1 / 49

Building a Beowulf: My Perspective and Experience

Building a Beowulf: My Perspective and Experience. Ron Choy Lab. for Computer Science MIT. Ver 1.02. First of all …. Why do we care? Make informed purchase decision Know how to evaluate Build one yourself!. Outline. History/Introduction Hardware aspects Software aspects

Télécharger la présentation

Building a Beowulf: My Perspective and Experience

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Building a Beowulf:My Perspective and Experience Ron Choy Lab. for Computer Science MIT Ver 1.02

  2. First of all … • Why do we care? • Make informed purchase decision • Know how to evaluate • Build one yourself!

  3. Outline • History/Introduction • Hardware aspects • Software aspects • Our class Beowulf • Beowulf design exercise

  4. What is a Beowulf ? • Massively parallel computer built out of COTS • Runs a free operating system (not Wolfpack, MSCS) • Connected by high speed interconnect • Compute nodes are dedicated (not Network of Workstations)

  5. Who uses Beowulfs? • Pharmaceutical companies • Investment firms • Animation makers • Me and you

  6. The Beginning • Thomas Sterling and Donald Becker CESDIS, Goddard Space Flight Center, Greenbelt, MD • Summer 1994: built an experimental cluster • Called their cluster Beowulf

  7. The First Beowulf • 16 x 486DX4, 100MHz processors • 16MB of RAM each, 256MB in total • Channel bonded Ethernet (2 x 10Mbps) • Not that different from our Beowulf

  8. The First Beowulf (2)

  9. Current Beowulfs • Faster processors, faster interconnect, but the idea remains the same • Cluster database: http://clusters.top500.org/db/Query.php3 • Top cluster: 1.920 11.06 TFLOPS peak

  10. Current Beowulfs (2)

  11. Why Beowulf? • It’s cheap! • Our Beowulf, 18 processors, 9GB RAM: $15000 • A Sun Enterprise 250 Server, 2 processors, 2GB RAM: $16000 • Everything in a Beowulf is open-source and open standard - easier to manage/upgrade

  12. Essential Components of a Beowulf • Processors • Memory • Interconnect • Software

  13. Processors • Major vendors: AMD, Intel • AMD: Athlon MP • Intel: Pentium 4

  14. Comparisons • Athlon MP and P4 are close in performance • One is better for some applications and the other is better for other applications – it all depends on what you want • However, Athlon MP is cheaper

  15. Comparisons (2) • P4 supports SSE2 instruction set, which perform SIMD operations on double precision data (2 x 64-bit) • Athlon MP supports only SSE, for single precision data (4 x 32-bit)

  16. Memory • DDR RAM (double data rate) – used mainly by Athlons, P4 can use them as well • RDRAM (Rambus DRAM) – used by P4s

  17. Memory Bandwidth • Good summary: http://www6.tomshardware.com/mainboard/02q1/020311/sis645dx-03.html • DDR beats out RDRAM in bandwidth, and is also cheaper

  18. Interconnect • The most important component • Factors to consider • Bandwidth • Latency • Price • Software support

  19. Ethernet • Relatively inexpensive, reasonably fast and very popular. • Developed by Bob Metcalfe and D.R. Boggs at Xerox PARC • A variety of flavors (10Mbps, 100Mbps, 1Gbps)

  20. Pictures of Ethernet Devices

  21. Myrinet • Developed by Myricom • “OS bypass”, the network card talks directly to host processes • Proprietary, but very popular because of its low latency and high bandwidth • Usually used in high-end clusters

  22. Myrinet pictures

  23. Comparison

  24. Cost Comparison • To equip our Beowulf with: • Fast ethernet: ~$1700 • Gigabit ethernet: ~ $5600 • Myrinet: ~$17300

  25. How to choose? • Depends on your application! • Requires really low latency e.g. QCD? Myrinet • Requires high bandwidth and can live with higher latency e.g. ScaLAPACK? Gigabit ethernet • Embarrassingly parallel? Anything

  26. What would you gain from fast interconnect? • Our cluster: Single fast ethernet (100Mbps) • 36.8 GFLOPS peak, HPL: ~12 GFLOPS • 32.6% efficiency • GALAXY: Gigabit ethernet • 20 GFLOPS peak, HPL: ~7 GFLOPS • 35% efficiency *old, slow tcp/ip stack!* • HELICS: Myrinet 2000 • 1.4 TFLOPS peak, HPL: ~864 GFLOPS • 61.7% efficiency

  27. My experience with hardware • How long did it take for me to assemble the 9 machines? 8 hours, nonstop

  28. Real issue 1 - space • Getting a Beowulf is great, but do you have the space to put it? • Often space is at a premium, and Beowulf is not as dense as traditional supercomputers • Rackmount? Extra cost! e.g. cabinet ~$1500, case for one node ~$400

  29. Real issue 2 – heat management • The nodes, with all the high powered processors and network cards, run hot • Especially true for Athlons - can reach 60°C • If not properly managed, the heat can cause crash or even hardware damage!

  30. Real issue 3 - power • Do you have enough power in your room? • UPS? Surge protection? • You don’t want a thunderstorm to fry your Beowulf! • For our case we have a managed machine room - lucky

  31. Real issue 4 - noise • Beowulfs are loud. Really loud. • You don’t want it on your desktop. Bad idea

  32. Noise (2) • You might want to consider using ‘quiet’ power supplies and a diskless architecture • Or a special purpose ‘personal cluster’

  33. Real issue 5 - cables • Color scheme your cables!

  34. Software • We’ll concentrate on the cluster management core • Three choices: • Vanilla Linux/FreeBSD • Free cluster management software (a very patched up Linux) • Commercial cluster management software (very very patched up Linux, with technical support)

  35. The issues • Beowulfs can get very large (100’s of nodes) • Compute nodes should setup themselves automatically • Software updates must be automated across all the nodes • Software coherency is an issue

  36. Vanilla Linux • Most customizable, easiest to make changes • Easiest to patch • Harder for someone else to inherit the cluster – a real issue • Need to know a lot about Linux to properly setup

  37. Free cluster management softwares • Oscar: http://oscar.sourceforge.net/ • Rocks: http://rocks.npaci.edu • MOSIX: http://www.mosix.org/ • (usually patched) Linux that comes with software for cluster management • Reduces dramatically the time needed to get things up and running • Open source, but if something breaks, you have one more piece of software to hack

  38. Commercial cluster management • Scyld: www.scyld.com - founded by Donald Becker • Scyld – father of Beowulf • Sells a heavily patched Linux distribution for clustering, free version available but old • Based on bProc, which is similar to MOSIX

  39. My experience/opinions • I chose Rocks because I needed the Beowulf up fast, and it’s the first cluster management software I came across • It was a breeze to setup • But now the pain begins … severe lack of documentations • I have actually stripped the cluster of all Rocks features now -> almost a plain RH7.1

  40. Experience (2) • Also a batch system like OpenPBS http://www.openpbs.org is a must in a multi-user environment

  41. Software (cont’d) • Note that I skipped a lot of details: e.g. file system choice (NFS? PVFS?), MPI choice (MPICH? LAM?), libraries to install … • I could talk forever about Beowulfs but it won’t fit in one talk

  42. Recipe we used for our Beowulf • Ingredients: $15000, 3 x 6 packs of coke, 1 grad student • Web surf for 1 weeks, try to focus on the Beowulf sites, decide on hardware • Spend 2 days filling in various forms for purchasing and obtaining “competitive quotes” • Wait 5 days for hardware to arrive, meanwhile web surf some more, and enjoy the last few days of free time in a while

  43. Recipe (cont’d) 4. Lock grad student, hardware (not money), and coke in an office. Ignore scream. The hardware should be ready after 8 hours. Office of the future

  44. Recipe (cont’d 2) 5. Move grad student and hardware to its final destination. By this time grad student will be emotionally attached to the hardware. This is normal. Have grad student set up software. This would take 2 weeks.

  45. Our Beowulf

  46. Things I would have done differently • Plain Linux • Color scheme the cables! • Try a diskless setup (saves on cost and management – but no local scratch space) • Get rackmount

  47. Design a $30000 Beowulf • One node (2 processors, 1GB RAM) costs $1400, with 4.6 GFLOPS peak • Should we get: • 16 nodes, with fast ethernet, or • 8 nodes, with Myrinet?

  48. Design (cont’d) • 16 nodes with fast ethernet: • 73.6 GFLOPS peak • 23.99 GFLOPS real (using the efficiency of our cluster) • 16 GB of RAM • 8 nodes with Myrinet • 36.8 GFLOPS peak • 22.7 GFLOPS real (using the efficiency of HELICS) • 8 GB of RAM

  49. Design (cont’d 2) • First choice is good if you work on embarrassingly parallel problems which does not require much communication • Second choice is more general purpose

More Related