SNLL
FNAL
ANL
AMES
Lab DC
Offices
PPPL
MAE-E
YUCCA MT
Equinix
OSC GTN
NNSA
PAIX-PA
Equinix, etc.
KCP
OSTI
LANL
ARM
GA
42 end user sites
Office Of Science Sponsored (22)
NNSA Sponsored (12)
Joint Sponsored (3)
Other Sponsored (NSF LIGO, NOAA)
Laboratory Sponsored (6)
commercial and R&E peering points
ESnet core hubs
JLAB
SNLA
ORNL
ORAU
NOAA
SRS
Allied
Signal
ESnet IP core: Packet over
SONET Optical Ring and Hubs
high-speed peering points with Internet2/Abilene
International (high speed)
10 Gb/s SDN core
10G/s IP core
2.5 Gb/s IP core
MAN rings (≥ 10 G/s)
OC12 ATM (622 Mb/s)
OC12 / GigEthernet
OC3 (155 Mb/s)
45 Mb/s and less
 DOE Office of Science Drivers for Networking
•
The role of ESnet is to provide networking for the Office of
Science Labs and their collaborators
• The large-scale science that is the mission of the Office of
Science is dependent on networks for
o
Sharing of massive amounts of data
o
Supporting thousands of collaborators world-wide
o
Distributed data processing
o
Distributed simulation, visualization, and computational steering
o
Distributed data management
• These issues were explored in two Office of Science
workshops that formulated networking requirements to meet
the needs of the science programs (see refs.)
3
CERN / LHC High Energy Physics Data Provides One of
Science’s Most Challenging Data Management Problems
(CMS is one of several experiments at LHC)
Online System
~PByte/sec
Tier 0 +1
human
~100
MBytes/sec
event
reconstruction
CERN LHC CMS detector
15m X 15m X 22m, 12,500 tons, $700M.
event
simulation
2.5-40 Gbits/sec
Tier 1
German
Regional
Center
French
Regional
Center
Tier2 Center
Tier2 Center
Tier2 Center
Tier2 Center
Tier2 Center
Tier 2
~0.6-2.5 Gbps
Tier 3
Institute
Institute Institute Institute
~0.25TIPS
Courtesy
Harvey
Newman,
CalTech
FermiLab, USA
Regional
Center
~0.6-2.5 Gbps
analysis
Physics data
cache
Italian Center
100 - 1000
Mbits/sec
Tier 4
Workstations
• 2000 physicists in 31 countries are
involved in this 20-year experiment in
which DOE is a major player.
• Grid infrastructure spread over the US
and Europe coordinates the data analysis
LHC Networking
•
This picture represents the MONARCH model – a
hierarchical, bulk data transfer model
•
Still accurate for Tier 0 (CERN) to Tier 1 (experiment
data centers) data movement
•
Probably not accurate for the Tier 2 (analysis) sites
5
Example: Complicated Workflow – Many Sites
6
Distributed Workflow
•
Distributed / Grid based workflow systems involve
many interacting computing and storage elements
that rely on “smooth” inter-element communication
for effective operation
•
The new LHC Grid based data analysis model will
involve networks connecting dozens of sites and
thousands of systems for each analysis “center”
7
Example: Multidisciplinary Simulation
Ecosystems
Species Composition
Ecosystem Structure
Energy
Water
Aerodynamics
Soil
Water
Snow
Intercepted
Water
Disturbance
Fires
Hurricanes
Vegetation
Ice Storms
Dynamics
Windthrows
(Courtesy Gordon Bonan, NCAR: Ecological Climatology: Concepts and Applications. Cambridge University Press, Cambridge, 2002.)
Years-To-Centuries
Watersheds
Surface Water
Subsurface Water
Geomorphology
Hydrologic
Cycle
Days-To-Weeks
Nutrient Availability
Minutes-To-Hours
A “complete”
Chemistry
Climate
CO2, CH4, N2O
Temperature, Precipitation,
approach to
ozone,
aerosols
Radiation, Humidity, Wind
climate
Heat
CO2 CH4
Moisture
N2O VOCs
modeling
Momentum
Dust
involves many
Biogeophysics
Biogeochemistry
Carbon Assimilation
interacting
Decomposition
models and data
Mineralization
Microclimate
that are provided
Canopy Physiology
by different
Phenology
Hydrology
groups at
Bud Break
Leaf Senescence
different
locationsEvaporation
Gross Primary
Species Composition
Transpiration
Production
Ecosystem Structure
(Tim Killeen,
Snow Melt
Plant Respiration
Nutrient Availability
Infiltration
Microbial Respiration
Water
NCAR) Runoff
8
Distributed Multidisciplinary Simulation
•
Distributed multidisciplinary simulation involves
integrating computing elements at several remote
locations
o
Requires co-scheduling of computing, data storage, and
network elements
o
Also Quality of Service (e.g. bandwidth guarantees)
o
There is not a lot of experience with this scenario yet, but
it is coming (e.g. the new Office of Science
supercomputing facility at Oak Ridge National Lab has a
distributed computing elements model)
9
Projected Science Requirements for Networking
Science Areas
considered in the
Workshop [1]
Today
End2End
Throughput
(not including Nuclear
Physics and
Supercomputing)
5 years
End2End
Documented
Throughput
Requirements
5-10 Years
End2End
Estimated
Throughput
Requirements
Remarks
High Energy
Physics
0.5 Gb/s
100 Gb/s
1000 Gb/s
high bulk throughput
with deadlines (Grid
based analysis
systems require QoS)
Climate (Data &
Computation)
0.5 Gb/s
160-200 Gb/s
N x 1000 Gb/s
high bulk throughput
SNS NanoScience
Not yet
started
1 Gb/s
1000 Gb/s
remote control and
time critical
throughput (QoS)
Fusion Energy
0.066 Gb/s
(500 MB/s
burst)
0.198 Gb/s
(500MB/
20 sec. burst)
N x 1000 Gb/s
time critical
throughput (QoS)
Astrophysics
0.013 Gb/s
(1 TBy/week)
N*N multicast
1000 Gb/s
computational
steering and
collaborations
Genomics Data &
Computation
0.091 Gb/s
(1 TBy/day)
100s of users
1000 Gb/s
high throughput and
steering
10
Feb, 05
Aug, 04
Feb, 04
Aug, 03
Feb, 03
Aug, 02
Feb, 02
Aug,01
Feb, 01
Aug, 00
Feb, 00
Aug, 99
Feb, 99
Aug, 98
Feb, 98
Aug, 97
Feb, 97
Aug, 96
Feb, 96
Aug, 95
Feb, 95
Aug, 94
500
Feb, 94
Aug, 93
Feb, 93
Aug, 92
Feb, 92
Aug, 91
Feb, 91
Aug, 90
Feb, 90
TByte/Month
TBytes/Month
Observed Drivers for the Evolution of ESnet
ESnet is Currently Transporting About 530 Terabytes/mo.
and this volume is increasing exponentially
600
ESnet Monthly Accepted Traffic
Feb., 1990 – May, 2005
400
300
200
100
0
11
Who Generates ESnet Traffic?
ESnet Inter-Sector Traffic Summary, Jan 03 / Feb 04/ Nov 04
72/68/62%
DOE sites
DOE is a net supplier
of data because
DOE facilities are
used by universities
and commercial
entities, as well as by
DOE researchers
21/14/10%
ESnet
~25/19/13%
14/12/9%
17/10/14%
10/13/16%
9/26/25%
4/6/13%
Note
• more than 90% of the ESnet traffic is OSC traffic
• less that 20% of the traffic is inter-Lab
R&E (mostly
universities)
Peering Points
53/49/50%
DOE collaborator traffic, inc. data
Commercial
International
(almost entirely
R&E sites)
Traffic coming into ESnet = Green
Traffic leaving ESnet = Blue
Traffic between ESnet sites
% = of total ingress or egress traffic
12
A Small Number of Science Users Account for
a Significant Fraction of all ESnet Traffic
ESnet Top 100 Host-to-Host Flows, Feb., 2005
Class 1: DOE LabInternational R&E
TBytes/Month
12
10
Class 2: Lab-U.S.
(domestic) R&E
8
6
Class 3: Lab-Lab
(domestic)
Total ESnet traffic
Feb., 2005 = 323 TBy
in approx.
6,000,000,000 flows
All other flows
(< 0.28 TBy/month
each)
4
2
Class 4: Lab-Comm.
(domestic)
Top 100 flows = 84 TBy
0
Notes: 1) This data does not include intra-Lab (LAN) traffic (ESnet ends at the Lab border
routers, so science traffic on the Lab LANs is invisible to ESnet
2) Some Labs have private links that are not part of ESnet - that traffic is not represented here.
Terabytes/Month
DOE Lab-International R&E
Lab-U.S. R&E (domestic)
12
12
10
10
8
8
6
6
4
4
2
2
0
SLAC (US)  RAL (UK)
Fermilab (US)  WestGrid (CA)
SLAC (US)  IN2P3 (FR)
LIGO (US)  Caltech (US)
SLAC (US)  Karlsruhe (DE)
LLNL (US)  NCAR (US)
SLAC (US)  INFN CNAF (IT)
Fermilab (US)  MIT (US)
Fermilab (US)  SDSC (US)
Fermilab (US)  Johns Hopkins
Fermilab (US)  Karlsruhe (DE)
IN2P3 (FR)  Fermilab (US)
LBNL (US)  U. Wisc. (US)
Fermilab (US)  U. Texas, Austin (US)
BNL (US)  LLNL (US)
BNL (US)  LLNL (US)
Fermilab (US)  UC Davis (US)
Qwest (US)  ESnet (US)
Fermilab (US)  U. Toronto (CA)
BNL (US)  LLNL (US)
BNL (US)  LLNL (US)
CERN (CH)  BNL (US)
NERSC (US)  LBNL (US)
DOE/GTN (US)  JLab (US)
U. Toronto (CA)  Fermilab (US)
NERSC (US)  LBNL (US)
NERSC (US)  LBNL (US)
NERSC (US)  LBNL (US)
NERSC (US)  LBNL (US)
CERN (CH)  Fermilab (US)
Source and Destination of the Top 30 Flows, Feb. 2005
Lab-Lab (domestic)
Lab-Comm. (domestic)
14
 Observed Drivers for ESnet Evolution
•
The observed combination of
o
o
exponential growth in ESnet traffic, and
large science data flows becoming a significant fraction of all ESnet
traffic
show that the projections of the science community are
reasonable and are being realized
•
The current predominance of international traffic is due to
high-energy physics
o
o
•
However, all of the LHC US tier-2 data analysis centers are at US
universities
As the tier-2 centers come on-line, the DOE Lab to US university
traffic will increase substantially
High energy physics is several years ahead of the other
science disciplines in data generation
o
Several other disciplines and facilities (e.g. climate modeling and the
supercomputer centers) will contribute comparable amounts of
additional traffic in the next few years
15
DOE Science Requirements for Networking
1) Network bandwidth must increase substantially, not
just in the backbone but all the way to the sites and
the attached computing and storage systems
2) A highly reliable network is critical for science –
when large-scale experiments depend on the
network for success, the network must not fail
3) There must be network services that can guarantee
various forms of quality-of-service (e.g., bandwidth
guarantees) and provide traffic isolation
4) A production, extremely reliable, IP network with
Internet services must support Lab operations and
the process of small and medium scale science
16
 ESnet’s Place in U. S. and International Science
•
ESnet and Abilene together provide most of the
nation’s transit networking for basic science
o
o
•
Abilene provides national transit networking for most of
the US universities by interconnecting the regional
networks (mostly via the GigaPoPs)
ESnet provides national transit networking for the DOE
Labs
ESnet differs from Internet2/Abliene in that
o
o
Abilene interconnects regional R&E networks – it does not
connect sites or provide commercial peering
ESnet serves the role of a tier 1 ISP for the DOE Labs
- Provides site connectivity
- Provides full commercial peering so that the Labs have full Internet
access
17
ESnet and GEANT
•
GEANT plays a role in Europe similar to Abilene and
ESnet in the US – it interconnects the European
National Research and Education Networks, to
which the European R&E sites connect
•
GEANT currently carries essentially all ESnet traffic
to Europe (LHC use of LHCnet to CERN is still
ramping up)
18
Ensuring High Bandwidth, Cross Domain Flows
•
ESnet and Abilene have recently established highspeed interconnects and cross-network routing
•
Goal is that DOE Lab ↔ Univ. connectivity should
be as good as Lab ↔ Lab and Univ. ↔ Univ.
•

Constant monitoring is the key

US LHC Tier 2 sites need to be incorporated
The Abilene-ESnet-GEANT joint monitoring
infrastructure is expected to become operational
over the next several months (by mid-fall, 2005)
19
Monitoring DOE Lab ↔ University Connectivity
AsiaPac
SEA
• Current monitor infrastructure (red&green) and target infrastructure
• Uniform distribution around ESnet and around Abilene
• All US LHC tier-2 sites will be added as monitors
CERN
CERN
Europe
Europe
LBNL
Abilene
FNAL
ESnet
OSU
Japan
Japan
CHI
NYC
DEN
SNV
DC
KC
BNL
IND
Japan
LA
NCS*
SDG
SDSC
ALB
ELP
HOU
DOE Labs w/ monitors
Universities w/ monitors
Initial site monitors
network hubs
high-speed cross connects: ESnet ↔ Internet2/Abilene (
ATL
ESnet
Abilene
*intermittent
scheduled for FY05)
20
One Way Packet Delays Provide a Fair Bit of Information
Normal: Fixed delay from one site to
another that is primary a function of
geographic separation
The result of a
congested tail circuit
to FNAL
The result of a problems
with the monitoring
system at CERN, not
the network
21
Strategy For The Evolution of ESnet
A three part strategy for the evolution of ESnet
1) Metropolitan Area Network (MAN) rings to provide
-
dual site connectivity for reliability
-
much higher site-to-core bandwidth
-
support for both production IP and circuit-based traffic
2) A Science Data Network (SDN) core for
-
provisioned, guaranteed bandwidth circuits to support large, high-speed
science data flows
-
very high total bandwidth
-
multiply connecting MAN rings for protection against hub failure
-
alternate path for production IP traffic
3) A High-reliability IP core (e.g. the current ESnet core) to address
-
general science requirements
-
Lab operational requirements
-
Backup for the SDN core
-
vehicle for science services
22
Strategy For The Evolution of ESnet:
Two Core Networks and Metro. Area Rings
CERN
Asia-Pacific
GEANT
(Europe)
Australia
Science Data Network Core
(SDN) (NLR circuits)
Aus.
Sunnyvale
New York
IP Core
Washington,
DC
LA
Albuquerque
San Diego
IP core hubs
Metropolitan
Area Rings
SDN/NLR hubs
Primary DOE Labs
New hubs
Production IP core (10-20 Gbps)
Science Data Network core (30-50 Gbps)
Metropolitan Area Networks (20+ Gbps)
Lab supplied (10+ Gbps)
International connections (10-40 Gbps)
ESnet MAN Architecture (e.g. Chicago)
core router
R&E peerings
International peerings
T320
ESnet
SDN
core
ESnet
production
IP core
core
router
switches managing
multiple lambdas
Starlight
Qwest
ESnet managed
λ / circuit services
ESnet production
IP service
ESnet
management and
monitoring
2-4 x 10 Gbps
channels
ANL
FNAL
monitor
site equip.
ESnet managed
λ / circuit services
tunneled through
the IP backbone
Site gateway router
Site LAN
monitor
Site gateway router
Site LAN
site equip.
24
First Two Steps in the Evolution of ESnet
1) The SF Bay Area MAN will provide to the five OSC
Bay Area sites
o
Very high speed site access – 20 Gb/s
o
Fully redundant site access
2) The first two segments of the second national
10 Gb/s core – the Science Data Network – are
San Diego to Sunnyvale to Seattle
25
ESnet SF Bay Area
MAN Ring (Sept., 2005)
SDN to Seattle
(NLR)
• 2 λs (2 X 10 Gb/s channels)
in a ring configuration, and
delivered as 10 GigEther
circuits
- 10-50X current site
bandwidth
Joint
Genome
Institute
λ2 SDN/circuits
NERSC
λ3 future
λ4 future
SF Bay
Area
• Will be used as a 10 Gb/s
production IP ring and
2 X 10 Gb/s paths (for
circuit services) to each site
• Project completion date is
9/2005
10 Gb/s
optical channels
λ1 production IP
LBNL
• Dual site connection
(independent “east” and
“west” connections) to each
site
• Qwest contract signed for
two lambdas 2/2005 with
options on two more
IP core to
Chicago (Qwest)
LLNL
SNLL
SLAC
DOE Ultra
Science Net
(research net)
Level 3
hub
Qwest /
ESnet hub
ESnet MAN
ring (Qwest
circuits)
ESnet hubs
and sites
SDN to San
Diego
NASA
Ames
IP core to El Paso
SF Bay Area MAN – Typical Site Configuration
max. of 2x10G
connections on any line
card to avoid switch
limitations
West
λ1 and λ2
0-10 Gb/s
drop-off
IP traffic
Site LAN
site
Site
ESnet
6509
nx1GE
or 10GE
IP
SF BA
MAN
1 or 2 x 10 GE
(provisioned circuits
via VLANS)
East
λ1 and λ2
0-20 Gb/s
VLAN traffic
0-10 Gb/s
pass-through
VLAN traffic
0-10 Gb/s
pass-through
IP traffic
= 24 x 1 GE line cards
= 4 x 10 GE line cards
(using 2 ports max. per
card)
27
Evolution of ESnet – Step One:
SF Bay Area MAN and West Coast SDN
CERN
Asia-Pacific
GEANT
(Europe)
Australia
Science Data Network Core
(SDN) (NLR circuits)
Sunnyvale
Aus.
New York
IP Core (Qwest)
Washington,
DC
LA
San Diego
IP core hubs
Albuquerque
Metropolitan
Area Rings
El Paso
SDN/NLR hubs
Primary DOE Labs
New hubs
In service by Sept., 2005
planned
Production IP core
Science Data Network core
Metropolitan Area Networks
Lab supplied
International connections
 ESnet Goal – 2009/2010
SEA
• 10 Gbps enterprise IP traffic
• 40-60 Gbps circuit based transport
Europe
CERN
Aus.
CERN
Europe
ESnet
Science Data Network
(2nd Core – 30-50 Gbps,
National Lambda Rail)
Japan
Japan
CHI
SNV
NYC
DEN
Europe
AsiaPac
DC
Metropolitan
Area
Rings
Aus.
ALB
SDG
ESnet IP Core
(≥10 Gbps)
ATL
ESnet hubs
New ESnet hubs
ELP
Metropolitan Area Rings
Major DOE Office of Science Sites
High-speed cross connects with Internet2/Abilene
Production IP ESnet core
Science Data Network core
Lab supplied
Major international
10Gb/s
10Gb/s
30Gb/s
40Gb/s
29
 Near-Term Needs for LHC Networking
•
The data movement requirements of the several
experiments at the CERN/LHC are considerable
•
Original MONARC model (CY2000 - Models of
Networked Analysis at Regional Centres for LHC
Experiments – Harvey Newman’s slide, above)
predicted
o Initial need for 10 Gb/s dedicated bandwidth for
LHC startup (2007) to each of the US Tier 1 Data
Centers
- By 2010 the number is expected to 20-40 Gb/s
per Center
o Initial need for 1 Gb/s from the Tier 1 Centers to
each of the associated Tier 2 centers
30
Near-Term Needs for LHC Networking
•
However, with the LHC commitment to Grid based
data analysis systems, the expected bandwidth
and network service requirements for the Tier 2
centers are much greater than the MONARCH bulk
data movement model
o
MONARCH still probably holds for the Tier0 (CERN) –
Tier 1 transfers
o
For widely distributed Grid workflow systems QoS is
considered essential
-
•
Without a smooth flow of data between workflow nodes the
overall system would likely be very inefficient due to stalling the
computing and storage elements
Both high bandwidth and QoS network services
must be addressed for LHC data analysis
31
Proposed LHC high-level architecture
LHC Network Operations Working Group,
LHC Computing Grid Project
32
Near-Term Needs for North American LHC Networking
•
Primary data paths from LHC Tier 0 to Tier 1
Centers will be dedicated 10Gb/s circuits
•
Backup paths must be provided
o
o
•
About day’s worth of data can be buffered at CERN
However, unless both the network and the analysis
systems are over-provisioned it may not be possible to
catch up even when the network is restored
Three level backup strategy
o
o
o
Primary: Dedicated 10G circuits provided by CERN and
DOE
Secondary: Preemptable10G circuits (e.g. ESnet’s SDN,
NSF’s IRNC links, GLIF, CA*net4)
Tertiary: Assignable QoS bandwidth on the production
networks (ESnet, Abilene, GEANT, CA*net4)
33
Proposed LHC high-level architecture
Tier2s
Tier1s
L3 Backbones
Tier0
Main connection
Backup connection
34
LHC Networking and ESnet, Abilene, and GEANT
•
USLHCnet (CERN+DOE funded) supports US participation in
the LHC experiments
o
•
Dedicated high bandwidth circuits from CERN to the U.S. transfer LHC
data to the US Tier 1 data centers (FNAL and BNL)
ESnet is responsible for getting the data from the transAtlantic connection points for the European circuits (Chicago
and NYC) to the Tier 1 sites
o
ESnet is also responsible for providing backup paths from the transAtlantic connection points to the Tier 1 sites
• Abilene is responsible for getting data from ESnet to the Tier 2
sites
•
The new ESnet architecture (Science Data Network) is
intended to accommodate the anticipated 20-40 Gb/s from
LHC to US (both US tier 1 centers are on ESnet)
35
ESnet Lambda Infrastructure and LHC T0-T1 Networking
CANARIE
Seattle
Toronto
Boise
BNL
Clev
New York
Denver
KC
Pitts
FNAL
Wash DC
GEANT-1
Sunnyvale
Chicago
CERN-1 CERN-2
CERN-3
TRIUMF Vancouver
Raleigh
Phoenix
Albuq.
Tulsa
San Diego
Atlanta
Dallas
NLR PoPs
Jacksonville
El Paso Las Cruces
ESnet IP core hubs
ESnet SDN/NLR hubs
Tier 1 Centers
Cross connects with Internet2/Abilene
New hubs
GEANT-2
LA
San Ant.
Houston
Pensacola
Baton Rouge
ESnet Production IP core (10-20 Gbps)
ESnet Science Data Network core (10G/link)
(incremental upgrades, 2007-2010)
Other NLR links
CERN/DOE supplied (10G/link)
International IP connections (10G/link)
36
Abilene* and LHC Tier 2, Near-Term Networking
Vancouver
CERN-1 CERN-2
CERN-3
TRIUMF
CANARIE
Seattle
Toronto
Boise
Clev
BNL
New York
Denver
KC
Pitts
FNAL
Wash DC
GEANT-1
Sunnyvale
Chicago
Raleigh
Phoenix
Albuq.
Tulsa
San Diego
Dallas
Atlanta
GEANT-2
LA
Jacksonville
El Paso Atlas Tier 2 Centers
Pensacola
Las
Cruces
NLR PoPs
• University
of Texas at Arlington
Baton Rouge
Houston
• University
CMS Tier
2 Centers
ESnet of
IP Oklahoma
core hubsNorman
San
Ant.
• University of New Mexico Albuquerque • MIT
ESnet Production IP core (10-20 Gbps)
• Langston
• University of Florida at Gainesville
ESnetUniversity
SDN/NLR hubs
< 10G
connections
to Abilene
ESnet Science Data
Network
core (10G/link)
• University of Chicago
• University of Nebraska at Lincoln
(incremental upgrades,
2007-2010)to USLHC or
Tier University
1 CentersBloomington
10G connections
• Indiana
• University of Wisconsin at Madison
Other NLR links
ESnet
• Boston
• Caltech
CrossUniversity
connects with Internet2/Abilene
CERN/DOE supplied (10G/link)
• Harvard University
• Purdue University
USLHCnet nodes
New
hubs
• University of Michigan
• University of California SanInternational
Diego
37
IP connections
(10G/link)
 QoS - New Network Service
•
New network services are critical for ESnet to meet
the needs of large-scale science like the LHC
•
Most important new network service is dynamically
provisioned virtual circuits that provide
o
Traffic isolation
- will enable the use of high-performance, non-standard transport
mechanisms that cannot co-exist with commodity TCP based
transport
(see, e.g., Tom Dunigan’s compendium
http://www.csm.ornl.gov/~dunigan/netperf/netlinks.html )
o
Guaranteed bandwidth
- the only way that we have currently to address deadline
scheduling – e.g. where fixed amounts of data have to reach sites
on a fixed schedule in order that the processing does not fall
behind far enough so that it could never catch up – very important
for experiment data analysis
38
OSCARS: Guaranteed Bandwidth Service
•
Must accommodate networks that are shared
resources
o
Multiple QoS paths
o
Guaranteed minimum level of service for best effort traffic
o
Allocation management
- There will be hundreds of contenders with different science
priorities
39
OSCARS: Guaranteed Bandwidth Service
•
Virtual circuits must be set up end-to-end across
ESnet, Abilene, and GEANT, as well as the
campuses
o
There are many issues that are poorly understood
o
To ensure compatibility the work is a collaboration with the
other major science R&E networks
- code is being jointly developed with Internet2's Bandwidth
Reservation for User Work (BRUW) project – part of the Abilene
HOPI (Hybrid Optical-Packet Infrastructure) project
- Close cooperation with the GEANT virtual circuit project
(“lightpaths – Joint Research Activity 3 project)
40
policer
authorization
user
system1
shaper
OSCARS: Guaranteed Bandwidth Service
resource
manager
bandwidth
broker
allocation
manager
site A
resource
manager
policer
• To address all of the
issues is complex
-There are many
potential restriction
points
-There are many users
that would like priority
service, which must be
rationed
resource
manager
user
system2
site B
41
Between ESnet, Abilene, GEANT, and the connected regional
R&E networks, there will be dozens of lambdas in production
networks that are shared between thousands of users who
want to use virtual circuits.
similar situation
in Europe
US R&E environment
Federated Trust Services
•
Remote, multi-institutional, identity authentication is
critical for distributed, collaborative science in order
to permit sharing computing and data resources,
and other Grid services
•
Managing cross site trust agreements among many
organizations is crucial for authentication in
collaborative environments
o
•
ESnet assists in negotiating and managing the cross-site,
cross-organization, and international trust relationships to
provide policies that are tailored to collaborative science
The form of the ESnet trust services are driven
entirely by the requirements of the science
community and direct input from the science
community
43
ESnet Public Key Infrastructure
•
ESnet provides Public Key Infrastructure and X.509
identity certificates that are the basis of secure,
cross-site authentication of people and Grid systems
•
These services (www.doegrids.org) provide
o
Several Certification Authorities (CA) with different uses
and policies that issue certificates after validating request
against policy

This service was the basis of the first routine sharing of
HEP computing resources between US and Europe
44
ESnet Public Key Infrastructure
•
ESnet provides Public Key Infrastructure and X.509
identity certificates that are the basis of secure,
cross-site authentication of people and Grid systems
•
The characteristics and policy of the several PKI
certificate issuing authorities are driven by the
science community and policy oversight (the Policy
Management Authority – PMA) is provided by the
science community + ESnet staff
•
These services (www.doegrids.org) provide
o

Several Certification Authorities (CA) with different uses
and policies that issue certificates after validating
certificate requests against policy
This service was the basis of the first routine sharing of
HEP computing resources between US and Europe
45
ESnet Public Key Infrastructure
• Root CA is kept off-line in a vault
• Subordinate CAs are kept in locked,
ESnet root CA
alarmed racks in an access controlled
machine room and have dedicated firewalls
• CAs with different policies as required by
the science community
DOEGrids CA
o DOEGrids CA has a policy tailored to
accommodate international science
collaboration
o NERSC CA policy integrates CA and
certificate issuance with NIM (NERSC
user accounts management services)
NERSC CA
FusionGrid CA
…… CA
o FusionGrid CA supports the FusionGrid
roaming authentication and
authorization services, providing
complete key lifecycle management
46
7500
7250
7000
6750
6500
6250
6000
5750
5500
5250
5000
4750
4500
4250
4000
3750
3500
3250
3000
2750
2500
2250
2000
1750
1500
1250
1000
750
500
250
0
User Certificates
Service Certificates
Expired(+revoked)
Certificates
Total Certificates Issued
Total Cert Requests
Ja
Fe n-0
b 3
M -0
ar 3
A -0
p 3
M r- 0
ay 3
Ju -03
n
Ju -0 3
A l-0
u 3
S g- 0
ep 3
O -0
c 3
N t- 0
ov 3
D -0
ec 3
Ja -0 3
Fe n-0
4
M b- 0
ar 4
A - 04
p
M r- 0
ay 4
Ju -04
n
Ju -0 4
A l-0
u 4
S g- 0
e 4
O p- 0
c 4
N t- 04
ov
D -0
ec 4
Ja -0
4
Fe n-0
b 5
M -0
ar 5
A - 05
M pr- 0
ay 5
Ju -05
n
Ju -0 5
l-0
5
No.of certificates or requests
DOEGrids CA (one of several CAs) Usage Statistics
Production service began in June 2003
User Certificates
1999 Total No. of Certificates
5479
Host & Service Certificates
3461 Total No. of Requests
7006
ESnet SSL Server CA Certificates
38
DOEGrids CA 2 CA Certificates (NERSC)
15
FusionGRID CA certificates
76
* Report as of Jun 15, 2005
47
DOEGrids CA Usage - Virtual Organization Breakdown
DOEGrids CA Statistics (5479)
ANL
3.5%
DOESG
0.3% ESG
0.8%
ESnet
0.4%
FusionGRID
4.8%
iVDGL
18.8%
*Others
41.2%
LBNL
1.2%
“Other” is mostly auto renewal certs (via
the Replacement Certificate interface) that
does not provide VO information
NERSC
3.2%
NCC-EPA
0.1%
LCG
0.6%
ORNL
0.6%
PNNL
0.4%
FNAL
8.9%
PPDG
15.2%
*DOE-NSF
collab.
48
North American Policy Management Authority
• The Americas Grid, Policy Management Authority
• An important step toward regularizing the management of trust in the
international science community
• Driven by European requirements for a single Grid Certificate Authority
policy representing scientific/research communities in the Americas
• Investigate Cross-signing and CA Hierarchies support for the science
community
• Investigate alternative authentication services
• Peer with the other Grid Regional Policy Management Authorities (PMA).
European Grid PMA [www.eugridpma.org ]
o Asian Pacific Grid PMA [www.apgridpma.org]
o
• Started in Fall 2004 [www.TAGPMA.org]
• Founding members
o
o
o
o
o
DOEGrids (ESnet)
Fermi National Accelerator Laboratory
SLAC
TeraGrid (NSF)
CANARIE (Canadian national R&E network)
49
References – DOE Network Related Planning Workshops

1) High Performance Network Planning Workshop, August 2002
http://www.doecollaboratory.org/meetings/hpnpw

2) DOE Science Networking Roadmap Meeting, June 2003
http://www.es.net/hypertext/welcome/pr/Roadmap/index.html
3) DOE Workshop on Ultra High-Speed Transport Protocols and Network
Provisioning for Large-Scale Science Applications, April 2003
http://www.csm.ornl.gov/ghpn/wk2003
4) Science Case for Large Scale Simulation, June 2003
http://www.pnl.gov/scales/
5) Workshop on the Road Map for the Revitalization of High End Computing, June
2003
http://www.cra.org/Activities/workshops/nitrd
http://www.sc.doe.gov/ascr/20040510_hecrtf.pdf (public report)
6) ASCR Strategic Planning Workshop, July 2003
http://www.fp-mcs.anl.gov/ascr-july03spw
7) Planning Workshops-Office of Science Data-Management Strategy, March &
May 2004
o
http://www-conf.slac.stanford.edu/dmw2004
50