That being said, you want a media server but you are not comfortable with these technologies, which might scare you or put you off. The applications used in this book are really straightforward to install and use; particularly, open source applications have been selected.

Additionally, we are going to use an embedded board instead of the heavy, large, and power-consuming PCs so that we can also drastically reduce the costs. While we will focus on BeagleBone Black in this book, the chapters have been written to use most of the boards available on the market, such as Raspberry Pi, WandBoard, CubieBoard, and some others, as long as the board supports Linux and has a network connection.

The main goal of this approach is that you learn to be independent enough so that the next time you receive a promotional e-mail with music streaming advertising, you'll throw it away, smiling proudly.

This chapter is about installing a MediaDrop server, which will be introduced shortly. As promised, it will be quite easy (really easy, in fact); no compilation or library will be invoked. Before the installation process itself, we will have a quick look at some situations that an embedded board is able to resolve. We will also find out why we should consider a server philosophy instead of a traditional computer. This will impact the remaining part of the book because the board will be accessed through an SSH connection; we will also have to take into consideration our available free space to store our applications and media contents. Then, we will start the installation part for the MediaDrop server itself. Management and security tasks have been split in additional chapters, so you can skip them and get back later if you want.

Welcome to the first step of your independence.

In this chapter, we will talk about the following topics:

BeagleBone Black has the advantage of booting either from an SD card or on-board flash (eMMC). The pros and cons of each of these is beyond the scope of this chapter. Generally, for this book, you should rely on the SD card support for the following reasons:

The SD card induces small additional latencies that you won't even notice, so it is better to use it as a sandbox. So feel free to explore and try crazy things so that you can learn without limitations and worry.

If you update or install anything right away after a boot, you will encounter problems related to disk space.

Indeed, if you use PuTTY (www.putty.org) or any other SSH software to connect yourself to the board (such as, 192.168.0.15) and look at the rootfs space information, you can guess that the available size will be quickly saturated.

The SD card has little space left the first time

So, as the server administrator, it is up to you to resolve the size constraints.

The next topic will describe how you can achieve this task easily.

The default free space is really small; for example, you can fill it completely with just one upgrade. As a result, the very first thing to do at boot is to resize the /root partition.

Resizing a partition is not recommended for first timers. The good news is that you won't have to enter many commands; you can make your life easy by relying on a dedicated script provided with the board, which will do all that for you.

Enter the following two commands:

debian@beaglebone:~$ cd /opt/scripts/tools/
debian@beaglebone:/opt/scripts/tools/$ git pull

In this directory, we have downloaded the last code from the official repository, so we can start the script thereafter:

debian@beaglebone:/opt/scripts/tools/$ sudo ./grow_partition.sh

While running, this script will display a lot of details, which, thanks to the author, you don't have to care about.

These are the commands that you won't have to learn

You can now go on to reboot your board (don't forget to press the user's button), as follows:

debian@beaglebone:/opt/scripts/tools/$ sudo reboot

Now, if you check, you'll see that the free space has been resized to the SD card's capabilities.

An 8-GB resized SD card

Now, you can continue with the update and upgrade…

Now that your system is ready to accept all the applications you want to install, let's begin with the MySQL part:

debian@beaglebone:~$ install mysql-server mysql-client

During this you'll be asked for the database's root password, then the remaining system files:

debian@beaglebone:~$ install libjpeg-dev zlib1g-dev libfreetype6-dev libmysqlclient-dev python-dev

Finally, the Python-related requirements that will help you in virtual environments can be installed using the following command:

debian@beaglebone:~$ install python-setuptools python-virtualenv

With system requirements in place, you can set up MediaDrop in just six installation steps. Before doing that, you need to define a database for it.

You need to create some credentials and assign a user to a new MediaDrop database. It requires you to type only a few commands. If this is something that you don't want to do, the script is available on the GitHub companion website at https://github.com/dlewin/BeagleboneBlack-Server-Book.

Just execute the following command:

./create_mediadropdb.sh

Instead, if you are like me and want to control your entire system, then open a MySQL console, and pay attention to the syntax of the following command (with the ;):

debian@beaglebone:~$ mysql –u root –p

Then, fill in the password you have defined in the installation steps.

You will now have access to the MySQL console for all the database operations that MediaDrop requires. Now, perform the following steps:

Here's a screenshot that shows these operations along with the related MySQL feedback you should get:

MediaDrop has been programmed in Python, a simple but powerful language that we'll also use in Chapter 6, Illuminate Your Imagination with Your Own Projects. As plugins are also in Python, you are free to extend the platform as you wish.

Let's see how to define a dedicated environment for this purpose.

Virtual environments introduce a functionality provided with Python that I like to use: actually, with this functionality, you can resolve the problems of software versions and incompatibility. The principle is to use as many virtual spaces with different packages or executables as you want without any confusion that might create a broken system. Thanks to the isolation provided by this tool, you are guaranteed that it is kept as you want it to be and thus you always have a safe system.

Currently, we are using the 0.10.3 version of MediaDrop, but nothing stops you from trying the next version in order to test any side effects on your architecture. Thus, instead of executing the current environment, you'll just have to start MediaDrop from the n+1 environment.

Virtual environment is presented just here, but you should create an environment each time you need different usages of an application or dependencies. It requires just less than a minute and can be done by performing the following steps:

That's all! You are in now. Cool, eh?

The command prompt is preceded by the virtual environment's name, so you can guess which environment is currently activated, which is shown in the screenshot that follows the code:

(venv)debian@beaglebone:~$

You can now install any package you want; it will only be accessible in the currently activated environment.

For installation, retrieve the last stable release from downloads:

(venv)debian@beaglebone:~$ mkdir MediaDrop
(venv)debian@beaglebone:~$ cd MediaDrop
(venv)debian@beaglebone:~/MediaDrop$ wget http://static.mediadrop.net/releases/MediaCore-0.10.3.tar.gz

Then, perform the following steps:

This will generate the production.ini file, which the server will use as a configuration file. So, we can adapt by editing it.

What we want to do here is tell which database we created and the related credentials. So, edit the production.ini file, with nano (or vim) in order to focus on the [app:main] section, as we want to change the following line:

sqlalchemy.url = mysql://username:pass@localhost/dbname?charset=utf8&use_unicode=0

The settings are shown in the following screenshot:

At the same time, as port 8080 is already used, let's change this to an available port, something such as 8082. As you'll see later on, many servers use the 8080 port for their configuration settings, and obviously only one application will be able to use this port, which means that according to your configurations, you will have to decide which application will use the 8080 port.

Now, save and quit the editor. We are done with the configuration part.

We have completed all the configuration steps, and we are now going to deploy all the website data content.

To do this, copy the data directory from the sources to be at the same level as your production.ini file:

(venv)debian@beaglebone:~/MediaDrop$ cp -a MediaCore-0.10.3/data/.

Give it the write permissions for the debian user; when some content is uploaded, it will be written here:

(venv)debian@beaglebone:~/MediaDrop$ chmod 666 data

It's now time to fill the MediaDrop database we've created previously with some tables and required data. We just have to call a predefined command:

(venv)debian@beaglebone:~/MediaDrop$ paster setup-app deployment.ini

Create some triggers that will allow you to have better searches than exact matches, as follows:

(venv)debian@beaglebone:~/MediaDrop$ mysql -u root -proot MediaDrop < MediaCore-0.10.3/setup_triggers.sql

Nothing else is remaining; we are in. Time to start our first server!

As an administrator, the first thing you need to do is change the default password; this is the most obvious thing that an attacker will try, so go to Users, as shown in the following screenshot:

You should have a unique user: admin. Click on this to go to the user's profile in order to change the password.

Now, click on Save.

Now, let's go through some of the settings together; click on the Settings button to access the settings that are not related to rights, as shown here:

In this section, you'll find everything related to the overall settings.

This setting is used for the following actions:

The customization is really interesting here because you can use your own CSS, HTML header, and footer, so you can have a personalized interface.

This part is important as you will need to enhance the provided categories for your needs.

While I won't explain what a category is, I can, however, say that there are useful combinations to create because a category can be related to another one as a child category, which lets you define subcategories. If all the users can use categories, only granted users will be able to define them. We will see how this is done in the next chapter.

In addition, an additional advantage is that categories are integrated during a textual search. This also includes entire sets of simple categories or more complex categories that can be made from a parent category, including many children categories.

The following functions are related to comments:

You can choose the type of events that you need to be warned about, as follows:

A list of players is provided with the server; these are capabilities that handle most of the media that a user will upload such as Vimeo, Youtube, DailyMotion, and so on.

I recommend that you allow all of them (click on the green light on the left-hand side) and define your preferred player using Priority, with arrows beside each player.

Each published video can be submitted to like/dislike votes. For your information and contrary to what one might think, popularity is not just the result of any addition/subtraction but it is a dedicated equation of the following form:

popularity_points = log_X(media.likes) + media.age/Y

This equation is clearly explained in the documentation at http://mediadrop.net/docs/user/admin/settings.html#popularity.

Both the decay exponent and lifetime are adaptable in this section.

Tags are related to categories, but are nevertheless different. You can use them when you post a micro description of one or two words (recommended), but they are also useful for searches. Moreover, when users browse contents tags, they also help to get an overall opinion.

As with Categories, you need to have granted access in order to define tags.

Here, you will find the setting that you might look after the most.

File size limit

Indeed, there is a default 300 MB size limit for each upload, so you might end up with the following error, shown in the screenshot:

Don't confuse size limits and directory rights. What is the common thread between the two, you might ask? It's that if you forget to set the /data directory, you will have the same upload error when posting a link to a file.

Storage engines

This is an important section under options that aims to list all the locations for the files, and mainly two of them need to be explained:

This points to the place where your content will be kept. This is really handy because you can modify the default partition scheme on BeagleBone if you have a local USB hard drive.

For a dedicated remote partition such as NFS or another type, there is:

Remote URLs

Here, you can specify the partition that accepts the files.