JarCastingMonix-Kafka
Monix integration with Kafka
Work in progress!
Table of Contents
- Getting Started with Kafka 1.0.x or above
- Getting Started with Kafka 0.11.x
- Getting Started with Kafka 0.10.x
- Getting Started with Kafka 0.9.x
- Getting Started with Kafka 0.8.x (no longer supported)
- Usage
- How can I contribute to Monix-Kafka?
- Maintainers
- License
Getting Started with Kafka 1.0.x or above
In SBT:
libraryDependencies += "io.monix" %% "monix-kafka-1x" % "1.0.0-RC6"
For kafka versions higher than 1.0.x also add a dependency override:
dependencyOverrides += "org.apache.kafka" % "kafka-clients" % "2.1.0"
Or in case you're interested in running the tests of this project, it now supports embedded kafka for integration testing. You can simply run:
sbt kafka1x/test
Getting Started with Kafka 0.11.x
In SBT:
libraryDependencies += "io.monix" %% "monix-kafka-11" % "1.0.0-RC6"
Or in case you're interested in running the tests of this project, it now supports embedded kafka for integration testing. You can simply run:
sbt kafka11/test
Getting Started with Kafka 0.10.x
In SBT:
libraryDependencies += "io.monix" %% "monix-kafka-10" % "1.0.0-RC6"
Or in case you're interested in running the tests of this project, it now supports embedded kafka for integration testing. You can simply run:
sbt kafka10/test
Getting Started with Kafka 0.9.x
Please note that EmbeddedKafka is not supported for Kafka 0.9.x
In SBT:
libraryDependencies += "io.monix" %% "monix-kafka-9" % "1.0.0-RC6"
Or in case you're interested in running the tests of this project, first download the Kafka server, version 0.9.x from their download page (note that 0.10.x or higher do not work with 0.9), then as the quick start section says, open a terminal window and first start Zookeeper:
bin/zookeeper-server-start.sh config/zookeeper.properties
Then start Kafka:
bin/kafka-server-start.sh config/server.properties
Create the topic we need for our tests:
bin/kafka-topics.sh --create --zookeeper localhost:2181 \
--replication-factor 1 --partitions 1 \
--topic monix-kafka-tests
And run the tests:
sbt kafka9/test
Getting Started with Kafka 0.8.x
Please note that support for Kafka 0.8.x is dropped and the last available version with this dependency is 0.14.
In SBT:
libraryDependencies += "io.monix" %% "monix-kafka-8" % "0.14"
Or in case you're interested in running the tests of this project, first download the Kafka server, version 0.8.x from their download page (note that 0.9.x or higher do not work with 0.8), then as the quick start section says, open a terminal window and first start Zookeeper:
bin/zookeeper-server-start.sh config/zookeeper.properties
Then start Kafka:
bin/kafka-server-start.sh config/server.properties
Create the topics we need for our tests:
bin/kafka-topics.sh --create --zookeeper localhost:2181 \
--replication-factor 1 --partitions 1 \
--topic monix-kafka-tests
bin/kafka-topics.sh --create --zookeeper localhost:2181 \
--replication-factor 1 --partitions 1 \
--topic monix-kafka-manual-commit-tests
And run the tests:
sbt kafka8/test
Usage
Producer
import monix.kafka._
import monix.execution.Scheduler
implicit val scheduler: Scheduler = monix.execution.Scheduler.global
// Init
val producerCfg = KafkaProducerConfig.default.copy(
bootstrapServers = List("127.0.0.1:9092")
)
val producer = KafkaProducer[String,String](producerCfg, scheduler)
// For sending one message
val recordMetadataF = producer.send("my-topic", "my-message").runToFuture
// For closing the producer connection
val closeF = producer.close().runToFuture
Calling producer.send returns a Task of Option[RecordMetadata] which can then be run and transformed into a Future.
If the Task completes with None it means that producer.send method was called after the producer was closed and that the message wasn't successfully acknowledged by the Kafka broker. In case of the failure of the underlying Kafka client the producer will bubble up the exception and fail the Task. All successfully delivered messages will complete with Some[RecordMetadata].
For pushing an entire Observable to Apache Kafka:
import monix.kafka._
import monix.execution.Scheduler
import monix.reactive.Observable
import org.apache.kafka.clients.producer.ProducerRecord
implicit val scheduler: Scheduler = monix.execution.Scheduler.global
// Initializing the producer
val producerCfg = KafkaProducerConfig.default.copy(
bootstrapServers = List("127.0.0.1:9092")
)
val producer = KafkaProducerSink[String,String](producerCfg, scheduler)
// Lets pretend we have this observable of records
val observable: Observable[ProducerRecord[String,String]] = ???
observable
// on overflow, start dropping incoming events
.whileBusyDrop
// buffers into batches if the consumer is busy, up to a max size
.bufferIntrospective(1024)
// consume everything by pushing into Apache Kafka
.consumeWith(producer)
// ready, set, go!
.runToFuture
Consumer
There are several ways for consuming from Apache Kafka (Version 0.11.x and above):
Consumer which commits offsets itself:
import monix.kafka._
val consumerCfg = KafkaConsumerConfig.default.copy(
bootstrapServers = List("127.0.0.1:9092"),
groupId = "kafka-tests"
// you can use this settings for At Most Once semantics:
// observableCommitOrder = ObservableCommitOrder.BeforeAck
)
val observable =
KafkaConsumerObservable[String,String](consumerCfg, List("my-topic"))
.take(10000)
.map(_.value())
Consumer which allows you to commit offsets manually:
import monix.kafka._
val consumerCfg = KafkaConsumerConfig.default.copy(
bootstrapServers = List("127.0.0.1:9092"),
groupId = "kafka-tests"
)
val observable =
KafkaConsumerObservable.manualCommit[String,String](consumerCfg, List("my-topic"))
.map(message => message.record.value() -> message.committableOffset)
.mapEval { case (value, offset) => performBusinessLogic(value).map(_ => offset) }
.bufferTimedAndCounted(1.second, 1000)
.mapEval(offsets => CommittableOffsetBatch(offsets).commitSync())
Enjoy!
Caveats
Issue#101 Starting from Kafka 0.10.1.0, there is max.poll.interval.ms setting:
The maximum delay between invocations of poll() when using consumer group management.
This places an upper bound on the amount of time that the consumer can be idle before
fetching more records. If poll() is not called before expiration of this timeout,
then the consumer is considered failed and the group will rebalance in order
to reassign the partitions to another member.
Since, monix-kafka backpressures until all records has been processed. This could be a problem if processing takes time. You can reduce max.poll.records if you are experiencing this issue.
How can I contribute to Monix-Kafka?
We welcome contributions to all projects in the Monix organization and would love for you to help build Monix-Kafka. See our contributor guide for more information about how you can get involed.
Maintainers
The current maintainers (people who can merge pull requests) are:
- Alexandru Nedelcu (alexandru)
- Piotr Gawryś (Avasil)
- Leandro Bolivar (leandrob13)
License
All code in this repository is licensed under the Apache License, Version 2.0. See LICENSE.txt.
