Data Persistence using Apache NIFI

This tutorial is an introduction to FIWARE Draco - an alternative generic enabler which is used to persist context data into third-party databases using Apache NIFI creating a historical view of the context. The in the same manner asthe previous tutorial, activates the dummy IoT sensors persists measurements from those sensors into a database for further analysis.

The tutorial uses cUrl commands throughout, but is also available as Postman documentation

このチュートリアルは日本語でもご覧いただけます。

Data Persistence
Architecture
Prerequisites
- Docker and Docker Compose
- Cygwin for Windows
Start Up
MongoDB - Persisting Context Data into a Database
PostgreSQL - Persisting Context Data into a Database
MySQL - Persisting Context Data into a Database
Multi-Agent - Persisting Context Data into a multiple Databases
Next Steps

Data Persistence using Apache NIFI

"Plots within plots, but all roads lead down the dragon’s gullet."

— George R.R. Martin (A Dance With Dragons)

FIWARE Draco is an alternative generic enabler which is able to persist historical context data to a series of databases. Like Cygnus - Draco is able subscribe to chnages of state from the Orion Context Broker and provide a funnel to process that data before persisting to a data sink.

As mentioned previously, persisting historical context data is useful for big data analysis or discovering trends or removing outliers. Which tool you use to do this will depend on your needs, and unlike Cygnus Draco offers a graphical interface to set up and monitor the procedure.

A summary of the differences can be seen below:

Draco	Cygnus
Offers an NGSI v2 interface for notifications	Offers an NGSI v1 interface for notifications
configurable subscription endpoint, but defaults to `/v2/notify`	subscription endpoint listens on `/notify`
listens on a single port	listens on separate ports for each input
Configured by a graphical interface	Configured via config files
Based on Apache NIFI	Based on Apache Flume
Draco is docummented here	Cygnus is documented here

Device Monitor

For the purpose of this tutorial, a series of dummy IoT devices have been created, which will be attached to the context broker. Details of the architecture and protocol used can be found in the IoT Sensors tutorial. The state of each device can be seen on the UltraLight device monitor web page found at: http://localhost:3000/device/monitor

Architecture

This application builds on the components and dummy IoT devices created in previous tutorials. It will make use of three FIWARE components - the Orion Context Broker, the IoT Agent for Ultralight 2.0 and introduce the Draco Generic Enabler for persisting context data to a database. Additional databases are now involved - both the Orion Context Broker and the IoT Agent rely on MongoDB technology to keep persistence of the information they hold, and we will be persisting our historical context data another database - either MySQL , PostgreSQL or MongoDB database.

Therefore the overall architecture will consist of the following elements:

Three FIWARE Generic Enablers:
- The FIWARE Orion Context Broker which will receive requests using NGSI-v2
- The FIWARE IoT Agent for Ultralight 2.0 which will receive northbound measurements from the dummy IoT devices in Ultralight 2.0 format and convert them to NGSI-v2 requests for the context broker to alter the state of the context entities
- FIWARE Draco which will subscribe to context changes and persist them into a database (MySQL , PostgreSQL or MongoDB)
One, two or three of the following Databases:
- The underlying MongoDB database :
  - Used by the Orion Context Broker to hold context data information such as data entities, subscriptions and registrations
  - Used by the IoT Agent to hold device information such as device URLs and Keys
  - Potentially used as a data sink to hold historical context data.
- An additional PostgreSQL database :
  - Potentially used as a data sink to hold historical context data.
- An additional MySQL database :
  - Potentially used as a data sink to hold historical context data.
Three Context Providers:
- The Stock Management Frontend is not used in this tutorial. It does the following:
  - Display store information and allow users to interact with the dummy IoT devices
  - Show which products can be bought at each store
  - Allow users to "buy" products and reduce the stock count.
- A webserver acting as set of dummy IoT devices using the Ultralight 2.0 protocol running over HTTP.
- The Context Provider NGSI proxy is not used in this tutorial. It does the following:
  - receive requests using NGSI-v2
  - makes requests to publicly available data sources using their own APIs in a proprietary format
  - returns context data back to the Orion Context Broker in NGSI-v2 format.

Since all interactions between the elements are initiated by HTTP requests, the entities can be containerized and run from exposed ports.

The specific architecture of each section of the tutorial is discussed below.

Prerequisites

Docker and Docker Compose

To keep things simple all components will be run using Docker. Docker is a container technology which allows to different components isolated into their respective environments.

To install Docker on Windows follow the instructions here
To install Docker on Mac follow the instructions here
To install Docker on Linux follow the instructions here

Docker Compose is a tool for defining and running multi-container Docker applications. A series of YAML files are used configure the required services for the application. This means all container services can be brought up in a single command. Docker Compose is installed by default as part of Docker for Windows and Docker for Mac, however Linux users will need to follow the instructions found here

You can check your current Docker and Docker Compose versions using the following commands:

docker-compose -v
docker version

Please ensure that you are using Docker version 20.10 or higher and Docker Compose 1.29 or higher and upgrade if necessary.

Cygwin for Windows

We will start up our services using a simple Bash script. Windows users should download cygwin to provide a command-line functionality similar to a Linux distribution on Windows.

Start Up

Before you start you should ensure that you have obtained or built the necessary Docker images locally. Please clone the repository and create the necessary images by running the commands as shown:

git clone https://github.com/fiware/tutorials.Historic-Context-NIFI.git
cd tutorials.Historic-Context-NIFI
git checkout NGSI-v2

./services create

Thereafter, all services can be initialized from the command-line by running the services Bash script provided within the repository:

./services <command>

Where <command> will vary depending upon the databases we wish to activate. This command will also import seed data from the previous tutorials and provision the dummy IoT sensors on startup.

ℹ️ Note: If you want to clean up and start over again you can do so with the following command:
./services stop

MongoDB - Persisting Context Data into a Database

Persisting historic context data using MongoDB technology is relatively simple to configure since we are already using a MongoDB instance to hold data related to the Orion Context Broker and the IoT Agent. The MongoDB instance is listening on the standard 27017 port and the overall architecture can be seen below:

MongoDB - Database Server Configuration

mongo-db:
    image: mongo:4.2
    hostname: mongo-db
    container_name: db-mongo
    ports:
        - "27017:27017"
    networks:
        - default

MongoDB - Draco Configuration

draco:
    image: ging/fiware-draco:1.1.0
    container_name: draco
    depends_on:
        - mongo-db
    environment:
        - NIFI_WEB_HTTP_PORT=9090
    ports:
        - "9090:9090"
        - "5050:5050"
    healthcheck:
        test: curl --fail -s http://localhost:9090/nifi-api/system-diagnostics || exit 1

The draco container is listening on two ports:

The Subscription Port for Draco - 5050 is where the service will be listening for notifications from the Orion context broker
The Web interface for Draco - 9090 is exposed purely for configuring the processors

MongoDB - Start up

To start the system with a MongoDB database only, run the following command:

./services mongodb

Then go to your browser and open Draco using this URL http://localhost:9090/nifi

Now go to the Components toolbar which is placed in the upper section of the NiFi GUI, find the template icon and drag and drop it inside the Draco user space. At this point, a popup should be displayed with a list of all the templates available. Please select the template MONGO-TUTORIAL.

Select all the processors (press shift and click on every processor) and start them by clicking on the start button. Now, you can see that the status icon of each processor turned from red to green.

Checking the Draco Service Health

Once Draco is running, you can check the status by making an HTTP request to the exposed draco port to /nifi-api/system-diagnostics. If the response is blank, this is usually because Draco is not running or is listening on another port.

1️⃣ Request:

curl -X GET \
  'http://localhost:9090/nifi-api/system-diagnostics'

Response:

The response will look similar to the following:

{
    "systemDiagnostics": {
        "aggregateSnapshot": {
            "totalNonHeap": "value",
            "totalNonHeapBytes": 0,
            "usedNonHeap": "value",
            "usedNonHeapBytes": 0,
            "freeNonHeap": "value",
            "freeNonHeapBytes": 0,
            "maxNonHeap": "value",
            "maxNonHeapBytes": 0,
            "nonHeapUtilization": "value",
            "totalHeap": "value",
            "totalHeapBytes": 0,
            "usedHeap": "value",
            "usedHeapBytes": 0,
            "freeHeap": "value",
            "freeHeapBytes": 0,
            "maxHeap": "value",
            "maxHeapBytes": 0,
            "heapUtilization": "value",
            "availableProcessors": 0,
            "processorLoadAverage": 0.0,
            "totalThreads": 0,
            "daemonThreads": 0,
            "uptime": "value",
            "flowFileRepositoryStorageUsage": {},
            "contentRepositoryStorageUsage": [{}],
            "provenanceRepositoryStorageUsage": [{}],
            "garbageCollection": [{}],
            "statsLastRefreshed": "value",
            "versionInfo": {}
        },
        "nodeSnapshots": [{}]
    }
}

Troubleshooting: What if the response is blank ?

To check that a docker container is running try
docker ps
You should see several containers running. If draco is not running, you can restart the containers as necessary.

Generating Context Data

For the purpose of this tutorial, we must be monitoring a system where the context is periodically being updated. The dummy IoT Sensors can be used to do this. Open the device monitor page at http://localhost:3000/device/monitor and unlock a Smart Door and switch on a Smart Lamp. This can be done by selecting an appropriate the command from the drop down list and pressing the send button. The stream of measurements coming from the devices can then be seen on the same page:

Subscribing to Context Changes

Once a dynamic context system is up and running, we need to inform Draco of changes in context.

This is done by making a POST request to the /v2/subscription endpoint of the Orion Context Broker.

The fiware-service and fiware-servicepath headers are used to filter the subscription to only listen to measurements from the attached IoT Sensors, since they had been provisioned using these settings
The idPattern in the request body ensures that Draco will be informed of all context data changes.
The notification url must match the configured Base Path and Listening port of the Draco Listen HTTP Processor
The throttling value defines the rate that changes are sampled.

2️⃣ Request:

curl -iX POST \
  'http://localhost:1026/v2/subscriptions' \
  -H 'Content-Type: application/json' \
  -H 'fiware-service: openiot' \
  -H 'fiware-servicepath: /' \
  -d '{
  "description": "Notify Draco of all context changes",
  "subject": {
    "entities": [
      {
        "idPattern": ".*"
      }
    ]
  },
  "notification": {
    "http": {
      "url": "http://draco:5050/v2/notify"
    }
  },
  "throttling": 5
}'

As you can see, the database used to persist context data has no impact on the details of the subscription. It is the same for each database. The response will be 201 - Created

If a subscription has been created, you can check to see if it is firing by making a GET request to the /v2/subscriptions endpoint.

3️⃣ Request:

curl -X GET \
  'http://localhost:1026/v2/subscriptions/' \
  -H 'fiware-service: openiot' \
  -H 'fiware-servicepath: /'

Response:

[
    {
        "id": "5b39d7c866df40ed84284174",
        "description": "Notify Draco of all context changes",
        "status": "active",
        "subject": {
            "entities": [
                {
                    "idPattern": ".*"
                }
            ],
            "condition": {
                "attrs": []
            }
        },
        "notification": {
            "timesSent": 158,
            "lastNotification": "2018-07-02T07:59:21.00Z",
            "attrs": [],
            "http": {
                "url": "http://draco:5050/v2/notify"
            },
            "lastSuccess": "2018-07-02T07:59:21.00Z"
        },
        "throttling": 5
    }
]

Within the notification section of the response, you can see several additional attributes which describe the health of the subscription

If the criteria of the subscription have been met, timesSent should be greater than 0. A zero value would indicate that the subject of the subscription is incorrect or the subscription has created with the wrong fiware-service-path or fiware-service header

The lastNotification should be a recent timestamp - if this is not the case, then the devices are not regularly sending data. Remember to unlock the Smart Door and switch on the Smart Lamp

The lastSuccess should match the lastNotification date - if this is not the case then Draco is not receiving the subscription properly. Check that the hostname and port are correct.

Finally, check that the status of the subscription is active - an expired subscription will not fire.

MongoDB - Reading Data from a database

To read MongoDB data from the command-line, we will need access to the mongo tool run an interactive instance of the mongo image as shown to obtain a command-line prompt:

docker run -it --network fiware_default  --entrypoint /bin/bash mongo

You can then log into to the running mongo-db database by using the command-line as shown:

mongo --host mongo-db

Show Available Databases on the MongoDB server

To show the list of available databases, run the statement as shown:

Query:

show dbs

Result:

admin          0.000GB
iotagentul     0.000GB
local          0.000GB
orion          0.000GB
orion-openiot  0.000GB
sth_openiot    0.000GB

The result include two databases admin and local which are set up by default by MongoDB, along with four databases created by the FIWARE platform. The Orion Context Broker has created two separate database instance for each fiware-service

The Store entities were created without defining a fiware-service and therefore are held within the orion database, whereas the IoT device entities were created using the openiot fiware-service header and are held separately. The IoT Agent was initialized to hold the IoT sensor data in a separate MongoDB database called iotagentul.

As a result of the subscription of Draco to Orion Context Broker, a new database has been created called sth_openiot. The default value for a Mongo DB database holding historic context consists of the sth_ prefix followed by the fiware-service header - therefore sth_openiot holds the historic context of the IoT devices.

Read Historical Context from the server

Query:

use sth_openiot
show collections

Result:

switched to db sth_openiot

sth_/_Door:001_Door
sth_/_Door:001_Door.aggr
sth_/_Lamp:001_Lamp
sth_/_Lamp:001_Lamp.aggr
sth_/_Motion:001_Motion
sth_/_Motion:001_Motion.aggr

Looking within the sth_openiot you will see that a series of tables have been created. The names of each table consist of the sth_ prefix followed by the fiware-servicepath header followed by the entity ID. Two table are created for each entity - the .aggr table holds some aggregated data which will be accessed in a later tutorial. The raw data can be seen in the tables without the .aggr suffix.

The historical data can be seen by looking at the data within each table, by default each row will contain the sampled value of a single attribute.

Query:

db["sth_/_Door:001_Door"].find().limit(10)

Result:

{ "_id" : ObjectId("5b1fa48630c49e0012f7635d"), "recvTime" : ISODate("2018-06-12T10:46:30.897Z"), "attrName" : "TimeInstant", "attrType" : "ISO8601", "attrValue" : "2018-06-12T10:46:30.836Z" }
{ "_id" : ObjectId("5b1fa48630c49e0012f7635e"), "recvTime" : ISODate("2018-06-12T10:46:30.897Z"), "attrName" : "close_status", "attrType" : "commandStatus", "attrValue" : "UNKNOWN" }
{ "_id" : ObjectId("5b1fa48630c49e0012f7635f"), "recvTime" : ISODate("2018-06-12T10:46:30.897Z"), "attrName" : "lock_status", "attrType" : "commandStatus", "attrValue" : "UNKNOWN" }
{ "_id" : ObjectId("5b1fa48630c49e0012f76360"), "recvTime" : ISODate("2018-06-12T10:46:30.897Z"), "attrName" : "open_status", "attrType" : "commandStatus", "attrValue" : "UNKNOWN" }
{ "_id" : ObjectId("5b1fa48630c49e0012f76361"), "recvTime" : ISODate("2018-06-12T10:46:30.836Z"), "attrName" : "refStore", "attrType" : "Relationship", "attrValue" : "Store:001" }
{ "_id" : ObjectId("5b1fa48630c49e0012f76362"), "recvTime" : ISODate("2018-06-12T10:46:30.836Z"), "attrName" : "state", "attrType" : "Text", "attrValue" : "CLOSED" }
{ "_id" : ObjectId("5b1fa48630c49e0012f76363"), "recvTime" : ISODate("2018-06-12T10:45:26.368Z"), "attrName" : "unlock_info", "attrType" : "commandResult", "attrValue" : " unlock OK" }
{ "_id" : ObjectId("5b1fa48630c49e0012f76364"), "recvTime" : ISODate("2018-06-12T10:45:26.368Z"), "attrName" : "unlock_status", "attrType" : "commandStatus", "attrValue" : "OK" }
{ "_id" : ObjectId("5b1fa4c030c49e0012f76385"), "recvTime" : ISODate("2018-06-12T10:47:28.081Z"), "attrName" : "TimeInstant", "attrType" : "ISO8601", "attrValue" : "2018-06-12T10:47:28.038Z" }
{ "_id" : ObjectId("5b1fa4c030c49e0012f76386"), "recvTime" : ISODate("2018-06-12T10:47:28.081Z"), "attrName" : "close_status", "attrType" : "commandStatus", "attrValue" : "UNKNOWN" }

The usual MongoDB query syntax can be used to filter appropriate fields and values. For example to read the rate at which the Motion Sensor with the id=Motion:001_Motion is accumulating, you would make a query as follows:

Query:

db["sth_/_Motion:001_Motion"].find({attrName: "count"},{_id: 0, attrType: 0, attrName: 0 } ).limit(10)

Result:

{ "recvTime" : ISODate("2018-06-12T10:46:18.756Z"), "attrValue" : "8" }
{ "recvTime" : ISODate("2018-06-12T10:46:36.881Z"), "attrValue" : "10" }
{ "recvTime" : ISODate("2018-06-12T10:46:42.947Z"), "attrValue" : "11" }
{ "recvTime" : ISODate("2018-06-12T10:46:54.893Z"), "attrValue" : "13" }
{ "recvTime" : ISODate("2018-06-12T10:47:00.929Z"), "attrValue" : "15" }
{ "recvTime" : ISODate("2018-06-12T10:47:06.954Z"), "attrValue" : "17" }
{ "recvTime" : ISODate("2018-06-12T10:47:15.983Z"), "attrValue" : "19" }
{ "recvTime" : ISODate("2018-06-12T10:47:49.090Z"), "attrValue" : "23" }
{ "recvTime" : ISODate("2018-06-12T10:47:58.112Z"), "attrValue" : "25" }
{ "recvTime" : ISODate("2018-06-12T10:48:28.218Z"), "attrValue" : "29" }

To leave the MongoDB client and leave interactive mode, run the following:

exit

exit

PostgreSQL - Persisting Context Data into a Database

To persist historic context data into an alternative database such as PostgreSQL, we will need an additional container which hosts the PostgreSQL server - the default Docker image for this data can be used. The PostgreSQL instance is listening on the standard 5432 port and the overall architecture can be seen below:

We now have a system with two databases, since the MongoDB container is still required to hold data related to the Orion Context Broker and the IoT Agent.

PostgreSQL - Database Server Configuration

postgres-db:
    image: postgres:latest
    hostname: postgres-db
    container_name: db-postgres
    expose:
        - "5432"
    ports:
        - "5432:5432"
    networks:
        - default
    environment:
        - "POSTGRES_PASSWORD=password"
        - "POSTGRES_USER=postgres"
        - "POSTGRES_DB=postgres"

The postgres-db container is listening on a single port:

Port 5432 is the default port for a PostgreSQL server. It has been exposed so you can also run the pgAdmin4 tool to display database data if you wish

The postgres-db container is driven by environment variables as shown:

Key	Value.	Description
POSTGRES_PASSWORD	`password`	Password for the PostgreSQL database user
POSTGRES_USER	`postgres`	Username for the PostgreSQL database user
POSTGRES_DB	`postgres`	The name of the PostgreSQL database

ℹ️ Note: Passing the Username and Password in plain text environment variables like this is a security risk. Whereas this is acceptable practice in a tutorial, for a production environment, you can avoid this risk by applying Docker Secrets

PostgreSQL - Draco Configuration

draco:
    image: ging/fiware-draco:1.1.0
    container_name: draco
    depends_on:
        - postgres-db
    environment:
        - NIFI_WEB_HTTP_PORT=9090
    ports:
        - "9090:9090"
        - "5050:5050"
    healthcheck:
        test: curl --fail -s http://localhost:9090/nifi-api/system-diagnostics || exit 1

The draco container is listening on two ports:

The Subscription Port for Draco - 5050 is where the service will be listening for notifications from the Orion context broker
The Web interface for Draco - 9090 is exposed purely for configuring the processors.

PostgreSQL - Start up

To start the system with a PostgreSQL database run the following command:

./services postgres

Then go to your browser and open Draco using this URL http://localhost:9090/nifi

Now go to the Components toolbar which is placed in the upper section of the NiFi GUI, find the template icon and drag and drop it inside the Draco user space. At this point, a popup should be displayed with a list of all the templates available. Please select the template POSTGRESQL-TUTORIAL.

Before starting the processors, you need to set your PostgreSQL password and enable the DBCConnectionPool controller. For doing that please follow the instructions:

Do right click on any part of the Draco GUI user space, and then click on configure.
Go to the Controller Services Tab, at this point a list of controllers should be displayed, locate the DBCConnectionPool controller.
Click on the configuration button of the "DBCPConnectionPool"
Go to the controller Properties tab and put "password" in the password field, then apply the changes.
Enable the processor by clicking on the thunder icon and then click on enable, then close the controller configuration page.

Select all the processors (press shift and click on every processor) and start them by clicking on the start button. Now, you can see that the status icon of each processor turned from red to green.

Checking the Draco Service Health

Once Draco is running, you can check the status by making an HTTP request to the exposed draco port to /nifi-api/system-diagnostics. If the response is blank, this is usually because Draco is not running or is listening on another port.

1️⃣ Request:

curl -X GET \
  'http://localhost:9090/nifi-api/system-diagnostics'

Response:

The response will look similar to the following:

{
    "systemDiagnostics": {
        "aggregateSnapshot": {
            "totalNonHeap": "value",
            "totalNonHeapBytes": 0,
            "usedNonHeap": "value",
            "usedNonHeapBytes": 0,
            "freeNonHeap": "value",
            "freeNonHeapBytes": 0,
            "maxNonHeap": "value",
            "maxNonHeapBytes": 0,
            "nonHeapUtilization": "value",
            "totalHeap": "value",
            "totalHeapBytes": 0,
            "usedHeap": "value",
            "usedHeapBytes": 0,
            "freeHeap": "value",
            "freeHeapBytes": 0,
            "maxHeap": "value",
            "maxHeapBytes": 0,
            "heapUtilization": "value",
            "availableProcessors": 0,
            "processorLoadAverage": 0.0,
            "totalThreads": 0,
            "daemonThreads": 0,
            "uptime": "value",
            "flowFileRepositoryStorageUsage": {},
            "contentRepositoryStorageUsage": [{}],
            "provenanceRepositoryStorageUsage": [{}],
            "garbageCollection": [{}],
            "statsLastRefreshed": "value",
            "versionInfo": {}
        },
        "nodeSnapshots": [{}]
    }
}

Troubleshooting: What if the response is blank ?

To check that a docker container is running try
docker ps
You should see several containers running. If draco is not running, you can restart the containers as necessary.

Generating Context Data

For the purpose of this tutorial, we must be monitoring a system where the context is periodically being updated. The dummy IoT Sensors can be used to do this. Open the device monitor page at http://localhost:3000/device/monitor and unlock a Smart Door and switch on a Smart Lamp. This can be done by selecting an appropriate the command from the drop down list and pressing the send button. The stream of measurements coming from the devices can then be seen on the same page:

Subscribing to Context Changes

Once a dynamic context system is up and running, we need to inform Draco of changes in context.

This is done by making a POST request to the /v2/subscription endpoint of the Orion Context Broker.

The fiware-service and fiware-servicepath headers are used to filter the subscription to only listen to measurements from the attached IoT Sensors, since they had been provisioned using these settings
The idPattern in the request body ensures that Draco will be informed of all context data changes.
The throttling value defines the rate that changes are sampled.

5️⃣ Request:

curl -iX POST \
  'http://localhost:1026/v2/subscriptions' \
  -H 'Content-Type: application/json' \
  -H 'fiware-service: openiot' \
  -H 'fiware-servicepath: /' \
  -d '{
  "description": "Notify Draco of all context changes",
  "subject": {
    "entities": [
      {
        "idPattern": ".*"
      }
    ]
  },
  "notification": {
    "http": {
      "url": "http://draco:5050/v2/notify"
    }
  },
  "throttling": 5
}'

As you can see, the database used to persist context data has no impact on the details of the subscription. It is the same for each database. The response will be 201 - Created

PostgreSQL - Reading Data from a database

To read PostgreSQL data from the command-line, we will need access to the postgres client, to do this, run an interactive instance of the postgresql-client image supplying the connection string as shown to obtain a command-line prompt:

docker run -it --rm  --network fiware_default jbergknoff/postgresql-client \
   postgresql://postgres:password@postgres-db:5432/postgres

Show Available Databases on the PostgreSQL server

To show the list of available databases, run the statement as shown:

Query:

\list

Result:

   Name    |  Owner   | Encoding |  Collate   |   Ctype    |   Access privileges
-----------+----------+----------+------------+------------+-----------------------
 postgres  | postgres | UTF8     | en_US.utf8 | en_US.utf8 |
 template0 | postgres | UTF8     | en_US.utf8 | en_US.utf8 | =c/postgres          +
           |          |          |            |            | postgres=CTc/postgres
 template1 | postgres | UTF8     | en_US.utf8 | en_US.utf8 | =c/postgres          +
           |          |          |            |            | postgres=CTc/postgres
(3 rows)

The result include two template databases template0 and template1 as well as the postgres database setup when the docker container was started.

To show the list of available schemas, run the statement as shown:

Query:

\dn

Result:

  List of schemas
  Name   |  Owner
---------+----------
 openiot | postgres
 public  | postgres
(2 rows)

As a result of the subscription of Draco to Orion Context Broker, a new schema has been created called openiot. The name of the schema matches the fiware-service header - therefore openiot holds the historic context of the IoT devices.

Read Historical Context from the PostgreSQL server

Once running a docker container within the network, it is possible to obtain information about the running database.

Query:

SELECT table_schema,table_name
FROM information_schema.tables
WHERE table_schema ='openiot'
ORDER BY table_schema,table_name;

Result:

 table_schema |    table_name
--------------+-------------------
 openiot      | door_001_door
 openiot      | lamp_001_lamp
 openiot      | motion_001_motion
(3 rows)

The table_schema matches the fiware-service header supplied with the context data:

To read the data within a table, run a select statement as shown:

Query:

SELECT * FROM openiot.motion_001_motion limit 10;

Result:

  recvtimets   |         recvtime         | fiwareservicepath |  entityid  | entitytype |  attrname   |   attrtype   |        attrvalue         |                                    attrmd
---------------+--------------------------+-------------------+------------+------------+-------------+--------------+--------------------------+------------------------------------------------------------------------------
 1528803005491 | 2018-06-12T11:30:05.491Z | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:30:05.423Z | []
 1528803005491 | 2018-06-12T11:30:05.491Z | /                 | Motion:001 | Motion     | count       | Integer      | 7                        | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:30:05.423Z"}]
 1528803005491 | 2018-06-12T11:30:05.491Z | /                 | Motion:001 | Motion     | refStore    | Relationship | Store:001                | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:30:05.423Z"}]
 1528803035501 | 2018-06-12T11:30:35.501Z | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:30:35.480Z | []
 1528803035501 | 2018-06-12T11:30:35.501Z | /                 | Motion:001 | Motion     | count       | Integer      | 10                       | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:30:35.480Z"}]
 1528803035501 | 2018-06-12T11:30:35.501Z | /                 | Motion:001 | Motion     | refStore    | Relationship | Store:001                | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:30:35.480Z"}]
 1528803041563 | 2018-06-12T11:30:41.563Z | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:30:41.520Z | []
 1528803041563 | 2018-06-12T11:30:41.563Z | /                 | Motion:001 | Motion     | count       | Integer      | 12                       | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:30:41.520Z"}]
 1528803041563 | 2018-06-12T11:30:41.563Z | /                 | Motion:001 | Motion     | refStore    | Relationship | Store:001                | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:30:41.520Z"}]
 1528803047545 | 2018-06-12T11:30:47.545Z | /

The usual PostgreSQL query syntax can be used to filter appropriate fields and values. For example to read the rate at which the Motion Sensor with the id=Motion:001_Motion is accumulating, you would make a query as follows:

Query:

SELECT recvtime, attrvalue FROM openiot.motion_001_motion WHERE attrname ='count'  limit 10;

Result:

         recvtime         | attrvalue
--------------------------+-----------
 2018-06-12T11:30:05.491Z | 7
 2018-06-12T11:30:35.501Z | 10
 2018-06-12T11:30:41.563Z | 12
 2018-06-12T11:30:47.545Z | 13
 2018-06-12T11:31:02.617Z | 15
 2018-06-12T11:31:32.718Z | 20
 2018-06-12T11:31:38.733Z | 22
 2018-06-12T11:31:50.780Z | 24
 2018-06-12T11:31:56.825Z | 25
 2018-06-12T11:31:59.790Z | 26
(10 rows)

To leave the Postgres client and leave interactive mode, run the following:

\q

You will then return to the command-line.

MySQL - Persisting Context Data into a Database

Similarly, to persisting historic context data into MySQL, we will again need an additional container which hosts the MySQL server, once again the default Docker image for this data can be used. The MySQL instance is listening on the standard 3306 port and the overall architecture can be seen below:

Once again we have a system with two databases, since the MongoDB container is still required to hold data related to the Orion Context Broker and the IoT Agent.

MySQL - Database Server Configuration

mysql-db:
    restart: always
    image: mysql:5.7
    hostname: mysql-db
    container_name: db-mysql
    expose:
        - "3306"
    ports:
        - "3306:3306"
    networks:
        - default
    environment:
        - "MYSQL_ROOT_PASSWORD=123"
        - "MYSQL_ROOT_HOST=%"

ℹ️ Note: Using the default root user and displaying the password in an environment variables like this is a security risk. Whereas this is acceptable practice in a tutorial, for a production environment, you can avoid this risk by setting up another user and applying Docker Secrets

The mysql-db container is listening on a single port:

Port 3306 is the default port for a MySQL server. It has been exposed so you can also run other database tools to display data if you wish

The mysql-db container is driven by environment variables as shown:

Key	Value.	Description
MYSQL_ROOT_PASSWORD	`123`.	specifies a password that is set for the MySQL `root` account.
MYSQL_ROOT_HOST	`postgres`	By default, MySQL creates the `root'@'localhost` account. This account can only be connected to from inside the container. Setting this environment variable allows root connections from other hosts

MySQL - Draco Configuration

draco:
    image: ging/fiware-draco:1.1.0
    container_name: draco
    depends_on:
        - mysql-db
    environment:
        - NIFI_WEB_HTTP_PORT=9090
    ports:
        - "9090:9090"
        - "5050:5050"
    healthcheck:
        test: curl --fail -s http://localhost:9090/nifi-api/system-diagnostics || exit 1

The draco container is listening on two ports:

The Subscription Port for Draco - 5050 is where the service will be listening for notifications from the Orion context broker
The Web interface for Draco - 9090 is exposed purely for configuring the processors

MySQL - Start up

To start the system with a MySQL database run the following command:

./services mysql

Then go to your browser and open Draco using this URL http://localhost:9090/nifi

Now go to the Components toolbar which is placed in the upper section of the NiFi GUI, find the template icon and drag and drop it inside the Draco user space. At this point, a popup should be displayed with a list of all the templates available. Please select the template MYSQL-TUTORIAL.

Before starting the processors, you need to set your MySQL password and enable the DBCConnectionPool controller. For doing that please follow the instructions:

Do right click on any part of the Draco GUI user space, and then click on configure.
Go to the Controller Services Tab, at this point a list of controllers should be displayed, locate the DBCConnectionPool controller.
Click on the configuration button of the "DBCPConnectionPool"
Go to the controller Properties tab and put "123" in the password field, then apply the changes.
Enable the processor by clicking on the thunder icon and then click on enable, then close the controller configuration page.

Select all the processors (press shift and click on every processor) and start them by clicking on the start button. Now, you can see that the status icon of each processor turned from red to green.

Checking the Draco Service Health

Once Draco is running, you can check the status by making an HTTP request to the exposed draco port to /system-diagnostics. If the response is blank, this is usually because Draco is not running or is listening on another port.

1️⃣ Request:

curl -X GET \
  'http://localhost:9090/system-diagnostics'

Response:

The response will look similar to the following:

{
    "systemDiagnostics": {
        "aggregateSnapshot": {
            "totalNonHeap": "value",
            "totalNonHeapBytes": 0,
            "usedNonHeap": "value",
            "usedNonHeapBytes": 0,
            "freeNonHeap": "value",
            "freeNonHeapBytes": 0,
            "maxNonHeap": "value",
            "maxNonHeapBytes": 0,
            "nonHeapUtilization": "value",
            "totalHeap": "value",
            "totalHeapBytes": 0,
            "usedHeap": "value",
            "usedHeapBytes": 0,
            "freeHeap": "value",
            "freeHeapBytes": 0,
            "maxHeap": "value",
            "maxHeapBytes": 0,
            "heapUtilization": "value",
            "availableProcessors": 0,
            "processorLoadAverage": 0.0,
            "totalThreads": 0,
            "daemonThreads": 0,
            "uptime": "value",
            "flowFileRepositoryStorageUsage": {},
            "contentRepositoryStorageUsage": [{}],
            "provenanceRepositoryStorageUsage": [{}],
            "garbageCollection": [{}],
            "statsLastRefreshed": "value",
            "versionInfo": {}
        },
        "nodeSnapshots": [{}]
    }
}

Troubleshooting: What if the response is blank ?

To check that a docker container is running try
docker ps
You should see several containers running. If draco is not running, you can restart the containers as necessary.

Generating Context Data

For the purpose of this tutorial, we must be monitoring a system where the context is periodically being updated. The dummy IoT Sensors can be used to do this. Open the device monitor page at http://localhost:3000/device/monitor and unlock a Smart Door and switch on a Smart Lamp. This can be done by selecting an appropriate the command from the drop down list and pressing the send button. The stream of measurements coming from the devices can then be seen on the same page:

Subscribing to Context Changes

Once a dynamic context system is up and running, we need to inform Draco of changes in context.

This is done by making a POST request to the /v2/subscription endpoint of the Orion Context Broker.

The fiware-service and fiware-servicepath headers are used to filter the subscription to only listen to measurements from the attached IoT Sensors, since they had been provisioned using these settings
The idPattern in the request body ensures that Draco will be informed of all context data changes.
The throttling value defines the rate that changes are sampled.

7️⃣ Request:

curl -iX POST \
  'http://localhost:1026/v2/subscriptions' \
  -H 'Content-Type: application/json' \
  -H 'fiware-service: openiot' \
  -H 'fiware-servicepath: /' \
  -d '{
  "description": "Notify Draco of all context changes",
  "subject": {
    "entities": [
      {
        "idPattern": ".*"
      }
    ]
  },
  "notification": {
    "http": {
      "url": "http://draco:5050/v2/notify"
    }
  },
  "throttling": 5
}'

As you can see, the database used to persist context data has no impact on the details of the subscription. It is the same for each database. The response will be 201 - Created

MySQL - Reading Data from a database

To read MySQL data from the command-line, we will need access to the mysql client, to do this, run an interactive instance of the mysql image supplying the connection string as shown to obtain a command-line prompt:

docker exec -it  db-mysql mysql -h mysql-db -P 3306  -u root -p123

Show Available Databases on the MySQL server

To show the list of available databases, run the statement as shown:

Query:

SHOW DATABASES;

Result:

+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| openiot            |
| performance_schema |
| sys                |
+--------------------+
5 rows in set (0.00 sec)

To show the list of available schemas, run the statement as shown:

Query:

SHOW SCHEMAS;

Result:

+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| openiot            |
| performance_schema |
| sys                |
+--------------------+
5 rows in set (0.00 sec)

As a result of the subscription of Draco to Orion Context Broker, a new schema has been created called openiot. The name of the schema matches the fiware-service header - therefore openiot holds the historic context of the IoT devices.

Read Historical Context from the MySQL server

Once running a docker container within the network, it is possible to obtain information about the running database.

Query:

SHOW tables FROM openiot;

Result:

 table_schema |    table_name
--------------+-------------------
 openiot      | door_001_door
 openiot      | lamp_001_lamp
 openiot      | motion_001_motion
(3 rows)

The table_schema matches the fiware-service header supplied with the context data:

To read the data within a table, run a select statement as shown:

Query:

SELECT * FROM openiot.Motion_001_Motion limit 10;

Result:

+---------------+-------------------------+-------------------+------------+------------+-------------+--------------+--------------------------+------------------------------------------------------------------------------+
| recvTimeTs    | recvTime                | fiwareServicePath | entityId   | entityType | attrName    | attrType     | attrValue                | attrMd                                                                       |
+---------------+-------------------------+-------------------+------------+------------+-------------+--------------+--------------------------+------------------------------------------------------------------------------+
| 1528804397955 | 2018-06-12T11:53:17.955 | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:53:17.923Z | []                                                                           |
| 1528804397955 | 2018-06-12T11:53:17.955 | /                 | Motion:001 | Motion     | count       | Integer      | 3                        | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:53:17.923Z"}] |
| 1528804397955 | 2018-06-12T11:53:17.955 | /                 | Motion:001 | Motion     | refStore    | Relationship | Store:001                | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:53:17.923Z"}] |
| 1528804403954 | 2018-06-12T11:53:23.954 | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:53:23.928Z | []                                                                           |
| 1528804403954 | 2018-06-12T11:53:23.954 | /                 | Motion:001 | Motion     | count       | Integer      | 5                        | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:53:23.928Z"}] |
| 1528804403954 | 2018-06-12T11:53:23.954 | /                 | Motion:001 | Motion     | refStore    | Relationship | Store:001                | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:53:23.928Z"}] |
| 1528804409970 | 2018-06-12T11:53:29.970 | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:53:29.948Z | []                                                                           |
| 1528804409970 | 2018-06-12T11:53:29.970 | /                 | Motion:001 | Motion     | count       | Integer      | 7                        | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:53:29.948Z"}] |
| 1528804409970 | 2018-06-12T11:53:29.970 | /                 | Motion:001 | Motion     | refStore    | Relationship | Store:001                | [{"name":"TimeInstant","type":"ISO8601","value":"2018-06-12T11:53:29.948Z"}] |
| 1528804446083 | 2018-06-12T11:54:06.83  | /                 | Motion:001 | Motion     | TimeInstant | ISO8601      | 2018-06-12T11:54:06.062Z | []                                                                           |
+---------------+-------------------------+-------------------+------------+------------+-------------+--------------+--------------------------+------------------------------------------------------------------------------+

The usual MySQL query syntax can be used to filter appropriate fields and values. For example to read the rate at which the Motion Sensor with the id=Motion:001_Motion is accumulating, you would make a query as follows:

Query:

SELECT recvtime, attrvalue FROM openiot.Motion_001_Motion WHERE attrname ='count' LIMIT 10;

Result:

+-------------------------+-----------+
| recvtime                | attrvalue |
+-------------------------+-----------+
| 2018-06-12T11:53:17.955 | 3         |
| 2018-06-12T11:53:23.954 | 5         |
| 2018-06-12T11:53:29.970 | 7         |
| 2018-06-12T11:54:06.83  | 12        |
| 2018-06-12T11:54:12.132 | 13        |
| 2018-06-12T11:54:24.177 | 14        |
| 2018-06-12T11:54:36.196 | 16        |
| 2018-06-12T11:54:42.195 | 18        |
| 2018-06-12T11:55:24.300 | 23        |
| 2018-06-12T11:55:30.350 | 25        |
+-------------------------+-----------+
10 rows in set (0.00 sec)

To leave the MySQL client and leave interactive mode, run the following:

\q

You will then return to the command-line.

Multi-Agent - Persisting Context Data into a multiple Databases

It is also possible to configure Draco to populate multiple databases simultaneously. We can combine the architecture from the three previous examples and configure Draco to store data in multiple sinks.

We now have a system with three databases, PostgreSQL and MySQL for data persistence and MongoDB for both data persistence and holding data related to the Orion Context Broker and the IoT Agent.

Multi-Agent - Draco Configuration for Multiple Databases

draco:
    image: ging/fiware-draco:1.1.0
    container_name: draco
    depends_on:
        - mysql-db
        - mongo-db
        - postgres-db
    environment:
        - NIFI_WEB_HTTP_PORT=9090
    ports:
        - "9090:9090"
        - "5050:5050"
    healthcheck:
        test: curl --fail -s http://localhost:9090/nifi-api/system-diagnostics || exit 1

The draco container is listening on two ports:

The Subscription Port for Draco - 5050 is where the service will be listening for notifications from the Orion context broker
The Web interface for Draco - 9090 is exposed purely for configuring the processors

Multi-Agent - Start up

To start the system with multiple databases run the following command:

./services multiple

Then go to your browser and open Draco using this URL http://localhost:9090/nifi

Now go to the Components toolbar which is placed in the upper section of the NiFi GUI, find the template icon and drag and drop it inside the Draco user space. At this point, a popup should be displayed with a list of all the templates available. Please select the template MULTIPLE-SINKS-TUTORIAL.

Now repeat the process for setting the password in the controller "DBCPConnectionPool" each connection MySQL and PostgreSQL

Select all the processors (press shift and click on every processor) and start them by clicking on the start button. Now, you can see that the status icon of each processor turned from red to green.

Checking the Draco Service Health

Once Draco is running, you can check the status by making an HTTP request to the exposed draco port to /system-diagnostics. If the response is blank, this is usually because Draco is not running or is listening on another port.

1️⃣ Request:

curl -X GET \
  'http://localhost:9090/nifi-api/system-diagnostics'

Response:

The response will look similar to the following:

{
    "systemDiagnostics": {
        "aggregateSnapshot": {
            "totalNonHeap": "value",
            "totalNonHeapBytes": 0,
            "usedNonHeap": "value",
            "usedNonHeapBytes": 0,
            "freeNonHeap": "value",
            "freeNonHeapBytes": 0,
            "maxNonHeap": "value",
            "maxNonHeapBytes": 0,
            "nonHeapUtilization": "value",
            "totalHeap": "value",
            "totalHeapBytes": 0,
            "usedHeap": "value",
            "usedHeapBytes": 0,
            "freeHeap": "value",
            "freeHeapBytes": 0,
            "maxHeap": "value",
            "maxHeapBytes": 0,
            "heapUtilization": "value",
            "availableProcessors": 0,
            "processorLoadAverage": 0.0,
            "totalThreads": 0,
            "daemonThreads": 0,
            "uptime": "value",
            "flowFileRepositoryStorageUsage": {},
            "contentRepositoryStorageUsage": [{}],
            "provenanceRepositoryStorageUsage": [{}],
            "garbageCollection": [{}],
            "statsLastRefreshed": "value",
            "versionInfo": {}
        },
        "nodeSnapshots": [{}]
    }
}

Troubleshooting: What if the response is blank ?

To check that a docker container is running try
docker ps
You should see several containers running. If draco is not running, you can restart the containers as necessary.

Generating Context Data

For the purpose of this tutorial, we must be monitoring a system where the context is periodically being updated. The dummy IoT Sensors can be used to do this. Open the device monitor page at http://localhost:3000/device/monitor and unlock a Smart Door and switch on a Smart Lamp. This can be done by selecting an appropriate the command from the drop down list and pressing the send button. The stream of measurements coming from the devices can then be seen on the same page:

Subscribing to Context Changes

Once a dynamic context system is up and running, we need to inform Draco of changes in context.

This is done by making a POST request to the /v2/subscription endpoint of the Orion Context Broker.

The fiware-service and fiware-servicepath headers are used to filter the subscription to only listen to measurements from the attached IoT Sensors
The idPattern in the request body ensures that Draco will be informed of all context data changes.
The throttling value defines the rate that changes are sampled.

9️⃣ Request:

curl -iX POST \
  'http://localhost:1026/v2/subscriptions' \
  -H 'Content-Type: application/json' \
  -H 'fiware-service: openiot' \
  -H 'fiware-servicepath: /' \
  -d '{
  "description": "Notify Draco of all context changes",
  "subject": {
    "entities": [
      {
        "idPattern": ".*"
      }
    ]
  },
  "notification": {
    "http": {
      "url": "http://draco:5050/v2/notify"
    }
  },
  "throttling": 5
}'

As you can see, the database used to persist context data has no impact on the details of the subscription. It is the same for each database. The response will be 201 - Created

Multi-Agent - Reading Persisted Data

To read persisted data from the attached databases, please refer to the previous sections of this tutorial.

Next Steps

Want to learn how to add more complexity to your application by adding advanced features? You can find out by reading the other tutorials in this series

Name		Name	Last commit message	Last commit date
Latest commit History 79 Commits
.github/workflows		.github/workflows
docker-compose		docker-compose
.dockerignore		.dockerignore
.env		.env
.gitignore		.gitignore
FIWARE Persisting Context Data using NIFI.postman_collection.json		FIWARE Persisting Context Data using NIFI.postman_collection.json
LICENSE		LICENSE
README.ja.md		README.ja.md
README.md		README.md
import-data		import-data
provision-devices		provision-devices
services		services

License

SBlechmann/tutorials.Historic-Context-NIFI

Folders and files

Latest commit

History

Repository files navigation

Contents

Data Persistence using Apache NIFI

Device Monitor

Architecture

Prerequisites

Docker and Docker Compose

Cygwin for Windows

Start Up

MongoDB - Persisting Context Data into a Database

MongoDB - Database Server Configuration

MongoDB - Draco Configuration

MongoDB - Start up

Checking the Draco Service Health

1️⃣ Request:

Response:

Generating Context Data

Subscribing to Context Changes

2️⃣ Request:

3️⃣ Request:

Response:

MongoDB - Reading Data from a database

Show Available Databases on the MongoDB server

Query:

Result:

Read Historical Context from the server

Query:

Result:

Query:

Result:

Query:

Result:

PostgreSQL - Persisting Context Data into a Database

PostgreSQL - Database Server Configuration

PostgreSQL - Draco Configuration

PostgreSQL - Start up

Checking the Draco Service Health

1️⃣ Request:

Response:

Generating Context Data

Subscribing to Context Changes

5️⃣ Request:

PostgreSQL - Reading Data from a database

Show Available Databases on the PostgreSQL server

Query:

Result:

Query:

Result:

Read Historical Context from the PostgreSQL server

Query:

Result:

Query:

Result:

Query:

Result:

MySQL - Persisting Context Data into a Database

MySQL - Database Server Configuration

MySQL - Draco Configuration

MySQL - Start up

Checking the Draco Service Health

1️⃣ Request:

Response:

Generating Context Data

Subscribing to Context Changes

7️⃣ Request:

MySQL - Reading Data from a database

Show Available Databases on the MySQL server

Query:

Result:

Query:

Result:

Read Historical Context from the MySQL server

Query:

Result:

Query:

Packages