|
|
@@ -0,0 +1,198 @@ |
|
|
# Vagrant crash course |
|
|
|
|
|
This is a 5-minute crash course to start using vagrant. We will host our vagrant files in ``/home/USER/vms``. That's |
|
|
my convention working with many vagrant machines. |
|
|
|
|
|
## Minute 0: Get vagrant and virtulabox |
|
|
|
|
|
vagrant: https://docs.vagrantup.com/v2/installation/ |
|
|
|
|
|
virtualbox: https://www.virtualbox.org/wiki/Downloads |
|
|
|
|
|
## Minute 1: Create an init file |
|
|
|
|
|
Think of vagrant as Python's ``virtualenv`` or making a new repo. You want to start off with |
|
|
a VM with some initial configuration. |
|
|
|
|
|
``` |
|
|
cd /home/USER/vms && mkdir master && cd master |
|
|
vagrant init |
|
|
``` |
|
|
|
|
|
Now we have a directory called **master** and is holding a vagrant configuration file called ``Vagrantfile``. |
|
|
|
|
|
## Minute 2: Download a vagrant box |
|
|
|
|
|
The point of using vagrant is so you can download a boostrapped VM (most of them are actually just vanilla |
|
|
VMs straight out of ``.iso`` image installation). |
|
|
|
|
|
Get a VM from http://vagrantbox.es/. But for now, let's just grab **Ubuntu Server Precise 12.04.3 amd64 |
|
|
Kernel is ready for Docker (Docker not included)**. |
|
|
|
|
|
To download a box, the syntax is ``vagrant box add {title} {url}``. In our case: |
|
|
|
|
|
``` |
|
|
vagrant box add 120403_64_docker http://nitron-vagrant.s3-website-us-east-1.amazonaws.com/vagrant_ubuntu_12.04.3_amd64_virtualbox.box |
|
|
``` |
|
|
|
|
|
I try to be as descriptive as possible. It's easy to read what it means. I think so... |
|
|
|
|
|
## Minute 3: Configure your ini file |
|
|
|
|
|
Wow. We got ourselves a box! Now let's actually use it in our init file. |
|
|
|
|
|
Use your favorite editor to open Vagrantfile and then replace ``"base"`` with ``"120403_64_docker"``. |
|
|
We are telling vagrant to launch a new VM based on the stock box ``120403_64_docker``. |
|
|
|
|
|
Easy right? |
|
|
|
|
|
## Minite 4: Launch your VM |
|
|
|
|
|
Simple! While you are in the **master** directory: |
|
|
|
|
|
``` |
|
|
vagrant up |
|
|
``` |
|
|
|
|
|
By default, Vagrant will launch the machine via Virtualbox and will use a private IP 10.0.2.15, aka a class A IP. You cannot SSH into this machine directly. |
|
|
|
|
|
After awhile your machine will be loaded and then |
|
|
|
|
|
``` |
|
|
vagrant ssh |
|
|
``` |
|
|
|
|
|
Yes! Now you are ssh-ed into the VM, you can start playing. Of course, you should do ``sudo apt-get update`` first. |
|
|
Notice your user is automatically created and is called **vagrant**. Guess what the password is? Yeah! **vagrant** too. |
|
|
|
|
|
|
|
|
## Minute 5: Shut down the VM |
|
|
|
|
|
When you want to shut down the VM you just go back to the **master** directory, and then do ``vagrant halt``. |
|
|
|
|
|
That's it. |
|
|
|
|
|
|
|
|
We still have time! So let's go over some advance stuff. |
|
|
|
|
|
## Commands |
|
|
|
|
|
``vagrant --help`` |
|
|
|
|
|
Some notable advance commands: |
|
|
|
|
|
* package: At some point if you have done enough work but you are about to mess up your VM yet you want to save the VM? |
|
|
You can either clone using your VM Manager (e.g. virtualbox), or using vagrant's ``package`` command. Or if you want to |
|
|
distribute a VM to your friend, you can just package your VM into a vagrant box. You can even package the Vagrantfile! |
|
|
|
|
|
* reload: the fancy thing about this is if you change anything with Vagrantfile and doesn't want full restart, |
|
|
you can try reload. |
|
|
|
|
|
* box list: you can use this command to find out a list of boxes you have installed. Did I tell you that |
|
|
boxes you've exported via ``package`` command can also be ``add`` to your vagrant environment? |
|
|
|
|
|
## Vagrant - insecured-ssh-key |
|
|
|
|
|
Vagrant comes with an insecured ssh key. That's how ``vagrant ssh`` works by default if you are using some default |
|
|
vagrant box. The location of this key is in ``~/.vagrant.d/insecure_private_key``, under the home directory |
|
|
of whichever user installed vagrant. |
|
|
|
|
|
## Vagrant - box location |
|
|
|
|
|
Vagrant saves all the boxes you've added to vagrant to ``~/.vagrant.d/boxes``. Sometimes, you should free them up! |
|
|
|
|
|
## Vagrantfile |
|
|
|
|
|
If you cd in **slaves** instead of **masters**, when you ``vagrant ssh`` you will ssh into the slave machine |
|
|
(provided that the machine is on). So remember, the vagrant command works with relative vagrant file. |
|
|
|
|
|
## Vagrantfile - init |
|
|
|
|
|
If you already have a box, you can actually create your Vagrantfile by doing ``vagrant init {box_name}`` |
|
|
so you don't have to edit the file just to use the right box. I also believe if the box |
|
|
comes with its own Vagrantfile during package time, that packaged Vagrantfile will be copied and created |
|
|
on your loca disk for you. This means, the settings will be carried over. I forgot, but probably a yes. |
|
|
|
|
|
## Vagrantfile - ssh |
|
|
|
|
|
You can specify the default user and the ssh private key to use for a given Vagrantfile. |
|
|
|
|
|
``` |
|
|
config.ssh.default.private_key_path = "meow_private_key" |
|
|
config.ssh.default.username = "cat" |
|
|
``` |
|
|
|
|
|
This will allow us to ssh into the machine (using ``vagrant ssh``) that was |
|
|
created by the Vagrantfile with this configuration |
|
|
as ``cat`` using the ``meow_private_key``. Note the key is relative to the location of Vagrantfile. |
|
|
|
|
|
## Vagrantfile - network |
|
|
|
|
|
You can enable the public network by uncommenting ``config.vm.network :public_network``. The so-called public |
|
|
network is simply a LAN network IP so that other machines on the same network can SSH into it directly). |
|
|
|
|
|
You can also enable another kind of private network other than the class A and the 192.168.1.* kind. Vagrant |
|
|
will assign static private network of ``192.168.33.*`` if you use ``config.vm.network :private_network``. By |
|
|
default, the last byte is the number 10 (192.168.33.10). |
|
|
|
|
|
It is best if you just use either the private or the "public network." Don't try to mess with the ``/etc/network/interface`` file. |
|
|
If you have to (sometimes when Vagrant isn't working well and you want to control the VM directly via Virtualbox), |
|
|
then you have to boot without vagrant with Virtualbox and remove the static and return to use DHCP. |
|
|
|
|
|
This is what vagrant does to the interface file: |
|
|
|
|
|
``` |
|
|
# The primary network interface |
|
|
auto eth0 |
|
|
iface eth0 inet dhcp |
|
|
pre-up sleep 2 |
|
|
#VAGRANT-BEGIN |
|
|
# The contents below are automatically generated by Vagrant. Do not modify. |
|
|
auto eth1 |
|
|
iface eth1 inet static |
|
|
address 192.168.33.10 |
|
|
netmask 255.255.255.0 |
|
|
#VAGRANT-END |
|
|
``` |
|
|
|
|
|
See [Vagrant networking](https://docs.vagrantup.com/v2/networking/index.html) for more information. |
|
|
|
|
|
## Vagrantfile - customize |
|
|
|
|
|
You can change how many cores and memory the VM will have in Vangrantfile. Remember |
|
|
the core of vagrant is to provide friendly commands to VM Manger. You can |
|
|
set these hardware settings directly from the VM Manager (e.g. virtualbox). |
|
|
|
|
|
I believe by default the total memory is around 378MB. Whatever the number is, it's pretty low. |
|
|
|
|
|
The option is in ``vb.customize``. |
|
|
|
|
|
To use it, you must enable ``config.vm.provider``. In our tutorial, the following code will set 600MB for memeory: |
|
|
|
|
|
``` |
|
|
config.vm.provider :virtualbox do |vb| |
|
|
# # Don't boot with headless mode |
|
|
# vb.gui = true |
|
|
# |
|
|
# # Use VBoxManage to customize the VM. For example to change memory: |
|
|
vb.customize ["modifyvm", :id, "--memory", "600"] |
|
|
end |
|
|
|
|
|
``` |
|
|
|
|
|
Refer to the following posts on how to set CPU cores and memory. |
|
|
|
|
|
|
|
|
* [Vagrant configurations] (https://docs.vagrantup.com/v2/virtualbox/configuration.html) |
|
|
* [Setup vagrant with 2 or more CPU cores on 64bits systems](http://www.ifusio.com/blog/setup-vagrant-with-2-or-more-cpu-cores-on-64bits-systems) |
|
|
|
|
|
|
|
|
## Vagrant provisioner - ansible |
|
|
|
|
|
Yes. You can hook up ansible with vagrant, but silex isn't working properly or well with this integration. |
|
|
But if you have time, play with it. |
|
|
|
|
|
You probably can use Vagrant for other tasks in other areas. So keep your mind open about this integration. |
|
|
|
|
|
* [Vagrant provisioning](https://docs.vagrantup.com/v2/provisioning/index.html) |
|
|
|
yeukhon revised this gist