Digital Ocean offers Virtual Private Servers (VPS) for a very low monthly fee. Recently I set up a server with them and I am quite satisfied. If you want to have a place where you can experiment with building and running web applications, this can be a perfect starting point.

In this article I'll explain how to set up a Perl Dancer based web site on a Digital Ocean droplet.

  1. Sign up to Digital Ocean and create a server
  2. Basic configuration of the server
  3. Installing packages
  4. Brewing Perl
  5. Installing Perl Modules from CPAN
  6. Creating a simple Dancer application
  7. Configuring Starman
  8. Configure nginx as a proxy
  9. Connecting a domain name

Sign up to Digital Ocean and create a server

First you need to visit Digital Ocean (This is an affiliate link, I'll earn some credits if you use this link to sign up.), provide your credit card information. Once you are in you need to create a server instance. They call these Droplets. There is a big green button to Create droplets.

You need to supply the Droplet hostname which can be any word. You can pick foo or web1 or whatever you like. I picked s12 for this server. This is the internal name.

Then you need to select the size. I go with the smallest: 512 Mb / 1 CPU / 20 Gb SSD disk / 1 TB transfer that costs $5 / month. I have not tried this yet, but if I understand correctly, I could create and destroy servers for much shorter periods and pay by the hour. This could be a lot of fun playing with setup.

The third thing to select is the region. Unless there is some really important reason to pick a specific data-center, you can go with any of those. I picked Amsterdam 1.

In the Select Image section, we need to select the Linux distribution we'll be using. You can pick your favorite one. I'll use Ubuntu 13.04 x64 for this article. This is the latest 64bit release of Ubuntu.

At the bottom you can then click on the big green button that says Create Droplet.

They will start building the server which takes about 60 seconds and send you an e-mail containing the IP address and the password of root. This is not the most secure process, but we'll change the password very soon so I don't think this is a big issue.

If you are worried they also allow you to provide a public key before you create the droplet and then they will install that key instead of sending you a password.

Basic configuration of the server

In the e-mail you will see instructions how to ssh to the server. If you are a command line user, you can use the ssh root@ command, with the IP address of your server.

If you run Windows, you will need to install an SSH client. I'd recommend installing putty. You just need to download the putty.exe file and you can double-click on it as it is. There is no need for any further "installation". Once the initial window of Putty has opened, there is a field for Host Name (or IP address) and a radio selector where you need to select SSH. Then press Open. This will get you to a window asking for Username:. There you type in root and press ENTER. It will then ask you for Password: and you type in the password you got in the e-mail.

The first thing we need to do is to change the password. So type in passwd and press ENTER. It will ask for a password and then it will ask you to repeat it. Please give a long password that you can remember. Something like The secret of a secure server is probably a good password. If you also add some numbers and some strange characters, it will become even stronger.

root@s12:~# passwd
Enter new UNIX password:
Retype new UNIX password:
passwd: password updated successfully

Once we set the new password we need to update the packages on the system. A Linux distribution consists of thousands of separate packages. Most Linux distributions constantly release fixes to these package and it is very likely that the droplet we created used an earlier set of these packages. So we should update and upgrade those packages:

aptitude update will download the most recent list of packages and their version numbers. This is usually done automatically, but we don't want to wait for that now.

root@s12:~# aptitude update

Get: 1 raring Release.gpg [933 B]
Get: 2 raring-updates Release.gpg [933 B]

aptitude safe-upgrade will first list the packages that need to be upgraded and ask: Do you want to continue? [Y/n/?]. If we press ENTER it will download these packages and install them on the system.

root@s12:~# aptitude safe-upgrade

Resolving dependencies...
The following NEW packages will be installed:
The following packages will be upgraded:
  apt apt-transport-https apt-utils apt-xapian-index bind9-host command-not-found
  language-selector-common libapt-inst1.5 libapt-pkg4.12 libasn1-8-heimdal libbind9-90
  libheimbase1-heimdal libheimntlm0-heimdal libhx509-5-heimdal libisc92 libisccc90
  libpython2.7-stdlib libpython3.3-minimal libpython3.3-stdlib libroken18-heimdal
  linux-image-virtual login lsb-base lsb-release openssl passwd plymouth
  python3.3-minimal rsyslog tzdata ubuntu-release-upgrader-core update-manager-core
74 packages upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
Need to get 51.9 MB of archives. After unpacking 35.2 MB will be used.
Do you want to continue? [Y/n/?]

If it asks questions about keeping the current version of various configuration files, you can usually just press ENTER. At this point this does not really matter.

Once the upgrade has finished it is recommended to reboot the system. Type in reboot. As the machine restarts this will disconnect you and you will need to connect to it via ssh (putty) again, after some time. (Maybe 30 sec?). Remember, you have changed the password!

Installing packages

As I mentioned every Linux distribution, and so Ubuntu too, comes with a lot of packages. Some of them are installed by default. Others need to be installed manually. Usually the way to install a package is to type aptitude install package-name. We will need a few of them so we type:

root@s12:~# aptitude install nginx

In some cases it will install the package right away, in other cases it might list what other packages need to be installed for the selected one to work (e.g. the dependencies on nginx) and it will ask you: Do you want to continue? [Y/n/?]. Just press ENTER.

We have just installed the nginx web server. We still need to start it for the first time. Type:

root@s12:~# service nginx start

Now you can open your browser and browse to the IP address of the server. Don't forget to add http:// in front of the IP address as some of the browsers will not add it automatically. So if your IP address is then type in the address bar of your server: You will see Welcome to nginx! or some similar message.

We will need a few more packages, so let's install those as well:

root@s12:~# aptitude install make gcc htop vim

In order to have a better security model we are going to run Starman our "application server" as a separate user we call starman. (We could have called it foobar, or any other name.) In order to create the user we run:

root@s12:~# adduser --gecos '' --disabled-password  starman

It will create the user starman and the only way to to access it will be via the user root. You can now switch to the starman user by typing:

root@s12:~# su - starman

Please note how the prompt changes from root@s12:~# to starman@s12:~$.

To return to the root user, just exit from the current user:

starman@s12:~$ exit

Brewing Perl

If you run perl -v you will see the server already has perl installed. It is usually referred to as system perl. As we are going to install all kinds of modules from CPAN some of those might interfere with modules used by the system. So the usual recommendation is to build a separate perl installation and use that for the application. That's what we are going to do. We will use Perlbrew for this to make the operation smooth.

Let's switch to the starman user and then follow the instructions on the Perlbrew web site and run:

root@s12:~# su - starman
starman@s12:~$ \curl -L | bash

This will install Perlbrew.

(BTW, if you wondering about the backslash prefix, it ensures that no local alias will interfere with the command. See here)

Then we need to add source ~/perl5/perlbrew/etc/bashrc to our start-up script which is the .bashrc file:

starman@s12:~$ echo "source ~/perl5/perlbrew/etc/bashrc" >> ~/.bashrc

We do this so every time we login to the machine it will be already configured. At this point you can either logout (exit) and login again, or you can load the updated version of .bashrc using:

starman@s12:~$ source ~/.bashrc

Now that perlbrew has been installed we can use it to install another version of Perl:

starman@s12:~$ perlbrew available

This will list the versions of Perl available from CPAN. As of the time of this writing this was the list:


5.19.4 is a development release (19, the second part of the version number is an odd number) and the most recent stable release is 5.18.1. So we'll install that version:

starman@s12:~$ perlbrew install perl-5.18.1

This will download the source of perl, compile it and test it. This can take quite some time. 10-20 minutes or even more. So this might be a good time to stretch a bit.

Once perl was built we can first list all the installed versions of perl:

starman@s12:~$ perlbrew list

and we can switch to it using the following command:

starman@s12:~$ perlbrew switch perl-5.18.1

If we run perl -v, we'll see it already report 5.18.1 We can also try which perl and it will report /home/starman/perl5/perlbrew/perls/perl-5.18.1/bin/perl.

Installing Perl Modules from CPAN

It is not enough to compile our own version of Perl we also need to install a few modules. For this we need a configured CPAN client. The traditional one is called cpan, but we are going to use the newer and slicker cpan minus, also known as cpanm. Perlbrew has its own command to install cpan minus:

starman@s12:~$ perlbrew install-cpanm

Once that's install we can install the required modules:

starman@s12:~$ cpanm Dancer2 Starman Daemon::Control

Creating a simple Dancer application

The command

starman@s12:~$ dancer2 -a Demo

will create a subdirectory called Demo and inside a skeleton of an application, called Demo. You can of course use any name as the name of your application, but it is usually recommended to start with a capital letter and then use lower case letters in the name.

We can change directory into the Demo directory and launch the web site with a development server on port 3000:

starman@s12:~$ cd Demo
starman@s12:~$ perl bin/

>> Dancer2 v0.10 server 29587 listening on

While it says as the IP address, you can actually use your own IP address. So if your IP address was then browse to and you will see the default page of Dancer.

You can go on and further improve the application by following the Getting started with Perl Dancer screencast, but for now we'll focus on configuring the server.

So let's stop the development server by pressing Ctrl-C on the console.

Configuring Starman

As user starman create /home/starman/ containing the following code:


use warnings;
use strict;
use Daemon::Control;

use Cwd qw(abs_path);

        name      => "Starman",
        lsb_start => '$syslog $remote_fs',
        lsb_stop  => '$syslog',
        lsb_sdesc => 'Starman Short',
        lsb_desc  => 'Starman controls the web sites.',
        path      => abs_path($0),

        program      => '/home/starman/perl5/perlbrew/perls/perl-5.18.1/bin/starman',
        program_args => [ '--workers', '3', '/home/starman/Demo/bin/' ],

        user  => 'starman',
        group => 'starman',

        pid_file    => '/tmp/',
        stderr_file => '/tmp/starman.err',
        stdout_file => '/tmp/starman.out',

        fork => 2,


set the executable bit using

starman@s12:~$ chmod +x /home/starman/

Then leave the starman user using exit and as user root create a symbolic link to the newly created file and then start the Starman service:

root@s12:~# ln -s /home/starman/ /etc/init.d/starman
root@s12:~# service starman start

In order to make sure Starman will start when the system boot we should add it to the boot scrips along with nginx. This is done by creating some more symbolic links.

This will list all the rc-directories where Nginx is listed. We should add the same set of links to Starman

root@s12:~# find /etc/ | grep rc| grep nginx | xargs ls -l

These commands will create the symbolic links:

root@s12:~# ln -s ../init.d/starman /etc/rc0.d/K20starman
root@s12:~# ln -s ../init.d/starman /etc/rc1.d/K20starman
root@s12:~# ln -s ../init.d/starman /etc/rc2.d/S20starman
root@s12:~# ln -s ../init.d/starman /etc/rc3.d/S20starman
root@s12:~# ln -s ../init.d/starman /etc/rc4.d/S20starman
root@s12:~# ln -s ../init.d/starman /etc/rc5.d/S20starman
root@s12:~# ln -s ../init.d/starman /etc/rc6.d/K20starman

Once you launched starman you can access the web application on the same hostname but on port 5000:

Configure nginx as a proxy

As user starman create the file /home/starman/nginx-demo.conf with the following content:


server {
  location / {
      proxy_set_header Host $http_host;
      proxy_set_header X-Forwarded-Host $http_host;
      proxy_set_header X-Real-IP $remote_addr;
      proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
      proxy_pass http://localhost:5000;

Then exit from the user, and as root remove the default configuration file, create a symbolic link to from the nginx configuration directory to the nginx-demo.conf file. Restart nginx.

root@s12:~# rm /etc/nginx/sites-enabled/default
root@s12:~# ln -s /home/starman/nginx-demo.conf /etc/nginx/sites-enabled/
root@s12:~# service nginx restart

Now you can visit replacing this with the IP address of your machine.

Connecting a domain name

If you already have a domain name registered, then you only need to configure it to make sure both and resolves to the IP address of your machine.

If you don't yet have one, you can register a domain name and then configure it.

No further changes are required to your configuration for this.

Further development

As you continue developing your application you will notice that the changes you make to files are not automatically reflected on the web site. This is normal. After all we just set up the deployment environment for your application.

If you want the new code to take effect, you'll need to restart Starman. As root run:

root@s12:~# service starman restart

Enjoy the development!