SlideShare ist ein Scribd-Unternehmen logo
1 von 22
Downloaden Sie, um offline zu lesen
This Presentation is protected by

Creative Common License v2.0
Agenda                                   Domain : High Performance Computing
                                         Presentation Level : Beginner
                                         Prerequisite : Familiarity with Linux
-What is a Supercomputer?
-Why do we need it?
-Types of Supercomputers
-The Recipe of building a cluster
          -Basic Concepts
          -Identifying the Application
          -Selection of Raw Materials
          -Preparation
          -Configuration
          -Deployment
          -Testing
          -Maintenance
-Home made vs. commercial clusters
-Making money from clusters
-Other resources and links
What is a Supercomputer?

“ An extremely fast computer that can perform hundreds of millions of instructions
per second.”

-A powerful system built from collection of special purpose hardware
-It is designed for a specific application
-Processing power range is very high
-There is no standard for supercomputer specification
-Works on parallel processing schema
Best Supercomputers of the world
Indian Supercomputers
PARAM Padma is C-DAC's next
generation high performance scalable
computing cluster, currently with a peak
computing power of One Teraflop.



                                   KABRU is a 144 node (Xeon DP) Linux cluster.
                                   Though it is a very fast supercomputer, it is not
                                   the fastest in the world. With a sustained
                                   performance of 1002.3 GFlops(reached on
                                   October 13th 2004) of double precision
                                   arithmetic it is the second fastest supercomputer
                                   in India and the fastest supercomputer in India
                                   belonging to an academic institution.
                                   --IMSc
Why do we need it?

“Obviously, we need it for more processing power!!!”

- We use it where computation can be parallel.
- We use it where “divide and conquer” is prominent in algorithms
- We use it for High Performance/Availability computing
- We use it for distributed computing


In India, CDAC uses supercomputers for research in:
Bioinfomatics, Computational Structural Mechanics , Computational
Atmospheric Science, Evolutionary Computing , Computational Chemistry etc.
Types of Supercomputers

“Two broad categories: Tightly coupled parallel systems & Loosely Coupled
Clusters”

Modern supercomputing clusters:
- High performance (HP) clusters
- Load-leveling clusters
- Web-service clusters
- Storage clusters
- Database clusters


           A special type is Single System Image (SSI) Clusters
The Recipe for building a cluster

“…before you make soup, you need hunger to enjoy it…”

Let’s brush up the basic concepts:
- Linux Installation Basics
- DHCP
- Network Boot (via PXE Boot or Etherboot)
- Interconnect
The Recipe for building a cluster
[Identifying the Application]


“Why would you need a supercomputer? Hey , I need it just for fun!”

Building a cluster for:
- High Performance (HP) need
- High Availability (HA) need
            or
-“just need it for fun”



“There is really lot of fun when you write & test your algorithms on a cluster…”
The Recipe for building a cluster
[Selection of raw material]


“Innovators build great things from non-great elements !”

Selection of Hardware:
- Few number of old mother boards
- Enough number of processors to sit on the boards
- Minimum of 32Mb RAM per board
- Network support via On-Board or External NIC cards
- At least one Hard disk & CD-ROM Drive
- Either BIOS support for Network Boot or A floppy drive for each board


“mean to say, just get few boxes from your friends if you feel lazy to build your own
hardware…”
The Recipe for building a cluster
[Selection of raw material]


“Innovators build great things from non-great elements !”

Selection of software:
- The OpenSSI (http://www.openssi.org)
- OSCAR (http://oscar.sourceforge.net/)
- TFTP
- Etherboot (http://rom-o-matic.net/5.2.4/ )


 “mean to say, just get few boxes from your friends if you feel lazy to build your own
hardware…”
The OpenSSI
                          “the most fantastic product I have ever seen !”




              It has internode communication, clusterwide process management, clusterwide devices, a cluster
filesystem, clusterwide IPC (pipes, fifos, msgqueues, semaphores, etc.) and clusterwide tcp/ip networking.
The Recipe for building a cluster
[Preparation]
“Question : Dedicated Cluster or Temporary Cluster ?”

- A clean install of base OS (Fedora Core 3)
 on PCs that has bootable device
- A clean Network Configuration
The Recipe for building a cluster
[Configuration]


- Download and unpack OpenSSI (http://www.openssi.org)
- Go through the DOCS
- ./install does everything for year
           Enter a clustername.
         Enter a node number between 1 and 125.
         Select a Network Interface Card (``NIC'') for the cluster interconnect.
Select (P)XE or (E)therboot as the network boot protocol for this node.
           Select whether you want to enable root filesystem failover.
- For Adding nodes, run openssi-config-node. Select ``Add a new node''.




                      “ Remember that node 1 is called the init node…”
The Recipe for building a cluster
[Configuration]


- Essentials
# cluster -v
(To check the membership of nodes in cluster with their status)
# bash-ll
(The shell that performs load leveling.
/etc/sysconfig/loadlevellist can also be added to specify special processes to be loadlevelled and run
service loadlevel restart )
# ssi-ksync
(To rebuild the ramdisk to include the driver and update the network boot images)
     # onnode <node_number>                      <command>
(To run a specific command on a specific node)
The Recipe for building a cluster
[Testing]


while(no_of_processes < 1000)
{
    if (fork()==0)
    {
        run_relevent_algorithm();
        exit(1);
    }
    no_of_processes++;
}


“…This piece of code can do wonders…Its fun to keep adding zeros in the loop...”
The Recipe for building a cluster
[Maintenance]


- Make sure there is no IP conflict if network is shared
- Perform regular efficiency audit on network
- Put proper firewall for security




                “ you actually don’t need to bother much for maintenance…”
Home made vs. commercial clusters

               “The obvious difference is in the looks…”
Making money from clusters

                          “Are’nt you interested in this…”


-Host web servers , file servers etc.
-Create supercomputer for fun, give access to yours friends
-Provide low-cost high performance computing facility to research institutes
-Convert offices and academic institutions into night-time research facility
Other resources and links

                          “go ahead find out more…”

Download this presentation and various other interesting things at:
http://www.parolkar.com/download.aspx


Other links:
http://www.openssi.org
http://www.beowulf.org
http://sourceforge.net/projects/ci-linux
http://linux-ha.org
http://www.openmosix.org
Questions


   ?
Thank You!!!

  I am just a click away
abhishek@parolkar.com

Weitere ähnliche Inhalte

Was ist angesagt?

Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster
Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster
Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster
Ceph Community
 
NGENSTOR_ODA_P2V_V5
NGENSTOR_ODA_P2V_V5NGENSTOR_ODA_P2V_V5
NGENSTOR_ODA_P2V_V5
UniFabric
 
ZFS for Databases
ZFS for DatabasesZFS for Databases
ZFS for Databases
ahl0003
 

Was ist angesagt? (20)

Ironic
IronicIronic
Ironic
 
How to Increase Performance of Your Hadoop Cluster
How to Increase Performance of Your Hadoop ClusterHow to Increase Performance of Your Hadoop Cluster
How to Increase Performance of Your Hadoop Cluster
 
Using Recently Published Ceph Reference Architectures to Select Your Ceph Con...
Using Recently Published Ceph Reference Architectures to Select Your Ceph Con...Using Recently Published Ceph Reference Architectures to Select Your Ceph Con...
Using Recently Published Ceph Reference Architectures to Select Your Ceph Con...
 
Walk Through a Software Defined Everything PoC
Walk Through a Software Defined Everything PoCWalk Through a Software Defined Everything PoC
Walk Through a Software Defined Everything PoC
 
Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster
Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster
Ceph Day Shanghai - CeTune - Benchmarking and tuning your Ceph cluster
 
Ceph Day Beijing - Ceph RDMA Update
Ceph Day Beijing - Ceph RDMA UpdateCeph Day Beijing - Ceph RDMA Update
Ceph Day Beijing - Ceph RDMA Update
 
Ceph Day Beijing - SPDK for Ceph
Ceph Day Beijing - SPDK for CephCeph Day Beijing - SPDK for Ceph
Ceph Day Beijing - SPDK for Ceph
 
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
Heterogeneous Computing on POWER - IBM and OpenPOWER technologies to accelera...
 
Ceph Day Melbourne - Walk Through a Software Defined Everything PoC
Ceph Day Melbourne - Walk Through a Software Defined Everything PoCCeph Day Melbourne - Walk Through a Software Defined Everything PoC
Ceph Day Melbourne - Walk Through a Software Defined Everything PoC
 
Modeling, estimating, and predicting Ceph (Linux Foundation - Vault 2015)
Modeling, estimating, and predicting Ceph (Linux Foundation - Vault 2015)Modeling, estimating, and predicting Ceph (Linux Foundation - Vault 2015)
Modeling, estimating, and predicting Ceph (Linux Foundation - Vault 2015)
 
NGENSTOR_ODA_P2V_V5
NGENSTOR_ODA_P2V_V5NGENSTOR_ODA_P2V_V5
NGENSTOR_ODA_P2V_V5
 
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA ArchitectureCeph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
 
MySQL Head-to-Head
MySQL Head-to-HeadMySQL Head-to-Head
MySQL Head-to-Head
 
Building reliable Ceph clusters with SUSE Enterprise Storage
Building reliable Ceph clusters with SUSE Enterprise StorageBuilding reliable Ceph clusters with SUSE Enterprise Storage
Building reliable Ceph clusters with SUSE Enterprise Storage
 
Virtualization Primer for Java Developers
Virtualization Primer for Java DevelopersVirtualization Primer for Java Developers
Virtualization Primer for Java Developers
 
Linux Performance Analysis and Tools
Linux Performance Analysis and ToolsLinux Performance Analysis and Tools
Linux Performance Analysis and Tools
 
ZFS for Databases
ZFS for DatabasesZFS for Databases
ZFS for Databases
 
GPUDirect RDMA and Green Multi-GPU Architectures
GPUDirect RDMA and Green Multi-GPU ArchitecturesGPUDirect RDMA and Green Multi-GPU Architectures
GPUDirect RDMA and Green Multi-GPU Architectures
 
Azure VM 101 - HomeGen by CloudGen Verona - Marco Obinu
Azure VM 101 - HomeGen by CloudGen Verona - Marco ObinuAzure VM 101 - HomeGen by CloudGen Verona - Marco Obinu
Azure VM 101 - HomeGen by CloudGen Verona - Marco Obinu
 
PostgreSQL High Availability in a Containerized World
PostgreSQL High Availability in a Containerized WorldPostgreSQL High Availability in a Containerized World
PostgreSQL High Availability in a Containerized World
 

Andere mochten auch (6)

microprocessor architecture
microprocessor architecture microprocessor architecture
microprocessor architecture
 
Meeting 1 types of computers ppt
Meeting 1 types of computers pptMeeting 1 types of computers ppt
Meeting 1 types of computers ppt
 
introduction to microprocessor and microcomputer
introduction to microprocessor and microcomputerintroduction to microprocessor and microcomputer
introduction to microprocessor and microcomputer
 
Supercomputers
SupercomputersSupercomputers
Supercomputers
 
Introduction History Significance of mainframe computer
Introduction History Significance of mainframe computerIntroduction History Significance of mainframe computer
Introduction History Significance of mainframe computer
 
Super computer
Super computerSuper computer
Super computer
 

Ähnlich wie Building SuperComputers @ Home

Ähnlich wie Building SuperComputers @ Home (20)

Develop Your Own Operating Systems using Cheap ARM Boards
Develop Your Own Operating Systems using Cheap ARM BoardsDevelop Your Own Operating Systems using Cheap ARM Boards
Develop Your Own Operating Systems using Cheap ARM Boards
 
2012-03-15 What's New at Red Hat
2012-03-15 What's New at Red Hat2012-03-15 What's New at Red Hat
2012-03-15 What's New at Red Hat
 
Cobbler, Func and Puppet: Tools for Large Scale Environments
Cobbler, Func and Puppet: Tools for Large Scale EnvironmentsCobbler, Func and Puppet: Tools for Large Scale Environments
Cobbler, Func and Puppet: Tools for Large Scale Environments
 
Cobbler, Func and Puppet: Tools for Large Scale Environments
Cobbler, Func and Puppet: Tools for Large Scale EnvironmentsCobbler, Func and Puppet: Tools for Large Scale Environments
Cobbler, Func and Puppet: Tools for Large Scale Environments
 
GPU and Deep learning best practices
GPU and Deep learning best practicesGPU and Deep learning best practices
GPU and Deep learning best practices
 
Sharing High-Performance Interconnects Across Multiple Virtual Machines
Sharing High-Performance Interconnects Across Multiple Virtual MachinesSharing High-Performance Interconnects Across Multiple Virtual Machines
Sharing High-Performance Interconnects Across Multiple Virtual Machines
 
Intel's Out of the Box Network Developers Ireland Meetup on March 29 2017 - ...
Intel's Out of the Box Network Developers Ireland Meetup on March 29 2017  - ...Intel's Out of the Box Network Developers Ireland Meetup on March 29 2017  - ...
Intel's Out of the Box Network Developers Ireland Meetup on March 29 2017 - ...
 
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
 
RISC V in Spacer
RISC V in SpacerRISC V in Spacer
RISC V in Spacer
 
Big Data Anti-Patterns: Lessons From the Front LIne
Big Data Anti-Patterns: Lessons From the Front LIneBig Data Anti-Patterns: Lessons From the Front LIne
Big Data Anti-Patterns: Lessons From the Front LIne
 
Sanger, upcoming Openstack for Bio-informaticians
Sanger, upcoming Openstack for Bio-informaticiansSanger, upcoming Openstack for Bio-informaticians
Sanger, upcoming Openstack for Bio-informaticians
 
Flexible compute
Flexible computeFlexible compute
Flexible compute
 
HPC and cloud distributed computing, as a journey
HPC and cloud distributed computing, as a journeyHPC and cloud distributed computing, as a journey
HPC and cloud distributed computing, as a journey
 
Build Your Private Cloud with Ezilla and Haduzilla
Build Your Private Cloud with Ezilla and HaduzillaBuild Your Private Cloud with Ezilla and Haduzilla
Build Your Private Cloud with Ezilla and Haduzilla
 
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-PremiseTackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
Tackling Network Bottlenecks with Hardware Accelerations: Cloud vs. On-Premise
 
OSDC 2017 | Open POWER for the data center by Werner Fischer
OSDC 2017 | Open POWER for the data center by Werner FischerOSDC 2017 | Open POWER for the data center by Werner Fischer
OSDC 2017 | Open POWER for the data center by Werner Fischer
 
OSDC 2017 - Werner Fischer - Open power for the data center
OSDC 2017 - Werner Fischer - Open power for the data centerOSDC 2017 - Werner Fischer - Open power for the data center
OSDC 2017 - Werner Fischer - Open power for the data center
 
OSDC 2017 | Linux Performance Profiling and Monitoring by Werner Fischer
OSDC 2017 | Linux Performance Profiling and Monitoring by Werner FischerOSDC 2017 | Linux Performance Profiling and Monitoring by Werner Fischer
OSDC 2017 | Linux Performance Profiling and Monitoring by Werner Fischer
 
Cloud-computing.ppt
Cloud-computing.pptCloud-computing.ppt
Cloud-computing.ppt
 
Setting up repositories
Setting up repositoriesSetting up repositories
Setting up repositories
 

Mehr von Abhishek Parolkar (6)

MyDuniya enterprise offering
MyDuniya enterprise offeringMyDuniya enterprise offering
MyDuniya enterprise offering
 
Ruby for soul of BigData Nerds
Ruby for soul of BigData NerdsRuby for soul of BigData Nerds
Ruby for soul of BigData Nerds
 
Nirvigna - Rendering Hi-Res graphics on commodity cluster
Nirvigna - Rendering Hi-Res graphics on commodity clusterNirvigna - Rendering Hi-Res graphics on commodity cluster
Nirvigna - Rendering Hi-Res graphics on commodity cluster
 
Building SMS Applications with Ruby-SMPP
Building SMS Applications with Ruby-SMPPBuilding SMS Applications with Ruby-SMPP
Building SMS Applications with Ruby-SMPP
 
Beyond Version Controlling Git By Parolkar
Beyond Version Controlling Git By ParolkarBeyond Version Controlling Git By Parolkar
Beyond Version Controlling Git By Parolkar
 
Canvas Tag By Abhishek Parolkar
Canvas Tag By Abhishek ParolkarCanvas Tag By Abhishek Parolkar
Canvas Tag By Abhishek Parolkar
 

Kürzlich hochgeladen

Why Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businessWhy Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire business
panagenda
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Safe Software
 

Kürzlich hochgeladen (20)

Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
 
Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024
 
AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of Terraform
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
 
Top 5 Benefits OF Using Muvi Live Paywall For Live Streams
Top 5 Benefits OF Using Muvi Live Paywall For Live StreamsTop 5 Benefits OF Using Muvi Live Paywall For Live Streams
Top 5 Benefits OF Using Muvi Live Paywall For Live Streams
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
Why Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businessWhy Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire business
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a Fresher
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CV
 

Building SuperComputers @ Home

  • 1. This Presentation is protected by Creative Common License v2.0
  • 2. Agenda Domain : High Performance Computing Presentation Level : Beginner Prerequisite : Familiarity with Linux -What is a Supercomputer? -Why do we need it? -Types of Supercomputers -The Recipe of building a cluster -Basic Concepts -Identifying the Application -Selection of Raw Materials -Preparation -Configuration -Deployment -Testing -Maintenance -Home made vs. commercial clusters -Making money from clusters -Other resources and links
  • 3. What is a Supercomputer? “ An extremely fast computer that can perform hundreds of millions of instructions per second.” -A powerful system built from collection of special purpose hardware -It is designed for a specific application -Processing power range is very high -There is no standard for supercomputer specification -Works on parallel processing schema
  • 5. Indian Supercomputers PARAM Padma is C-DAC's next generation high performance scalable computing cluster, currently with a peak computing power of One Teraflop. KABRU is a 144 node (Xeon DP) Linux cluster. Though it is a very fast supercomputer, it is not the fastest in the world. With a sustained performance of 1002.3 GFlops(reached on October 13th 2004) of double precision arithmetic it is the second fastest supercomputer in India and the fastest supercomputer in India belonging to an academic institution. --IMSc
  • 6. Why do we need it? “Obviously, we need it for more processing power!!!” - We use it where computation can be parallel. - We use it where “divide and conquer” is prominent in algorithms - We use it for High Performance/Availability computing - We use it for distributed computing In India, CDAC uses supercomputers for research in: Bioinfomatics, Computational Structural Mechanics , Computational Atmospheric Science, Evolutionary Computing , Computational Chemistry etc.
  • 7. Types of Supercomputers “Two broad categories: Tightly coupled parallel systems & Loosely Coupled Clusters” Modern supercomputing clusters: - High performance (HP) clusters - Load-leveling clusters - Web-service clusters - Storage clusters - Database clusters A special type is Single System Image (SSI) Clusters
  • 8. The Recipe for building a cluster “…before you make soup, you need hunger to enjoy it…” Let’s brush up the basic concepts: - Linux Installation Basics - DHCP - Network Boot (via PXE Boot or Etherboot) - Interconnect
  • 9. The Recipe for building a cluster [Identifying the Application] “Why would you need a supercomputer? Hey , I need it just for fun!” Building a cluster for: - High Performance (HP) need - High Availability (HA) need or -“just need it for fun” “There is really lot of fun when you write & test your algorithms on a cluster…”
  • 10. The Recipe for building a cluster [Selection of raw material] “Innovators build great things from non-great elements !” Selection of Hardware: - Few number of old mother boards - Enough number of processors to sit on the boards - Minimum of 32Mb RAM per board - Network support via On-Board or External NIC cards - At least one Hard disk & CD-ROM Drive - Either BIOS support for Network Boot or A floppy drive for each board “mean to say, just get few boxes from your friends if you feel lazy to build your own hardware…”
  • 11. The Recipe for building a cluster [Selection of raw material] “Innovators build great things from non-great elements !” Selection of software: - The OpenSSI (http://www.openssi.org) - OSCAR (http://oscar.sourceforge.net/) - TFTP - Etherboot (http://rom-o-matic.net/5.2.4/ ) “mean to say, just get few boxes from your friends if you feel lazy to build your own hardware…”
  • 12. The OpenSSI “the most fantastic product I have ever seen !” It has internode communication, clusterwide process management, clusterwide devices, a cluster filesystem, clusterwide IPC (pipes, fifos, msgqueues, semaphores, etc.) and clusterwide tcp/ip networking.
  • 13. The Recipe for building a cluster [Preparation] “Question : Dedicated Cluster or Temporary Cluster ?” - A clean install of base OS (Fedora Core 3) on PCs that has bootable device - A clean Network Configuration
  • 14. The Recipe for building a cluster [Configuration] - Download and unpack OpenSSI (http://www.openssi.org) - Go through the DOCS - ./install does everything for year Enter a clustername. Enter a node number between 1 and 125. Select a Network Interface Card (``NIC'') for the cluster interconnect. Select (P)XE or (E)therboot as the network boot protocol for this node. Select whether you want to enable root filesystem failover. - For Adding nodes, run openssi-config-node. Select ``Add a new node''. “ Remember that node 1 is called the init node…”
  • 15. The Recipe for building a cluster [Configuration] - Essentials # cluster -v (To check the membership of nodes in cluster with their status) # bash-ll (The shell that performs load leveling. /etc/sysconfig/loadlevellist can also be added to specify special processes to be loadlevelled and run service loadlevel restart ) # ssi-ksync (To rebuild the ramdisk to include the driver and update the network boot images) # onnode <node_number> <command> (To run a specific command on a specific node)
  • 16. The Recipe for building a cluster [Testing] while(no_of_processes < 1000) { if (fork()==0) { run_relevent_algorithm(); exit(1); } no_of_processes++; } “…This piece of code can do wonders…Its fun to keep adding zeros in the loop...”
  • 17. The Recipe for building a cluster [Maintenance] - Make sure there is no IP conflict if network is shared - Perform regular efficiency audit on network - Put proper firewall for security “ you actually don’t need to bother much for maintenance…”
  • 18. Home made vs. commercial clusters “The obvious difference is in the looks…”
  • 19. Making money from clusters “Are’nt you interested in this…” -Host web servers , file servers etc. -Create supercomputer for fun, give access to yours friends -Provide low-cost high performance computing facility to research institutes -Convert offices and academic institutions into night-time research facility
  • 20. Other resources and links “go ahead find out more…” Download this presentation and various other interesting things at: http://www.parolkar.com/download.aspx Other links: http://www.openssi.org http://www.beowulf.org http://sourceforge.net/projects/ci-linux http://linux-ha.org http://www.openmosix.org
  • 22. Thank You!!! I am just a click away abhishek@parolkar.com