Emulated IoT (ELIOT) Platform

The Emulated IoT (ELIOT) platform enables emulating simple IoT devices on top of Docker. It is based on libraries, coap-node and Leshan, which implement device management over CoAP and LWM2M. Devices consisting of simple sensors and actuators can be built using IPSO objects. The current implementation provides ready-to-use devices, such as weather observer, presence detector, light controller and radiator.

More detailed information about ELIOT can be found in the wiki.

This module implements the following LWM2M interfaces & operations:

Bootstrap	Register	Device Management & Service Enablement	Information Reporting
Bootstrap	Register	Read (Text/JSON/TLV)	Observe (Text/JSON/TLV)

      | Update     | Write (Text/JSON/TLV)                  | Notify (Text/JSON/TLV)
      | Deregister | Create (JSON/TLV)                      | Cancel
      |            | Execute                                |
      |            | Delete                                 |
      |            | Write attributes

Usage with Docker

1. Run the LWM2M Server:

docker run --rm -ti --name ms corfr/leshan

2. Run the LWM2M Bootstrap Server:

docker run --rm -ti --name bss --link ms corfr/leshan bootstrap

3. Run the devices (PresenceDetector | WeatherObserver | LightController | Radiator):

Without Bootstrap Server

docker run -it --link ms riccap/eliot <jsfile_name> ms [OPTIONS]

With Bootstrap Server

docker run -it --link bss --link ms riccap/eliot <jsfile_name> bss [OPTIONS]

JS-files:
* presence.js
* weather.js
* light_control.js
* radiator.js

Options:
-b          Bootstrap mode
-t          Real Weather data (only WeatherObserver)

Note: instead of using the name ms/bss you can use the IP address without the --link flags.

4. Run multiple clients with docker-compose. The basic configuration docker-compose spins-up the LWM2Mbootstrap server and the LWM2M server plus the IoT devices. The default configuration includes also monitoring capabilities: cadvisor for containers monitoring, node exporter for machine monitoring, prometheus to collect the various metrics, and Grafana to show them.

Note the -d flag to run docker-compose in detached mode.

docker-compose up -d

docker-compose scale weather=X presence=X radiator=X light=X

If you do not want to include the monitoring services just run

docker-compose -f docker-compose.yml up -d

or simply rename the docker-compose.override.yml file (e.g. to docker-compose.monitoring.yml), and then use it later through

docker-compose -f docker-compose.yml -f docker-compose.monitoring.yml up -d

Monitoring

An extension of the project allows you to monitor resources used by the system through various components such as Prometheus, Grafana, cadvisor and node exporter. A Grafana dashboard has been implemented to give a general overview of resource utilisation for ELIOT; the dashboard works only with the K8s deployment (not with docker-compose). A snapshot of the dashboard can be downloaded from Grafana website using the dashboard id 12627.

Monitoring ELIOT in docker-compose

The default deployment of Prometheus and Grafana stores data and configurations in docker volumes. To check where the data is stored:

docker volume ls
docker volume inspect NAME_OF_THE_VOLUME | jq -r '.[0].Mountpoint'

You can store the data on a directory of your local host by commenting out the docker volume and using the absolute path in docker-compose.override.yml as shown below.

With docker volumes:

#- /tmp/eliot/grafana-storage:/var/lib/grafana # shared with K8s
- grafana-storage:/var/lib/grafana # only docker-compose
- grafana-config:/etc/grafana

With absolute path:

- /tmp/eliot/grafana-storage:/var/lib/grafana # shared with K8s
# - grafana-storage:/var/lib/grafana # only docker-compose
# - grafana-config:/etc/grafana

Monitoring ELIOT in k8s

Please refer to the README of the K8s folder.