IBM Dynamic System Analysis (DSA) v3.10 (Portable) for RHEL 5 - IBM BladeCenter and System x

IBM DSA Portable Edition is a system information collection and analysis tool that is used by IBM xSeries(R) Service and Support personnel to aid in the diagnosis of system problems.

1. Please go to: http://www-947.ibm.com/systems/support/supportsite.wss/docdisplay?lndocid=MIGR-5081460&brandind=5000008
2. Open a linux shell prompt (or command line session)
3. Login ar root (by using "su" or "sudo")
4. Change the file permission to become executable
   # chmod +x thefileyoudownloadded.bin
5. Run the ".bin" file and follow the instructions
   #./thefileyoudownloadded.bin
6. Please locate the folder /var/log/IBM_SUPPORT and copy the ".gz" file in the folder and send it back to IBM

Encountering VGL Error - Could not open display 0

[VGL] NOTICE: Automatically setting VGL_CLIENT environment variable to
[VGL] xxx.xxx.xxx.xxx, the IP address of your SSh client.
No protocol specified
[VGL] ERROR: Could not open display :0.

Do note that for VirtualGL has to connect :0 to access the 3D Graphic Card. Apparently, VGL cannot get access to the display :0



Possible resolution 1:
Is the X Server up? X Server should be up to allow connection to display :0



Possible resolution 2:
Granting access the 3D X Server Read  Granting Access to the 3D X Server from VirtualGL Documentation



Possible resolution 3:
After configuring access to the 3D X Server and booting to init 5, you should log off completely and log in for the VirtualGL test. I did not log-off and was wondering why I have the above error.

There may be other resolutions.....Will document when it comes along

Using xdpyinfo to display information utility for X

According to the manual page, xdpyinfo is a useful utility for diaplaying information about an X server. It can be used to examine

1. Capabilities of the server
2. Predefined values for various parameters in communicating between clients and the server
3. Different types of screen and visuals available

Commands to use

# xdpyinfo -display :0

Using TurboVNC 0.6 and VirtualGL 2.1.4 to run OpenGL Application Remotely on CentOS

For more information on Using TurboVNC 0.6 and VirtualGL 2.1.4 to run OpenGL Application Remotely on CentOS on Linux Cluster

“VirtualGL is an open source package which gives any Unix or Linux remote display software the ability to run OpenGL applications with full 3D hardware acceleration…..With VirtualGL, the OpenGL commands and 3D data are instead redirected to a 3D graphics accelerator on the application server, and only the rendered 3D images are sent to the client machine……” For more information see What is VirtualGL? from Project Website"

Using glxgears to report Frame Per Second

glxgears is an OpenGL program that reports FPS (frames per second) numbers. I find this program useful to test out OpeGL Renderer such as simple comparison between VNC, VirtualGL and of course the Direct-From-The-Machine itself.

For TurboVNC (unencrypted) churned out about on a Local LAN

# glxgears
6028 frames in 5.1 seconds = 1186.369 FPS
6000 frames in 5.1 seconds = 1178.267 FPS
5760 frames in 5.1 seconds = 1132.999 FPS
5760 frames in 5.1 seconds = 1133.431 FPS
5520 frames in 5.0 seconds = 1096.104 FPS
5147 frames in 5.0 seconds = 1027.200 FPS
5160 frames in 5.1 seconds = 1016.540 FPS
5160 frames in 5.1 seconds = 1016.316 FPS
5160 frames in 5.1 seconds = 1016.578 FPS
5160 frames in 5.1 seconds = 1016.974 FPS
5160 frames in 5.1 seconds = 1018.695 FPS

For SSH and Xming churned out about on a Local LAN. But the glxgears image refresh is definitely slower than TurboVNC even though the FPS is similar.

# glxgears
6761 frames in 5.0 seconds = 1345.283 FPS
7022 frames in 5.2 seconds = 1343.998 FPS
7021 frames in 5.2 seconds = 1337.664 FPS
6632 frames in 5.1 seconds = 1288.340 FPS
6632 frames in 5.1 seconds = 1297.666 FPS
6760 frames in 5.2 seconds = 1309.137 FPS
6632 frames in 5.2 seconds = 1266.929 FPS
7022 frames in 5.3 seconds = 1330.755 FPS
6631 frames in 5.1 seconds = 1309.255 FPS
6632 frames in 5.0 seconds = 1314.089 FPS
6891 frames in 5.3 seconds = 1305.207 FPS
6631 frames in 5.1 seconds = 1291.670 FPS

For Direct-From-The-Machine. This is the best performance

# glxgears
61237 frames in 5.0 seconds = 12247.371 FPS
65314 frames in 5.0 seconds = 13062.798 FPS
60291 frames in 5.0 seconds = 12058.135 FPS
64627 frames in 5.0 seconds = 12925.391 FPS
64976 frames in 5.0 seconds = 12995.109 FPS
64949 frames in 5.0 seconds = 12989.766 FPS
64961 frames in 5.0 seconds = 12992.117 FPS
64953 frames in 5.0 seconds = 12990.459 FPS
64920 frames in 5.0 seconds = 12983.886 FPS

For VirtualGL on VGL Image Transport with a Direct X11 Connection. Performance is very good near Direct-From-The-Machine

6559 frames in 5.0 seconds = 1311.784 FPS
6573 frames in 5.0 seconds = 1314.600 FPS
6548 frames in 5.0 seconds = 1309.470 FPS
6547 frames in 5.0 seconds = 1309.399 FPS
6545 frames in 5.0 seconds = 1308.945 FPS
6544 frames in 5.0 seconds = 1308.787 FPS
6527 frames in 5.0 seconds = 1305.391 FPS
6527 frames in 5.0 seconds = 1305.336 FPS

Hadoop Communites in Facebook

Hadoop Communites in Facebook

1. Hadoop Group in Facebook
2. Singapore Hadoop Group in Facebook

Welcome.conf and Apache Default web Page

In CentOS and Fedora, the default Document Root is /var/www/html. If the user does not have a index.htm page in their directory, Apache will show a web page "Apache 2 Test Page" Power by CentOS

I think you may ask where is the web page coming from since there is no starting web page. To locate this "
"starting page", go to

# less /etc/httpd/conf.d/welcome.conf

The easier way to "remove" the Default Web Page is by creating a starting web page or a redirect web page

There are unfinished transactions remaining. You might consider running yum-complete-transaction first to finish them

When I was using yum to install a package, I've received a

"There are unfinished transactions remaining. You might consider running yum-complete-transaction first to finish them. The program yum-complete-transaction is found in the yum-utils package."

Apparently, there were aborted or incomplete yum transactions on the system.

First thing First Make sure you have yum-utils package

# yum install yum-utils

Next,run the yum-complete-transaction. You may want to run the yum-transaction --cleanup which is a safe option

# yum-complete-transaction --cleanup-only

Encountering CHECK_NRPE: Error - Could not complete SSL handshake ...

My Linux Platform is CentOS 5.x.

When you encounter CHECK_NRPE: Error - Could not complete SSL handshake ... If you encounter this error when using the nagios check_nrpe plugin and the monitored server, it could be due to the following issues.

1. You will have only an SSL problem if the SSL version used by the check_nrpe and the NRPE daemon are incompatible
2. Access permission for nrpe.cfg on the monitored server is blocking access. It should have read access rw_r__r__
3. Permission error within the  /etc/nagios/nrpe.cfg file.
   

   allowed_hosts=127.0.0.1,192.168.x.x

4. You will also receive the error if the Nagios Server cannot access the NRPE serice at all through the xinetd.d service. Check the /etc/xinetd.d/nrpe service
   
   

   {
   ....
   user = nagios
   group = nagios
   server = /usr/sbin/nrpe
   server_args = -c /etc/nagios/nrpe.cfg --inetd
   log_on_failure += USERID
   disable = no
   only_from = 127.0.0.1,192.168.x.x
   }

5. If all else fail, check the /var/log/messages
   

   # grep nrpe /var/log/messages

Once successful, if you do a ./check_nrpe

# cd /usr/lib64/nagios/plugins
# ./check_nrpe -H monitored_server, 
NRPE v2.12

How to know Hyperthreading is activated from CPU Information generated by dmidecode

According to Intel® Hyper-Threading Technology

"Intel® Hyper-Threading Technology (Intel® HT Technology) is available on laptop, desktop, server, and workstation systems. Look for systems with the Intel® HT Technology logo which your system vendor has verified utilize Intel® HT Technology".

As System Administrator, how do we detect the Hyperthreading in Linux Server? Which CPU are physical and which is logical? For those running HPC, this could be important as you are more interested in number of physical CPU rather than Logical CPU count.

For more information on dmidecode. See dmidecode - Finding the hardware details remotely

Well here come dmidecode to the rescue:

Physical CPU Result (Notice the Status: "Status: Populated, Enabled")

Physical CPU
Handle 0x000C, DMI type 4, 32 bytes
Processor Information
Socket Designation: Socket 1 CPU 1
Type: Central Processor
Family: Xeon
Manufacturer: GenuineIntel
ID: 43 0F 00 00 01 03 00 00
Signature: Type 0, Family 15, Model 4, Stepping 3
Flags:
FPU (Floating-point unit on-chip)
CX8 (CMPXCHG8 instruction supported)
APIC (On-chip APIC hardware supported)
Version: Intel Xeon
Voltage: 1.5 V
External Clock: 200 MHz
Max Speed: 4000 MHz
Current Speed: 3800 MHz
Status: Populated, Enabled
Upgrade: ZIF Socket
L1 Cache Handle: 0x0004
L2 Cache Handle: 0x0005
L3 Cache Handle: Not Provided

Logical CPU Result (Notice the Status: "Status: Unpopulated")

Handle 0x000D, DMI type 4, 32 bytes
Processor Information
Socket Designation: Socket 2 CPU 2
Type: Unknown
Family: Unknown
Manufacturer: Not Specified
ID: 00 00 00 00 00 00 00 00
Version: Not Specified
Voltage: 1.5 V
External Clock: 200 MHz
Max Speed: 4000 MHz
Current Speed: 3800 MHz
Status: Unpopulated
Upgrade: ZIF Socket
L1 Cache Handle: 0x0006
L2 Cache Handle: 0x0007
L3 Cache Handle: Not Provided

Installing and checking /proc/mount using Nagios Plugins on CentOS 5.x

This Plugin is very useful to capture all the values of /proc/mount and display in % free. And all can be display with a single command plugin-in. For more information, take a look at Installing and checking /proc/mount using Nagios Plugins on CentOS 5

Incorporating PNP 0.4.x (PNP is not Perfparse) with Nagios 3 and CentOS 5

This Blog Entry is on Incorporating PNP 0.4.x (PNP is not Perfparse) with Nagios 3 and CentOS 5. PNP.

PNP4Nagios (English) is an addon to nagios which analyzes performance data provided by plugins and stores them automatically into RRD-databases.

For more information, read on Incorporating PNP 0.4.x (PNP is not Perfparse) with Nagios 3 and CentOS 5 on Linux Cluster

Installing Check Disk IO Plugins via NRPE on CentOS 5.x

This Blog Entry is on installing Check Disk IO Plugins for Nagios 3.x. You can learn how to monitor Disk I/O on your remote Server. For more information go to Installing Check Disk IO Plugins via NRPE on CentOS 5.x at Linux Cluster

Issues when installing NagVis to Nagios on CentOS 5.x

Installing Nagvis on Nagios 3.x is quite a challenge on CentOS 5.x. Still in the process of doing it. But I thought I record some of the notes during installation of Nagvis prerequisties.

Before you install the Nagvis prerequisties, do install the following:

1. Install mySQL database

# yum install mysql mysql-devel mysql-server gcc-c++

2. Install PHP and required PHP modules

# yum install php* php-gd php php-mbstring php-mysql

It is useful to install NDOutils as Nagios' NDOutils store all configuration and event data from Nagios into a database, which will help facilidate quicker retrieval and processing event data. Lots of other Nagios-related tools (e.g. NagVis rely on the MySQL backend made available by NDOUtils. Do refer to this excellent article NDOutils on CentOS


3. Compiling Nagvis on CentOS 5.x
Finally, here the unfinished portion, first go to Nagvis Download site
Untar the Nagvis package and install

# tar -zxvf nagvis-1.5.1.tar.gz
# cd nagvis-1.5.1
# ./install.sh -B /usr/bin/nagios

You want to take a look at Install and Configure NagVis although the instruction is for Debian Derivative

/usr/bin/gxl2dot conflicts between attempted installs of graphviz-2.12-8.el5.i386 and graphviz-2.22.0-4.el5.rf.x86_64

When I was installing GraphViz. Using the information taken from Installing Graphviz on CentOS and Fedora

# yum install graphviz

Transaction Error: file /usr/bin/gxl2dot conflicts between attempted installs of graphviz-2.12-8.el5.i386 and graphviz-2.22.0-4.el5.rf.x86_64

To resolve the above error, install the

# yum install graphviz.86_64

Install Graphviz on CentOS and Fedora

Graphviz is open source graph visualization software.

Graph visualization is a way of representing structural information as diagrams of abstract graphs and networks. Automatic graph drawing has many important applications in software engineering, database and web design, networking, and in visual interfaces for many other domains.

To install GraphViz, download the correct repository according to the distribution of your choice. The supported distribution is Fedora, RHEL, Ubuntu, Windows, Mac

The easiest way to install and maintain graphviz on RHEL or Centos is to use yum. To set up yum, download the graphviz-rhel.repo file

To install  Graphviz,

# yum install graphviz*

Install and use Eclipse for Linux on POWER

Learn to write C++and  Java on Eclipse

"Learn how to install and use Eclipse specifically for Linux® running on IBM® POWER™ processor-based systems. Learn, also, how to use Eclipse to compile and run applications through sample Java™ and C programs."

For more information read the article "Install and use Eclipse for Linux on POWER" by IBM

Using Intel Compilers with Eclipse IDE on Linux

Intel has come up with a dated documentation on Using Intel Compilers with Eclipse IDE on Linux. Still useful to give an idea of how to integrate the intel compilers to Eclipse

Intel® C++ Compiler for Linux* - Using Intel® Compilers with the Eclipse* IDE [PDF]

Users encountering Invalid Configuration Location on Eclipse

After installing the Eclipse software as according to Eclipse Integrated Development Environment ( IDE ) for Linux, your users may encounter error after they type "eclipse"

The error found Locking is not possible in the directory

"/usr/local/eclipse/configuration/org.eclipse.osgi. A common reason is that the file system or Runtime Environment does not support file locking for that location. Please choose a different location,or disable file locking passing "=Dosgi.lockling=none" as a VM argument. 'usr/local/eclipse/configuration/org.eclipse.orgi/.manager/.fileTableLock (Permission denied)"

To solve the issue, you simply have to initiate a flag "-configuration" and point the location to your home directory

$ eclipse -configuration < location >

Eclipse Integrated Development Environment ( IDE ) for Linux

What is Eclipse?

1. It is a vendor-neutral open source workbench for multi-language development.
2. Extensible platform for tool integration
3. Plugin-in based framework to create, integrate and utilise software tools.

It has several packages which includes

1. Eclipse IDE for Java EE Developers
2. Eclipse IDE for Java PHP Developers
3. Eclipse IDE for C/C++ Developers
4. Eclipse IDE for JavaScript Web Developers
5. Eclipse IDE for C/C++ Linux Developers (includes Incubating components)
6. ....and many more

Installing eclipse on CentOS 5.x. First you need Java 1.5 and above

# yum install java*

Download the Linux version from Eclipse Download. In my case, I downloaded the "Eclipse IDE for C/C++ Developers"

# tar -zxvf eclipse-cpp-helios-linux-gtk-x86_64.tar.gz

If you wish to put eclipse in a common location for common use, put the extracted directory to /usr/local/ or any place of your choice. After placing the eclipse, remember to path it

# touch /etc/profile.d/eclipse.sh

At /etc/profile.d/eclipse.sh

# PATH=$PATH:/usr/local/eclipse

Source eclipse.sh

# source /etc/profile.d/eclipse.sh

Launch the eclipse application

# eclipse

Alacarte Gnome Menu Editor

Alarcarte is Gnome Menu Editor. A sniffy little tool to help you manage your Gnome Menu. To install, just type

# yum install alacarte

You will be able to manipulate menus

Reading generated logs in Real Time

Step 1: To help troubleshoot quicker, you may want to view the log in real time

# less /var/log/messages

Step 2: Press "F"

.....
.....
Waiting for data... (interrupt to abort)

Step 3: To exit, press
Ctrl+C
Alternatively, you can scroll forward or backward using "d" and "b" respectively.

Unable to move albums for Gallery

I was trying to move albums by clicking "move album" link and the pop-up windows show nothing at least for Internet Explorer. there was no change of website name or any other configuration change. The first thing I did was to take a look at the log file at

# less /var/log/httpd/error_log

*Click F so that you can see the log in real time

You will see an error

[client 192.168.1.2] PHP Fatal error: Allowed memory size of 16777216 bytes exhausted.....

To solve the issue,

# vim /etc/php.ini

memory_limit = 32M

Restart httpd

# service httpd restart

Article on hdparm

An  interesting article Speeding up Linux Using hdparm by Rob Flickenger

A interesting article on the history, roadmap and architecture of xCAT

From Clusters To Clouds: xCAT 2 Is Out Of The Bag from Linux Magazine written by Egan Ford

Installing Wireless Driver on HP Mini 2140 with Fedora 13

Installing Fedora 13 on HP Mini Note 2140 is a breeze except that for the Broadcom 4312G 802.11b/g Wi-Fi Adapter will not be installed. To install the driver for Broadcom 4312G 802.11b/g Wi-Fi Adapter will not be installed, follow the excellent tutorial on Fedora Broadcom Linux STA Driver .

Step 1: Get a wired connection to the Internet.

Step 2: Configure RPM Fusion. Download both free and non-free repository

# rpm -Uvh http://download1.rpmfusion.org/free/fedora/rpmfusion-free-release-stable.noarch.rpm

# rpm -Uvh http://download1.rpmfusion.org/nonfree/fedora/rpmfusion-nonfree-release-stable.noarch.rpm 

Step 3: Install the Broadcom wireless driver package

# yum install broadcom-wl

Step 4: Finally Try loading the driver

# modprobe wl

If the modprobe doesn't work, just reboot

# reboot

Using GParted to resize partition

The GParted application is a graphical partition editor for

1. Creating, reorganizing, and deleting disk partitions. GParted allows you to perform actions on partition such as resize or move (while preserving data) , create, labelling and deleting of partition.
2. The supported file system includes ext3, ext4, NTFS, linux-swap, FAT16/FAT32
3. Enable and Disable partition

GParted can be found on most distribution. For Fedora, you can install by

# yum install GParted

Using GParted is quite easy to use. If you need more information, you can use GParted Live Howto Guides

1. How to put GParted live on USB
2. How to put GParted live on a PXE server
3. How to put GParted live on a hard drive
4. How to create GParted live from scratch

You can also use GParted partitioning software - Full tutorial

Accessing Linux Ext2 and Ext3 from Windows

Explore2fs is a GUI explorer tool for accessing ext2 and ext3 filesystems. It can read ext2 and ext3 filesystem

Using TestDisk to conduct emergency recovery for Linux and Windows

Testdisk by CGSecurity is a useful toolkit to help. According to the TestDisk Website,

"TestDisk is a powerful free data recovery software! It was primarily designed to help recover lost partitions and/or make non-booting disks bootable again when these symptoms are caused by faulty software, certain types of viruses or human error (such as accidentally deleting a Partition Table). Partition table recovery using TestDisk is really easy."

"In addition, it can

1. Fix partition table, recover deleted partition
2. Recover FAT32 boot sector from its backup
3. Rebuild FAT12/FAT16/FAT32 boot sector
4. Fix FAT tables
5. Rebuild NTFS boot sector
6. Recover NTFS boot sector from its backup
7. Fix MFT using MFT mirror
8. Locate ext2/ext3 Backup SuperBlock
9. Undelete files from FAT, NTFS and ext2 filesystem
10. Copy files from deleted FAT, NTFS and ext2/ext3 partitions. "

Testdisk is very useful expecially if you are dual-booting from both Linux and Windows and if accidentally deletion. a few listed tutorial from a much larger list.

1. Recover Damaged Hard Disks (bad sectors)
2. TestDisk Step-by-Step
3. Recovery of CD/DVD 

Useful Voltaire and IBM Links for OFED Drivers

Useful Voltaire and IBM Links for OFED Drivers

1. Voltaire GridStack - IBM BladeCenter
2. Voltaire OFED Version 1.4 User Manual (pdf)
3. Adding Infinifband to a stateless node by xCAT 

Useful Information to compile Infiniband with OpenMPI

Taken from the excellent OpenMPI FAQ. "How do I build Open MPI with support for Open IB (Infiniband), mVAPI (Infiniband), GM (Myrinet), and/or MX (Myrinet)?"

To compile OpenMPI against Infiniband
you only have to specify the directory where its support header files and libraries were installed to Open MPI's configure script

--with-openib=
Build support for OpenFabrics (previously known as "Open IB", for Infiniband and iWARP networks -- note that iWARP support was added in the v1.3 series).

# ./configure --with-openib=/path/to/openib/installation

For more information on myrinet etc, do go to the FAQs.