=== Notes on using Mender on Buildroot
======================================
Mender is an open source over-the-air (OTA) software updater for
embedded Linux devices. Mender comprises a client running at the
embedded device, as well as a server that manages deployments across
many devices. There is also various tooling around the Mender project,
such as 'mender-artifact' which is used to create Mender Artifacts
that are compatible with the Mender client and server.
Mender aims to address this challenge with a robust and easy to use
updater for embedded Linux devices, which is open source and available
to anyone.
Robustness is ensured with atomic image-based deployments using a dual
A/B rootfs partition layout. This makes it always possible to roll
back to a working state, even when losing power at any time during the
update process.
The official documentation is a good resource to get an in depth
understanding of how Mender works:
https://docs.mender.io
In Buildroot the following packages are provided:
- BR2_PACKAGE_MENDER
- This will install the client on target rootfs
- BR2_PACKAGE_HOST_MENDER_ARTIFACT
- This will install the 'mender-artifact' tool in host rootfs.
To fully utilize atomic image-based deployments using the A/B update
strategy, additional integration is required in the bootloader. This
integration is board specific.
Currently supported bootloaders are GRUB and U-boot, and for reference
integrations please visit:
https://github.com/mendersoftware/buildroot-mender
Default configurations files
----------------------------
Buildroot comes with a default configuration and there a couple of
files that need your attention:
- /etc/mender/mender.conf
- main configuration file for the Mender client
- https://docs.mender.io/client-configuration/configuration-file/configuration-options
- /etc/mender/artifact_info
- The name of the image or update that will be built. This is what the
device will report that it is running, and different updates must have
different names
- /var/lib/mender/device_type
- A string that defines the type of device
Mender server configuration
---------------------------
The Mender server can be setup in different ways, and how you
configure the Mender client differs slightly depending on which server
environment is used.
- Mender demo environment
This is if you have followed the Getting started documentation where
you launch a Mender server locally and to configure your environment
to connect to this local server you need to provide the IP address of
the server on the local network.
By default the demo environment will connect to 'docker.mender.io' and
's3.docker.mender.io' and we need to make sure that these are resolved
to the local IP address of the running server by adding the following
entry to '/etc/hosts'
<ip address of demo environment> docker.mender.io s3.docker.mender.io
This is required because the communication between client and server
is utilizing TLS and the provided demo server certificate (server.crt)
is only valid for 'docker.mender.io' and 's3.docker.mender.io'
domains.
- Hosted Mender
To authenticate the Mender client with the Hosted Mender server you
need a tenant token.
To get your tenant token:
- log in to https://hosted.mender.io
- click your email at the top right and then “My organization”
- press the “COPY TO CLIPBOARD”
- assign content of clipboard to TenantToken
Example mender.conf options for Hosted Mender:
{
...
"ServerURL": "https://hosted.mender.io",
"TenantToken": "<paste tenant token here>"
...
}
Creating Mender Artifacts
-------------------------
To create Mender Artifacts based on Buildroot build output you must
include BR2_PACKAGE_HOST_MENDER_ARTIFACT in your configuration, and
then you would typically create the Mender Artifact in a post image
script (BR2_ROOTFS_POST_IMAGE_SCRIPT). Below is an example of such a
script:
#!/bin/sh
set -e
set -x
device_type=$(cat ${TARGET_DIR}/var/lib/mender/device_type | sed 's/[^=]*=//')
artifact_name=$(cat ${TARGET_DIR}/etc/mender/artifact_info | sed 's/[^=]*=//')
if [ -z "${device_type}" ] || [ -z "${artifact_name}" ]; then
echo "missing files required by Mender"
exit 1
fi
${HOST_DIR}/usr/bin/mender-artifact write rootfs-image \
--update ${BINARIES_DIR}/rootfs.ext4 \
--output-path ${BINARIES_DIR}/${artifact_name}.mender \
--artifact-name ${artifact_name} \
--device-type ${device_type}
As you can see some properties are extracted from target rootfs, and
this is because these values are used for compatibility checks,
meaning that the information must be present in both rootfs and in
Mender Artifact meta data.
- device_type - must be an exact match between rootfs and Mender
Artifact meta-data to apply update. You can set an
array of devices here as well, e.g if your image is
compatible with multiple hardware revisions
- artifact_name - must be an exact match between rootfs and Mender
Artifact meta-data to apply update.
Configuring Mender with certificates
------------------------------------
Mender uses TLS to communicate with the management server, and if you
use a CA-signed certificate on the server, you must include
BR2_PACKAGE_CA_CERTIFICATES in your configuration to authenticate TLS
connections.
b5056518cd