Tag: Docker

How to set up a Z-Wave network for your home automation

If you have been following along, you now have a Hass.io installation setup up and ready to begin creating automations.

As I said in my article about deciding on your home automation setup, you now need to set up the Z-Wave antenna so your various home automation equipment can talk to your Raspberry Pi and the Home Assistant software.

I recommend the Aeotec Z-Stick Gen5, Z-Wave Plus USB (affiliate link), but you theoretically can get any Z-Wave antenna and it will work.

  • Plug this USB device into your Raspberry Pi (and restart just to be safe).
  • Login to the Hass.io configuration using Putty
  • You need to figure out what the USB drive is called on your installation. Run the following command.
ls /dev/ttyACM*
/dev/ttyACM0     <-- this is the most common name, but yours may differ
  • You need to tell Home Assistant that it should expect to see Z-Wave communication on this device.
  • Open the configuration.yaml file using vim
vim configuration.yaml
  • Scroll down to the bottom of the file and enter the following command. Save and close vim (:wq)
zwave: !include zwave.yaml

This command will tell Home Assistant to look enable the Z-Wave component and to look for its configuration in a separate file in the same directory as configuration.yaml called zwave.yaml.

  • Create a new file using vim and enter the name of the Z-Wave USB stick you found before. Save and close vim.
vim zwave.yaml

usb_path: /dev/ttyACM0

Now, you will begin to “test” your configuration before actually running it. It is very easy to mess up the format of the .yaml files and cause Home Assistant to not start.

Luckily, Home Assistant has an easy way to check to see if configuration files are valid.

hass --script check_config -c /config

This tells Home Assistant to scan all of the .yaml files and see if there are any configuration issues that would cause it to not be able to restart and load. In typical Linux philosophy, if it reports no status, everything is good. This is because Linux “pipes” the output from one command to another and “chatty” output messages make it harder to pipe commands together.

You can also do this in the Home Assistant web portal.

  • Navigate to the web portal
  • Click on Configuration
  • Click on General
  • Click on Check Config
  • You will get

If everything is good, you will see a green “Configuration valid” message.

You can either reboot the Docker container or use the GUI. Note, as you add more and more automations, this will slow down (several minutes to reboot). This is why the GUI has options for just reloading pieces of the system.

exit            

whoami <-- make sure you are the "pi" user outside the Docker container
pi

sudo docker restart homeassistant
homeassistant

After Docker reports success (by printing out homeassistant), you will have to wait a few minutes for Home Assistant to finish loading.

  • You need to push you new changes to Github so they are saved. Log back into the Docker container and check the status of the files
sudo docker exec -it homeassistant /bin/bash
  • Check the status of your modified configuration files.
git status
On branch master
Your branch is up to date with 'origin/master'.

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git checkout -- <file>..." to discard changes in working directory)

        modified:   configuration.yaml
        modified:   zwave.yaml

no changes added to commit (use "git add" and/or "git commit -a")
  • Add these files (stage them) to your git changset so you can commit them.
git add *

git status
On branch master
Your branch is up to date with 'origin/master'.

Changes to be committed:
  (use "git reset HEAD <file>..." to unstage)

        modified:   configuration.yaml
        modified:   zwave.yaml
  • Commit the changes
git commit -m "Added zwave"
  • Push your changes to Github
git push

You are now finally ready to begin adding devices to your home automation!

How to set up Git to version control your Home Assistant configuration

Congratulations on making it this far!

You now have an installation of Hass.io running on your Raspberry Pi. This is the crucial point where you want to begin treating your installation with more care and professionalism.

There is a phrase in the DevOps world that applies here. “Treat your servers like cattle, not pets”. This means that you want your server (i.e. Raspberry Pi) to be homogeneous and easily replaceable (like cattle). You don’t want your server to be hand-built and hand-maintained (like pets).

This means that you should be able to throw away your server (i.e. flash it with a clean install) and use Infrastructure as Code (IaS) to configure it.

In the cloud development world, we would provision everything, including the server itself with code. In our installation, we have an actual piece of hardware (i.e. Raspberry Pi) and so we can’t do a complete IaS implementation.

However, we can version control our Hass.io configuration. Hass.io runs on a series of YAML files. These are simple JSON documents that declarative describe what they system should do. Hass.io determines how to do it based upon what is in the YAML files.

Hass.io does have a built-in way to “backup” your configuration; snapshots. On the Hass.io page, under the “SNAPSHOTS” tab, there is a button to take a snapshot. The downside of this is that it is still stored locally. You would have to then setup a system to copy this snapshot somewhere else. Therefore, I do not recommend this as the primary way you back up your configuration.

The best way, in my humble opinion, is to use Git & Github to back up your configuration files. Git is the de-facto standard for source control management and Github is a free, online repository for backing up, contributing and sharing out source code with the world.

Note that this is a rehash of the excellent instructions on the Home Assistant website.

Be warned, however, Git is not a simple system to learn and there are a huge number of options. However, since you will be doing very little multi-user editing of the configuration files, you are unlikely to run into the big issues with understanding git’s branching, merging, rebasing, etc. tools. A few simple commands should be sufficient to effectively use Git.

Setting up Github

First, we need to set up a Github account so we can store our configuration files. Navigate to the Github signup page and create an account.

Once you have an account, click on the New Repository button to create a new repo to store your configuration files.

Give it a meaningful name (example: homeassistant-config). The standard naming convention for Git repos is all lowercase letters separated by a dash. This is not required, but will make typing it easier.

Choose Public or Private. This is really up to you. If you want to make it easier to share your configuration with others, you can make it Public. If you don’t want to do this, you can make it Private.

Warning!!! Because you will be uploading your configuration to a “public” site, you must be careful not to expose secrets. We will go into this later, but setting the repo to “Private” is not enough to protect your secrets (passwords, IP addresses, etc.). Hass.io has built-in ways to deal with these.

Do not check the “Initialize this repository with a README” checkbox. Since we will already have a repo set up on our local machine, we want to upload it to a clean Github repo.

Do not add a .gitignore or license file yet. We want a clean, empty repo to start with and we will have a custom .gitignore file anyway.

Initialize the Git repo

Now that we can login to our Raspberry Pi and see our configuration files, we need to upload them to Git.

  • Launch PuTTY and connect to your Raspberry Pi (if you need a refresher, please see my previous blog post).
  • Now we are in the directory where the Hass.io configuration files are located inside the Docker image (if you are in the “home” directory for the Raspberry Pi, either see my previous blog post).
  • Make sure git is already installed (or install if needed)
which git
/usr/bin/git       <-- already installed
                   <-- not installed (will not return anything)

sudo apt-get update         <-- update your packages
sudo apt-get install git    <-- install git if needed

You will now need to create a .gitignore file. This is the most important file, both to prevent publishing important/sensitive data (IP addresses, passwords, etc) and to reduce what you upload to Github to a minimum.

whoami         <-- which user are you logged in as
root

pwd
/config         <-- make sure you are in this directory

vim .gitignore

Here is a copy of the excellent .gitignore file from the Home Assistant website. Feel free to customize as needed.

# Example .gitignore file for your config dir.
# A * ensures that everything will be ignored.
*

# You can whitelist files/folders with !, these will not be ignored.
!*.yaml
!.gitignore
!*.md

# Ignore folders.
.storage
.cloud
.google.token

# Ensure these YAML files are ignored, otherwise your secret data/credentials will leak.
ip_bans.yaml
secrets.yaml
known_devices.yaml

Now we need to “initialize” your git repository. This means we are going to tell git that this directory and all of its contents should be tracked as version controlled files (excluding the files/directories mentioned in the .gitignore file, of course)

git init       <-- initialize as a version controlled directory

git status     <-- see what files have been modified
On branch master
Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git checkout -- <file>..." to discard changes in working directory)

        modified:   configuration.yaml
        modified:   groups.yaml
        modified:   scene.yaml
        modified:   sensor.yaml
        modified:   switch.yaml

You are now seeing a listing of all the files and directories that git sees that have not been committed yet (your list will be different depending on what was set up in the directory).

We need to set some configuration options so git knows how to annotate our changes

git config user.email "dwight.k.schrute@dunder-mifflin.com" 
git config user.name "dwight"   

We need to “add” all the files and directories to the changeset so git saves their current state.

git add *

Now commit your changes so git will save them.

git commit -m "Initial checkin"

Excellent, now you have an initial git repo with a known setup that you can always refer to (or revert to if you break something).

Generate SSH key to connect between Raspberry Pi and Github

Now we need to “push” our changes to a central server to back them up. Git is a distributed version control system. This means it tracks and saves changes locally, but does not “upload” them somewhere else unless you tell it to. This was originally designed to allow lots of people to work independently (even when disconnected) and make it easy to upload and share code when network connectivity was restored.

The most secure way to connect to Github is via SSH key. Here is the link to the help documents on the Github site. We need to generate an SSH key locally, store it in the SSH agent on the Raspberry Pi and then upload it to Github.

Note: we have to generate and store the SSH key in the /config directory inside the Docker container. If it is stored anywhere else (like the default location, ~/.ssh/id_rsa, it will get deleted the next time the container is restarted.

  • Make sure you are in the /config directory
pwd
/config
  • Make a new directory to store the key, generate the ssh key and save it to disk. Save the file to .ssh/id_rsa. Enter a passphrase if you want to (you will be required to enter it each time unless you also set up ssh-agent, I’m not doing this as I feel it is overkill at this time).
mkdir .ssh       <- create a new "hidden" directory to store the key in

ssh-keygen -t rsa -b 4096 -C "dwight.k.schrute@dunder-mifflin.com"
Generating public/private rsa key pair.
Enter file in which to save the key (/root/.ssh/id_rsa): .ssh/id_rsa
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in .ssh/id_rsa.
Your public key has been saved in .ssh/id_rsa.pub.

Now we need to upload the SSH key to Github so we can use it to authenticate our git repo pushes.You need to copy the public key from your local Raspberry Pi to Github (actual key redacted). If you are using Putty, you can just highlight the text starting with ssh-rsa and ending with your email address and it will be copied to the clipboard.

cat .ssh/id_rsa.pub
ssh-rsa .......................................................................................................................................................................................................................................== dwight.k.schrute@dunder-mifflin.com
  • In your Github.com account you created earlier, click on your profile picture in the upper-right hand corner and select Settings.
  • Select SSH and GPG keys
  • Click on New SSH key
  • Enter a Title: Home Assistant (or whatever will remind you what this key is used for)
  • Enter the Key: ssh-rsa…
  • Click Add SSH key
  • Enter your Github.com password to confirm

Push from Raspberry Pi to Github

Now you can “push” your configuration files from your Docker container to Github so they are backed up.

  • Navigate to your Github repo (click on the left-hand icon and select the repo you created earlier)
  • Click on the Clone or download button
  • Click on the Use SSH hyperlink in the right-hand corner if it is not already selected (the title of the dialog should say ‘Clone with SSH’)
  • Copy the url (git@github.com:username/repoName.git)
  • Set the remote repo where git will push your changes
git remote set-url origin git@github.com:dw/homeassistant-config.git
  • If you try to push now, you will get a permission denied error (-u says to use a new upstream repo for this branch, origin is the name of this upstream repo in git, master is the branch we are on). This is because you have stored your SSH key in a non-standard location. We need to tell git where to find the SSH key.
git push -u origin master
git@github.com: Permission denied (publickey).
fatal: Could not read from remote repository.

Please make sure you have the correct access rights
and the repository exists.
  • Git allows you to store configuration files locally, inside the repo itself. This is especially useful for our Docker container since we don’t want these changes to be lost when the container restarts. We need to tell Git where to find our SSH file
git config core.sshCommand "ssh -i /usr/share/hassio/homeassistant/.ssh/id_rsa -F /dev/null"

git push -u origin master           <-- push your changes to Github.com
Enumerating objects: 19, done.
Counting objects: 100% (18/18), done.
Delta compression using up to 4 threads
Compressing objects: 100% (11/11), done.
Writing objects: 100% (12/12), 1.26 KiB | 143.00 KiB/s, done.
Total 12 (delta 8), reused 0 (delta 0)
remote: Resolving deltas: 100% (8/8), completed with 5 local objects.

To github.com:dw/homeassistant-config.git
 * [new branch]      master -> master
Branch 'master' set up to track remote branch 'master' from 'origin'.

Hurray! You have successfully uploaded your initial configuration to Github!