Cloud Computing Bof

Cloud Computing BOF OGF22 Birds of a Feather Session Hyatt Regency Cambridge February 27 2008 Geoffrey Fox Indiana University [email protected]

© 2007 Open Grid Forum

Cloud Agenda • • • • •

Geoffrey Fox (Indiana U.) Remarks on Cloud Computing Martin Swany (Internet2) Clouds and Dynamic Networking Steven Newhouse (Microsoft) Personal View on Clouds Kate Keahey (Argonne, Chicago) First Steps in the Clouds Next Steps

© 2007 Open Grid Forum

2

What are Clouds? • Clouds are “Virtual Clusters” (“Virtual Grids”) of possibly “Virtual Machines” • They may cross administrative domains or may “just be a single cluster”; the user cannot and does not want to know

• Clouds support access (lease of) computer instances

• Instances accept data and job descriptions (code) and return results that are data and status flags

• Each Cloud is a “Narrow” (perhaps internally proprietary) Grid • When does Cloud concept work

• Parameter searches, LHC style data analysis .. • Common case (most likely success case for clouds) versus corner case?

• Clouds can be built from Grids • Grids can be built from Clouds © 2007 Open Grid Forum

3

Cloud References • http://en.wikipedia.org/wiki/Cloud_computing

• Includes references to Amazon, Apple, Dell, Enomalism, Globus, Google, IBM, KnowledgeTreeLive, Nature, New York Times, Zimdesk • Others like Microsoft Windows Live Skydrive important

• http://en.wikipedia.org/wiki/Amazon_Elastic_Compute_Cloud • http://uc.princeton.edu/main/index.php?option=com_content&task Policy Issues • http://www.cra.org/ccc/home.article.bigdata.html • Hadoop (MapReduce) and “Data Intensive Computing” • See Data intensive computing minitrack at HICSS-42 January 2009

• http://ianfoster.typepad.com/blog/2008/01/theres-grid-in.html • OGF Thought Leadership blog

• OGF22 talks by Charlie Catlett and Irving Wladawsky-Berger © 2007 Open Grid Forum

4

Big-Data Computing Study Group

CCC Role Versus OGF?

© 2007 Open Grid Forum

Hadoop and MapReduce are “just” workflow?

5

Google MapReduce Simplified Data Processing on Clusters/Clouds • http://labs.google.com/papers/mapreduce.html • This is a dataflow model between services where services can do useful document oriented data parallel applications including reductions • The decomposition of services onto cluster engines (clouds) is automated • The large I/O requirements of datasets changes efficiency analysis in favor of dataflow • Services (count words in example) can obviously be extended to general parallel applications • There are many alternatives to language expressing either dataflow and/or parallel operations and/or workflow

© 2007 Open Grid Forum

6

Technical Questions about Clouds I • What is performance overhead? • On individual CPU • On system including data and program transfer

• What is cost gain • From size efficiency; “green” location (rumor that Google has purchased the Niagara Falls including Canada!)

• Is Cloud Security adequate: can clouds be trusted? • Can one can do parallel computing on clouds? • Looking at “capacity” not “capability” i.e. lots of modest sized jobs • Marine corps will use Petaflop machines – they just need ssh and a.out © 2007 Open Grid Forum

7

Technical Questions about Clouds II • How is data compute affinity tackled in clouds? • Co-locate data and compute clouds? • Lots of optical fiber i.e. “just” move the data?

• What happens in clouds when demand for resources exceeds capacity – is there a multi-day job input queue? • Are there novel cloud scheduling issues?

• Do we want to link clouds (or ensembles as atomic clouds); if so how and with what protocols • Is there an intranet cloud e.g. “cloud in a box” software to manage personal (cores on my future 128 core laptop) department or enterprise cloud? © 2007 Open Grid Forum

8

Standards for Compute and Storage Clouds

• We no longer need interoperability of services and messages (SOAP) but rather interoperability of clouds • Maybe each cloud so big that interoperability between clouds not so critical

• Interoperability certainly for application specific data and perhaps also for job specifications • WFS, GML for Geo-data; IVOA standards; DST LHC experiment formats • JSDL, BES etc.

• Each Cloud will be proprietary but they might want raw infrastructure standards so they can easily swap in and out different vendor’s disk drives • Clouds very very loosely coupled; services loosely coupled © 2007 Open Grid Forum

9

MSI Challenge Problem

• There are > 330 MSI’s – Minority Serving Institutions • 2 examples

• ECSU is a small state university in North Carolina

• HBCU with 4000 students • Working on PolarGrid (Sensors in Arctic/Antarctic linked to “TeraGrid”)

• Navajo Tech in Crown Point NM is community college with technology leadership for Navajo Nation • “Internet to the Hogan and Dine Grid” links Navajo communities by wireless • Wish to integrate TeraGrid science into Navajo Nation education curriculum

• Current Grid technology too complicated if you are not an R1 institution • Hard to deploy campus grids broadly into MSI’s • Clouds provide virtual campus resources? © 2007 Open Grid Forum

10

Next Steps at OGF • Clouds are just starting and build on/are related to Grids • Clear need for best practice in use and technology • Likely to be need for new standards and novel use of existing/projected standards • New Cloud Community Group? • Chairs, participants?

• Workshop? • OGF23 activity? • Identify key players not currently involved with OGF? © 2007 Open Grid Forum

11