Some lessons learned on Vagrant with Fedora19
For details, see : the “repackage” command. http://docs.vagrantup.com/v2/cli/box.html
To rescue an old box, if you lost the url, you can run the repackage command (i.e.):
vagrant box repackage vagrant-fedora19B virtualbox
This will create a new box from the original box which you pulled down. I think (not sure) that it will be identical to what was recorded in your “vmboxurl” parameter.
2) There are two vagrant syntax’s.
The old one:
Vagrant::Config.run do |config|
And the new one:
Vagrant.configure(“2”) do |config|
In any case, be careful that your variables match the version of vagrant you are using. For example, there is a new syntrax for shared folders.
3) There is (unfortunately) no current way to specify your Vagrantfile.
Ever have two “similar” implementations of your vagrant setup? It would be nice to run
vagrant up Vagrantfile2
But alas, you cant currently do this.
4) Static IPs : If they don’t work out of the box ~ try another box ~ 🙂 Preferably, one hosted on vagrantbox.es !
To set up a private internal network between VMs, its quite easy:
gluster2.vm.network “private_network”, ip: “10.10.10.12”
But some .box implementations for one reason or another will result in a every odd “eth1” error. Im not a networking geek so I can’t say why. But in any case, after alot of failed attempts to fix my networking with the fedora19 box downloaded from this post http://readonlyfriday.com/blog/2013/07/05/fedora-19-with-vagrant/, I found that I could get everything to work with the vagrant fedora 19 box from http://www.vagrantbox.es/ did not have this problem.
5) Passwordless ssh is RIDICULOUSLY easy. Just use the standard pre made vagrant insecure private key.
This is easy in vagrant. Vagrant actually creates a insecure private/public key pair, and adds the public key to all vagrant boxes on startup. All you have to do to add one vagrant box ssh to the other is enable private networking and paste:
echo “—–BEGIN RSA PRIVATE KEY—–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—–END RSA PRIVATE KEY—–” >> /home/vagrant/.ssh/id_rsa
into your provisioner. For example, for me, its literally the first line of my provision.sh script.
6) Post provisioning: A hack around.
There is no notion of a “post-provision” task in the Vagrantfile (for example: when all my machines are spun up – a task that runs afterwards, maybe, for example pinging them or updating /etc/hosts on them or whatever).
So instead ~ if you have static ips enabled, you can just add an if block to the end of your provisioning script which only runs if the “last” ip is pingable. This is indirectly saying “only run if you are the last machine and your all ready to go”
#Say my machines are 10.10.10.1 and 10.10.10.12. I can just add this..
if ping -c 1 10.10.10.12 > /dev/null 2>&1 ; then
ssh 10.10.10.1 “echo \”hi\” >> /tmp/a”
ssh 10.10.10.12 “echo \”hi\” >> /tmp/a”
fi
Anyways… If you need any help setting up vagrant on f19 just ping ..
7) https://github.com/fgrehm/vagrant-cachier is awesome.
If you are using Yum or Apt to setup your VMs after they are provisioned, and you have alot of dependencies, these can make setup super fast, especially in a cluster environment where you have many VMs. The way to confirm that its working as expected is to list the contents of the vagrant cache in /tmp. Its easy to see that, on all VMs (i.e. in the cluster scenario), the timestamp of the writes for the *rpm files which were yum downloaded are the same.