Arbeiten mit dem Firehose-Delivery-Stream

Datenstreaming mit AWS Kinesis und Lambda

Maksim Pecherskiy

Data Engineer

Bereit zum Erstellen des Streams

Role-ARN abrufen

boto3-Client initialisieren

import boto3

firehose = boto3.client('firehose', 
                        aws_access_key_id=AWS_KEY_ID, 
                        aws_secret_access_key=AWS_SECRET, 
                        region_name='us-east-1')

Stream erstellen!

res = firehose.create_delivery_stream(

    DeliveryStreamName = "gps-delivery-stream",

    DeliveryStreamType = "DirectPut",

    S3DestinationConfiguration = {

        "RoleARN": "arn:aws:iam::0000000:role/firehoseDeliveryRole",

        "BucketARN": "arn:aws:s3:::sd-vehicle-data"

  }

)

Antwort: Stream erstellt

print(res['DeliveryStreamARN'])

# ARN des neuen Streams
"arn:aws:firehose:us-east-1:0000000:deliverystream/gps-delivery-stream"

Stream ist bereit

In den Stream schreiben

Telematik-Hardware

Telematikdaten senden

Einzelner Datensatz

{

    'record_id': '939ed1d1-1740-420c-8906-445278573c7f', # <-- Eindeutige Datensatz-ID

    'timestamp': '4:25:06.000', # <-- Messzeit

    'vin': '4FTEX4944AK844294', # <-- Fahrzeug-ID

    'lon': 106.9447146, # <-- Längengrad des Fahrzeugs

    'lat': -6.3385652, # <-- Breitengrad des Fahrzeugs

    'speed': 25 # <-- Fahrzeuggeschwindigkeit

}

Eingehende Datensätze

Weiterer Use Case

Muster

Datensatz senden

res = firehose.put_record(

    DeliveryStreamName='gps-delivery-stream',

    Record = {
        'Data': payload 
    }

)

Datensatz senden

    Record = {
        'Data': payload 
    }

Datensatz senden

So sieht unser Datensatz aus

record = {
 'record_id': '939ed1d1-1740-420c-8906-445278573c7f',
 'timestamp': '4:25:06.000','vin': '4FTEX4944AK844294',
 'lon': 106.9447146,'lat': -6.338565200000001,
 'speed': 25}

Was wir senden wollen (ein String)

"939ed1d1-1740-420c-8906-445278573c7f 4:25:06.000 
4FTEX4944AK844294 106.9447146 -6.338565200000001 25"

Datensatz senden

payload = " ".join(
    str(value) for value in record.values()
)

print(payload)

"939ed1d1-1740-420c-8906-445278573c7f 4:25:06.000 
4FTEX4944AK844294 106.9447146 -6.338565200000001 25"

Alles zusammenführen

record = {
 'record_id': '939ed1d1-1740-420c-8906-445278573c7f',
 'timestamp': '4:25:06.000','vin': '4FTEX4944AK844294',
 'lon': 106.9447146,'lat': -6.338565200000001, 'speed': 25}

payload = " ".join(
    str(value) for value in record.values()
)

#"939ed1d1-1740-420c-8906-445278573c7f 4:25:06.000 4FTEX4944AK844294 106.9447146 -6.338565200000001 25"

Alles zusammenführen

res = firehose.put_record(
    DeliveryStreamName='gps-delivery-stream',
    Record = {
        'Data': payload + "\n" #<-- Zeilenumbruch!
    }
)

Erstellte Dateien

Beispieldaten

939ed1d1-1740-420c-8906-445278573c7f 4:25:06.000 4FTEX4944AK844294 106.9447146 -6.338565200000001 25
f29a5b3d-d0fa-43c0-9e1a-e2a5cdb8be7a 8:10:47.000 3FTEX1G5XAK844393 108.58068100000001 34.79925 37
ff8e7131-408d-463b-8d07-d016419b0656 20:26:44.000 2LAXX1C8XAK844292 114.39239199999999 36.097577 90
bc75da5f-1bf6-444c-80ad-49c180e1b8de 23:16:06.000 3FTEX1G5XAK844393 -76.6990172 2.481207 40
7bdcf779-444e-4313-83da-140461933aeb 22:28:44.000 5FTEX1MAXAK844295 -47.0145295 -21.4649238 40

Erstellte Dateien

S3-Client erstellen

# Create boto3 S3 client.
s3 = boto3.client('s3', 
                  aws_access_key_id=AWS_KEY_ID, 
                  aws_secret_access_key=AWS_SECRET, 
                  region_name='us-east-1')

Daten in DataFrame laden

# Get the object from S3
obj_data = s3.get_object(Bucket='sd-vehicle-data', Key=KEY_YOU_COPIED)

# Read read the object into a DataFrame
vehicle_data = pd.read_csv(
    data['Body'], 
    delimiter = " ", 
    names=["record_id", "timestamp", "vin", "lon", "lat", "speed"]))

vehicle_data

	record_id	timestamp	vin	lon	lat	speed
0	939ed1d1...	4:25:06.000	4FTEX4944AK844294	106.945	-6.33857	25
1	f29a5b3d...	8:10:47.000	3FTEX1G5XAK844393	108.581	34.7993	37
2	ff8e7131...	20:26:44.000	2LAXX1C8XAK844292	114.392	36.0976	90
3	bc75da5f...	23:16:06.000	3FTEX1G5XAK844393	-76.699	2.48121	40
4	7bdcf779...	22:28:44.000	5FTEX1MAXAK844295	-47.0145	-21.4649	40

Rückblick

res = firehose.create_delivery_stream(

  DeliveryStreamName =  "gps-delivery-stream",

  DeliveryStreamType = "DirectPut",

  S3DestinationConfiguration = {

    "RoleARN": "arn:aws:iam::0000000:role/firehoseDeliveryRole",

    "BucketARN": "arn:aws:s3:::sd-vehicle-data",

  }

)

Rückblick

	record_id	timestamp	vin	lon	lat	speed
0	939ed1d1...	4:25:06.000	4FTEX4944AK844294	106.945	-6.33857	25
1	f29a5b3d...	8:10:47.000	3FTEX1G5XAK844393	108.581	34.7993	37
2	ff8e7131...	20:26:44.000	2LAXX1C8XAK844292	114.392	36.0976	90
3	bc75da5f...	23:16:06.000	3FTEX1G5XAK844393	-76.699	2.48121	40
4	7bdcf779...	22:28:44.000	5FTEX1MAXAK844295	-47.0145	-21.4649	40

Lass uns üben!

Datenstreaming mit AWS Kinesis und Lambda