- Introduction
- IoT Agent settings -
config.js
- Provisioning a Config Group
- Provisioning an Individual Device
- Receiving measures from devices
In this guide we will be using the IoT Agent JSON (which is the reference IoTAgent using the IoTAgent Library) as an example to demonstrate how to provision config groups, devices and how to receive measures from devices.
Be aware that every IoT Agent which uses the library is different, but the concepts for provisioning IoT devices remain the same regardless of protocol.
The IoT Agent JSON is a simple IoT Agent which uses JSON payloads to send and receive data. It is a good starting point for understanding how an IoT Agent works since it uses JSON payloads to send and receive data.
The config.js
holds common information about the interactions between the Agent and the Context Broker. Additional
custom settings may also be required dependent upon the actual IoT Agent used. The following is an example of a typical
config.js
file:
config = {
logLevel: 'DEBUG',
contextBroker: {
host: 'orion',
port: '1026'
},
server: {
port: 4041,
host: '0.0.0.0'
},
deviceRegistry: {
type: 'mongodb'
},
mongodb: {
host: 'localhost',
port: '27017',
db: 'iotagent'
},
service: 'openiot',
subservice: '/',
providerUrl: 'http://iot-agent:4041',
defaultType: 'Thing'
};
In this case the context broker hostname is orion
and is listening on port 1026
, the IoT Agent can be provisioned by
sending requests to port 4041
which is also the port used to receive NGSI requests. The IoT Agent is using the
iotagent
database from a MongoDB instance at localhost:27017
to store needed information (provisioned groups and
devices, etc.).
The remaining settings help to define the NGSI interactions - the IoT Agent will be using the fiware-service=openiot
and fiware-service-path=/
. The default type
for each created entity is Thing
, although this can be overridden as
shown below. Devices will be registered for a period of one month and the IoT Agent will receive registration callbacks
at the URL http://iot-agent:4041
.
All configuration settings can also updated using Docker environment variables. You can find more information about the available configuration parameters and environment variables in the Administration Guide.
Settings which are common to a group of devices can be passed to the IoT Agent using the Config Group API. Each config
group may override values previously defined in the global configuration if necessary. When using the config group API,
the fiware-service
and fiware-servicepath
headers will define the service and subservice to which the configuration
will be applied. Additionally, the apikey
field is used to identify the configuration group. An example of a basic
config group is shown below:
curl -iX POST \
'http://localhost:4041/iot/groups' \
-H 'Content-Type: application/json' \
-H 'fiware-service: openiot' \
-H 'fiware-servicepath: /' \
-d '{
"groups": [
{
"apikey": "4jggokgpepnvsb2uv4s40d59ov",
"entity_type": "Device",
"resource": "/iot/json",
"attributes": [
{ "object_id": "t", "name": "temperature", "type": "Number" }
]
}
]
}'
In this case an apiKey
for identifying devices has been created and all interactions to the path /iot/d
which
present this apiKey
will be created as entities of type=Device
rather than using the configuration default of
type=Thing
.
Additionally, the group has defined an attribute mapping for a measurement t
to be mapped to temperature
attribute
when receiving data from devices.
The config group would usual hold additional attribute mappings, commands and common static attributes as well.
Settings which are specific to an individual device can be passed to the IoT Agent using the Device API. The configuration provided in the Device API will override any settings defined in the Config Group API and the global configuration as well. An example of a basic device configuration is shown below:
curl -iX POST \
'http://localhost:4041/iot/devices' \
-H 'Content-Type: application/json' \
-H 'fiware-service: openiot' \
-H 'fiware-servicepath: /' \
-d '{
"devices": [
{
"device_id": "motion001",
"entity_name": "urn:ngsi-ld:Motion:001",
"entity_type": "Motion",
"attributes": [
{ "object_id": "c", "name": "count", "type": "Number" }
],
"static_attributes": [
{ "name":"refStore", "type": "Relationship", "value": "urn:ngsi-ld:Store:001"}
]
}
]
}
'
The device motion001
has been provisioned to persist data to the Context Broker as an entity of type=Motion
(instead
of the default type=Thing
). The destination entity is identified by the entity_name
field, which is set to
urn:ngsi-ld:Motion:001
. The device has a single attribute mapping for a measurement c
to be mapped to count
attribute, additionally to one defined in the group mapping (temperature
). The device also has a static attribute
refStore
which is a relationship to the entity urn:ngsi-ld:Store:001
.
This information is combined with the common config group information whenever a measurement is received at the IoT Agent and used to create or update the relevant entity in the Context Broker.
In order to see the complete process from provisioning the groups and devices to receiving measures, we will show how a
device can send a measure to the IoT Agent. In this case, we will use the IoTA JSON as an example, using the HTTP
transport protocol. To reproduce the measure sending, you can use the following curl
commands.
The device measures are sent to the South port of the IoT Agent which is listening in the port 7896
.
In this case, the device has been provisioned previously. We will use the motion001
device defined in the previous
example. To simulate the device sending a measure, the following request is sent to the South port:
curl -iX POST \
'http://localhost:7896/iot/json?k=4jggokgpepnvsb2uv4s40d59ov&i=motion001' \
-H 'Content-Type: application/json' \
-d '{"t":23,"c":1}'
The IoT Agent South port is listening to the path defined in the config group, and the API key is recognized to match.
Because the device_id
is also recognized, the provisioned device configuration is used and its settings are combined
with the config group.
Mapping has been found to use the c
measurement as count
and the t
measurement as temperature
attributes values.
The following context entity is created in the context broker:
{
"id": "urn:ngsi-ld:Motion:001",
"type": "Motion",
"temperature": { "value": 23, "type": "Number" },
"count": { "value": 1, "type": "Number" },
"refStore": { "value": "urn:ngsi-ld:Store:001", "type": "Relationship" }
}
When receiving a measure, it is not necessary to have the device provisioned. In this case, the IoT Agent will use the config group configuration to create the device and the entity. This process is called "autoprovision" and it is enabled by default in provisioned groups (for further information, review the Autoprovision section in the API documentation).
Take as an example the following request from an anonymous device:
curl -iX POST \
'http://localhost:7896/iot/json?k=4jggokgpepnvsb2uv4s40d59ov&i=dev001' \
-H 'Content-Type: application/json' \
-d '{"t":13,"c":4}'
The IoT Agent South port is listen to the path defined in the config group, and the API key is recognized to match, so
the config group configuration will be used. No device has been provisioned with the device_id=dev001
, so the IoT
Agent will only use the config group configuration.
A new entity will be created in the Context Broker with the id
Device:dev001
and the type
Device
. Only the t
measurement will be mapped to the temperature
attribute, as defined in the config group. The remaining measurements
will be created as attributes with the same name. The following context entity will be created in the context broker:
{
"id": "Device:dev001",
"type": "Device",
"temperature": { "value": 13, "type": "Number" },
"c": { "value": 4, "type": "Number" }
}