Reactive Streams

Since Camel 2.19

Both producer and consumer are supported

The Reactive Streams component allows you to exchange messages with reactive stream processing libraries compatible with the reactive streams standard.

The component supports backpressure and has been tested using the reactive streams technology compatibility kit (TCK).

The Camel module provides a reactive-streams component that allows users to define incoming and outgoing streams within Camel routes, and a direct client API that allows using Camel endpoints directly into any external reactive framework.

Camel uses an internal implementation of the reactive streams Publisher and Subscriber, so it’s not tied to any specific framework. The following reactive frameworks have been used in the integration tests: Reactor Core 3, RxJava 2.

Maven users will need to add the following dependency to their pom.xml for this component:

<dependency>
    <groupId>org.apache.camel</groupId>
    <artifactId>camel-reactive-streams</artifactId>
    <version>x.x.x</version>
    <!-- use the same version as your Camel core version -->
</dependency>

URI format

reactive-streams://stream?[options]

Where stream is a logical stream name used to bind Camel routes to the external stream processing systems.

Configuring Options

Camel components are configured on two separate levels:

  • component level

  • endpoint level

Configuring Component Options

The component level is the highest level which holds general and common configurations that are inherited by the endpoints. For example a component may have security settings, credentials for authentication, urls for network connection and so forth.

Some components only have a few options, and others may have many. Because components typically have pre configured defaults that are commonly used, then you may often only need to configure a few options on a component; or none at all.

Configuring components can be done with the Component DSL, in a configuration file (application.properties|yaml), or directly with Java code.

Configuring Endpoint Options

Where you find yourself configuring the most is on endpoints, as endpoints often have many options, which allows you to configure what you need the endpoint to do. The options are also categorized into whether the endpoint is used as consumer (from) or as a producer (to), or used for both.

Configuring endpoints is most often done directly in the endpoint URI as path and query parameters. You can also use the Endpoint DSL as a type safe way of configuring endpoints.

A good practice when configuring options is to use Property Placeholders, which allows to not hardcode urls, port numbers, sensitive information, and other settings. In other words placeholders allows to externalize the configuration from your code, and gives more flexibility and reuse.

The following two sections lists all the options, firstly for the component followed by the endpoint.

Component Options

The Reactive Streams component supports 9 options, which are listed below.

Name Description Default Type

threadPoolMaxSize (common)

The maximum number of threads used by the reactive streams internal engine.

int

threadPoolMinSize (common)

The minimum number of threads used by the reactive streams internal engine.

int

threadPoolName (common)

The name of the thread pool used by the reactive streams internal engine.

String

bridgeErrorHandler (consumer)

Allows for bridging the consumer to the Camel routing Error Handler, which mean any exceptions occurred while the consumer is trying to pickup incoming messages, or the likes, will now be processed as a message and handled by the routing Error Handler. By default the consumer will use the org.apache.camel.spi.ExceptionHandler to deal with exceptions, that will be logged at WARN or ERROR level and ignored.

boolean

backpressureStrategy (producer)

The backpressure strategy to use when pushing events to a slow subscriber.

Enum values:

  • BUFFER

  • OLDEST

  • LATEST

ReactiveStreamsBackpressureStrategy

lazyStartProducer (producer)

Whether the producer should be started lazy (on the first message). By starting lazy you can use this to allow CamelContext and routes to startup in situations where a producer may otherwise fail during starting and cause the route to fail being started. By deferring this startup to be lazy then the startup failure can be handled during routing messages via Camel’s routing error handlers. Beware that when the first message is processed then creating and starting the producer may take a little time and prolong the total processing time of the processing.

boolean

autowiredEnabled (advanced)

Whether autowiring is enabled. This is used for automatic autowiring options (the option must be marked as autowired) by looking up in the registry to find if there is a single instance of matching type, which then gets configured on the component. This can be used for automatic configuring JDBC data sources, JMS connection factories, AWS Clients, etc.

boolean

reactiveStreamsEngineConfiguration (advanced)

To use an existing reactive stream engine configuration.

ReactiveStreamsEngineConfiguration

serviceType (advanced)

Set the type of the underlying reactive streams implementation to use. The implementation is looked up from the registry or using a ServiceLoader, the default implementation is DefaultCamelReactiveStreamsService.

String

Endpoint Options

The Reactive Streams endpoint is configured using URI syntax:

reactive-streams:stream

with the following path and query parameters:

Path Parameters (1 parameters)

Name Description Default Type

stream (common)

Name of the stream channel used by the endpoint to exchange messages.

String

Query Parameters (10 parameters)

Name Description Default Type

bridgeErrorHandler (consumer)

Allows for bridging the consumer to the Camel routing Error Handler, which mean any exceptions occurred while the consumer is trying to pickup incoming messages, or the likes, will now be processed as a message and handled by the routing Error Handler. By default the consumer will use the org.apache.camel.spi.ExceptionHandler to deal with exceptions, that will be logged at WARN or ERROR level and ignored.

boolean

concurrentConsumers (consumer)

Number of threads used to process exchanges in the Camel route.

int

exchangesRefillLowWatermark (consumer)

Set the low watermark of requested exchanges to the active subscription as percentage of the maxInflightExchanges. When the number of pending items from the upstream source is lower than the watermark, new items can be requested to the subscription. If set to 0, the subscriber will request items in batches of maxInflightExchanges, only after all items of the previous batch have been processed. If set to 1, the subscriber can request a new item each time an exchange is processed (chatty). Any intermediate value can be used.

double

forwardOnComplete (consumer)

Determines if onComplete events should be pushed to the Camel route.

boolean

forwardOnError (consumer)

Determines if onError events should be pushed to the Camel route. Exceptions will be set as message body.

boolean

maxInflightExchanges (consumer)

Maximum number of exchanges concurrently being processed by Camel. This parameter controls backpressure on the stream. Setting a non-positive value will disable backpressure.

Integer

exceptionHandler (consumer (advanced))

To let the consumer use a custom ExceptionHandler. Notice if the option bridgeErrorHandler is enabled then this option is not in use. By default the consumer will deal with exceptions, that will be logged at WARN or ERROR level and ignored.

ExceptionHandler

exchangePattern (consumer (advanced))

Sets the exchange pattern when the consumer creates an exchange.

Enum values:

  • InOnly

  • InOut

  • InOptionalOut

ExchangePattern

backpressureStrategy (producer)

The backpressure strategy to use when pushing events to a slow subscriber.

Enum values:

  • BUFFER

  • OLDEST

  • LATEST

ReactiveStreamsBackpressureStrategy

lazyStartProducer (producer)

Whether the producer should be started lazy (on the first message). By starting lazy you can use this to allow CamelContext and routes to startup in situations where a producer may otherwise fail during starting and cause the route to fail being started. By deferring this startup to be lazy then the startup failure can be handled during routing messages via Camel’s routing error handlers. Beware that when the first message is processed then creating and starting the producer may take a little time and prolong the total processing time of the processing.

boolean

Usage

The library is aimed to support all the communication modes needed by an application to interact with Camel data:

  • Get data from Camel routes (In-Only from Camel)

  • Send data to Camel routes (In-Only towards Camel)

  • Request a transformation to a Camel route (In-Out towards Camel)

  • Process data flowing from a Camel route using a reactive processing step (In-Out from Camel)

Getting data from Camel

In order to subscribe to data flowing from a Camel route, exchanges should be redirected to a named stream, like in the following snippet:

from("timer:clock")
.setBody().header(Exchange.TIMER_COUNTER)
.to("reactive-streams:numbers");

Routes can also be written using the XML DSL.

In the example, an unbounded stream of numbers is associated to the name numbers. The stream can be accessed using the CamelReactiveStreams utility class.

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

// Getting a stream of exchanges
Publisher<Exchange> exchanges = camel.fromStream("numbers");

// Getting a stream of Integers (using Camel standard conversion system)
Publisher<Integer> numbers = camel.fromStream("numbers", Integer.class);

The stream can be used easily with any reactive streams compatible library. Here is an example of how to use it with RxJava 2 (although any reactive framework can be used to process events).

Flowable.fromPublisher(integers)
    .doOnNext(System.out::println)
    .subscribe();

The example prints all numbers generated by Camel into System.out.

Getting data from Camel using the direct API

For short Camel routes and for users that prefer defining the whole processing flow using functional constructs of the reactive framework (without using the Camel DSL at all), streams can also be defined using Camel URIs.

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

// Get a stream from all the files in a directory
Publisher<String> files = camel.from("file:folder", String.class);

// Use the stream in RxJava
Flowable.fromPublisher(files)
    .doOnNext(System.out::println)
    .subscribe();

Sending data to Camel

When an external library needs to push events into a Camel route, the Reactive Streams endpoint must be set as consumer.

from("reactive-streams:elements")
.to("log:INFO");

A handle to the elements stream can be obtained from the CamelReactiveStreams utility class.

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

Subscriber<String> elements = camel.streamSubscriber("elements", String.class);

The subscriber can be used to push events to the Camel route that consumes from the elements stream.

Here is an example of how to use it with RxJava 2 (although any reactive framework can be used to publish events).

Flowable.interval(1, TimeUnit.SECONDS)
    .map(i -> "Item " + i)
    .subscribe(elements);

String items are generated every second by RxJava in the example and they are pushed into the Camel route defined above.

Sending data to Camel using the direct API

Also in this case, the direct API can be used to obtain a Camel subscriber from an endpoint URI.

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

// Send two strings to the "seda:queue" endpoint
Flowable.just("hello", "world")
    .subscribe(camel.subscriber("seda:queue", String.class));

Request a transformation to Camel

Routes defined in some Camel DSL can be used within a reactive stream framework to perform a specific transformation (the same mechanism can be also used to eg. just send data to a http endpoint and continue).

The following snippet shows how RxJava functional code can request the task of loading and marshalling files to Camel.

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

// Process files starting from their names
Flowable.just(new File("file1.txt"), new File("file2.txt"))
    .flatMap(file -> camel.toStream("readAndMarshal", String.class))
    // Camel output will be converted to String
    // other steps
    .subscribe();

In order this to work, a route like the following should be defined in the Camel context:

from("reactive-streams:readAndMarshal")
.marshal() // ... other details

Request a transformation to Camel using the direct API

An alternative approach consists in using the URI endpoints directly in the reactive flow:

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

// Process files starting from their names
Flowable.just(new File("file1.txt"), new File("file2.txt"))
    .flatMap(file -> camel.to("direct:process", String.class))
    // Camel output will be converted to String
    // other steps
    .subscribe();

When using the to() method instead of the toStream, there is no need to define the route using "reactive-streams:" endpoints (although they are used under the hood).

In this case, the Camel transformation can be just:

from("direct:process")
.marshal() // ... other details

Process Camel data into the reactive framework

While a reactive streams Publisher allows exchanging data in a unidirectional way, Camel routes often use a in-out exchange pattern (eg. to define REST endpoints and, in general, where a reply is needed for each request).

In these circumstances, users can add a reactive processing step to the flow, to enhance a Camel route or to define the entire transformation using the reactive framework.

For example, given the following route:

from("timer:clock")
.setBody().header(Exchange.TIMER_COUNTER)
.to("direct:reactive")
.log("Continue with Camel route... n=${body}");

A reactive processing step can be associated to the "direct:reactive" endpoint:

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

camel.process("direct:reactive", Integer.class, items ->
    Flowable.fromPublisher(items) // RxJava
        .map(n -> -n)); // make every number negative

Data flowing in the Camel route will be processed by the external reactive framework then continue the processing flow inside Camel.

This mechanism can also be used to define a In-Out exchange in a completely reactive way.

CamelReactiveStreamsService camel = CamelReactiveStreams.get(context);

// requires a rest-capable Camel component
camel.process("rest:get:orders", exchange ->
                    Flowable.fromPublisher(exchange)
                            .flatMap(ex -> allOrders())); // retrieve orders asynchronously

See Camel examples (camel-example-reactive-streams) for details.

Advanced Topics

Controlling Backpressure (producer side)

When routing Camel exchanges to an external subscriber, backpressure is handled by an internal buffer that caches exchanges before delivering them. If the subscriber is slower than the exchange rate, the buffer may become too big. In many circumstances this must be avoided.

Considering the following route:

from("jms:queue")
.to("reactive-streams:flow");

If the JMS queue contains a high number of messages and the Subscriber associated with the flow stream is too slow, messages are dequeued from JMS and appended to the buffer, possibly causing a "out of memory" error. To avoid such problems, a ThrottlingInflightRoutePolicy can be set in the route.

ThrottlingInflightRoutePolicy policy = new ThrottlingInflightRoutePolicy();
policy.setMaxInflightExchanges(10);

from("jms:queue")
.routePolicy(policy)
.to("reactive-streams:flow");

The policy limits the maximum number of active exchanges (and so the maximum size of the buffer), keeping it lower than the threshold (10 in the example). When more than 10 messages are in flight, the route is suspended, waiting for the subscriber to process them.

With this mechanism, the subscriber controls the route suspension/resume automatically, through backpressure. When multiple subscribers are consuming items from the same stream, the slowest one controls the route status automatically.

In other circumstances, eg. when using a http consumer, the route suspension makes the http service unavailable, so using the default configuration (no policy, unbounded buffer) should be preferable. Users should try to avoid memory issues by limiting the number of requests to the http service (eg. scaling out).

In contexts where a certain amount of data loss is acceptable, setting a backpressure strategy other than BUFFER can be a solution for dealing with fast sources.

from("direct:thermostat")
.to("reactive-streams:flow?backpressureStrategy=LATEST");

When the LATEST backpressure strategy is used, only the last exchange received from the route is kept by the publisher, while older data is discarded (other options are available).

Controlling Backpressure (consumer side)

When Camel consumes items from a reactive-streams publisher, the maximum number of inflight exchanges can be set as endpoint option.

The subscriber associated with the consumer interacts with the publisher to keep the number of messages in the route lower than the threshold.

An example of backpressure-aware route:

from("reactive-streams:numbers?maxInflightExchanges=10")
.to("direct:endpoint");

The number of items that Camel requests to the source publisher (through the reactive streams backpressure mechanism) is always lower than 10. Messages are processed by a single thread in the Camel side.

The number of concurrent consumers (threads) can also be set as endpoint option (concurrentConsumers). When using 1 consumer (the default), the order of items in the source stream is maintained. When this value is increased, items will be processed concurrently by multiple threads (so not preserving the order).

Spring Boot Auto-Configuration

When using reactive-streams with Spring Boot make sure to use the following Maven dependency to have support for auto configuration:

<dependency>
  <groupId>org.apache.camel.springboot</groupId>
  <artifactId>camel-reactive-streams-starter</artifactId>
  <version>x.x.x</version>
  <!-- use the same version as your Camel core version -->
</dependency>

The component supports 10 options, which are listed below.

Name Description Default Type

camel.component.reactive-streams.autowired-enabled

Whether autowiring is enabled. This is used for automatic autowiring options (the option must be marked as autowired) by looking up in the registry to find if there is a single instance of matching type, which then gets configured on the component. This can be used for automatic configuring JDBC data sources, JMS connection factories, AWS Clients, etc.

true

Boolean

camel.component.reactive-streams.backpressure-strategy

The backpressure strategy to use when pushing events to a slow subscriber.

ReactiveStreamsBackpressureStrategy

camel.component.reactive-streams.bridge-error-handler

Allows for bridging the consumer to the Camel routing Error Handler, which mean any exceptions occurred while the consumer is trying to pickup incoming messages, or the likes, will now be processed as a message and handled by the routing Error Handler. By default the consumer will use the org.apache.camel.spi.ExceptionHandler to deal with exceptions, that will be logged at WARN or ERROR level and ignored.

false

Boolean

camel.component.reactive-streams.enabled

Whether to enable auto configuration of the reactive-streams component. This is enabled by default.

Boolean

camel.component.reactive-streams.lazy-start-producer

Whether the producer should be started lazy (on the first message). By starting lazy you can use this to allow CamelContext and routes to startup in situations where a producer may otherwise fail during starting and cause the route to fail being started. By deferring this startup to be lazy then the startup failure can be handled during routing messages via Camel’s routing error handlers. Beware that when the first message is processed then creating and starting the producer may take a little time and prolong the total processing time of the processing.

false

Boolean

camel.component.reactive-streams.reactive-streams-engine-configuration

To use an existing reactive stream engine configuration. The option is a org.apache.camel.component.reactive.streams.engine.ReactiveStreamsEngineConfiguration type.

ReactiveStreamsEngineConfiguration

camel.component.reactive-streams.service-type

Set the type of the underlying reactive streams implementation to use. The implementation is looked up from the registry or using a ServiceLoader, the default implementation is DefaultCamelReactiveStreamsService.

String

camel.component.reactive-streams.thread-pool-max-size

The maximum number of threads used by the reactive streams internal engine.

10

Integer

camel.component.reactive-streams.thread-pool-min-size

The minimum number of threads used by the reactive streams internal engine.

Integer

camel.component.reactive-streams.thread-pool-name

The name of the thread pool used by the reactive streams internal engine.

CamelReactiveStreamsWorker

String