Skip to content
Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.

RxJava 2.x

View at Github page

also available for reactor-core


Reactive Streams

Reactive Streams is a programming concept for handling asynchronous data streams in a non-blocking manner while providing backpressure to stream publishers. It has evolved into a specification that is based on the concept of Publisher<T> and Subscriber<T>. A Publisher is the source of events T in the stream, and a Subscriber is the consumer for those events. A Subscriber subscribes to a Publisher by invoking a "factory method" in the Publisher that will push the stream items <T> starting a new Subscription:

public interface Publisher<T> {
    public void subscribe(Subscriber<? super T> s);

When the Subscriber is ready to start handling events, it signals this via a request to that Subscription

public interface Subscription {
    public void request(long n); //request n items
    public void cancel();

Upon receiving this signal, the Publisher begins to invoke Subscriber::onNext(T) for each event T. This continues until either completion of the stream (Subscriber::onComplete()) or an error occurs during processing (Subscriber::onError(Throwable)).

public interface Subscriber<T> {
    //signals to the Publisher to start sending events
    public void onSubscribe(Subscription s);     
    public void onNext(T t);
    public void onError(Throwable t);
    public void onComplete();

Flowable and Observable

RxJava provides more types of event publishers:

  • Flowable Publisher that emits 0..N elements, and then completes successfully or with an error

  • Observable like Flowables but without a backpressure strategy. They were introduced in RxJava 1.x

  • Single a specialized emitter that completes with a value successfully either an error.(doesn't have onComplete callback, instead onSuccess(val))

  • Maybe a specialized emitter that can complete with / without a value or complete with an error.

  • Completable a specialized emitter that just signals if it completed successfully or with an error.

Code is available at

Simple operators to create Streams

Flowable<Integer> flowable = Flowable.just(1, 5, 10);
Flowable<Integer> flowable = Flowable.range(1, 10);
Flowable<String> flowable = Flowable.fromArray(new String[] {"red", "green", "blue"});
Flowable<String> flowable = Flowable.fromIterable(List.of("red", "green", "blue"));

Flowable from Future

CompletableFuture<String> completableFuture = CompletableFuture
            .supplyAsync(() -> { //starts a background thread the ForkJoin common pool
          "CompletableFuture work starts");  
                    return "red";

Single<String> single = Single.from(completableFuture);
single.subscribe(val ->"Stream completed successfully : {}", val));

Creating your own stream

We can use Flowable.create(...) to implement the emissions of events by calling onNext(val), onComplete(), onError(throwable)

When subscribing to the Observable / Flowable with flowable.subscribe(...) the lambda code inside create(...) gets executed. Flowable.subscribe(...) can take 3 handlers for each type of event - onNext, onError and onCompleted.

When using Observable.create(...) you need to be aware of backpressure and that Observables created with 'create' are not BackPressure aware

Observable<Integer> stream = Observable.create(subscriber -> {"Started emitting");"Emitting 1st");
    subscriber.onNext(1);"Emitting 2nd");


//Flowable version same Observable but with a BackpressureStrategy
//that will be discussed separately.
Flowable<Integer> stream = Flowable.create(subscriber -> {"Started emitting");"Emitting 1st");
    subscriber.onNext(1);"Emitting 2nd");

}, BackpressureStrategy.MISSING);

       val ->"Subscriber received: {}", val),
       err -> log.error("Subscriber received error", err),
       () ->"Subscriber got Completed event")

Streams are lazy

Streams are lazy meaning that the code inside create() doesn't get executed without subscribing to the stream. So event if we sleep for a long time inside create() method(to simulate a costly operation), without subscribing to this Observable, the code is not executed and the method returns immediately.

public void observablesAreLazy() {
    Observable<Integer> observable = Observable.create(subscriber -> {"Started emitting but sleeping for 5 secs"); //this is not executed
[main] - Finished

Multiple subscriptions to the same Observable / Flowable

When subscribing to an Observable/Flowable, the create() method gets executed for each Subscriber, the events inside create(..) are re-emitted to each subscriber independently.

So every subscriber will get the same events and will not lose any events - this behavior is named 'cold observable' See Hot Publishers to understand sharing a subscription and multicasting events.

Observable<Integer> observable = Observable.create(subscriber -> {"Started emitting");"Emitting 1st event");
   subscriber.onNext(1);"Emitting 2nd event");

});"Subscribing 1st subscriber");
observable.subscribe(val ->"First Subscriber received: {}", val));"=======================");"Subscribing 2nd subscriber");
observable.subscribe(val ->"Second Subscriber received: {}", val));

will output

[main] - Subscribing 1st subscriber
[main] - Started emitting
[main] - Emitting 1st event
[main] - First Subscriber received: 1
[main] - Emitting 2nd event
[main] - First Subscriber received: 2
[main] - =======================
[main] - Subscribing 2nd subscriber
[main] - Started emitting
[main] - Emitting 1st event
[main] - Second Subscriber received: 1
[main] - Emitting 2nd event
[main] - Second Subscriber received: 2

Observable / Flowable lifecycle


Between the source Observable / Flowable and the Subscriber there can be a wide range of operators and RxJava provides lots of operators to chose from. Probably you are already familiar with functional operations like filter and map. so let's use them as example:

Flowable<Integer> stream = Flowable.create(subscriber -> {
    }, BackpressureStrategy.MISSING);
    .filter(val -> val < 10)
    .map(val -> val * 10)
    .subscribe(val ->"Received: {}", val));

When we call Flowable.create() you might think that we're calling onNext(..), onComplete(..) on the Subscriber at the end of the chain, not the operators between them.

This is not true because the operators themselves are decorators for their source wrapping it with the operator behavior like an onion's layers. When we call .subscribe() at the end of the chain, Subscription propagates through the layers back to the source, each operator subscribing itself to it's wrapped source Observable / Flowable and so on to the original source, triggering it to start producing/emitting items.

Flowable.create calls ---> filterOperator.onNext(val) which if val > 10 calls ---> mapOperator.onNext(val) does val = val * 10 and calls ---> subscriber.onNext(val).

Found a nice analogy with a team of house movers, with every mover doing it's thing before passing it to the next in line until it reaches the final subscriber.


Canceling subscription

Inside the create() method, we can check is there are still active subscribers to our Flowable/Observable.

There are operators that also unsubscribe from the stream so the source knows to stop producing events.
It's a way to prevent to do extra work(like for ex. querying a datasource for entries) if no one is listening In the following example we'd expect to have an infinite stream, but because we stop if there are no active subscribers, we stop producing events.

take(limit) is a simple operator. It's role is to count the number of events and then unsubscribes from it's source once it received the specified amount and calls onComplete() to it's subscriber.

Observable<Integer> observable = Observable.create(subscriber -> {

    int i = 1;
    while(true) {
        if(subscriber.isDisposed()) {

        //registering a callback when the downstream subscriber unsubscribes
        subscriber.setCancellable(() ->"Subscription canceled"));

    .take(5) //unsubscribes after the 5th event
    .subscribe(val ->"Subscriber received: {}", val),
               err -> log.error("Subscriber received error", err),
               () ->"Subscriber got Completed event") //The Complete event 
               //is triggered by 'take()' operator

[main] - Subscriber received: *1*
[main] - Subscriber received: *2*
[main] - Subscriber received: *3*
[main] - Subscriber received: *4*
[main] - Subscriber received: *5*
[main] - Subscriber got Completed event
[main] - Subscription canceled

Simple Operators

Code is available at


Delay operator - the Thread.sleep of the reactive world, it's pausing each emission for a particular increment of time.

CountDownLatch latch = new CountDownLatch(1);
Flowable.range(0, 2)
        .doOnNext(val ->"Emitted {}", val))
        .delay(5, TimeUnit.SECONDS)
        .subscribe(tick ->"Tick {}", tick),
                   (ex) ->"Error emitted"),
                   () -> {

14:27:44 [main] - Starting
14:27:45 [main] - Emitted 0
14:27:45 [main] - Emitted 1
14:27:50 [RxComputationThreadPool-1] - Tick 0
14:27:50 [RxComputationThreadPool-1] - Tick 1
14:27:50 [RxComputationThreadPool-1] - Completed

The .delay(), .interval() operators uses a Scheduler by default which is why we see it executing on a different thread RxComputationThreadPool-1 which actually means it's running the operators and the subscribe operations on another thread and so the test method will terminate before we see the text from the log unless we wait for the completion of the stream. This is the role of the CountdownLatch.


Periodically emits a number starting from 0 and then increasing the value on each emission."Starting");
Flowable.interval(5, TimeUnit.SECONDS)
       .subscribe(tick ->"Subscriber received {}", tick),
                  (ex) ->"Error emitted"),
                  () ->"Subscriber got Completed event"));

12:17:56 [main] - Starting
12:18:01 [RxComputationThreadPool-1] - Subscriber received: 0
12:18:06 [RxComputationThreadPool-1] - Subscriber received: 1
12:18:11 [RxComputationThreadPool-1] - Subscriber received: 2
12:18:16 [RxComputationThreadPool-1] - Subscriber received: 3
12:18:21 [RxComputationThreadPool-1] - Subscriber received: 4
12:18:21 [RxComputationThreadPool-1] - Subscriber got Completed event


Takes an initial value and a function(accumulator, currentValue). It goes through the events sequence and combines the current event value with the previous result(accumulator) emitting downstream the function's result for each event(the initial value is used for the first event)

Flowable<Integer> numbers = 
                Flowable.just(3, 5, -2, 9)
                    .scan(0, (totalSoFar, currentValue) -> {
                     "TotalSoFar={}, currentValue={}", 
                                            totalSoFar, currentValue);
                               return totalSoFar + currentValue;

16:09:17 [main] - Subscriber received: 0
16:09:17 [main] - TotalSoFar=0, currentValue=3
16:09:17 [main] - Subscriber received: 3
16:09:17 [main] - TotalSoFar=3, currentValue=5
16:09:17 [main] - Subscriber received: 8
16:09:17 [main] - TotalSoFar=8, currentValue=-2
16:09:17 [main] - Subscriber received: 6
16:09:17 [main] - TotalSoFar=6, currentValue=9
16:09:17 [main] - Subscriber received: 15
16:09:17 [main] - Subscriber got Completed event


reduce operator acts like the scan operator but it only passes downstream the final result (doesn't pass the intermediate results downstream) so the subscriber receives just one event

Flowable<Integer> numbers = Flowable.just(3, 5, -2, 9)
                            .reduce(0, (totalSoFar, val) -> {
                               "totalSoFar={}, emitted={}",
                                                        totalSoFar, val);
                                         return totalSoFar + val;
17:08:29 [main] - totalSoFar=0, emitted=3
17:08:29 [main] - totalSoFar=3, emitted=5
17:08:29 [main] - totalSoFar=8, emitted=-2
17:08:29 [main] - totalSoFar=6, emitted=9
17:08:29 [main] - Subscriber received: 15
17:08:29 [main] - Subscriber got Completed event


collect operator acts similar to the reduce operator, but while the reduce operator uses a reduce function which returns a value, the collect operator takes a container supplier and a function which doesn't return anything(a consumer). The mutable container is passed for every event and thus you get a chance to modify it in this collect consumer function.

Flowable<List<Integer>> numbers = Flowable.just(3, 5, -2, 9)
                                        .collect(ArrayList::new, (container, value) -> {
                                  "Adding {} to container", value);
                                            //notice we don't need to return anything
17:40:18 [main] - Adding 3 to container
17:40:18 [main] - Adding 5 to container
17:40:18 [main] - Adding -2 to container
17:40:18 [main] - Adding 9 to container
17:40:18 [main] - Subscriber received: [3, 5, -2, 9]
17:40:18 [main] - Subscriber got Completed event

because the usecase to store to a List container is so common, there is a .toList() operator that is just a collector adding to a List.


An easy way to switch from a blocking method to a reactive Single/Flowable is to use .defer(() -> blockingOp()).

Simply using Flowable.just(blockingOp()) would still block, as Java needs to resolve the parameter when invoking Flux.just(param) method, so blockingOp() method would still be invoked(and block).

Flowable<String> flowableBlocked = Flowable.just((blockingOp())); //blocks on this line

In order to get around this problem, we can use Flowable.defer(() -> blockingOp()) and wrap the blockingOp() call inside a lambda which will be invoked lazy at subscribe time.

Flowable<String> stream = Flowable.defer(() -> Flowable.just(blockingOperation())); 
stream.subscribe(val ->"Val " + val)); //only now the code inside defer() is executed

Merging Streams

Operators for working with multiple streams Code at


Zip operator operates sort of like a zipper in the sense that it takes an event from one stream and waits for an event from another other stream. Once an event for the other stream arrives, it uses the zip function to merge the two events.

This is an useful scenario when for example you want to make requests to remote services in parallel and wait for both responses before continuing. It also takes a function which will produce the combined result of the zipped streams once each has emitted a value.


Zip operator besides the streams to zip, also takes as parameter a function which will produce the combined result of the zipped streams once each stream emitted its value

Single<Boolean> isUserBlockedStream = 
                    Single.fromFuture(CompletableFuture.supplyAsync(() -> {
                            return Boolean.FALSE;

Single<Integer> userCreditScoreStream = 
                    Single.fromFuture(CompletableFuture.supplyAsync(() -> {
                            return 5;

Single<Pair<Boolean, Integer>> userCheckStream =, userCreditScoreStream, 
                      (blocked, creditScore) -> new Pair<Boolean, Integer>(blocked, creditScore));

userCheckStream.subscribe(pair ->"Received " + pair));

Even if the 'isUserBlockedStream' finishes after 200ms, 'userCreditScoreStream' is slow at 2.3secs, the 'zip' method applies the combining function(new Pair(x,y)) after it received both values and passes it to the subscriber.

Another good example of 'zip' is to slow down a stream by another basically implementing a periodic emitter of events:

Flowable<String> colors = Flowable.just("red", "green", "blue");
Flowable<Long> timer = Flowable.interval(2, TimeUnit.SECONDS);

Flowable<String> periodicEmitter =, timer, (key, val) -> key);

Since the zip operator needs a pair of events, the slow stream will work like a timer by periodically emitting with zip setting the pace of emissions downstream every 2 seconds.

Zip is not limited to just two streams, it can merge 2,3,4,.. streams and wait for groups of 2,3,4 'pairs' of events which it combines with the zip function and sends downstream.


Merge operator combines one or more stream and passes events downstream as soon as they appear.


Flowable<String> colors = periodicEmitter("red", "green", "blue", 2, TimeUnit.SECONDS);

Flowable<Long> numbers = Flowable.interval(1, TimeUnit.SECONDS)
//notice we can't say Flowable<String> or Flowable<Long> as the return stream o the merge operator since 
//it can emit either a color or number.                  
Flowable flowable = Flowable.merge(colors, numbers);                

21:32:15 - Subscriber received: 0
21:32:16 - Subscriber received: red
21:32:16 - Subscriber received: 1
21:32:17 - Subscriber received: 2
21:32:18 - Subscriber received: green
21:32:18 - Subscriber received: 3
21:32:19 - Subscriber received: 4
21:32:20 - Subscriber received: blue


Concat operator appends another streams at the end of another concat

Flowable<String> colors = periodicEmitter("red", "green", "blue", 2, TimeUnit.SECONDS);

Flowable<Long> numbers = Flowable.interval(1, TimeUnit.SECONDS)

Flowable events = Flowable.concat(colors, numbers);

22:48:23 - Subscriber received: red
22:48:25 - Subscriber received: green
22:48:27 - Subscriber received: blue
22:48:28 - Subscriber received: 0
22:48:29 - Subscriber received: 1
22:48:30 - Subscriber received: 2
22:48:31 - Subscriber received: 3

Even if the 'numbers' streams should start early, the 'colors' stream emits fully its events before we see any 'numbers'. This is because 'numbers' stream is actually subscribed only after the 'colors' complete. Should the second stream be a 'hot' emitter, its events would be lost until the first one finishes and the seconds stream is subscribed.

Hot Publishers

We've seen that with 'cold publishers', whenever a subscriber subscribes, each subscriber will get it's version of emitted values independently, the exact set of data indifferently when they subscribe. But cold publishers only produce data when the subscribers subscribes, however there are cases where the events happen independently from the consumers regardless if someone is listening or not and we don't have control to request more. So you could say we have 'cold publishers' for pull scenarios and 'hot publishers' which push.


Subjects are one way to handle hot observables. Subjects keep reference to their subscribers and allow 'multicasting' an event to them.

for (Disposable<T> s : subscribers.get()) {

Subjects besides being traditional Observables you can use the same operators and subscribe to them, are also an Observer(interface like Subscriber from reactive-streams, implementing the 3 methods onNext, onError, onComplete), meaning you can invoke subject.onNext(value) from different parts in the code, which means that you publish events which the Subject will pass on to their subscribers.

Subject<Integer> subject = ReplaySubject.create()
                     .subscribe(); //


remember for

Observable.create(subscriber -> {


ReplaySubject keeps a buffer of events that it 'replays' to each new subscriber, first he receives a batch of missed and only later events in real-time.

Subject<Integer> subject = ReplaySubject.createWithSize(50);"Pushing 0");
subject.onNext(0);"Pushing 1");
subject.onNext(1);"Subscribing 1st");
subject.subscribe(val ->"Subscriber1 received {}", val), 
                            logError(), logComplete());"Pushing 2");
subject.onNext(2);"Subscribing 2nd");
subject.subscribe(val ->"Subscriber2 received {}", val), 
                            logError(), logComplete());"Pushing 3");


[main] - Pushing 0
[main] - Pushing 1
[main] - Subscribing 1st
[main] - Subscriber1 received 0
[main] - Subscriber1 received 1
[main] - Pushing 2
[main] - Subscriber1 received 2
[main] - Subscribing 2nd
[main] - Subscriber2 received 0
[main] - Subscriber2 received 1
[main] - Subscriber2 received 2
[main] - Pushing 3
[main] - Subscriber1 received 3
[main] - Subscriber2 received 3
[main] - Subscriber got Completed event
[main] - Subscriber got Completed event

ConnectableObservable / ConnectableFlowable and resource sharing

There are cases when we want to share a single subscription between subscribers, meaning while the code that executes on subscribing should be executed once, the events should be published to all subscribers.

For ex. when we want to share a connection between multiple Observables / Flowables. Using a plain Observable would just reexecute the code inside .create() and opening / closing a new connection for each new subscriber when it subscribes / cancels its subscription.

ConnectableObservable are a special kind of Observable. No matter how many Subscribers subscribe to ConnectableObservable, it opens just one subscription to the Observable from which it was created.

Anyone who subscribes to ConnectableObservable is placed in a set of Subscribers(it doesn't trigger the .create() code a normal Observable would when .subscribe() is called). A .connect() method is available for ConnectableObservable. As long as connect() is not called, these Subscribers are put on hold, they never directly subscribe to upstream Observable

ConnectableObservable<Integer> connectableObservable = 
                                  Observable.<Integer>create(subscriber -> {"Inside create()");

     /* A JMS connection listener example
         Just an example of a costly operation that is better to be shared **/

     /* Connection connection = connectionFactory.createConnection();
        Session session = connection.createSession(true, AUTO_ACKNOWLEDGE);
        MessageConsumer consumer = session.createConsumer(orders);
        consumer.setMessageListener(subscriber::onNext); */

        subscriber.setCancellable(() ->"Subscription cancelled"));"Emitting 1");
        subscriber.onNext(1);"Emitting 2");


       .subscribe((val) ->"Subscriber1 received: {}", val), 
                    logError(), logComplete());

       .subscribe((val) ->"Subscriber2 received: {}", val), 
                    logError(), logComplete());"Now connecting to the ConnectableObservable");


share() operator

Another operator of the ConnectableObservable .refCount() allows to do away with having to manually call .connect(), instead it invokes the .create() code when the first Subscriber subscribes while sharing this single subscription with subsequent Subscribers. This means that .refCount() basically keeps a count of references of it's subscribers and subscribes to upstream Observable (executes the code inside .create() just for the first subscriber), but multicasts the same event to each active subscriber. When the last subscriber unsubscribes, the ref counter goes from 1 to 0 and triggers any unsubscribe callback associated.
If another Subscriber subscribes after that, counter goes from 0 to 1 and the process starts over again.

ConnectableObservable<Integer> connectableStream = Observable.<Integer>create(subscriber -> {"Inside create()");
   //Simulated MessageListener emits periodically every 500 milliseconds
   ResourceConnectionHandler resourceConnectionHandler = new ResourceConnectionHandler() {
        public void onMessage(Integer message) {
   "Emitting {}", message);

   //when the last subscriber unsubscribes it will invoke disconnect on the resourceConnectionHandler

//publish().refCount() have been joined together in the .share() operator
Observable<Integer> observable = connectableObservable.refCount();

CountDownLatch latch = new CountDownLatch(2);
      .subscribe((val) ->"Subscriber1 received: {}", val), 
                    logError(), logComplete(latch));

Helpers.sleepMillis(1000);"Subscribing 2nd");
//we're not seing the code inside .create() reexecuted
      .subscribe((val) ->"Subscriber2 received: {}", val), 
                    logError(), logComplete(latch));

//waiting for the streams to complete

//subscribing another after previous Subscribers unsubscribed
latch = new CountDownLatch(1);"Subscribing 3rd");
     .subscribe((val) ->"Subscriber3 received: {}", val), logError(), logComplete(latch));

private abstract class ResourceConnectionHandler {

   ScheduledExecutorService scheduledExecutorService;

   private int counter;

   public void openConnection() {"**Opening connection");

      scheduledExecutorService = periodicEventEmitter(() -> {
            counter ++;
      }, 500, TimeUnit.MILLISECONDS);

   public abstract void onMessage(Integer message);

   public void disconnect() {"**Shutting down connection");

14:55:23 [main] INFO BaseTestObservables - Inside create()
14:55:23 [main] INFO BaseTestObservables - **Opening connection
14:55:23 [pool-1-thread-1] INFO BaseTestObservables - Emitting 1
14:55:23 [pool-1-thread-1] INFO BaseTestObservables - Subscriber1 received: 1
14:55:24 [pool-1-thread-1] INFO BaseTestObservables - Emitting 2
14:55:24 [pool-1-thread-1] INFO BaseTestObservables - Subscriber1 received: 2
14:55:24 [pool-1-thread-1] INFO BaseTestObservables - Emitting 3
14:55:24 [pool-1-thread-1] INFO BaseTestObservables - Subscriber1 received: 3
14:55:24 [main] INFO BaseTestObservables - Subscribing 2nd
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Emitting 4
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Subscriber1 received: 4
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Subscriber2 received: 4
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Emitting 5
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Subscriber1 received: 5
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Subscriber got Completed event
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Subscriber2 received: 5
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - **Shutting down connection
14:55:25 [pool-1-thread-1] INFO BaseTestObservables - Subscriber got Completed event
14:55:25 [main] INFO BaseTestObservables - Subscribing 3rd
14:55:25 [main] INFO BaseTestObservables - Inside create()
14:55:25 [main] INFO BaseTestObservables - **Opening connection
14:55:25 [pool-2-thread-1] INFO BaseTestObservables - Emitting 1
14:55:25 [pool-2-thread-1] INFO BaseTestObservables - Subscriber3 received: 1
14:55:25 [pool-2-thread-1] INFO BaseTestObservables - **Shutting down connection
14:55:25 [pool-2-thread-1] INFO BaseTestObservables - Subscriber got Completed event

The share() operator of Observable / Flowable is an operator which basically does publish().refCount().


RxJava provides some high level concepts for concurrent execution, like ExecutorService we're not dealing with the low level constructs like creating the Threads ourselves. Instead we're using a Scheduler which create Workers who are responsible for scheduling and running code. By default RxJava will not introduce concurrency and will run the operations on the subscription thread.

There are two methods through which we can introduce Schedulers into our chain of operations:

  • subscribeOn allows to specify which Scheduler invokes the code contained in the lambda code for Observable.create()
  • observeOn allows control to which Scheduler executes the code in the downstream operators

RxJava provides some general use Schedulers:

  • Schedulers.computation() - to be used for CPU intensive tasks. A threadpool. Should not be used for tasks involving blocking IO.
  • - to be used for IO bound tasks
  • Schedulers.from(Executor) - custom ExecutorService
  • Schedulers.newThread() - always creates a new thread when a worker is needed. Since it's not thread pooled and always creates a new thread instead of reusing one, this scheduler is not very useful

Although we said by default RxJava doesn't introduce concurrency. Notice how we are not doing anything on another thread than the subscribing thread 'main' and the Test doesn't end until the complete event is processed:

public void byDefaultRxJavaDoesntIntroduceConcurrency() {"Starting");

   Observable.<Integer>create(subscriber -> {"Someone subscribed");

   .map(val -> {"Mapping {}", val);
         //what if we do some Thread.sleep here 
         return val * 10;
11:23:49 [main] INFO BaseTestObservables - Starting
11:23:50 [main] INFO BaseTestObservables - Someone subscribed
11:23:50 [main] INFO BaseTestObservables - Mapping 1
11:23:50 [main] INFO BaseTestObservables - Subscriber received: 10
11:23:50 [main] INFO BaseTestObservables - Mapping 2
11:23:50 [main] INFO BaseTestObservables - Subscriber received: 20

now let's enable that Thread.sleep(2000) above.

11:42:12 [main] INFO BaseTestObservables - Starting
11:42:12 [main] INFO BaseTestObservables - Someone subscribed
11:42:12 [main] INFO BaseTestObservables - Mapping 1
11:42:14 [main] INFO BaseTestObservables - Subscriber received: 10
11:42:14 [main] INFO BaseTestObservables - Mapping 2
11:42:16 [main] INFO BaseTestObservables - Subscriber received: 20

as expected nothing changes, just that we receive the events in the Subscriber delayed by 2 secs. To prevent this, lots of RxJava operators that involve waiting as delay,interval, zip run on a Scheduler, otherwise they would just block the subscribing thread. By default Schedulers.computation() is used, but the Scheduler can be passed as a parameter to those methods.

Ok so how can we provide different threads to run the different parts of the code.


As stated above subscribeOn allows to specify on which Scheduler thread the subscribtion is made - which thread invokes the code contained in the lambda for Observable.create() - (it's not abouth the thread for where the code in .subscribe((val) -> {...}) gets executed). Since the operators are lazy and nothing happens until subscription, where the .subscribeOn() is called doesn't make any difference. Also calling .subscribeOn() multiple times at different positions doesn't have any effect, only the first .subscribeOn() Scheduler is considered.

public void testSubscribeOn() {"Starting");

   Observable<Integer> observable = Observable.create(subscriber -> { 
       //code that will execute inside the IO ThreadPool"Starting slow network op");
       Helpers.sleepMillis(2000);"Emitting 1st");


   observable = observable
                .subscribeOn( //Specify execution on the IO Scheduler
                .map(val -> {
                    int newValue = val * 10;
          "Mapping {} to {}", val, newValue);
                    return newValue;

   /** Since we are switching the subscription thread we now need to wait 
   * for the Thread to complete so again we are using the CountDownLatch "trick" to do it.
   CountDownLatch latch = new CountDownLatch(1);

13:16:31 [main] INFO BaseTestObservables - Starting
13:16:31 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Starting slow network op
13:16:33 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Emitting 1st
13:16:33 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Mapping 1 to 10
13:16:33 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Subscriber received: 10
13:16:33 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Subscriber got Completed event

Notice how the code and also the flow down the operators like .map() is switched to this new Scheduler that was specified.


observeOn allows control to which Scheduler executes the code in the downstream operators. So by using observeOn() we changed the Scheduler for the map operator, but notice how the last .observeOn(Schedulers.newThread()) we also influence the code received by the subscriber, while .subscribeOn() just had a part on the code executed before we changed with .observeOn()"Starting");

Observable<Integer> observable = 
        Observable.create(subscriber -> { 
                    //code that will execute inside the IO Scheduler
   "Emitting 1st");
   "Emitting 2nd");
        .map(val -> {
              int newValue = val * 10;
    "Mapping {} to {}", val, newValue);
              return newValue;

   CountDownLatch latch = new CountDownLatch(1);

19:35:01 [main] INFO BaseTestObservables - Starting
19:35:01 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Started emitting
19:35:01 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Emitting 1st
19:35:01 [RxCachedThreadScheduler-1] INFO BaseTestObservables - Emitting 2nd
19:35:01 [RxComputationThreadPool-1] INFO BaseTestObservables - Mapping 1 to 10
19:35:01 [RxNewThreadScheduler-1] INFO BaseTestObservables - Subscriber received: 10
19:35:01 [RxComputationThreadPool-1] INFO BaseTestObservables - Mapping 2 to 20
19:35:01 [RxNewThreadScheduler-1] INFO BaseTestObservables - Subscriber received: 20
19:35:01 [RxNewThreadScheduler-1] INFO BaseTestObservables - Subscriber got Completed event

back to blocking world

How about when we want to switch back to a blocking flow. We saw above how we need to explicitly use latching to keep the [main] thread. Say we're incrementally switching from legacy code and we have a Service method Collection<String> findUsers() inside this method we can still be reactive but to the caller of the method we still need to block until we get all the elements of the Collection. Using blockingIterable will block our Test thread till the Flow completes, waiting for the events to be emitted(we're sleeping just to show it's not completing by chance)."Starting");

Flowable<String> flowable = simpleFlowable()
                .map(val -> {
                    String newValue = "^^" + val + "^^";
          "Mapping new val {}", newValue);
                    return newValue;

Iterable<String> iterable = flowable.blockingIterable(); //this call will block until
//the stream completes
iterable.forEach(val ->"Received {}", val));

17:48:13 [RxCachedThreadScheduler-1] - Started emitting
17:48:13 [RxCachedThreadScheduler-1] - Emitting 1st
17:48:13 [RxCachedThreadScheduler-1] - Mapping new val ^^1^^
17:48:14 [RxCachedThreadScheduler-1] - Emitting 2nd
17:48:14 [main] - Received ^^1^^
17:48:14 [RxCachedThreadScheduler-1] - Mapping new val ^^2^^
17:48:14 [main] - Received ^^2^^
17:48:14 [main] - Finished blockingIterable

we can see the events being received back on the [main] thread.

    //block until the stream completes or throws an error.
    flowable.blockingSubscribe(val ->"Subscriber received {}", val));

Flatmap operator

The flatMap operator is so important and has so many different uses it deserves it's own category to explain it. Code at

I like to think of it as a sort of fork-join operation because what flatMap does is it takes individual stream items and maps each of them to an Observable(so it creates new Streams from each object) and then 'flattens' the events from these Streams back as coming from a single stream.

Why this looks like fork-join because for each element you can fork some jobs that keeps emitting results, and these results are emitted back as elements to the subscribers downstream

Rules of thumb to consider before getting comfortable with flatMap:

  • When you have an 'item' T and a method T -< Flowable<X>, you need flatMap. Most common example is when you want to make a remote call that returns an Observable / Flowable . For ex if you have a stream of customerIds, and downstream you want to work with actual Customer objects:

  • When you have Observable<Observable<T>>(aka stream of streams) you probably need flatMap. Because flatMap means you are subscribing to each substream.

We use a simulated remote call that returns asynchronous events. This is a most common scenario to make a remote call for each stream element, (although in non reactive world we're more likely familiar with remote operations returning Lists T -> List<X>). Our simulated remote operation produces as many events as the length of the color string received as parameter every 200ms, so for example red : red0, red1, red2

private Flowable<String> simulateRemoteOperation(String color) {
  return Flowable.intervalRange(1, color.length(), 0, 200, TimeUnit.MILLISECONDS)
             .map(iteration -> color + iteration);

If we have a stream of color names:

Flowable<String> colors = Flowable.just("orange", "red", "green")

to invoke the remote operation:

Flowable<String> colors = Flowable.just("orange", "red", "green")
         .flatMap(colorName -> simulatedRemoteOperation(colorName));

colors.subscribe(val ->"Subscriber received: {}", val));         

16:44:15 [Thread-0]- Subscriber received: orange0
16:44:15 [Thread-2]- Subscriber received: green0
16:44:15 [Thread-1]- Subscriber received: red0
16:44:15 [Thread-0]- Subscriber received: orange1
16:44:15 [Thread-2]- Subscriber received: green1
16:44:15 [Thread-1]- Subscriber received: red1
16:44:15 [Thread-0]- Subscriber received: orange2
16:44:15 [Thread-2]- Subscriber received: green2
16:44:15 [Thread-1]- Subscriber received: red2
16:44:15 [Thread-0]- Subscriber received: orange3
16:44:15 [Thread-2]- Subscriber received: green3
16:44:16 [Thread-0]- Subscriber received: orange4
16:44:16 [Thread-2]- Subscriber received: green4
16:44:16 [Thread-0]- Subscriber received: orange5

Notice how the results are coming intertwined(mixed) and it might not be as you expected it.This is because flatMap actually subscribes to it's inner Observables returned from 'simulateRemoteOperation'. You can specify the concurrency level of flatMap as a parameter. Meaning you can say how many of the substreams should be subscribed "concurrently" - after onComplete is triggered on the substreams, a new substream is subscribed-.

By setting the concurrency to 1 we don't subscribe to other substreams until the current one finishes:

Flowable<String> colors = Flowable.just("orange", "red", "green")
                     .flatMap(val -> simulateRemoteOperation(val), 1); //

Notice now there is a sequence from each color before the next one appears

17:15:24 [Thread-0]- Subscriber received: orange0
17:15:24 [Thread-0]- Subscriber received: orange1
17:15:25 [Thread-0]- Subscriber received: orange2
17:15:25 [Thread-0]- Subscriber received: orange3
17:15:25 [Thread-0]- Subscriber received: orange4
17:15:25 [Thread-0]- Subscriber received: orange5
17:15:25 [Thread-1]- Subscriber received: red0
17:15:26 [Thread-1]- Subscriber received: red1
17:15:26 [Thread-1]- Subscriber received: red2
17:15:26 [Thread-2]- Subscriber received: green0
17:15:26 [Thread-2]- Subscriber received: green1
17:15:26 [Thread-2]- Subscriber received: green2
17:15:27 [Thread-2]- Subscriber received: green3
17:15:27 [Thread-2]- Subscriber received: green4

There is actually an operator which is basically this flatMap with 1 concurrency called concatMap.

Inside the flatMap we can operate on the substream with the same stream operators

Observable<Pair<String, Integer>> colorsCounted = colors
    .flatMap(colorName -> {
               Observable<Long> timer = Observable.interval(2, TimeUnit.SECONDS);

               return simulateRemoteOperation(colorName) // <- Still a stream
                              .zipWith(timer, (val, timerVal) -> val)
                              .map(counter -> new Pair<>(colorName, counter));

We can also use switchIfEmpty to provide some values when the original Publisher doesn't return anything, just completes.

Flowable<String> colors = Flowable.just("red", "", "blue")
                            .flatMap(colorName -> simulateRemoteOperation(colorName)

13:11:02  Subscriber received: red0
13:11:02  Subscriber received: red1
13:11:02  Subscriber received: red2
13:11:03  Subscriber received: NONE
13:11:03  Subscriber received: blue0
13:11:03  Subscriber received: blue1
13:11:03  Subscriber received: blue2
13:11:03  Subscriber received: blue3
13:11:03  Subscriber got Completed event

flatMapIterable is just an easy way to pass each of the elements of a collection as a stream

Flowable<String> colors = Flowable.just(1)
                .flatMapIterable(it -> generateColors());

private List<String> generateColors() {
   return Arrays.asList("red", "green", "blue");

switchMap operator also prevents inter-leavings as only one of stream is subscribed at a time, but this is controlled from upstream. If a new value comes from upstream, the current subscribed inner-stream gets canceled and a new subscription is made for the new value. The current stream will remain subscribed as long as there are no new values from upstream.

Flowable<String> colors = Flowable.interval(0,400, TimeUnit.MILLISECONDS)
         .zipWith(Arrays.asList("EUR", "USD", "GBP"), (it, currency) -> currency)
         .doOnNext(ev ->"Emitting {}", ev))
         .switchMap(currency -> simulateRemoteOperation(currency)
                      .doOnSubscribe((subscription) ->"Subscribed new"))
                      .doOnCancel(() ->"Unsubscribed {}", currency))

17:45:16 [RxComputationThreadPool-1] INFO BaseTestObservables - Emitting EUR 17:45:16 [RxComputationThreadPool-1] INFO BaseTestObservables - Subscribed new 17:45:16 [RxComputationThreadPool-2] INFO BaseTestObservables - Subscriber received: EUR1 17:45:16 [RxComputationThreadPool-2] INFO BaseTestObservables - Subscriber received: EUR2 17:45:16 [RxComputationThreadPool-1] INFO BaseTestObservables - Emitting USD 17:45:16 [RxComputationThreadPool-1] INFO BaseTestObservables - Unsubscribed EUR 17:45:16 [RxComputationThreadPool-1] INFO BaseTestObservables - Subscribed new 17:45:16 [RxComputationThreadPool-3] INFO BaseTestObservables - Subscriber received: USD1 17:45:16 [RxComputationThreadPool-3] INFO BaseTestObservables - Subscriber received: USD2 17:45:17 [RxComputationThreadPool-1] INFO BaseTestObservables - Emitting GBP 17:45:17 [RxComputationThreadPool-1] INFO BaseTestObservables - Unsubscribed USD 17:45:17 [RxComputationThreadPool-1] INFO BaseTestObservables - Subscribed new 17:45:17 [RxComputationThreadPool-3] INFO BaseTestObservables - Subscriber received: USD3 17:45:17 [RxComputationThreadPool-4] INFO BaseTestObservables - Subscriber received: GBP1 17:45:17 [RxComputationThreadPool-4] INFO BaseTestObservables - Subscriber received: GBP2 17:45:17 [RxComputationThreadPool-4] INFO BaseTestObservables - Subscriber received: GBP3 17:45:17 [RxComputationThreadPool-4] INFO BaseTestObservables - Subscriber got Completed event

Error handling

Code at

Exceptions are for exceptional situations. The Reactive Streams specification says that exceptions are terminal operations. That means in case an error occurs, it triggers an unsubscription upstream and the error travels downstream to the Subscriber, invoking the 'onError' handler:

Observable<String> colors = Observable.just("green", "blue", "red", "yellow")
       .map(color -> {
              if ("red".equals(color)) {
                        throw new RuntimeException("Encountered red");
              return color + "*";
       .map(val -> val + "XXX");

         val ->"Subscriber received: {}", val),
         exception -> log.error("Subscriber received error '{}'", exception.getMessage()),
         () ->"Subscriber completed")


23:30:17 [main] INFO - Subscriber received: green*XXX
23:30:17 [main] INFO - Subscriber received: blue*XXX
23:30:17 [main] ERROR - Subscriber received error 'Encountered red'

After the map() operator encounters an error it unsubscribes(cancels the subscription) from upstream (therefore 'yellow' is not even emitted). The error travels downstream and triggers the error handler in the Subscriber.

There are operators to deal with error flow control:


The 'onErrorReturn' operator replaces an exception with a value:

Flowable<Integer> numbers = Flowable.just("1", "3", "a", "4", "5", "c")
                            .doOnCancel(() ->"Subscription canceled"))

Subscriber received: 1
Subscriber received: 3
Subscription canceled
Subscriber received: 0
Subscriber got Completed event

Notice though how it didn't prevent map() operator from unsubscribing from the Flowable, but it did trigger the normal onNext callback instead of onError in the subscriber.

Let's introduce a more realcase scenario of a simulated remote request that fails whenever it's invoked with "red" and "black" color parameters otherwise just add some *s.

private Observable<String> simulateRemoteOperation(String color) {
    return Observable.<String>create(subscriber -> {
         if ("red".equals(color)) {
    "Emitting RuntimeException for {}", color);
              throw new RuntimeException("Color red raises exception");
         if ("black".equals(color)) {
    "Emitting IllegalArgumentException for {}", color);
              throw new IllegalArgumentException("Black is not a color");

         String value = "**" + color + "**";
"Emitting {}", value);

Flowable<String> colors = Flowable.just("green", "blue", "red", "white", "blue")
                .flatMap(color -> simulateRemoteOperation(color))
                .onErrorReturn(throwable -> "-blank-");


22:15:51 [main] INFO - Emitting **green**
22:15:51 [main] INFO - Subscriber received: **green**
22:15:51 [main] INFO - Emitting **blue**
22:15:51 [main] INFO - Subscriber received: **blue**
22:15:51 [main] INFO - Emitting RuntimeException for red
22:15:51 [main] INFO - Subscriber received: -blank-
22:15:51 [main] INFO - Subscriber got Completed event

flatMap encounters an error when it subscribes to 'red' substreams and thus still unsubscribe from 'colors' stream and the remaining colors are not longer emitted

Flowable<String> colors = Flowable.just("green", "blue", "red", "white", "blue")
                .flatMap(color -> simulateRemoteOperation(color)
                                    .onErrorReturn(throwable -> "-blank-")

onErrorReturn() is applied to the flatMap substream and thus translates the exception to a value and so flatMap continues on with the other colors after red


22:15:51 [main] INFO - Emitting **green**
22:15:51 [main] INFO - Subscriber received: **green**
22:15:51 [main] INFO - Emitting **blue**
22:15:51 [main] INFO - Subscriber received: **blue**
22:15:51 [main] INFO - Emitting RuntimeException for red
22:15:51 [main] INFO - Subscriber received: -blank-
22:15:51 [main] INFO - Emitting **white**
22:15:51 [main] INFO - Subscriber received: **white**
22:15:51 [main] INFO - Emitting **blue**
22:15:51 [main] INFO - Subscriber received: **blue**
22:15:51 [main] INFO - Subscriber got Completed event


onErrorResumeNext() returns a stream instead of an exception, useful for example to invoke a fallback method that returns an alternate Stream

Observable<String> colors = Observable.just("green", "blue", "red", "white", "blue")
     .flatMap(color -> simulateRemoteOperation(color)
                        .onErrorResumeNext(th -> {
                            if (th instanceof IllegalArgumentException) {
                                return Observable.error(new RuntimeException("Fatal, wrong arguments"));
                            return fallbackRemoteOperation();

private Observable<String> fallbackRemoteOperation() {
        return Observable.just("blank");



Timeout operator raises exception when there are no events incoming before it's predecessor in the specified time limit.


retry() - resubscribes in case of exception to the Observable

Flowable<String> colors = Flowable.just("red", "blue", "green", "yellow")
       .concatMap(color -> delayedByLengthEmitter(TimeUnit.SECONDS, color) 
                             //if there are no events flowing in the timeframe   
                             .timeout(6, TimeUnit.SECONDS)  



12:40:16 [main] INFO - Received red delaying for 3 
12:40:19 [main] INFO - Subscriber received: red
12:40:19 [RxComputationScheduler-2] INFO - Received blue delaying for 4 
12:40:23 [main] INFO - Subscriber received: blue
12:40:23 [RxComputationScheduler-4] INFO - Received green delaying for 5 
12:40:28 [main] INFO - Subscriber received: green
12:40:28 [RxComputationScheduler-6] INFO - Received yellow delaying for 6 
12:40:34 [RxComputationScheduler-7] INFO - Received yellow delaying for 6 
12:40:40 [RxComputationScheduler-1] INFO - Received yellow delaying for 6 
12:40:46 [main] INFO - Subscriber received: blank
12:40:46 [main] INFO - Subscriber got Completed event

When you want to retry considering the thrown exception type:

Observable<String> colors = Observable.just("blue", "red", "black", "yellow")
         .flatMap(colorName -> simulateRemoteOperation(colorName)
                .retry((retryAttempt, exception) -> {
                           if (exception instanceof IllegalArgumentException) {
                               log.error("{} encountered non retry exception ", colorName);
                               return false;
                 "Retry attempt {} for {}", retryAttempt, colorName);
                           return retryAttempt <= 2;
                .onErrorResumeNext(Observable.just("generic color"))
13:21:37 [main] INFO - Emitting **blue**
13:21:37 [main] INFO - Emitting RuntimeException for red
13:21:37 [main] INFO - Retry attempt 1 for red
13:21:37 [main] INFO - Emitting RuntimeException for red
13:21:37 [main] INFO - Retry attempt 2 for red
13:21:37 [main] INFO - Emitting RuntimeException for red
13:21:37 [main] INFO - Retry attempt 3 for red
13:21:37 [main] INFO - Emitting IllegalArgumentException for black
13:21:37 [main] ERROR - black encountered non retry exception 
13:21:37 [main] INFO - Emitting **yellow**
13:21:37 [main] INFO - Subscriber received: **blue**
13:21:37 [main] INFO - Subscriber received: generic color
13:21:37 [main] INFO - Subscriber received: generic color
13:21:37 [main] INFO - Subscriber received: **yellow**
13:21:37 [main] INFO - Subscriber got Completed event


A more complex retry logic like implementing a backoff strategy in case of exception This can be obtained with retryWhen(exceptionObservable -> Observable)

retryWhen resubscribes when an event from an Observable is emitted. It receives as parameter an exception stream

we zip the exceptionsStream with a .range() stream to obtain the number of retries, however we want to wait a little before retrying so in the zip function we return a delayed event - .timer()

The delay also needs to be subscribed to be effected so we also flatMap

Observable<String> colors = Observable.just("blue", "green", "red", "black", "yellow");

colors.flatMap(colorName -> 
                      .retryWhen(exceptionStream -> exceptionStream
                                    .zipWith(Observable.range(1, 3), (exc, attempts) -> {
                                        //don't retry for IllegalArgumentException
                                        if(exc instanceof IllegalArgumentException) {
                                             return Observable.error(exc);

                                        if(attempts < 3) {
                                             return Observable.timer(2 * attempts, TimeUnit.SECONDS);
                                        return Observable.error(exc);
                                    .flatMap(val -> val)
                      .onErrorResumeNext(Observable.just("generic color")
15:20:23 [main] INFO - Emitting **blue**
15:20:23 [main] INFO - Emitting **green**
15:20:23 [main] INFO - Emitting RuntimeException for red
15:20:23 [main] INFO - Emitting IllegalArgumentException for black
15:20:23 [main] INFO - Emitting **yellow**
15:20:23 [main] INFO - Subscriber received: **blue**
15:20:23 [main] INFO - Subscriber received: **green**
15:20:23 [main] INFO - Subscriber received: generic color
15:20:23 [main] INFO - Subscriber received: **yellow**
15:20:25 [RxComputationScheduler-1] INFO - Emitting RuntimeException for red
15:20:29 [RxComputationScheduler-2] INFO - Emitting RuntimeException for red
15:20:29 [main] INFO - Subscriber received: generic color
15:20:29 [main] INFO - Subscriber got Completed event

retryWhen vs repeatWhen With similar names it worth noting the difference.

  • repeat() resubscribes when it receives onCompleted().
  • retry() resubscribes when it receives onError().

Example using repeatWhen() to implement periodic polling

remoteOperation.repeatWhen(completed -> completed
                                     .delay(2, TimeUnit.SECONDS))                                                       


It can be the case of a slow consumer that cannot keep up with the producer that is producing too many events that the subscriber cannot process.

Backpressure relates to a feedback mechanism through which the subscriber can signal to the producer how much data it can consume and so to produce only that amount.

The reactive-streams section above we saw that besides the onNext, onError and onComplete handlers, the Subscriber has an onSubscribe(Subscription), Subscription through which it can signal upstream it's ready to receive a number of items and after it processes the items request another batch.

public interface Subscriber<T> {
    //signals to the Publisher to start sending events
    public void onSubscribe(Subscription s);     
    public void onNext(T t);
    public void onError(Throwable t);
    public void onComplete();

The methods exposed by Subscription through which the subscriber comunicates with the upstream:

public interface Subscription {
    public void request(long n); //request n items
    public void cancel();

So in theory the Subscriber can prevent being overloaded by requesting an initial number of items. The Publisher would send those items downstream and not produce any more, until the Subscriber would request more. We say in theory because until now we did not see a custom onSubscribe(Subscription) request being implemented. This is because if not specified explicitly, there is a default implementation which requests of Long.MAX_VALUE which basically means "send all you have".

Neither did we see the code in the producer that takes consideration of the number of items requested by the subscriber.

Flowable.create(subscriber -> {"Started emitting");

      for(int i=0; i < 300; i++) {
           if(subscriber.isCanceled()) {
 "Emitting {}", i);

}, BackpressureStrategy.BUFFER); //BackpressureStrategy will be explained further bellow

Looks like it's not possible to slow down production based on request(as there is no reference to the requested items), we can at most stop production if the subscriber canceled subscription.

This can be done if we extend Flowable so we can pass our custom Subscription type to the downstream subscriber:

private class CustomRangeFlowable extends Flowable<Integer> {

        private int startFrom;
        private int count;

        CustomRangeFlowable(int startFrom, int count) {
            this.startFrom = startFrom;
            this.count = count;

        public void subscribeActual(Subscriber<? super Integer> subscriber) {
            subscriber.onSubscribe(new CustomRangeSubscription(startFrom, count, subscriber));

        class CustomRangeSubscription implements Subscription {

            volatile boolean cancelled;
            boolean completed = false;
            private int count;
            private int currentCount;
            private int startFrom;

            private Subscriber<? super Integer> actualSubscriber;

            CustomRangeSubscription(int startFrom, int count, Subscriber<? super Integer> actualSubscriber) {
                this.count = count;
                this.startFrom = startFrom;
                this.actualSubscriber = actualSubscriber;

            public void request(long items) {
      "Downstream requests {} items", items);
                for(int i=0; i < items; i++) {
                    if(cancelled || completed) {

                    if(currentCount == count) {
                        completed = true;
                        if(cancelled) {


                    int emitVal = startFrom + currentCount;

            public void cancel() {
                cancelled = true;

Now lets see how we can custom control how many items we request from upstream, to simulate an initial big request, and then a request for other smaller batches of items as soon as the subscriber finishes and is ready for another batch.

Flowable<Integer> flowable = new CustomRangeFlowable(5, 10);

flowable.subscribe(new Subscriber<Integer>() {

       private Subscription subscription;
       private int backlogItems;

       private final int BATCH = 2;
       private final int INITIAL_REQ = 5;

       public void onSubscribe(Subscription subscription) {
                this.subscription = subscription;
                backlogItems = INITIAL_REQ;

      "Initial request {}", backlogItems);

            public void onNext(Integer val) {
      "Subscriber received {}", val);
                backlogItems --;

                if(backlogItems == 0) {
                    backlogItems = BATCH;

            public void onError(Throwable throwable) {
      "Subscriber encountered error");

            public void onComplete() {
      "Subscriber completed");
Initial request 5
Downstream requests 5 items
Subscriber received 5
Subscriber received 6
Subscriber received 7
Subscriber received 8
Subscriber received 9
Downstream requests 2 items
Subscriber received 10
Subscriber received 11
Downstream requests 2 items
Subscriber received 12
Subscriber received 13
Downstream requests 2 items
Subscriber received 14
Subscriber completed        

Returning to the Flowable.create() example since it's not taking any account of the requested items by the subscriber, does it mean it might overwhelm a slow Subscriber?

private Flowable<Integer> createFlowable(int items,
                     BackpressureStrategy backpressureStrategy) {

return Flowable.create(subscriber -> {"Started emitting");

        for (int i = 0; i < items; i++) {
            if(subscriber.isCancelled()) {
  "Emitting {}", i);

}, backpressureStrategy); //can be BackpressureStrategy.DROP, BUFFER, LATEST,..

This is where the 2nd parameter BackpressureStrategy comes in that allows you to specify what to do in the case.

  • BackpressureStrategy.BUFFER buffer in memory the events that overflow. Of course is we don't drop over some threshold, it might lead to OufOfMemory.
  • BackpressureStrategy.DROP just drop the overflowing events
  • BackpressureStrategy.LATEST keep only recent event and discards previous unconsumed events.
  • BackpressureStrategy.ERROR we get an error in the subscriber immediately
  • BackpressureStrategy.MISSING means we don't care about backpressure(we let one of the downstream operators onBackpressureXXX handle it -explained further down-)

Still what does it mean to 'overwhelm' the subscriber? It means to emit more items than requested by downstream subscriber. But we said that by default the subscriber requests Long.MAX_VALUE since the code flowable.subscribe(onNext(), onError, onComplete) uses a default onSubscribe:

(subscription) -> subscription.request(Long.MAX_VALUE);

so unless we override it like in our custom Subscriber above, it means it would never overflow. But between the Publisher and the Subscriber you'd have a series of operators. When we subscribe, a Subscriber travels up through all operators to the original Publisher and some operators override the requested items upstream. One such operator is observeOn() which makes it's own request to the upstream Publisher(256 by default), but can take a parameter to specify the request size.

Flowable<Integer> flowable = createFlowable(5, BackpressureStrategy.DROP)
                .observeOn(, false, 3);
flowable.subscribe((val) -> {
                     "Subscriber received: {}", val);
                           }, logError(), logComplete());
[main] - Started emitting
[main] - Emitting 0
[main] - Emitting 1
[main] - Emitting 2
[main] - Emitting 3
[main] - Emitting 4
[RxCachedThreadScheduler-1] - Subscriber received: 0
[RxCachedThreadScheduler-1] - Subscriber received: 1
[RxCachedThreadScheduler-1] - Subscriber received: 2
[RxCachedThreadScheduler-1] - Subscriber got Completed event  

This is expected, as the subscription travels upstream through the operators to the source Flowable, while initially the Subscriber requesting Long.MAX_VALUE from the upstream operator observeOn, which in turn subscribes to the source and it requests just 3 items from the source instead. Since we used BackpressureStrategy.DROP all the items emitted outside the expected 3, get discarded and thus never reach our subscriber.

You may wonder what would have happened if we didn't use observeOn. We had to use it if we wanted to be able to produce faster than the subscriber(it wasn't just to show a limited request operator), because we'd need a separate thread to produce events faster than the subscriber processes them.

Also you can transform an Observable to Flowable by specifying a BackpressureStrategy, otherwise Observables just throw exception on overflowing(same as using BackpressureStrategy.DROP in Flowable.create()).

Flowable flowable = observable.toFlowable(BackpressureStrategy.DROP)

so can a hot Publisher be converted to a Flowable:

PublishSubject<Integer> subject = PublishSubject.create();

Flowable<Integer> flowable = subject

There are also specialized operators to handle backpressure the onBackpressureXXX operators: onBackpressureBuffer, onBackpressureDrop, onBackpressureLatest

These operators request Long.MAX_VALUE(unbounded amount) from upstream and then take it upon themselves to manage the requests from downstream. In the case of onBackpressureBuffer it adds in an internal queue and send downstream the events as requested, onBackpressureDrop just discards events that are received from upstream more than requested from downstream, onBackpressureLatest also drops emitted events excluding the last emitted event(most recent).

Flowable<Integer> flowable = createFlowable(10, BackpressureStrategy.MISSING)
                .onBackpressureBuffer(5, () ->"Buffer has overflown"));

flowable = flowable
                .observeOn(, false, 3);

[main] - Started emitting
[main] - Emitting 0
[main] - Emitting 1
[RxCachedThreadScheduler-1] - Subscriber received: 0
[main] - Emitting 2
[main] - Emitting 3
[main] - Emitting 4
[main] - Emitting 5
[main] - Emitting 6
[main] - Emitting 7
[main] - Emitting 8
[main] - Emitting 9
[main] - Buffer has overflown
[RxCachedThreadScheduler-1] ERROR - Subscriber received error 'Buffer is full'                

We create the Flowable with BackpressureStrategy.MISSING saying we don't care about backpressure but let one of the onBackpressureXXX operators handle it. Notice however

Chaining together multiple onBackpressureXXX operators doesn't actually make sense Using something like

Flowable<Integer> flowable = createFlowable(10, BackpressureStrategy.MISSING)
                 .onBackpressureDrop((val) ->"Dropping {}", val))
flowable = flowable
                .observeOn(, false, 3);

is not behaving as probably you'd expected - buffer 5 values, and then dropping overflowing events-. Because onBackpressureDrop subscribes to the previous onBackpressureBuffer operator signaling it's requesting Long.MAX_VALUE(unbounded amount) from it. Thus onBackpressureBuffer will never feel its subscriber is overwhelmed and never "trigger", meaning that the last onBackpressureXXX operator overrides the previous one if they are chained.

Of course for implementing an event dropping strategy after a full buffer, there is the special overrided version of onBackpressureBuffer that takes a BackpressureOverflowStrategy.

Flowable<Integer> flowable = createFlowable(10, BackpressureStrategy.MISSING)
                .onBackpressureBuffer(5, () ->"Buffer has overflown"),

flowable = flowable
                .observeOn(, false, 3);


[main] - Started emitting
[main] - Emitting 0
[main] - Emitting 1
[RxCachedThreadScheduler-1] - Subscriber received: 0
[main] - Emitting 2
[main] - Emitting 3
[main] - Emitting 4
[main] - Emitting 5
[main] - Emitting 6
[main] - Emitting 7
[main] - Emitting 8
[main] - Buffer has overflown
[main] - Emitting 9
[main] - Buffer has overflown
[RxCachedThreadScheduler-1] - Subscriber received: 1
[RxCachedThreadScheduler-1] - Subscriber received: 2
[RxCachedThreadScheduler-1] - Subscriber received: 5
[RxCachedThreadScheduler-1] - Subscriber received: 6
[RxCachedThreadScheduler-1] - Subscriber received: 7
[RxCachedThreadScheduler-1] - Subscriber received: 8
[RxCachedThreadScheduler-1] - Subscriber received: 9
[RxCachedThreadScheduler-1] - Subscriber got Completed event

onBackpressureXXX operators can be added whenever necessary and it's not limited to cold publishers and we can use them on hot publishers also.

Articles and books for further reading

Reactive Programming with RxJava


RxJava playground



No releases published


No packages published