InfluxDB

The Apache Pekko Connectors InfluxDb connector provides Apache Pekko Streams integration for InfluxDB.

For more information about InfluxDB, please visit the InfluxDB Documentation

Artifacts

sbt

val PekkoVersion = "1.0.3"
libraryDependencies ++= Seq(
  "org.apache.pekko" %% "pekko-connectors-influxdb" % "1.0.2",
  "org.apache.pekko" %% "pekko-stream" % PekkoVersion
)

Maven

<properties>
  <pekko.version>1.0.3</pekko.version>
  <scala.binary.version>2.13</scala.binary.version>
</properties>
<dependencies>
  <dependency>
    <groupId>org.apache.pekko</groupId>
    <artifactId>pekko-connectors-influxdb_${scala.binary.version}</artifactId>
    <version>1.0.2</version>
  </dependency>
  <dependency>
    <groupId>org.apache.pekko</groupId>
    <artifactId>pekko-stream_${scala.binary.version}</artifactId>
    <version>${pekko.version}</version>
  </dependency>
</dependencies>

Gradle

def versions = [
  PekkoVersion: "1.0.3",
  ScalaBinary: "2.13"
]
dependencies {
  implementation "org.apache.pekko:pekko-connectors-influxdb_${versions.ScalaBinary}:1.0.2"
  implementation "org.apache.pekko:pekko-stream_${versions.ScalaBinary}:${versions.PekkoVersion}"
}

The table below shows direct dependencies of this module and the second tab shows all libraries it depends on transitively.

Set up InfluxDB client

Sources, Flows and Sinks provided by this connector need a prepared org.influxdb.InfluxDB to access to InfluxDB.

Scala

copysourceinfluxDB = InfluxDBFactory.connect(INFLUXDB_URL, USERNAME, PASSWORD);
influxDB.setDatabase(DatabaseName);
influxDB.query(new Query("CREATE DATABASE " + DatabaseName, DatabaseName));

Java

copysourcefinal InfluxDB influxDB = InfluxDBFactory.connect(INFLUXDB_URL, USERNAME, PASSWORD);
influxDB.setDatabase(databaseName);
influxDB.query(new Query("CREATE DATABASE " + databaseName, databaseName));
return influxDB;

InfluxDB as Source and Sink

Now we can stream messages from or to InfluxDB by providing the InfluxDB to the InfluxDbSource InfluxDbSource or the InfluxDbSink. InfluxDbSink.

Scala

copysource@Measurement(name = "cpu", database = "InfluxDbSpec")
public class InfluxDbSpecCpu extends Cpu {
    public InfluxDbSpecCpu() {
    }

    public InfluxDbSpecCpu(Instant time, String hostname, String region, Double idle, Boolean happydevop, Long uptimeSecs) {
        super(time, hostname, region, idle, happydevop, uptimeSecs);
    }

    public InfluxDbSpecCpu cloneAt(Instant time) {
        return new InfluxDbSpecCpu(time, getHostname(), getRegion(), getIdle(), getHappydevop(), getUptimeSecs());
    }
}

Java

copysource@Measurement(name = "cpu", database = "InfluxDbTest")
public class InfluxDbCpu extends Cpu {

  public InfluxDbCpu() {}

  public InfluxDbCpu(
      Instant time,
      String hostname,
      String region,
      Double idle,
      Boolean happydevop,
      Long uptimeSecs) {
    super(time, hostname, region, idle, happydevop, uptimeSecs);
  }

  public InfluxDbCpu cloneAt(Instant time) {
    return new InfluxDbCpu(
        time, getHostname(), getRegion(), getIdle(), getHappydevop(), getUptimeSecs());
  }
}

With typed source

Use InfluxDbSource.typed and InfluxDbSink.typed to create source and sink. The data is converted by InfluxDBMapper. The data is converted by InfluxDBMapper.

Scala

copysourceval f1 = InfluxDbSource
  .typed(classOf[InfluxDbSpecCpu], InfluxDbReadSettings(), influxDB, query)
  .map { (cpu: InfluxDbSpecCpu) =>
    {
      val clonedCpu = cpu.cloneAt(cpu.getTime.plusSeconds(60000))
      List(InfluxDbWriteMessage(clonedCpu))
    }
  }
  .runWith(InfluxDbSink.typed(classOf[InfluxDbSpecCpu]))

Java

copysourceCompletionStage<Done> completionStage =
    InfluxDbSource.typed(InfluxDbCpu.class, InfluxDbReadSettings.Default(), influxDB, query)
        .map(
            cpu -> {
              InfluxDbCpu clonedCpu = cpu.cloneAt(cpu.getTime().plusSeconds(60000l));
              return InfluxDbWriteMessage.create(clonedCpu, NotUsed.notUsed());
            })
        .groupedWithin(10, Duration.of(50l, ChronoUnit.MILLIS))
        .runWith(InfluxDbSink.typed(InfluxDbCpu.class, influxDB), system);

With QueryResult source

Use InfluxDbSource.create and InfluxDbSink.create to create source and sink.

Scala

copysourceval query = new Query("SELECT * FROM cpu", DatabaseName);

val f1 = InfluxDbSource(influxDB, query)
  .map(resultToPoints)
  .runWith(InfluxDbSink.create())

Java

copysourceQuery query = new Query("SELECT * FROM cpu", DATABASE_NAME);

CompletionStage<Done> completionStage =
    InfluxDbSource.create(influxDB, query)
        .map(queryResult -> points(queryResult))
        .mapConcat(i -> i)
        .groupedWithin(10, Duration.of(50l, ChronoUnit.MILLIS))
        .runWith(InfluxDbSink.create(influxDB), system);

TODO

Writing to InfluxDB

You can also build flow stages. InfluxDbFlow. InfluxDbFlow. The API is similar to creating Sinks.

Scala

copysourceval result = Source(
  List(
    List(validMessage))).via(InfluxDbFlow.create())
  .runWith(Sink.seq)
  .futureValue

Java

copysourceCompletableFuture<List<List<InfluxDbWriteResult<Point, NotUsed>>>> completableFuture =
    Source.single(Collections.singletonList(influxDbWriteMessage))
        .via(InfluxDbFlow.create(influxDB))
        .runWith(Sink.seq(), system)
        .toCompletableFuture();

Passing data through InfluxDbFlow

When streaming documents from Kafka, you might want to commit to Kafka AFTER the document has been written to InfluxDB.

Scala

copysource// We're going to pretend we got messages from kafka.
// After we've written them to InfluxDB, we want
// to commit the offset to Kafka

case class KafkaOffset(offset: Int)
case class KafkaMessage(cpu: InfluxDbFlowCpu, offset: KafkaOffset)

val messagesFromKafka = List(
  KafkaMessage(new InfluxDbFlowCpu(Instant.now().minusSeconds(1000), "local_1", "eu-west-2", 1.4d, true, 123L),
    KafkaOffset(0)),
  KafkaMessage(new InfluxDbFlowCpu(Instant.now().minusSeconds(2000), "local_2", "eu-west-1", 2.5d, false, 125L),
    KafkaOffset(1)),
  KafkaMessage(new InfluxDbFlowCpu(Instant.now().minusSeconds(3000), "local_3", "eu-west-4", 3.1d, false, 251L),
    KafkaOffset(2)))

var committedOffsets = List[KafkaOffset]()

def commitToKafka(offset: KafkaOffset): Unit =
  committedOffsets = committedOffsets :+ offset

val f1 = Source(messagesFromKafka)
  .map { (kafkaMessage: KafkaMessage) =>
    val cpu = kafkaMessage.cpu
    println("hostname: " + cpu.getHostname)

    InfluxDbWriteMessage(cpu).withPassThrough(kafkaMessage.offset)
  }
  .groupedWithin(10, 50.millis)
  .via(
    InfluxDbFlow.typedWithPassThrough(classOf[InfluxDbFlowCpu]))
  .map { (messages: Seq[InfluxDbWriteResult[InfluxDbFlowCpu, KafkaOffset]]) =>
    messages.foreach { message =>
      commitToKafka(message.writeMessage.passThrough)
    }
  }
  .runWith(Sink.ignore)

Java

copysource// We're going to pretend we got metrics from kafka.
// After we've written them to InfluxDb, we want
// to commit the offset to Kafka

/** Just clean the previous data */
influxDB.query(new Query("DELETE FROM cpu"));

List<Integer> committedOffsets = new ArrayList<>();
List<MessageFromKafka> messageFromKafka =
    Arrays.asList(
        new MessageFromKafka(
            new InfluxDbCpu(
                Instant.now().minusSeconds(1000), "local_1", "eu-west-2", 1.4d, true, 123L),
            new KafkaOffset(0)),
        new MessageFromKafka(
            new InfluxDbCpu(
                Instant.now().minusSeconds(2000), "local_2", "eu-west-1", 2.5d, false, 125L),
            new KafkaOffset(1)),
        new MessageFromKafka(
            new InfluxDbCpu(
                Instant.now().minusSeconds(3000), "local_3", "eu-west-4", 3.1d, false, 251L),
            new KafkaOffset(2)));

Consumer<KafkaOffset> commitToKafka =
    kafkaOffset -> committedOffsets.add(kafkaOffset.getOffset());

Source.from(messageFromKafka)
    .map(
        kafkaMessage -> {
          return InfluxDbWriteMessage.create(
              kafkaMessage.influxDbCpu, kafkaMessage.kafkaOffset);
        })
    .groupedWithin(10, Duration.ofMillis(10))
    .via(InfluxDbFlow.typedWithPassThrough(InfluxDbCpu.class, influxDB))
    .map(
        messages -> {
          messages.stream()
              .forEach(
                  message -> {
                    KafkaOffset kafkaOffset = message.writeMessage().passThrough();
                    commitToKafka.accept(kafkaOffset);
                  });
          return NotUsed.getInstance();
        })
    .runWith(Sink.seq(), system)
    .toCompletableFuture()
    .get(10, TimeUnit.SECONDS);