Getting started tutorial

Run the following command to enable fleet indexing and specify the data sources to search from.

aws iot update-indexing-configuration \
--thing-indexing-configuration "thingIndexingMode=REGISTRY_AND_SHADOW,customFields=[{name=attributes.temperature,type=Number},{name=attributes.rackId,type=String},{name=attributes.stateNormal,type=Boolean}],thingConnectivityIndexingMode=STATUS" \

The preceding example CLI command enables fleet indexing to support searching registry data, shadow data, and thing connectivity status using the AWS_Things index.

The configuration change can take a few minutes to complete. Verify that your fleet indexing is enabled before you create fleet metrics.

To check if your fleet indexing has been enabled, run the following CLI command:

aws --region us-east-1 iot describe-index --index-name "AWS_Things"

For more information, see Enable thing indexing.

Run the following bash script to create ten things and describe them.

# Bash script. Type `bash` before running in other shells.

Temperatures=(70 71 72 73 74 75 47 97 98 99)
Racks=(Rack1 Rack1 Rack2 Rack2 Rack3 Rack4 Rack5 Rack6 Rack6 Rack6)
IsNormal=(true true true true true true false false false false)

for ((i=0; i < 10; i++))
do
  thing=$(aws iot create-thing --thing-name "TempSensor$i" --attribute-payload attributes="{temperature=${Temperatures[@]:$i:1},rackId=${Racks[@]:$i:1},stateNormal=${IsNormal[@]:$i:1}}")
  aws iot describe-thing --thing-name "TempSensor$i"
done

This script creates ten things to represent ten sensors. Each thing has attributes of temperature, rackId, and stateNormal as described in the following table:

The output of this script contains ten JSON files. One of the JSON file looks like the following:

{
    "version": 1, 
    "thingName": "TempSensor0", 
    "defaultClientId": "TempSensor0", 
    "attributes": {
        "rackId": "Rack1", 
        "stateNormal": "true", 
        "temperature": "70"
    }, 
    "thingArn": "arn:aws:iot:region:account:thing/TempSensor0", 
    "thingId": "example-thing-id"
}

For more information, see Create a thing.

Run the following command to create a fleet metric named high_temp_FM. You create the fleet metric to monitor the number of sensors with temperatures exceeding 80 degrees Fahrenheit in CloudWatch.

aws iot create-fleet-metric --metric-name "high_temp_FM" --query-string "thingName:TempSensor* AND attributes.temperature >80" --period 60 --aggregation-field "attributes.temperature" --aggregation-type name=Statistics,values=count

--metric-name

Data type: string. The --metric-name parameter specifies a fleet metric name. In this example, you're creating a fleet metric named high_temp_FM.

--query-string

Data type: string. The --query-string parameter specifies the query string. In this example, the query string means to query all the things with names starting with TempSensor and with temperatures higher than 80 degrees Fahrenheit. For more information, see Query syntax.

--period

Data type: integer. The --period parameter specifies the time to retrieve the aggregated data in seconds. In this example, you specify that the fleet metric you're creating retrieves the aggregated data every 60 seconds.

--aggregation-field

Data type: string. The --aggregation-field parameter specifies the attribute to evaluate. In this example, the temperature attribute is to be evaluated.

--aggregation-type

The --aggregation-type parameter specifies the statistical summary to display in the fleet metric. For your monitoring tasks, you can customize the aggregation query properties for the different aggregation types (Statistics, Cardinality, and Percentile). In this example, you specify count for the aggregation type and Statistics to return the count of devices that have attributes that match the query, in other words, to return the count of the devices with names starting with TempSensor and with temperatures higher than 80 degrees Fahrenheit. For more information, see Querying for aggregate data.

The output of this command looks like the following:

{
    "metricArn": "arn:aws:iot:region:111122223333:fleetmetric/high_temp_FM", 
    "metricName": "high_temp_FM"
}

To learn more about how to create a fleet metric, read Managing fleet metrics.

If you can't create a fleet metric, read Troubleshooting fleet metrics.

(Optional) Run the following command to describe your fleet metric named high_temp_FM:

aws iot describe-fleet-metric --metric-name "high_temp_FM"

The output of this command looks like the following:

{
    "queryVersion": "2017-09-30", 
    "lastModifiedDate": 1625249775.834, 
    "queryString": "*", 
    "period": 60, 
    "metricArn": "arn:aws:iot:region:111122223333:fleetmetric/high_temp_FM", 
    "aggregationField": "registry.version", 
    "version": 1, 
    "aggregationType": {
        "values": [
            "count"
        ], 
        "name": "Statistics"
    }, 
    "indexName": "AWS_Things", 
    "creationDate": 1625249775.834, 
    "metricName": "high_temp_FM"
}

Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

On the CloudWatch menu on the left panel, choose Metrics to expand the submenu and then choose All metrics. This opens the page with the upper half to display the graph and the lower half containing four tabbed sections.

The first tabbed section All metrics lists all the metrics that you can view in groups, choose IoTFleetMetrics. This contains all of your fleet metrics.

On the Aggregation type section of the All metrics tab, choose Aggregation type to view all the fleet metrics you created.

Choose the fleet metric to display graph on the left of the Aggregation type section. You will see the value count to the left of your Metric name, and this is the value of the aggregation type that you specified in the Create fleet metrics section of this tutorial.

Choose the second tab named Graphed metrics to the right of the All metrics tab to view the fleet metric you chose from the previous step.

You should be able to see a graph that displays the number of sensors with temperatures higher than 80 degrees Fahrenheit like the following:

(Optional) You can set a metric alarm.