You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 48 Next »

Status

Current state: [ UNDER DISCUSSION ]

Discussion thread<link to mailing list DISCUSS thread>

JIRA Unable to render Jira issues macro, execution error.

Released: 

Problem:

Samza today provides various api's like high level (fluent/application) for chaining complex jobs, low level (task) with configuration like Samza-Yarn, Samza as a library (Samza standalone) and in coming future Samza will support In-memory system for in memory streams that a user can produce/conusme from. Still Samza users have to get their hands dirty with too much configuration details and dependency on external system(like kafka) to write integration tests. In order to alleviate this we intend to provide a fluent test api to users. This will help in setting-up quick tests against their system without depending on underlying incoming stream from another system e.g Kafka Topic.

The goal of the proposed SEP we will provide a standardized framework to set-up continuous integration test for High Level Api, Low Level Api with Single Container configuration & Multi Container Configuration (Samza standalone using Zookeeper) with ease. Users won’t have to worry about the underlying Samza config details or even understand implementation details. This abstraction will increase developer productivity and help client's make their Samza system robust.

 Targeted audience for this SEP are two sets of users

  • Samza Job Owners: To make their Samza jobs robust
  • Samza Dev Teams: To enable dev teams make their samza infra robust

Motivation

Addition of this Continuous Integration Test Framework will alleviate:

  1. Lack of a standardized way to set up integration tests for users

  2. Lack of brevity in code and much detailed understanding of low level configs to set up samza systems

  3. Lack of a pluggable Test System with any Samza configuration (single & multi-container)

  4. Comprehensive set of tests for Samza api

Assumptions 

 

  1. System depends on In Memory system to spin up in-memory system consumers and producers for in memory data streams

  2. Single Container mode runs on SingleContainerGrouperFactory

  3. Multi Container system leverages Samza as a library (Samza standalone) using Zookeeper as coordination service for setting up test

  4. In order to provide a testing api with least config management, basic configs are set up for users, but flexibility to add/override some custom config is provided

  5. Testing is always supposed to be done using bounded streams using EndOfStreamMessage

  6. System should directly consume from File stream or Local Kafka stream without any additional configs, File Stream implementation is out of scope for this SEP, although this SEP suggests what the design for file stream should look like

  7. Bounded message streams from Kafka (either using Latch or EndOfStreamMessage) is beyond the scope of this SEP

  8. For Stateful testing system may use RocksDb as it ships with Samza or maintain the state in memory

Design

We propose a three step process for configuration of integration test for your samza job. Three facets are data injection, transformation and validation.

  • Data Injection phase means configuring the input source for Samza job.
  • Data Transformation phase means api logic of samza job(low-level/high-level). 
  • Data Validation phase asserts the expected results to the actual results computed after running a Samza job.

 

Flow New

Data Injection:

In the context of this Test framework an input stream means a stream that Samza job may consume from and an output stream means a stream that Samza job can produce to. The initial source of data stream is always supposed to be bounded since this is a test, and that may originate as one of the following sources listed below.

  • In Memory Data Stream
    Users have ability to produce to & consume from in memory system partitions after SEP-8. In Memory Data system alleviates the need of serialization/deserialization of data since it does not persist data. We take advantage of this and provide very succinct Stream classes to serve as input data sources, which are the following:
    • Collection Stream
      Users can plug a collection (either List or Map) to create an in-memory input stream
      e.g:  CollectionStream.of(...,{1,2,3,4})
    • Event Builder Stream
      Event builder helps a user to mimic runtime samza processing environment in its tests, for example adding an exception in the steam, advancing time for window functions.
  • File Stream
    Users can create an input stream reading and writing from local file
    e.g:  FileStream.of("/path/to/file")
  • Local Kafka Stream
    Users can also consume bounded streams from a kafka topic which serves as initial input. Samza already provide api's to consume from and produce to kafka. For kafka streams we will need serde configurations, however implementation of bounded streams in kafka (using Latche Message or EndOfStreamMessage) is beyond scope of this SEP

For in memory streams the api actually initializes the stream by spinning up a Samza producer using an InMemorySystemProducer to write the stream, this is how a collection of data or events gets initialized as a steam. It also configures any output stream if the user has added any.

Data Transformation:

This is the Samza job you write, this can be either done using Low Level Api or the fluent High Level Api. Test frameworks provides api to set up test for both the cases. Test framework supports both the apis with Single container and Multi-container mode. Users implement StreamTask/Async Stream/Stream Application in the same way, as they do for their Samza job (or pass an instance of Samza job).

Data Validation: 

For the low level api & high level api once user runs the job, users can assert data from any streams the job produces to. StreamAssert spins up a consumer for the system under test, gets the messages in the stream and compares this result with the expected value. Various assertion functions provided are:

 

StreamAssertStateAssert
containscontains
containsInAnyOrdersatisfies
inWindowPane 
satifies 

 

Data Types & Partitions:

The framework will provide complete flexibility for usage of primitive and derived data types. Serdes are required for local kafka stream and file stream but in memory streams dont require any Serde configuration. Serdes will be also required if users want to maintain State. Test framework will provide api's for initialization of input streams (both single and multi-partition), and also data validation on single partition and multi-partition of the bounded streams 

 

Running Config

Traditionally we ask users to set up config for any samza job, for test purposes we set up basic config boiler plate for users and provide them a flexible option to still add any custom config (rarely needed), api exposes functions to configure single container or multi container mode (using Zookeeper). It also provides functions to configure concurrency semantics for their Samza job. 

 

Stateful & Stateless Testing:

Test framework will support stateful and stateless testing, satateful testing can be either done using RocksDB or maintaining state in memory

Future Changes with Stream Descriptors:

The test framework is designed in a way which asks users to do none or minimal Samza configs, in future we intend to use StreamDescriptors in the test framework to do Samza configs. With StreamDescriptor and High Level api, another wrapper can be provided over the job to run the configuration in a test environment

Test Matrix (Plan)


Samza APIConcurrencyPartitionsContainerExpected Result
StreamTask
task.max.concurrency = 1
job.container.thread.pool.size = n
1 <= p <= n1 / nin order processing
   
task.max.concurrency > 1
job.container.thread.pool.size = n
1 <= p <= n1 / nout of order processing
   
AsyncStreamTask
task.max.concurrency = 11 <= p <= n1 / nin order processing
task.max.concurrency = n1 <= p <= n1 / nout of order processing
Windowable TaskN/A1 <= p <= n1 / nexpecting processing n messages for messages window of time t
Initiable TaskN/A1 <= p <= n1 / nstateful testing (assertions on kv store)
Closable TaskN/A1 <= p <= n1 / n?

Map / Flatmap / Filter /

Partition By / Merge /

SendTo / Sink / Join /

Window

 

task.max.concurrency = 1

job.container.thread.pool.size = n

1 <= p < n 1 / nin order processing

task.max.concurrency > 1

job.container.thread.pool.size = n

out of order processing
Metadata StreamsTests
Changelog StreamAssertions on Changelog stream in Stateful tests
Checkpoint StreamAsserts on Checkpoint Stream (Kafka or InMemory)
Coordinator StreamAsserts on Coordinator stream (Kafka or InMemory)



Public Interfaces

 

TestRunner
/**
* TestRunner provides static factory for quick setup of Samza environment for testing Low level api and High level api , users can configure input streams 
* they consume from and output streams they produce to, users pass in their Samza job api logic via StreamTask/AsyncStreamTask/StreamApplication
*/ 

public class TestRunner {
  
  // Static Factory to config & create runner for low level api 
  public static TestRunner of(StreamTask task) {...}
  public static TestRunner of(AsyncStreamTask task) {...}
  public static TestRunner of(StreamApplication app) {...}
  
  // Add/Ovveride any custom configs
  public TestRunner addOverrideConfigs(Map<String,String> configs) {...}
  public TestRunner addOverrideConfigs(String configUri) {...}
  public TestRunner addOverrideConfig(String key, String val){...}
 
  // Set container mode either single container or multi container
  public TestRunner setContainerMode(Mode mode) {...}
  
  // Multithreading configs for users
  public TestRunner setTaskMaxConcurrency(Integer value) {...}
  public TestRunner setTaskCallBackTimeoutMS(Integer value) {...}
  public TestRunner setJobContainerThreadPoolSize(Integer value) {...}
 
  // Configure state for application
  public TestRunner addState(String storeName) {...}  
  
  // Configure an input stream for samza system, that job can consume from
  public TestRunner addInputStream(CollectionStream stream) {...}
  // Configures an output stream for samza system, that job can producer to
  public TestRunner addOutputStream(CollectionStream stream) {...}
  
  // Run the app
  public void run() {...}
}

 

 

Types of Input Streams

 

CollectionStream
/**
* CollectionStream provides utilities to build an in memory input stream of collections(list, map). It also supports initialization of 
* multiple partitions for an input stream 
*/
 
public class CollectionStream<T> { 
  // Create an empty stream that a Samza job can produce to
  public static <T> CollectionStream<T> empty(String steamId) {...}
  
  // Create a stream of messages from input list with single partition
  public static <T> CollectionStream<T> of(String steamId, Iterable<T> collection){...}
  
  // Create a stream of messages from input list with multiple partition, key of partitions map is partitionId
  public static <T> CollectionStream<T> of(String steamId, Map<Integer,Iterable<T>> partitions){...}
  
}
EventStream
/**
* EventStream provides utilities to build an in memory input stream of events. It helps mimic run time environment of your job, 
* advancing time for windowing functions  
*/
 
public class EventStream<T> {
  public static abstract class Builder<T> {
    public abstract Builder addElement();
    public abstract Builder addException();
    public abstract Builder advanceTimeTo(long time);
    public abstract EventStream<T> build(); 
 }

 public static <T> Builder<T> builder() {...}
}
FileStream
/**
* FileStream provides utilities to build a stream of messages from a file on disk
*/
 
public class FileStream<T> {
  public static <T> FileStream<T> of(String fileUri) {...}
}


Examples Usages of Test Api:


Simple StreamTask Test
/**
* Simple Test case using a collection as an input for a low level application. It demonstrates set up and comparison of a test with 
* minimal(eg. here none) configs and it reads an input stream of integers and multiplies each integer with 10
*/
 
// Create a StreamTask
StreamTask myTask = new StreamTask() {
  @Override
  public void process(IncomingMessageEnvelope envelope, MessageCollector collector, TaskCoordinator coordinator) throws Exception {
    Integer obj = (Integer) envelope.getMessage();
    collector.send(new OutgoingMessageEnvelope(new SystemStream("test-samza","output"), obj*10));
  }
};
 
// Initialize and run the test framework
TestRunner
    .of(myTask)
    .setJobContainerThreadPoolSize(4)
    .addInputStream(CollectionStream.of("test-samza.input", {1,2,3,4}))
    .addOutputStream(CollectionStream.empty("test-samza.output"))
    .run();
 
// Assertions on the outputs
StreamAssert.that("test-samza.output").contains({10,30,20,40});
Simple StreamTask with StreamDescriptors
/**
* Simple Test case using a collection as an input for a low level application. It demonstrates set up and comparison of a test with 
* minimal(eg. here none) configs and it reads an input stream of integers and multiplies each integer with 10
*/
 
class MyStreamTaskFactory implements StreamTaskFactory {
    @Override
    public StreamTask createInstance() {
      return (envelope, collector, coordinator) -> {
        return;
      };
    }
  }
}

StreamDescriptor.Input input = StreamDescriptor.<String, String>input("input")
    .withKeySerde(new StringSerde("UTF-8"))
    .withMsgSerde(new StringSerde("UTF-8"));

StreamDescriptor.Output output = StreamDescriptor.<String, String>output("output")
    .withKeySerde(new StringSerde("UTF-8"))
    .withMsgSerde(new StringSerde("UTF-8"));

InMemoryTestSystem testingSystem = InMemoryTestSystem.create("samza-test")
    .addInput(input, {1,2,3,4,5})
    .addOutput("output"/);

StreamTaskApplication app = StreamApplications.createStreamTaskApp(config, new MyStreamTaskFactory());


app
  .addInputs(input)
  .addOutputs(output)
  .runAsTest();
 
// Assertions on the outputs
Assert.assertThat(testingSystem.getStreamState(output), IsIterableContainingInOrder.contains({10,20,30,40,50})));



Simple AsyncStreamTask Test
/**
* Simple Test case using a collection as an input for a low level application in the async mode
*/

// 
public class AsyncRestTask implements AsyncStreamTask, InitableTask, ClosableTask {
  private Client client;
  private WebTarget target; 

  @Override
  public void init(Config config, TaskContext taskContext) throws Exception {
    // Your initialization of web client code goes here
    client = ClientBuilder.newClient();
    target = client.target("http://example.com/resource/").path("hello");
  }

  @Override
  public void processAsync(IncomingMessageEnvelope envelope, MessageCollector collector,
      TaskCoordinator coordinator, final TaskCallback callback) {
    target.request().async().get(new InvocationCallback<Response>() {
      @Override
      public void completed(Response response) {
        Integer obj = (Integer) envelope.getMessage();
        collector.send(new OutgoingMessageEnvelope(new SystemStream("test-samza","output"), obj*10));
        callback.complete();
      }

      @Override
      public void failed(Throwable throwable) {
        System.out.println("Invocation failed.");
        callback.failure(throwable);
      }
    });
  }

  @Override
  public void close() throws Exception {
    client.close();
  }
}
 

// Initialize and run the test framework
TestRunner
    .of(new AsyncRestTask())
	.setTaskCallBackTimeoutMS(200)
	.setTaskMaxConcurrency(4)
	.addInputStream(CollectionStream.of("test-samza.Input", {1,2,3,4}))
	.addOutputStream(CollectionStream.empty("test-samza.output"))
	.run();
 
// Assertions on the outputs
StreamAssert.that("test-samza.output").containsInAnyOrder({10,20,30,40});
Simple High Level Api Test
/**
* Simple Test case using a collection as an input for a High level application
*/
 
public class MyStreamApplication implements StreamApplication {
  @Override
  public void init(StreamGraph graph, Config config) {
    MessageStream<Integer> pageViews = graph.getInputStream(“test-samza.page-views”);

    pageViews.map(s -> "processed " + s)
             .sendTo(graph.getOutputStream(“test-samza.output”));
 }
}


 
// Initialize and run the test framework
TestRunner
	.of(new MyStreamApplication());
    .addInputStream(CollectionStream.of("test-samza.input", {1,2,3}))
    .addOutputStream(CollectionStream.empty("test-samza.output"))
    .run();
 
// Assertions on the outputs
StreamAssert.that("test-samza.Output").contains({"processed 1", "processed 2", "processed 4"});

Implementation and Test Plan

  • Introduce the new interface and api for test framework
  • Use the same api to write comprehensive integration test for: 
    • Samza Low level Async Api (Single and Multi-Container Mode)
    • Samza Low level Synchronous Api (Single and Multi-Container Mode)
    • Samza High Level Api (Single and Multi-Container Mode)

Compatibility, Deprecation, and Migration Plan

As this is a new feature, no plans are required for compatibility, deprecation and migration.

Rejected Alternatives

 

 

  • No labels