Transcript Slides

Networking for LHC
Status Report 4/2005
Klaus Ullmann, DFN Verein
TNLC, EGEE April 21th 2005
Contents
• Task for national and international research
networks (NRENs/GEANT)
• Proposed high-level architecture for LHC
networking
• Some details
• Conclusions for EGEE1+2
Thanks go to...
• E.-J. Bos (Surfnet) who volunteered to draft
the report,
• E. Martelli and P. Moroni (both CERN-IT)
who worked as „chief editor“ besides their
technical contributions ...
• and the whole task force who worked hard
since Jan 2005
History and mission for LHC Network
• January 20 & 21, 2005 meeting in
Amsterdam chaired by D. Foster:
– Presentations by the experiments
– Presentations by some network orgs
– Conclusion: Move from bottom up to top down
design
– Consensus on small task force for proposing
LHC high-level network architecture
– Restriction to T0-T1 because no requirements
from T2 available
First steps to the architecture
• Assumptions:
– High-volume data streams
– Continuous data streams
– Keep It Simple
• Stay as low in the stack as you can
• Security Issues
– Important to address security concerns already in the
design phase
– architecture will be kept as protected as possible from
external access
– at least in the beginning, access from trusted sources
(i.e. LHC prefixes) will not be restricted
Proposed architecture (1)
• Optical Private Network, consisting of
dedicated 10G paths between T0 and each
T1, two flavors:
– “Light path T1”
– “Routed T1”
• Special measures for back-up for T0-T1, to
be filled-in later
• T0 preferred interface is 10Gbps Ethernet
LAN-PHY
Proposed architecture (2)
Proposed architecture (3)
Current Status of T1s
Name of T1 LP/Routed Networks
ASCC
BNL
Light Path
CNAF
FNAL
IN2P3
Light Path
Routed
Routed
Light Path
ASNet, NetherLight, GÉANT2 ?
ESnet, LHCnet*
GÉANT2, GARR
ESnet, LHCnet*
RENATER3
Ligth Path
GridKa
SARA
Light Path
NorduGrid Light Path
GÉANT2, X-WiN
GÉANT2, SURFnet6
GÉANT2, NORDUnet, Nordic NRNs
PIC
RAL
TRIUMF
GÉANT2, RedIRIS
GÉANT2, UKLight, SuperJANET4
CA*net 4, Netherlight
Light Path
Light Path
Light Path
* = CALTECH-CERN transatlantic links
Fibre in the LHC network in Europe
-SE
NIKHEF-NL
RAL-UK
GridKa-DE
CERN
df
-ES
IN2P3-FR
-IT
A word on future growth
• Some light path math:
– 10 Gbit/s ~ 1014 byte/day or 100 Tbyte/day
– Eleven 10G light paths -> more than 1
petabyte/day or roughly half an exabyte/year
• In case a 10G is not sufficient:
– Order a second 10G between T0 and T1
– Preferably on a separate physical path
– Architecture fully allows for this
Planning
• Start date for LHC production traffic is 06/07
• T1s are encouraged to proceed with
provisioning well before that date, ideally
already within 2005
• Nevertheless, T1s must be ready at full
bandwidth not later than Q1 2006
GÉANT2 - Big Users (Examples)
• LHC
• 11 Tier1 sites
» 7 in Europe
» 4 outside Europe (US, Canada and Taiwan)
• EVN (European VLBI Network)
• 15 sites
» 5 already connected
• MUPBED
• DEISA
Conclusions for GEANT2
- GÉANT2 is designed to meet demands of
emerging new high BW applications.
- It aims to support e2e, multi-domain MBS
[“lightpath”] signalling and setup.
- It will need policy adjustments during the
next years according to new scenarios put in
place.
Items for further discussion
• Agree with T0 about the physical interface for the T0-T1 link
• Verify that the proposed addressing set-up is compatible
with the grid software (e.g. can the servers be grouped in
the same CIDR block?)
• Inform T0 about the AS number used
• Check if it is possible to establish an environment without
default route
• Verify if the proposed security model is compatible with the
Grid applications
• Decide a backup strategy in case an alternate path at full
speed is not available: tolerate a few hours stop or prefer
low performance on general purpose research backbones.
Conclusions for EGEE1+2
Benefits of using the new fiber networks:
• closed fibre rings enable possibility of
protection on router level
• possibility of using specific transport
protocols
• security problems are easy to solve
• „Heavy“ Grid traffic „decoupled“ from the
general purpose IP network - much easier to
manage
• economic efficient