The document discusses extensions to Apache Flink's DataStream API, specifically focusing on the implementation of ProcessFunction for low-level stream processing and asynchronous I/O operations. It provides examples of using the ProcessFunction to manage state and register timers for handling streams of events, including how to maintain counts and emit results. Additionally, the document outlines the structure of an asynchronous function and its integration with the DataStream API to enhance performance through non-blocking I/O operations.

1. 1
Kostas Kloudas
@KLOUBEN_K
Flink Forward San Francisco
April 11, 2017
Extending Flink’s Streaming APIs

2. 2
Original creators of Apache
Flink®
Providers of the
dA Platform, a supported
Flink distribution

3. Extensions to the DataStream API
3

4. Extensions to the DataStream API
4
 ProcessFunction for Low-level Operations
 Support for Asynchronous I/O

5. ProcessFunction
5

6. Stream Processing
6
Computation
Computations on
never-ending
“streams” of events

7. Distributed Stream Processing
7
Computation
Computation
spread across
many machines
Computation Computation

8. Stateful Stream Processing
8
Computation
State
Result depends
on history of
stream

9. Stream Processing Engines
 Time:
• handle infinite streams
• with out-of-order events
 State:
• guarantee fault-tolerance (distributed)
• guarantee consistency (infinite streams)
9

10.  Gives access to all basic building blocks:
• Events
• Fault-tolerant, Consistent State
• Timers (event- and processing-time)
• Side Outputs
10
ProcessFunction

11. Common Usecase Skeleton A
 On each incoming element:
• update some state
• register a callback for a moment in the future
 When that moment comes:
• Check a condition and perform a certain
action, e.g. emit an element
11

12.  Use built-in windowing:
• +Expressive
• +A lot of functionality out-of-the-box
• - Not always intuitive
• - An overkill for simple cases
 Write your own operator:
• - Too many things to account for
12
Before the ProcessFunction

13.  Simple yet powerful API:
13
/**
* Process one element from the input stream.
*/
void processElement(I value, Context ctx, Collector<O> out) throws Exception;
/**
* Called when a timer set using {@link TimerService} fires.
*/
void onTimer(long timestamp, OnTimerContext ctx, Collector<O> out) throws Exception;
ProcessFunction

14.  Simple yet powerful API:
14
/**
* Process one element from the input stream.
*/
void processElement(I value, Context ctx, Collector<O> out) throws Exception;
/**
* Called when a timer set using {@link TimerService} fires.
*/
void onTimer(long timestamp, OnTimerContext ctx, Collector<O> out) throws Exception;
A collector to emit result
values
ProcessFunction

15.  Simple yet powerful API:
15
/**
* Process one element from the input stream.
*/
void processElement(I value, Context ctx, Collector<O> out) throws Exception;
/**
* Called when a timer set using {@link TimerService} fires.
*/
void onTimer(long timestamp, OnTimerContext ctx, Collector<O> out) throws Exception;
1. Get the timestamp of the element
2. Register and use side outputs
3. Interact with the TimerService to:
• query the current time
• register timers
1. Do the above
2. Query if we are on Event or
Processing time
ProcessFunction

16.  Requirements:
• maintain counts per incoming key, and
• emit the key/count pair if no element came for
the key in the last 100 ms (in event time)
16
ProcessFunction: example

17. 17
 Implementation sketch:
• Store the count, key and last mod timestamp in
a ValueState (scoped by key)
• For each record:
• update the counter and the last mod timestamp
• register a timer 100ms from “now” (in event time)
• When the timer fires:
• check the timer’s timestamp against the last mod time for that
key and
• emit the key/count pair if they differ by 100ms
ProcessFunction: example

18. 18
public class MyProcessFunction extends
ProcessFunction<Tuple2<String, String>, Tuple2<String, Long>> {
// define your state descriptors
@Override
public void processElement(Tuple2<String, Long> value, Context ctx,
Collector<Tuple2<String, Long>> out) throws Exception {
// update our state and register a timer
}
@Override
public void onTimer(long timestamp, OnTimerContext ctx,
Collector<Tuple2<String, Long>> out) throws Exception {
// check the state for the key and emit a result if needed
}
}
ProcessFunction: example

19. 19
public class MyProcessFunction extends
ProcessFunction<Tuple2<String, String>, Tuple2<String, Long>> {
// define your state descriptors
private final ValueStateDescriptor<CounterWithTS> stateDesc =
new ValueStateDescriptor<>("myState", CounterWithTS.class);
}
ProcessFunction: example

20. 20
public class MyProcessFunction extends
ProcessFunction<Tuple2<String, String>, Tuple2<String, Long>> {
@Override
public void processElement(Tuple2<String, String> value, Context ctx,
Collector<Tuple2<String, Long>> out) throws Exception {
ValueState<MyStateClass> state = getRuntimeContext().getState(stateDesc);
CounterWithTS current = state.value();
if (current == null) {
current = new CounterWithTS();
current.key = value.f0;
}
current.count++;
current.lastModified = ctx.timestamp();
state.update(current);
ctx.timerService().registerEventTimeTimer(current.lastModified + 100);
}
}
ProcessFunction: example

21. 21
public class MyProcessFunction extends
ProcessFunction<Tuple2<String, String>, Tuple2<String, Long>> {
@Override
public void onTimer(long timestamp, OnTimerContext ctx,
Collector<Tuple2<String, Long>> out) throws Exception {
CounterWithTS result = getRuntimeContext().getState(stateDesc).value();
if (timestamp == result.lastModified + 100) {
out.collect(new Tuple2<String, Long>(result.key, result.count)); }
}
}
ProcessFunction: example

22. 22
stream.keyBy(”key”)
.process(new MyProcessFunction())
ProcessFunction: example

23. ProcessFunction: Side Outputs
 Additional (to the main) output streams
 No type limitations
• each side output can have its own type
23

24.  Requirements:
• maintain counts per incoming key, and
• emit the key/count pair if no element came for
the key in the last 100 ms (in event time)
• in other case, if the count > 10, send the key
to a side-output named gt10
24
ProcessFunction: example+

25. 25
final OutputTag<String> outputTag = new OutputTag<String>(”gt10"){};
SingleOutputStreamOperator<Tuple2<String, Long>> mainStream = input.process(
new ProcessFunction<Tuple2<String, String>, Tuple2<String, Long>>() {
@Override
public void onTimer(long timestamp, OnTimerContext ctx,
Collector<Tuple2<String, Long>> out) throws Exception {
CounterWithTS result = getRuntimeContext().getState(adStateDesc).value();
if (timestamp == result.lastModified + 100) {
out.collect(new Tuple2<String, Long>(result.key, result.count));
} else if (result.count > 10) {
ctx.output(outputTag, result.key);
}
}
DataStream<String> sideOutputStream = mainStream.getSideOutput(outputTag);
ProcessFunction: example+

26. 26
final OutputTag<String> outputTag = new OutputTag<String>(”gt10"){};
SingleOutputStreamOperator<Tuple2<String, Long>> mainStream = input.process(
new ProcessFunction<Tuple2<String, String>, Tuple2<String, Long>>() {
@Override
public void onTimer(long timestamp, OnTimerContext ctx,
Collector<Tuple2<String, Long>> out) throws Exception {
CounterWithTS result = getRuntimeContext().getState(adStateDesc).value();
if (timestamp == result.lastModified + 100) {
out.collect(new Tuple2<String, Long>(result.key, result.count));
} else if (result.count > 10) {
ctx.output(outputTag, result.key);
}
}
DataStream<String> sideOutputStream = mainStream.getSideOutput(outputTag);
ProcessFunction: example+

27. 27
 Applicable to Keyed streams
 For Non-Keyed streams:
 group on a dummy key if you need the timers
 BEWARE: parallelism of 1
 Use it directly without the timers
 CoProcessFunction for low-level joins:
• Applied on two input streams
ProcessFunction

28. Asynchronous I/O
28

29. Common Usecase Skeleton B
29
 On each incoming element:
• extract some info from the element (e.g. key)
• query an external storage system (DB or KV-
store) for additional info
• emit an enriched version of the input element

30.  Write a MapFunction that queries the DB:
• +Simple
• - Slow (synchronous access) or/and
• - Requires high parallelism (more tasks)
 Write your own operator:
• - Too many things to account for
30
Before the AsuncIO support

31.  Write a MapFunction that queries the DB:
• +Simple
• - Slow (synchronous access) or/and
• - Requires high parallelism (more tasks)
 Write your own operator:
• - Too many things to account for
31
Before the AsyncIO support

32. 32
Synchronous Access

33. 33
Communication delay can
dominate application
throughput and latency
Synchronous Access

34. 34
Asynchronous Access

35.  Requirement:
• a client that supports asynchronous requests
 Flink handles the rest:
• integration of async IO with DataStream API
• fault-tolerance
• order of emitted elements
• correct time semantics (event/processing time)
35
AsyncFunction

36.  Simple API:
/**
* Trigger async operation for each stream input.
*/
void asyncInvoke(IN input, AsyncCollector<OUT> collector) throws Exception;
 API call:
/**
* Example async function call.
*/
DataStream<...> result = AsyncDataStream.(un)orderedWait(stream,
new MyAsyncFunction(), 1000, TimeUnit.MILLISECONDS, 100);
36
AsyncFunction

37. 37
Emitter
P2P3 P1P4
AsyncWaitOperator
E5
AsyncWaitOperator:
• a queue of “Promises”
• a separate thread (Emitter)
AsyncFunction

38. 38
Emitter
P2P3 P1P4
AsyncWaitOperator
• Wrap E5 in a “promise” P5
• Put P5 in the queue
• Call asyncInvoke(E5, P5)
E5
P5
asyncInvoke(E5, P5)P5
AsyncFunction

39. 39
Emitter
P2P3 P1P4
AsyncWaitOperator
E5
P5
asyncInvoke(E5, P5)P5
asyncInvoke(value, asyncCollector):
• a user-defined function
• value : the input element
• asyncCollector : the collector of the
result (when the query returns)
AsyncFunction

40. 40
Emitter
P2P3 P1P4
AsyncWaitOperator
E5
P5
asyncInvoke(E5, P5)P5
asyncInvoke(value, asyncCollector):
• a user-defined function
• value : the input element
• asyncCollector : the collector of the
result (when the query returns)
Future<String> future = client.query(E5);
future.thenAccept((String result) -> { P5.collect(
Collections.singleton(
new Tuple2<>(E5, result)));
});
AsyncFunction

41. 41
Emitter
P2P3 P1P4
AsyncWaitOperator
E5
P5
asyncInvoke(E5, P5)P5
asyncInvoke(value, asyncCollector):
• a user-defined function
• value : the input element
• asyncCollector : the collector of the
result (when the query returns)
Future<String> future = client.query(E5);
future.thenAccept((String result) -> { P5.collect(
Collections.singleton(
new Tuple2<>(E5, result)));
});
AsyncFunction

42. 42
Emitter
P2P3 P1P4
AsyncWaitOperator
E5
P5
asyncInvoke(E5, P5)P5
Emitter:
• separate thread
• polls queue for completed
promises (blocking)
• emits elements downstream
AsyncFunction

43. 43
DataStream<Tuple2<String, String>> result =
AsyncDataStream.(un)orderedWait(stream,
new MyAsyncFunction(),
1000, TimeUnit.MILLISECONDS,
100);
 our asyncFunction
 a timeout: max time until considered failed
 capacity: max number of in-flight requests
AsyncFunction

44. 44
DataStream<Tuple2<String, String>> result =
AsyncDataStream.(un)orderedWait(stream,
new MyAsyncFunction(),
1000, TimeUnit.MILLISECONDS,
100);
AsyncFunction

45. 45
DataStream<Tuple2<String, String>> result =
AsyncDataStream.(un)orderedWait(stream,
new MyAsyncFunction(),
1000, TimeUnit.MILLISECONDS,
100);
P2P3 P1P4E2E3 E1E4
Ideally... Emitter
AsyncFunction

46. 46
DataStream<Tuple2<String, String>> result =
AsyncDataStream.unorderedWait(stream,
new MyAsyncFunction(),
1000, TimeUnit.MILLISECONDS,
100);
P2P3 P1P4E2E3 E1E4
Reallistically... Emitter
...output ordered based on which request finished first
AsyncFunction

47. 47
P2P3 P1P4E2E3 E1E4
Emitter
 unorderedWait: emit results in order of completion
 orderedWait: emit results in order of arrival
 Always: watermarks never overpass elements and vice versa
AsyncFunction

48. Documentation
 ProcessFunction:
https://ci.apache.org/projects/flink/flink-docs-release-
1.2/dev/stream/process_function.html
https://ci.apache.org/projects/flink/flink-docs-release-
1.3/dev/stream/process_function.html
 AsyncIO:
https://ci.apache.org/projects/flink/flink-docs-release-1.2/dev/stream/asyncio.html
48

49. 4
Thank you!
@KLOUBEN_K
@ApacheFlink
@dataArtisans

50. 50
Stream Processing
and Apache Flink®'s
approach to it
@StephanEwen
Apache Flink PMC
CTO @ data ArtisansFLINKFORWARD IS COMING BACKTO BERLIN
SEPTEMBER11-13, 2017
BERLIN.FLINK-FORWARD.ORG -

51. We are hiring!
data-artisans.com/careers