In order to do this, you need to create a new virtual host in your Apache config. Create a new file in the directory /etc/apache2/sites-available and open it in your favourite editor. It doesn’t matter what the file is called, but it’s best to keep it descriptive. We’ll call this project “mysite”, so the file can be called “mysite”. In the file we need to configure the Apache virtual host.

<VirtualHost 127.0.0.1>
ServerName mysite.localhost
DocumentRoot /var/www/mysite/public/
</VirtualHost>

In the VirtualHost tag, you put the IP, seeing as I only want this for local loopback (for development) I have just put 127.0.0.1. The ServerName is the URL that you use to connect to the site and the DocumentRoot is where the public documents are stored. This is a very basic set up, so there are many more options you can add.

To make the site enabled, you create a symbolic link to the file from the sites-enabled directory.

cd /etc/apache2/sites-enabled
ln -s ../sites-available/mysite mysite

You now need to add the subdomain (mysite.localhost) to the list of hosts, so open /etc/hosts in your favourite editor and append the line:

127.0.0.1 mysite.localhost

And then restart Apache:

sudo /etc/init.d/apache2 restart

Now you should be able to visit http://mysite.localhost on the local machine (assuming the directory does actually exist).

This should also be similar on MacOS and other linux Distros, but the file locations (particularly for Apache) will vary.

… when I boot up, I get the Ubuntu boot screen, which shows it’s connecting to the terminal server, however it then fails with an errror saying

Error: Failed to connect to NBD server

And I get sent to a basic busybox shell.

There were two reasons for this.

PXE… booted

First of all, because I obtain the DHCP separately from the network boot, I need to treat it as if it’s a static IP. LTSP can handle static IPs, but this posed a couple of problems. I would need to specify a separate config file for each MAC address in the pxelinux.cfg/ directory. Secondly it would require each MAC address to be given the same IP each time (this was not going to happen).

So instead of getting gPXE to PXE boot, I completely bypass pxelinux.0 and use my own boot script. In this script I pass in the IP and other information. gPXE has some environment variables which can be used for this, so I wrote a script (The ‘x’s should be replaced by the terminal server’s IP address).

#!gpxe
dhcp net0
kernel tftp://xxx.xxx.xxx.xxx/ltsp/i386/vmlinuz ip=${ip}:xxx.xxx.xxx.xxx:${gateway}:${netmask}:${hostname}:eth0:none nbdroot=xxx.xxx.xxx.xxx:2000
initrd tftp://xxx.xxx.xxx.xxx/ltsp/i386/initrd.img
boot vmlinuz

This script retrieves the kernel, and passes as parameters the environment variables (which were set by the dhcp) the IP, gateway, netmask and hostname. Another parameter is the nbd server location and port. The we retrieve the initial ramdisk (initrd) and boot.

More haste less speed

After that was fixed, on my test machines it still didn’t connect to the NBD server. This is because my test machines are core2duo 3ghz with 4gb RAM; they were so fast at booting up, that it didn’t get a response from the NBD server in time. I diagnosed this by adding the parameter:

break=mount

To the kernel line in the script above. This stopped the boot, then it got a response from the server, and when I pressed ctrl+D (to continue the boot) it booted up fine. This is a bug in the ltsp_nbd script.

I solved this by logging into the terminal server and opening the file /opt/ltsp/i386/usr/share/initramfs-tools/scripts/ltsp_nbd, then added the line:

sleep 5

after the line:

ip link set lo up

This meant that the script paused for 5 seconds to allow the NBD server to respond.

Once edited the initramfs needs updating, as do the kernels:

chroot /opt/ltsp/i386 update-initramfs -u
ltsp-update-kernels

I will be writing a part 3 to this sometime soon talking about some of the customisations I will be adding.

At work we have Ubuntu terminal servers which the students currently connect to using NX on ThinStation in the labs. This is not an ideal setup, especially because the ThinStation kernel is too old to run on (even relatively) modern hardware. It is designed to setup an old computer as a thin client, and allow it to access a terminal server which will do all the hard work.

The other issue which we have is that the network is not owned or managed by my department (computer science), but by the central IT department of the university. This means I do not have access to the DHCP server which is required to PXE boot the computers. On top of this, the computers need to dual boot between the thin client/terminal server and a local installation of Windows.

This is where gPXE comes to the rescue! What gPXE allows me to do is retrieve an IP address from the university DHCP server, and then create my own network boot script completely separately.

LTSP Installation on Ubuntu

This was (relatively) painless on Ubuntu Lucid. As I’m currently just setting up a test server, I used the ltsp-server-standalone package.

sudo apt-get install ltsp-server-standalone

This installs all the relevant packages (including DHCP) to set up a terminal server. Once that had installed I needed to make the thin client OS. Also, as the terminal server is running Ubuntu 64 and the clients are 32 bit, I needed to use the –arch parameter.

sudo ltsp-build-client --arch i386

For some reason I had trouble with the default tftp server (tftpd-hpa), so I ended up installing at atftpd.

sudo apt-get install atftpd

This seemed to solve the problem.

gPXE

The easiest way to test gPXE is to download the ISO from rom-o-matic, burn it to a CD and boot off it. During the boot sequence it tells you to press ctrl+B to enter the console, if you do that you have a minimal CLI.

What I wanted to do was to retrieve an IP from the main DHCP server, and then specify the next-server and boot file (which are normally specified by the DHCP server).

gpxe> dhcp net0
gpxe> set next-server xxx.xxx.xxx.xxx
gpxe> set filename ltsp/i386/pxelinux.0
gpxe> autoboot

Line 1 tells it to retrieve an IP for the network adapted net0. This may differ if you have multiple network cards installed. Line 2 sets where it should go to next (which contains the tftp server location to download the pxe boot image). Line 3 sets the path/filename for the pxelinux boot image which is the default location for aftpd. Finally we tell it to attempt to boot with those parameters.

Problems

As stated at the top, this is still very much work in progress, and I plan to edit it as I find fixes. I am still have trouble with this (as I stated at the top this is work in progress, and will be editing it as I go along). The first is that gPXE has trouble working with a batch of network cards which we have in some of the labs. These cards are on-board Intel 82566DM [8086/104a] (rev 2) network cards. For some reason gPXE cannot interface with them. I am going to try and overcome this by putting other network cards into those machines.

Also, when I boot up, I get the Ubuntu boot screen, which shows it’s connecting to the terminal server, however it then fails with an errror saying

Error: Failed to connect to NBD server

And I get sent to a basic busybox shell.

This (I am hoping) is nothing to do with the gPXE boot, but with the LTSP setup. So I will do further investigation into that and report back.

I ran iotop and analysed what process was continually writing to the disk. I noticed this process was updatedb.mlocate. This is the Ubuntu indexing process, which goes through all the files indexing them to make searching much quicker. As this backup server uses backuppc and has years worth of backups, it has many thousands of files to be indexed and meant that updatedb.mlocate (which is run daily) was taking longer than 24 hours to index.

To solve this problem I edited the file /etc/updatedb.conf and added the backuppc directory to the line that tells updatedb.mlocate not to index files within a certain path. This is simply done by adding “/var/lib/backuppc” to the end of the line that starts off with “PRUNEPATHS=”.

So mine now reads: PRUNEPATHS=”/tmp /var/spool /media /var/lib/backuppc”

If you have installed a standard LAMP server (there are many guides on how to do this), you now need to install the suphp package for apache, this is called libapache2-mod-suphp:

$ sudo apt-get install libapache2-mod-suphp

Once that has been installed open the file /etc/suphp/suphp.conf in your favourite text editor. Find the line that has the docroot on, and change the docroot so that it is just “/”. This means that the suphp engine will parse anywhere in the file system, and not just in the standard html directory, thus allowing users to have their own.

docroot=/

You may also want to change the security options as appropriate, just change the “false” to “true” of the applicable ones the enable them. This is worth experimenting with. Further down you will want to set the “check_vhost_docroot” is set to false, this again is to do with the fact that userdirs are not in the vhost’s document root.

check_vhost_docroot=false

Finally you have the min_uid and min_gid properties. These are worth altering if you still want to be able to have a website running as www-data (such as the default website). If this is the case, change them both to the uid and gid of www-data (33 by default). It is not recommended to allow suphp to run as root, so do not set it to 0.

min_uid=33
min_gid=33

Finally, you need to enable the mods suphp and userdir, and disable the mod php5, this is done with two commands, and then restart apache2:

sudo a2dismod php5
sudo a2enmod userdir suphp
sudo /etc/init.d/apache2 restart

This should then allow you to run php scripts as the user who created them. To test this, create a new php file that contains:

<?php
system(id);

This will give you information of the user that the php process is running as. I recommend changing the ownership and retrying it, just make sure sure suphp is running as it should be.

Also I am working on getting suPHP to work with userdir, to allow users of a server to have their own webspace in a secure fashion. I will be blogging about this when it is done.

However, I have found a few things that may be of interest:
Things you probably didn’t know about PHP – Some really interesting and neat things you can do with PHP.

PHP Object Generator – Generates class structures and objects, so you don’t have to

who | wc -l

However, if you want to perform a line count of the NX list, you have 6 lines at the top which are not part of the count. In order to perform this subtraction, we need to do a calculation. To do this in bash we use a $(()):

echo $((2 + 2))

So, if we were to use this when performing the command on nxserver, this is how it would look in the end:

echo $(($(/usr/NX/bin/nxserver --list | wc -l) - 6))

Or, if you’re not running this script as root (and using sudo):

echo $(($(sudo /usr/NX/bin/nxserver --list | wc -l) - 6))

Download the latest version from here.

At the same time I have added a script to the section, which automatically downloads and installs FreeNX onto an Ubuntu system. The script and further details are available from here.

I can relate this to the two things of which I am most passionate about; music and computing.

Let’s start with computing. I am a Linux user, I have been for a while. That’s not to say I hate Windows, I just find I am able to do things in Linux which I am not able to do in Windows. Therefore it is a tool which suits me best. However, invert the situation and ask some (not all, I don’t want to make huge sweeping generalisations myself) Windows users why they have never considered Linux. The answer on the whole is “Linux is hard”. Not only is that statement a broad generalisation, it is actually untrue. People have been using Windows for a number of years (even the exact same OS for 7 years!) and are used to it. That doesn’t mean it’s easier. These people have probably never used Linux in their life, or at least not properly. Yet they dismiss it.

I have never really touched FreeBSD, and have only played with Solaris on a relatively small scale. I don’t use them. I would not, however, go around slandering or libeling them. At the same time, I can understand why people would use Windows (application support, gaming etc..), but I think people need to open their minds to the fact there is an alternative and not immediately dismiss it.

My second point is music. Anyone who knows me knows that I am a metal fan (as well as rock, classical, jazz…). Something that irritates me are the people who dismiss it completely (particularly those snobs who think of it as a lesser art form) by saying comments such as “it’s just noise”. Have they heard every band? I’m a metal fan, and I wouldn’t claim to know all the different types. It is a huge genre. These people are again dismissing through ignorance. They think they know what it is like, and it all must be the same, therefore they don’t like any of it.

I realise in this case, labelling has a lot to do with it. Yes, all these music types come under the “umbrella” metal, but in fact there are many subgenres, and even within subgenres there are differences (eg. I really like Freedom Call but don’t like Hammerfall). I am making this point not because I think this is something that’s unique to metal music, but because it is something that isn’t. Everyday people/ideas/art forms are being judged because they fall into a category.

I am not trying to change people’s mind in this post, all I am trying to achieve is make people think twice before they completely dismiss something and to provoke thought.

Rant over….

…for now.