Upload
ian-foster
View
831
Download
1
Tags:
Embed Size (px)
DESCRIPTION
Keynote talk at the 3rd International Conference on Supercomputing in Mexico: www.isum.mx. A great group of people!A substantially revised version of a talk with the same title given on previous occasions.
Citation preview
www.ci.anl.govwww.ci.uchicago.edu
Accelerating data-driven discoveryby outsourcing the mundane
Ian Foster
www.ci.anl.govwww.ci.uchicago.edu
The data deluge
www.ci.anl.govwww.ci.uchicago.edu
3
The data deluge in biology
x105 in 6 years
x10 in 6 years
www.ci.anl.govwww.ci.uchicago.edu
4
Number of sequencing machines
http://omicsmaps.com/
www.ci.anl.govwww.ci.uchicago.edu
5
18 ordersof magnitudein 5 decades!12 orders of
magnitudein 6 decades
Moore’s Law for X-ray sources
Credit: Linda Young
www.ci.anl.govwww.ci.uchicago.edu
6
Exploding data volumes in astronomy
100,000 TB
MACHO et al.: 1 TB
Palomar: 3 TB2MASS: 10 TBGALEX: 30 TBSloan: 40 TB
Pan-STARRS: 40,000 TB
www.ci.anl.govwww.ci.uchicago.edu
7
Exploding data volumes in climate science
Climate model intercomparisonproject (CMIP) of the IPCC
2004: 36 TB
2012: 2,300 TB
www.ci.anl.govwww.ci.uchicago.edu
8
Big science has been successful
All build on NSF OCI (& DOE)-supported Globus Toolkit software
LIGO: 1 PB data in last science run, distributed worldwide
ESG: 1.2 PB climate datadelivered to 23,000 users; 600+ pubs
OSG: 1.4M CPU-hours/day, >90 sites, >3000 users, >260 pubs in 2010
Robust production solutionsSubstantial teams and expenseSustained, multi-year effortApplication-specific solutions, built on common technology
www.ci.anl.govwww.ci.uchicago.edu
9
Small science is struggling
More data, more complex dataAd-hoc solutionsInadequate software, hardwareData plan mandates
www.ci.anl.govwww.ci.uchicago.edu
10
Dark data in the long tail of science
Awarded Amount 2007
$0
$1,000,000
$2,000,000
$3,000,000
$4,000,000
$5,000,000
$6,000,000
$7,000,000
1 586 1171 1756 2341 2926 3511 4096 4681 5266 5851 6436 7021 7606 8191 8776
NSF grant awards, 2007 (Bryan Heidorn)
www.ci.anl.govwww.ci.uchicago.edu
11
The challenge of staying competitive
"Well, in our country," said Alice … "you'd generally get to somewhere else — if you run very fast for a long time, as we've been doing.”
"A slow sort of country!" said the Queen. "Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!"
www.ci.anl.govwww.ci.uchicago.edu
12
A crisis that demands new approaches
• We have exceptional infrastructure for the 1% (e.g., supercomputers, Large Hadron Collider, …)
• But not for the 99% (e.g., the vast majority of the 1.8M publicly funded researchers in the EU)
We need new approaches to providing research cyberinfrastructure, that:— Reduce barriers to entry— Are cheaper— Are sustainable
www.ci.anl.govwww.ci.uchicago.edu
13
You can run a company from a coffee shop
www.ci.anl.govwww.ci.uchicago.edu
14
Because businesses outsource their IT
Web presence Email (hosted Exchange) Calendar Telephony (hosted VOIP) Human resources and payroll Accounting Customer relationship mgmt
Software as a Service
(SaaS)
www.ci.anl.govwww.ci.uchicago.edu
15
And often their large-scale computing too
Web presence Email (hosted Exchange) Calendar Telephony (hosted VOIP) Human resources and payroll Accounting Customer relationship mgmt Data analytics Content distribution
Infrastructure as a Service
(IaaS)
Software as a Service
(SaaS)
www.ci.anl.govwww.ci.uchicago.edu
16
Let’s rethink how we provide research IT
Accelerate discovery and innovation worldwide by providing research IT as a service
Leverage the cloud to• provide millions of researchers with
unprecedented access to powerful tools; • enable a massive shortening of cycle times in
time-consuming research processes; and• reduce research IT costs dramatically via
economies of scale
www.ci.anl.govwww.ci.uchicago.edu
17grail.cs.washington.edu
www.ci.anl.govwww.ci.uchicago.edu
18
Cloud layers
18
Software as a Service: SaaS
Platform as a Service: PaaS
Infrastructure as a Service: IaaS
www.ci.anl.govwww.ci.uchicago.edu
19
Common research data management steps
• Dark Energy Survey• Galaxy genomics• LIGO observatory
• SBGrid structural biology consortium• NCAR climate data applications• Land use change; economics
www.ci.anl.govwww.ci.uchicago.edu
20
Common research data management steps
• Dark Energy Survey• Galaxy genomics• LIGO observatory
• SBGrid structural biology consortium• NCAR climate data applications• Land use change; economics
www.ci.anl.govwww.ci.uchicago.edu
21
Scientific data delivery, 2012
• “[A] majority of users at BES facilities … physically transport data to a home institution using portable media … data volumes are going to increase significantly in the next few years (to 70 TB/day or more) – data must be transferred over the network”
• “the effectiveness of data transfer middleware [is] not just on the transfer speed, but also the time and interruption to other work required to supervise and check on the success of large data transfers”
• “It took two weeks and email traffic between network specialists at NERSC and ORNL, sys-admins at NERSC, … and combustion staff at ORNL and SNL to move 10 TB from NERSC to ORNL”
[ESNet Network Requirements Workshops, 2007-2010]
Major usability, productivity, performance problems
1980
www.ci.anl.govwww.ci.uchicago.edu
22
The challenge: Moving big data easily
What should be trivial …
… can be painfully tedious and time-consuming
“I need my data over there – at my _____”
( supercomputing center, campus server,
etc.)
Data Source
Data Destination
! Config issues
! Unexpected failure = manual retry
Data Source
Data Destination
“GAAAH!
%&@#&” ! Firewall issues
GO PICTURE
www.ci.anl.govwww.ci.uchicago.edu
24
Globus Online: Data transfer as SaaS• Reliable file transfer.
– Easy “fire-and-forget” transfers– Automatic fault recovery– High performance– Across multiple security domains
• No IT required.– Software as a Service (SaaS)
o No client software installationo New features automatically available
– Consolidated support & troubleshooting– Works with existing GridFTP servers– Globus Connect solves “last mile problem”
• >4000 registered users, >3 Petabytes moved
Recommended by XSEDE, NERSC, Blue Waters, and many campuses
www.ci.anl.govwww.ci.uchicago.edu
25
Dark Energy Survey use of Globus Online• Dark Energy Survey
receives 100,000 files each night in Illinois
• They transmit files to Texas for analysis … then move results back to Illinois
• Process must be reliable, routine, and efficient
• They outsource this task to Globus Online
Image credit: Roger Smith/NOAO/AURA/NSF
Blanco 4m on Cerro Tololo
www.ci.anl.govwww.ci.uchicago.edu
26
www.ci.anl.govwww.ci.uchicago.edu
27
www.ci.anl.govwww.ci.uchicago.edu
28
Integration with Earth System Grid
28
High-speed transfersAutomated retriesWorks behind firewallsCredential managementTransfer monitoring
www.ci.anl.govwww.ci.uchicago.edu
29
Globus Online under the covers
User Hub manages user identities and profilesGroup Hub manages groups and policiesResource Hub for resource definitions
www.ci.anl.govwww.ci.uchicago.edu
30
Globus Online under the covers
User Hub manages user identities and profilesGroup Hub manages groups and policiesResource Hub for resource definitions
Monitoring and controlAuto-tuning of transfer parametersDetection & attempted correction of errorsManual intervention when required
www.ci.anl.govwww.ci.uchicago.edu
31
Globus Online under the covers
User Hub manages user identities and profilesGroup Hub manages groups and policiesResource Hub for resource definitions
Monitoring and controlAuto-tuning of transfer parametersDetection & attempted correction of errorsManual intervention when required
Reliable cloud-based infrastructureEC2 for transfer managementS3 for system stateSimpleDB for lock managementReplication across availability zones
www.ci.anl.govwww.ci.uchicago.edu
32
Globus Online under the covers
User Hub manages user identities and profilesGroup Hub manages groups and policiesResource Hub for resource definitions
Monitoring and controlAuto-tuning of transfer parametersDetection & attempted correction of errorsManual intervention when required
Reliable cloud-based infrastructureEC2 for transfer managementS3 for system stateSimpleDB for lock managementReplication across availability zones
www.ci.anl.govwww.ci.uchicago.edu
33
• Dark Energy Survey• Galaxy genomics• LIGO observatory
• SBGrid structural biology consortium• NCAR climate data applications• Land use change; economics
Towards “research IT as a service”
www.ci.anl.govwww.ci.uchicago.edu
34
Towards “research IT as a service”
www.ci.anl.govwww.ci.uchicago.edu
35
Commercial storage service
provider
National research center
Campus computing
center
Globus Storage: For when you want to …
• Place your data where you want
• Access it from anywhere via different protocols
• Update it, version it,and take snapshots
• Share versions with who you want
• Synchronize among locations
Globus Storage volume
GridFTP, HTTP, WebDAV
www.ci.anl.govwww.ci.uchicago.edu
36
Globus Collaborate: For when you want to
Join with a few or many people to:• Share documents• Track tasks• Send email• Share data • Do whateverWith:• Common groups• Delegated mgmt
www.ci.anl.govwww.ci.uchicago.edu
37
Globus Integrate: For when you want to
Write programs that access/manage user identities, profiles, groups, resources—and data …
… via REST APIs and command line programs
Globus Integrate• Transfer API available• User profile, group APIs in alpha• APIs for Storage, Collaborate
planned after app release
Globus Connect Multi User
Globus Connect
Globus Transfer• In production use• Service and Web
UI enhancements continue
Globus Storage• Early release
available in March• Generally
available in Q3
Globus Collaborate
• Initial projects starting in March
• Early release sometime in Q3
www.ci.anl.govwww.ci.uchicago.edu
38
Other innovative science SaaS projects
www.ci.anl.govwww.ci.uchicago.edu
39
Other innovative science SaaS projects
www.ci.anl.govwww.ci.uchicago.edu
40
Other innovative science SaaS projects
www.ci.anl.govwww.ci.uchicago.edu
41
Other innovative science SaaS projects
www.ci.anl.govwww.ci.uchicago.edu
42
Realizing the benefits of cloud services
• Understand what services researchers really need
• Acquire and sustain the expertise required to create and operate useful services
• Incentivize those who produce services that are widely adopted
• Provide excellent network connectivity
www.ci.anl.govwww.ci.uchicago.edu
43
On the importance of networks
“80 percent of success is showing up”
www.ci.anl.govwww.ci.uchicago.edu
44
Time required to move 10 Terabytes
10 30 100 300 1000 3000 10000 30000 100000 300000 10000000.01
0.10
1.00
10.00
100.00
1,000.00
10,000.00
Network speed in Megabits/sec
Hou
rs to
tran
sfer
10
Tera
byte
s
www.ci.anl.govwww.ci.uchicago.edu
45
Time required to move 10 Terabytes
10 30 100 300 1000 3000 10000 30000 100000 300000 10000000.01
0.10
1.00
10.00
100.00
1,000.00
10,000.00
Network speed in Megabits/sec
Hou
rs to
tran
sfer
10
Tera
byte
s
2 hours US R1 Universities
www.ci.anl.govwww.ci.uchicago.edu
46
Time required to move 10 Terabytes
10 30 100 300 1000 3000 10000 30000 100000 300000 10000000.01
0.10
1.00
10.00
100.00
1,000.00
10,000.00
Network speed in Megabits/sec
Hou
rs to
tran
sfer
10
Tera
byte
s
2 hours US R1 Universities10 mins Upgrade
www.ci.anl.govwww.ci.uchicago.edu
47
Time required to move 10 Terabytes
10 30 100 300 1000 3000 10000 30000 100000 300000 10000000.01
0.10
1.00
10.00
100.00
1,000.00
10,000.00
Network speed in Megabits/sec
Hou
rs to
tran
sfer
10
Tera
byte
s
2 hours US R1 Universities10 mins Upgrade
1 month Cinvestav Langebio
www.ci.anl.govwww.ci.uchicago.edu
48
A 21st C research cyberinfrastructure
LL
LL
L
LL
L
LL
L
LL
L
LL
L
LL
L
LL
L
LL
L
LP P P P
Research data management Collaboration, computationResearch administration
• To providemore capability formore people at less cost …
• Create cloud-based services– Robust and universal– Economies of scale– Positive returns to scale
• Via the creative use of– Aggregation (“cloud”)– Federation (“grid”)
• Powered by networks
Small and medium laboratories and projects
aaS
P
www.ci.anl.govwww.ci.uchicago.edu
49
Questions for you
• How much “dark data” exists in your field? How important is that data?
• Can you quantify the scale, in your field, of– Wasted resources due to duplicated effort– Delays in research progress due to inadequate
infrastructure?• If you could do one thing to accelerate adoption
of advanced computing within your field, what would it be?
www.ci.anl.govwww.ci.uchicago.edu
50
Acknowledgments
Colleagues at UChicago and ArgonneSteve Tuecke, Ravi Madduri, Kyle Chard, Tanu Malik, Rachana Ananthakrisnan, Raj Kettimuthu,and others listed at www.globusonline.org/about/goteam/
NSF Office of CyberinfrastructureDOE Office of Advanced Scientific Computing Res.National Institutes of Health
www.ci.anl.govwww.ci.uchicago.edu
51
For more information
Attend GlobusWorld in Chicago, April 10-12, 2012• www.globusonline.org• Twitter: @globusonline, Globus Online on Facebook• Foster, I. Globus Online: Accelerating and democratizing
science through cloud-based services. IEEE Internet Computing(May/June):70-73, 2011.
• Allen, B., Bresnahan, J., Childers, L., Foster, I., Kandaswamy, G., Kettimuthu, R., Kordas, J., Link, M., Martin, S., Pickett, K. and Tuecke, S. Software as a Service for Data Scientists. Communications of the ACM, Feb, 2012.
www.ci.anl.govwww.ci.uchicago.edu
Thank you!
[email protected]@anl.gov
www.globusonline.orgTwitter: @globusonline, @ianfoster