A CentOS NetInstall, or network installation, is the process of installing a full working CentOS machine using a very small image, around 10MB, downloaded from the CentOS website. In this post I will show you the process to follow for an easy and complete CentOS Netinstall .

Get the ISO

The first step is, of course, downloading the ISO needed for the CentOS NetInstall, you can get it directly from CentOS website here the complete mirror list for x64 based images and here the 32bit version, the file is named CentOS-5.5-x86_64-netinstall.iso.

Prepare to install

Once you’ve downloaded the CentOS NetInstall image you should burn it to a CD, I will be using VmWare for this post, so that you will be able to boot your machine from it.

Install the Server

Once you have booted the server you will be presented with the standard boot screen on which just hit Enter and proceed to select various options (Language to use, Keyboard layout etc.) :

After choosing the basic options you will be asked to choose the installation method to use, of course we are not going to use the local media which contains just the essentials file but a network protocol like HTTP in this example :

Next you will be asked to configure your network interfaces/settings, if you are using a DHCP in your infrastructure you can simply hit Enter and accept defaults :

Once your machine has received the correct IP configuration you will be asked to give instruction on the FTP or HTTP site to be used to get the installation files. Of course here I’m using an example IP address and directory, in a production environment this should point to the actual server and directory hosting the full CentOS image used for NetInstall :

Once you click ok the installer will try to contact the server distribution point and will download the needed files to start the installation process :

From this point on the rest is like a standard installation via CD/DVD so I won’t describe the process here. The total time it took me to install CentOS on the virtual machine was around 30 minutes, keep in mind the machine I’ve been using is a pretty slow one, and went on without any particular problem.

I’m a big fan of Network installation as it usually saves me the hassle to find the installation media and put into the cd swearing if for some reasons it does not work

Hope you’ll find the article useful and please retweet if it was so!

Cheer Lethe.

In the continuous effort of serving the community with resources to prepare for RHCE exam I’m releasing a new quiz/exercise on the RHCE practice lab series. This RHCE practice lab will focus on the troubleshooting part of the exam which is the one scaring most candidates.

RHCE Practice Lab – Troubleshoot boot issues

I’ve created a small script that will render your system unbootable, so please use caution and don’t use this on a production system just on a virtual machine that you can “sacrifice”.

At the end of the post you’ll find a small Python script, today I felt like using Python instead of Bash, launch it and wait for the reboot of the machine which should hang (this is planned), force a reboot of the computer and try to make it work again.

For this particular RHCE practice Lab you don’t have any time constraints but you should be able to have the system up and running in around 20 minutes.

I will repeat myself once again this script is to be used only as an educational resource as part of the RHCE practice lab posts series, it could and will render your system unusable.

To make the simulation more realistic try to solve the problems introduced byt the script without checking its code.

How to use the RHCE Practice Lab

Downaload the attached files, for security reasons I had to rename its extension, save it and rename to break.py make it executable and launch it or simply issue :

python break.py

Once the script has done it will create fixme.txt under the /root/ home folder with some instructions in it and it will reboot your machine.

As always please let me know if you face any problem with solving this RHCE practice, in the following days I will publish the solution to this practice lab.

Cheer Lethe.

This is something I’ve always told, never got tired of it, to my students “Guys always measure twice and cut once. It’s better to spend one more week in the planning phase than having to troubleshoot problems once you have implemented something”. I think this is a basic principal of IT that can easily be applied to every other job or task out there.

It’s pretty simple, if you spend time planning something the right way you’ll get back 10 x the same time when something goes wrong, as if you are reading this blog you know stuff in IT can and will go wrong.

So here’s the story, a customer with a DHCP server running in Linux in a plain flat LAN, no VLANs no IGP nothing, basically one of the many “grown systems” where you start just with a bunch of clients and maybe 1 or 2 servers and one day you discover you have hundreds of clients and tens of server in your infrastructure… how do you call this? For me it’s lack of planning…

There is no easy escape from this kind of situation, usually the “faster” way out is just stop a second and rethink everything from scratch. I know this could sound like as the hardest and longest way to reach a stable solution but it is the one that will lead you to success without too much pain.

Back to the story, via had to decommission this setup to migrate everything under Windows, it’s a pretty standard and easy task when you have proper documentation and people that actually know how their network is setup… It is a nightmare if you work in a blind way without any documentation and without knowledge of what is hiding behind the corner. You switch off a service and BAM everything falls on you and you are forced to quickly restore the previous condition.

I’ll make this short, some people think planning and writing documentation is boring, trust me it is not if you approach it with the right attitude, or not needed “I can remember anything” but trust me the difference between success and disaster is just that PROPER planning. I’ve been in the industry for so much time I’ve seen situations like this too many times and it is never funny.

Think about it the next time someone asks you to write some documentation, it could save your life.

Take this as a free advice from some who have learned this the hard way.

Cheer Lethe.

Good morning all, i'm sorry for the long silence due to work problems and to my relocation to a new house, I'm almost back in my usual workplace (I've been out of town for some time) and finally had back my internet connection in the new house, what a shame as a network engineer I had to wait more than 2 weeks for a local supplier to finally send me my DSL router without which I could not connect :S

Hopefully between today and tomorrow I'll finally be able to post the third part of the series on how to emulate Junos on GNS3 so stay tuned!

Cheer Lethe.

  Copyright secured by Digiprove © 2010

Originally posted 2010-02-23 09:40:06. Republished by Blog Post Promoter

So you want to start monitoring your growing network and heard about Nagios but have no idea how to install it, great I will show you. I won’t cover the installation from source code as it is already well described in the official documentation.

Configure Repositories

The first step will be installing the right repositories to make our life easier and yum priorities so that we’ll not run into any problem with the official reposittories. Open a console and input :

 yum install yum-priorities 

Go to /etc/yum/pluginconf.d/priorities.conf and make sure that is contains the following lines :

Now open all the .repo files and configure priorities by adding the line :

 priority=N 

Where N is an arbitrary value between 1 and 99, below you can find the suggested values for each repository but feel free to test and “explore” yourself :

<pre>[base], [addons], [updates], [extras] ... priority=1
[centosplus],[contrib] ... priority=2
Third Party Repos such as rpmforge ... priority=N  (where N is > 10 and based on your preference) 

Install RPMforge repositories

Ok once you've completed this it's time to actually install and configure the RPMforge repositories which can be found here (I'm listing both x86 and x64 packages) :

• i386 http://packages.sw.be/rpmforge-release/rpmforge-release-0.5.1-1.el5.rf.i386.rpm
• x86_64 http://packages.sw.be/rpmforge-release/rpmforge-release-0.5.1-1.el5.rf.x86_64.rpm

Once installed the packages just issue the following command to make sure everything is working as intended :

 yum check-update 

Install Nagios

Ok now that you have installed the required repositories we can actually start the installation of Nagios, I'll leave up to you configuring services for automatic startup and firewall rules, with the following command :

 yum install nagios nagios-plugins nagios-plugins-setuid httpd php gcc glibc glibc-common gd gd-devel 

Sit back and enjoy while the system takes care of installing everything and putting the configuration files in the right palce.

Configure Nagios Web Interface

As I've already said I'm leaving the service configuration of services etc. to you, but to get you going with the Web interface be sure to configure at least the default Nagios Web access user with the following command :

 htpasswd -c /etc/nagios/htpasswd.users nagiosadmin 

Start monitoring

At this point you're good to go, start HTTPD and Nagios services and point your browser to <nagios_server>\nagios and login with the credentials configured in the previous step.

Conclusion

This is a very simple and high level guide on how to get Nagios installed on your Centos machine, this means there is still a lot to configure and to take care of to have a fully featured Nagios monitoring infrastructure but I will take care of this in another series of posts.

Hope you'll find this small guide usefull, please take some time to share or retweet it!

Cheer Lethe.

Q. When I try to launch vim to edit a file I get a command not found error. How can I solve this?

A: You need to manually install VIM on a system running CentOS.

Vim, or VI enhanced, is the successor of the famous vi editor which is found almost in every Linux distribution even on a recovery environment and has some nice features like syntax highlight for files and other useful stuff.

To install it and use in CentOS all you have to do is input the following :

 yum install vim-X11 vim-common vim-enhanced vim-minimal 

Once done you can enjoy the full benefits of a full featured and very powerful text editor.

Hope you’ll find this useful!

Cheer Lethe.

Professional Linux Administration posts.

Good morning all, I’ve received in the past a few comments and email about a series of posts describing the in and out of professional Linux administration. The fact is a lot of fresher in the Linux field out there often find problems after certifications like RHCE and so  n, and if you are not lucky enough to work in the field every day there are a lot of things that books or courses won’t tell you.

Everyone of us is learning some new bits everyday, I too learn something new every day and try to put all the interesting things on the blog but, as some people told me on the comments, most often than not I talk about problems you could face on your home or client Linux workstation. What about Professional Linux Administration?

Of course the server installation, management, monitoring or configuration is a totally different beast. For example which file system should i use for file server installation? How can I implement a centralized logging facility for all my Linux server? How can I implement a redundant NAS on Linux?

These are all common questions that maybe you’ve been googling for and I will try to, in a series of posts, to cover all that is needed for correctly plan, implement and configure correctly Linux Servers.

If you have any suggestion or would like to see some content featured on the blog please let me know, I will be more than glad to add to the list of planned posts!

Cheer Lethe.

Continuing my series on posts on the small, but yet annoying bugs, on Fedora 13 here’s another. When you try to launch VLC media player it will silently fail to start, read on and I will show you how to solve this.

Install VLC Media player.

First step is, of course, install VLC media player on our Fedora distribution. Open a terminal and become root, or use sudo, and input the following command :

 yum install vlc 

Yum will take care of resolving dependencies and installing all required packages. At this point if you try to start VLC, clicking on a file or from the menu, it will silently fail to start without any error message. If you try to start it from the command line you’ll get error messages similar to the following :

 VLC media player 1.0.6 Goldeneye
*** glibc detected *** vlc: free(): invalid pointer: 0x00aa5ff0 ***
======= Backtrace: =========
/lib/libc.so.6(+0x33b1e1)[0x5371e1]
/usr/lib/libstdc++.so.6(_ZdlPv+0x22)[0xa6d4b2]
/usr/lib/libstdc++.so.6(_ZNSbIwSt11char_traitsIwESaIwEE4_Rep10_M_destroyERKS1_+0x1e)[0xa6432e]
/usr/lib/libstdc++.so.6(_ZNSbIwSt11char_traitsIwESaIwEE7reserveEj+0x9a)[0xa65b3a]
/usr/lib/libstdc++.so.6(_ZNSbIwSt11char_traitsIwESaIwEE6appendEjw+0x76)[0xa65c56]
/usr/lib/libstdc++.so.6(_ZNSbIwSt11char_traitsIwESaIwEE6resizeEjw+0x7d)[0xa671ed]
/usr/lib/libtag.so.1(_ZN6TagLib6StringC2EPKcNS0_4TypeE+0xc2)[0x6dd8762]
/usr/lib/libtag.so.1(+0x2b4f7)[0x6db74f7]
/usr/lib/libtag.so.1(+0x6d71d)[0x6df971d]
/usr/lib/libtag.so.1(+0x23c7c)[0x6dafc7c]
/lib/ld-linux.so.2[0x2b8da0]
/lib/ld-linux.so.2[0x2bd13f]
/lib/ld-linux.so.2[0x2b8966]
/lib/ld-linux.so.2[0x2bc986]
/lib/libdl.so.2[0x477c3b]
/lib/ld-linux.so.2[0x2b8966]
/lib/libdl.so.2[0x47803c]
/lib/libdl.so.2(dlopen+0x41)[0x477b71]
/usr/lib/libvlccore.so.2(+0x90c6b)[0x420c6b]
======= Memory map: ========
00110000-00127000 r-xp 00000000 08:03 266493	 /lib/libpthread-2.12.so
00127000-00128000 r--p 00016000 08:03 266493	 /lib/libpthread-2.12.so
00128000-00129000 rw-p 00017000 08:03 266493	 /lib/libpthread-2.12.so 

For brevity I omitted part of the output.

Make VLC work on Fedora 13

The above is a known bug with a pretty easy solution, once you know it, just open a terminal window and input the following command in it :

 export MALLOC_CHECK_=1 

Now if you try to launch VLC again, both from the command line or from the Menu, it will run without any problem. you need to launch this command only once and it will fix your VLC forever. Lethe’s Certified as I had exactly the same problem on my Fedora installation

Hope you’ll find this useful, please feel free to share the post or ask me any question if you need.

Cheer Lethe.

As you know Skype has available both Linux and windows version of their popular software, problem is the Linux version development has never been following the same pace as the Windows version and with  Fedora 13 coming out the situation is not any better, in this post I’ll guide you through the whole process.

Skype Installation

Of course the first step is that of installing Skype, you can easily do this going to Skype’s official site and downloading the Linux version (as per this writing latest version is 2.1 beta 2), if you’re lazy here’s the direct link. Once you’ve downloaded the rpm just install it double clicking on it, Fedora will take care of downloading almost all the dependencies, or issue the following command as root :

 rpm -Uvh skype-2.1.0.81-fc10.i586.rpm 

Once the installer has done his job you should be able to see a link to Skype under Applications –> Internet –> Skype

Make Skype work under Feodra 13

As I’ve written when you install Skype’s RPM it will take care of installing ALMOST all dependencies, if you try to launch it via the gui nothing will happen but if you try to launch it via the command line you will see some errors like the following :

 [lethe@box ~]$ skype
skype: error while loading shared libraries: libXv.so.1: cannot open shared object file: No such file or directory 

This is due to poor packaging of the RPM which will not handle the installation of all dependencies on your machine, the solution is pretty simple. Open a terminal window and as root input the following :

 yum install qt-x11.i686 libXScrnSaver.i686 

This will take care of installing all the required, and at this point missing, packages/libraries.

Once the process has completed you’ll be able to launch and use Skype for all your calls.

Hope you’ll find this useful, cheer Lethe.

Well it finally happened, yesterday my blog completely crashed and I was forced to reinstall everything from scratch. I’ve lost just a  few posts (that I’ve been able to successfully restore from an old backup) but I’m still in the process of recovering the old settings, features and plugins.

Probably during the next few days you’ll notice theme changes and missing functionalities, this is planned and I’m working on this part but please if you notice something missing from the old blog just let me know, it will help me out spotting stuff that I forgot.

I think I’ll just stick with this theme as it is easier to configure/maintain and I want to fully concentrate on content of the posts rather than on tweaking the design which for me is a time consuming process.

Sorry for this major down.

Cheer Lethe.