Thursday, July 28, 2011

Intel to Acquire Fulcrum Microsystems

Important news! Intel to Acquire Fulcrum Microsystems

Wow. This is a very good move by Intel to acquire Fulcrum Microsystems. I have been very impressed with Fulcrum Microsystems as its Ethernet Switch Chip scores the lowest latency and I'm very impressed as it is closing the gap with the infiniband

Read my other blog entry Watch out Infiniband! Low Latency Ethernet Switch Chips are closing the gap

Wednesday, July 27, 2011

National Institute of Standards and Technology (NIST) Definition of Cloud Computing (Draft)

National Institute of Standards and Technology (NIST) Definition of Cloud Computing (Draft) can be found in this website from NIST. This definition is very clear.
  1. The NIST Definition of Cloud Computing (Draft) (pdf)
The essential services are
  1. On-demand self-service
  2. Broad network access 
  3. Resource pooling. 
  4. Rapid elasticity. 
  5. Measured Service. 
Service Models:
  1. Cloud Software as a Service (SaaS).
  2. Cloud Platform as a Service (PaaS).
  3. Cloud Infrastructure as a Service (IaaS).
Deployment Models:
  1. Private cloud
  2. Community cloud
  3. Public cloud
  4. Hybrid cloud
Read the article for more information

Monday, July 25, 2011

Training Materials for HPC, MPI, Blue-Gene, IBM POWER and more by Livermore Computing Center

Livermore Computing Center has provided a series of very useful online Training Manual information for HPC users and administrator. See Livermore Computing Center High Performance Computing Training

The topics includes

  1. Introduction To Parallel Computing
  2. Introduction to Livermore Computing Resources
  3. Moab
  4. Using the Dawn BGP System
  5. BG/L Tutorials
  6. Using ASC Purple
  7. IBM POWER Systems Overview
  8. Linux Clusters Overview
  9. Message Passing Interface (MPI)
  10. Advanced MPI
  11. POSIX Threads Programming
  12. OpenMP
  13. TotalView Debugger
  14. ASC Tri-Lab Code Development Tools Workshop
  15. IBM HPC Toolkit
  16. A Holistic Approach Towards Automatic Performance Analysis and Tuning
  17. Performance Analysis Tools and Topics for LC'S IBM SP Systems
  18. MPI Performance Topics 
  19. Performance Analysis Tools 
  20. Red Storm / XT3 Workshop (LLNL internal only)
  21. Using Scientific Python
  22. Scientific Python II
  23. Using Python Numeric
  24. MPI Parallel Programming In Python

Sunday, July 24, 2011

Adding Color to the Linux Man Pages on CentOS 5

If you wish to add color to the Linux Manual Page, you have to install a program called "most". To install "most", you have to ensure you have RPMForge in your repository. For more information, see Installing RPMForge

Once you have install RPMForge, just do a
# yum install most

 Package      Arch           Version                     Repository        Size
 most         x86_64         4.10.2-1.2.el5.rf           rpmforge          55 k

Transaction Summary
Install       1 Package(s)
Upgrade       0 Package(s)

Once done, update your .bash_profile
export MANPAGER="/usr/bin/most"

Source the .bashrc_profile
# source .bashrc_profile

Now call us the Linux Man Page, for example "man ls", you will see color on the manual pages

Saturday, July 23, 2011

Setting and changing Torque Server Configuration

After creating a Torque Default Queue as described in the Setting up Torque Server on xCAT 2.x 
, you may want to  change some specific configuration later, you can change it using the "qmgr -c" command. For example, you may want to change the default walltime,the issue this command at your prompt

# qmgr -c "set default_qeue resource_default.walltime=3600:00:00

For more information, you may want to see Torque Resource Manager - 1.4 Manual Setup of Initial Server Configuration

Monday, July 18, 2011

Basic Kickstart on CentOS 5

A simple readable tutorial from my other blog (Linux Cluster) on how to configure Redhat / CentOS kickstart to create another copy of your configuration on the next server

  1. Basic Kickstart on CentOS 5 (Part 1)
  2. Basic Kickstart on CentOS 5 (Part 2)

Happy reading.... :)

Friday, July 15, 2011

How to UnInstall Intel Compilers versions 11 and 12

If you need to reinstall Intel Compilers, you need to uninstall first,

Composer XE 2011 Update  2:
#  /opt/intel/composerxe-2011.2.137/bin/ --default

Compiler Pro 11.1: 
# /opt/intel/Compiler/11.1/038/bin/ia32/ --default

For more information, see Intel Compilers for Linux Versions 11.1 and 12.0 Silent Installation Guide

Wednesday, July 13, 2011

File Transfer Server could not be started or it exited unexpectedly. Exit value 0 was returned

One of our researcher was trying to sftp from his Windows Box to a Linux Server and he encounter this error "File Transfer Server could not be started or it exited unexpectedly. Exit value 0 was returned"

When someone launched a sFTP session, the user start-up scripts such as .bashrc and .bash_profile is evoked. Often the root cause of the error is erronous statement in the .bashrc or .bash_profile.
  1. Often, the users evoked a csh or tcsh while on the default interpreter is bash.
  2. Sometimes it is also due to erronous message which are meant for interative logon like output text messages etc.
With a bit of correction, your sftp will works.

See similar problems like
  1. Resolution for bash orted command not found
  2. Unable to source path during non-interactive ssh session  
  3. bash: scp: command not found

For other notes, see
  1. SSH FAQ on sftp corruption

Tuesday, July 12, 2011

CentOS 6 is out!

The CentOS team is pleased to announce the immediate availability of CentOS-6.0 for i386 and x86_64 Architectures.................

Since upstream has a 6.1 version already released, we will be using a Continous Release repository for 6.0 to bring all 6.1 and post 6.1 security updates to all 6.0 users, till such time as CentOS-6.1 is released itself.

Now for the Cent OS mirrors. Happy Downloading

Friday, July 8, 2011

Watch out Infiniband! Low Latency Ethernet Switch Chips are closing the gap

One of the most important component that affect Ethernet Switch Latency is the Ethernet Switch Chip. You got to pay attention to the tremendous stide made by GE especially 10GE. The market leader and 1 that I'm personally enthuastic of is actually the switch chips from Fulcrum MicroSystem Inc.

Finally a Low Latency Ethernet Switch Chip that is going to put the heat on Infiniband and will help Managers move towards the Single Fabric at the Data Centre.

Here is some extracts:

FocalPoint, a breakthrough 10G and 40G Ethernet switch chip family, adds several new dimensions to ubiquitous Ethernet, including low latency, fabric convergence, standards adaptability, and scalability. The device family delivers the industry's lowest latency: 200nS as a layer-2 switch; 300nS as a full router with all features enabled.

Combined this with Chelsio Communication cutting edge adapter technology and you have a serious contender to Infiniband

Thursday, July 7, 2011

Building HPC Clusters with 10Gigabit Ethernet (White Paper)

This document presents data comparing the performance of two well-known HPC applications and the impact of performance and bench-marked performance between 1GE, 10GE and InfiniBand. This paper is from ARISTA.
  1. Building HPC Clusters with 10Gigabit Ethernet (White Paper)

Wednesday, July 6, 2011

High Performance Cluster for Biomedical Research using 10GB Ethernet iWARP Fabric

This white paper from Intel is on a large research institute that has achieve 36 TeraFLOPS with at least 84% efficiency using HPL benchmark on a cluster of 4032 cores using 10GB Ethernet iWARP Fabric and ARISTA 10GB switches

For more information, see  High Performance Cluster for Biomedical Research using 10GB Ethernet iWARP Fabric

Tuesday, July 5, 2011

Clock skew detected. Your build may be incomplete

Sometimes when you compiling a program in your /home directoryand when you run a make you receive this error

make: :  Clock skew detected.  Your build may be incomplete

This occurs when the Server you are logging on and the NFS Server has discrepancy in their time. The solution is to synchronize their time, best with a time server. For more information, see Configuring NTP Server and Client on CentOS 5.x on how to set up a NTP Server and NTP Client

Monday, July 4, 2011

bash: scp: command not found

I've receive this error when I was trying to scp to a remote machine using root account. I've got this error.
bash: scp: command not found
lost connection

This error is similar to the
  1. Resolution for bash orted command not found and
  2. Unable to source path during non-interactive ssh session
If I interactive directly with the remote machine like ssh into the machine using root, I can enter into the machine. But not with scp.

I'm assuming you have in your /etc/ssh/sshd_config already allows
PermitRootlogin yes

Step 1: To solve this issue, find out where is your scp command resides in your remote computer
# which scp

Step 2: Put the path into your root .bashrc 
export PATH=/usr/bin:{$PATH}