| <!DOCTYPE html> |
| <!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]--> |
| <!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]--> |
| <head> |
| <meta charset="utf-8"> |
| <meta http-equiv="X-UA-Compatible" content="IE=edge"> |
| <meta name="viewport" content="width=device-width, initial-scale=1.0"> |
| |
| |
| |
| <title>Kafka Input - Apache Apex Malhar Documentation</title> |
| |
| |
| <link rel="shortcut icon" href="../../favicon.ico"> |
| |
| |
| |
| <link href='https://fonts.googleapis.com/css?family=Lato:400,700|Roboto+Slab:400,700|Inconsolata:400,700' rel='stylesheet' type='text/css'> |
| |
| <link rel="stylesheet" href="../../css/theme.css" type="text/css" /> |
| <link rel="stylesheet" href="../../css/theme_extra.css" type="text/css" /> |
| <link rel="stylesheet" href="../../css/highlight.css"> |
| |
| |
| <script> |
| // Current page data |
| var mkdocs_page_name = "Kafka Input"; |
| var mkdocs_page_input_path = "operators/kafkaInputOperator.md"; |
| var mkdocs_page_url = "/operators/kafkaInputOperator/"; |
| </script> |
| |
| <script src="../../js/jquery-2.1.1.min.js"></script> |
| <script src="../../js/modernizr-2.8.3.min.js"></script> |
| <script type="text/javascript" src="../../js/highlight.pack.js"></script> |
| <script src="../../js/theme.js"></script> |
| |
| |
| </head> |
| |
| <body class="wy-body-for-nav" role="document"> |
| |
| <div class="wy-grid-for-nav"> |
| |
| |
| <nav data-toggle="wy-nav-shift" class="wy-nav-side stickynav"> |
| <div class="wy-side-nav-search"> |
| <a href="../.." class="icon icon-home"> Apache Apex Malhar Documentation</a> |
| <div role="search"> |
| <form id ="rtd-search-form" class="wy-form" action="../../search.html" method="get"> |
| <input type="text" name="q" placeholder="Search docs" /> |
| </form> |
| </div> |
| </div> |
| |
| <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation"> |
| <ul class="current"> |
| |
| <li> |
| <li class="toctree-l1 "> |
| <a class="" href="../..">Apache Apex Malhar</a> |
| |
| </li> |
| <li> |
| |
| <li> |
| <ul class="subnav"> |
| <li><span>Operators</span></li> |
| |
| |
| |
| <li class="toctree-l1 current"> |
| <a class="current" href="./">Kafka Input</a> |
| |
| <ul> |
| |
| <li class="toctree-l3"><a href="#kafka-input-operator">KAFKA INPUT OPERATOR</a></li> |
| |
| <li><a class="toctree-l4" href="#introduction-about-kafka-input-operator">Introduction: About Kafka Input Operator</a></li> |
| |
| <li><a class="toctree-l4" href="#why-is-it-needed">Why is it needed ?</a></li> |
| |
| <li><a class="toctree-l4" href="#abstractkafkainputoperator">AbstractKafkaInputOperator</a></li> |
| |
| <li><a class="toctree-l4" href="#kafkaconsumer">KafkaConsumer</a></li> |
| |
| <li><a class="toctree-l4" href="#pre-requisites">Pre-requisites</a></li> |
| |
| <li><a class="toctree-l4" href="#offsetmanager">OffsetManager</a></li> |
| |
| <li><a class="toctree-l4" href="#partitioning">Partitioning</a></li> |
| |
| <li><a class="toctree-l4" href="#abstractsingleportkafkainputoperator">AbstractSinglePortKafkaInputOperator</a></li> |
| |
| <li><a class="toctree-l4" href="#concrete-classes">Concrete Classes</a></li> |
| |
| <li><a class="toctree-l4" href="#application-example">Application Example</a></li> |
| |
| |
| </ul> |
| |
| </li> |
| |
| |
| |
| <li class="toctree-l1 "> |
| <a class="" href="../file_splitter/">File Splitter</a> |
| |
| </li> |
| |
| |
| |
| <li class="toctree-l1 "> |
| <a class="" href="../block_reader/">Block Reader</a> |
| |
| </li> |
| |
| |
| |
| <li class="toctree-l1 "> |
| <a class="" href="../file_output/">File Output</a> |
| |
| </li> |
| |
| |
| </ul> |
| <li> |
| |
| </ul> |
| </div> |
| |
| </nav> |
| |
| <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"> |
| |
| |
| <nav class="wy-nav-top" role="navigation" aria-label="top navigation"> |
| <i data-toggle="wy-nav-top" class="fa fa-bars"></i> |
| <a href="../..">Apache Apex Malhar Documentation</a> |
| </nav> |
| |
| |
| <div class="wy-nav-content"> |
| <div class="rst-content"> |
| <div role="navigation" aria-label="breadcrumbs navigation"> |
| <ul class="wy-breadcrumbs"> |
| <li><a href="../..">Docs</a> »</li> |
| |
| |
| |
| <li>Operators »</li> |
| |
| |
| |
| <li>Kafka Input</li> |
| <li class="wy-breadcrumbs-aside"> |
| |
| </li> |
| </ul> |
| <hr/> |
| </div> |
| <div role="main"> |
| <div class="section"> |
| |
| <h1 id="kafka-input-operator">KAFKA INPUT OPERATOR</h1> |
| <h3 id="introduction-about-kafka-input-operator">Introduction: About Kafka Input Operator</h3> |
| <p>This is an input operator that consumes data from Kafka messaging system for further processing in Apex. Kafka Input Operator is an fault-tolerant and scalable Malhar Operator.</p> |
| <h3 id="why-is-it-needed">Why is it needed ?</h3> |
| <p>Kafka is a pull-based and distributed publish subscribe messaging system, topics are partitioned and replicated across |
| nodes. Kafka input operator is needed when you want to read data from multiple |
| partitions of a Kafka topic in parallel in an Apex application.</p> |
| <h3 id="abstractkafkainputoperator">AbstractKafkaInputOperator</h3> |
| <p>This is the abstract implementation that serves as base class for consuming messages from Kafka messaging system. This class doesn’t have any ports.</p> |
| <p><img alt="AbstractKafkaInput.png" src="../images/kafkainput/image00.png" /></p> |
| <h4 id="configuration-parameters">Configuration Parameters</h4> |
| <p><table> |
| <col width="25%" /> |
| <col width="75%" /> |
| <tbody> |
| <tr class="odd"> |
| <td align="left"><p>Parameter</p></td> |
| <td align="left"><p>Description</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>maxTuplesPerWindow</p></td> |
| <td align="left"><p>Controls the maximum number of messages emitted in each streaming window from this operator. Minimum value is 1. Default value = MAX_VALUE </p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>idempotentStorageManager</p></td> |
| <td align="left"><p>This is an instance of IdempotentStorageManager. Idempotency ensures that the operator will process the same set of messages in a window before and after a failure. For example, let's say the operator completed window 10 and failed somewhere between window 11. If the operator gets restored at window 10 then it will process the same messages again in window 10 which it did in the previous run before the failure. Idempotency is important but comes with higher cost because at the end of each window the operator needs to persist some state with respect to that window. Default Value = com.datatorrent.lib.io.IdempotentStorageManager.<br>NoopIdempotentStorageManager</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>strategy</p></td> |
| <td align="left"><p>Operator supports two types of partitioning strategies, ONE_TO_ONE and ONE_TO_MANY.</p> |
| <p>ONE_TO_ONE: If this is enabled, the AppMaster creates one input operator instance per Kafka topic partition. So the number of Kafka topic partitions equals the number of operator instances.</p> |
| <p>ONE_TO_MANY: The AppMaster creates K = min(initialPartitionCount, N) Kafka input operator instances where N is the number of Kafka topic partitions. If K is less than N, the remaining topic partitions are assigned to the K operator instances in round-robin fashion. If K is less than initialPartitionCount, the AppMaster creates one input operator instance per Kafka topic partition. For example, if initialPartitionCount = 5 and number of Kafka partitions(N) = 2 then AppMaster creates 2 Kafka input operator instances. |
| Default Value = ONE_TO_ONE</p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>msgRateUpperBound</p></td> |
| <td align="left"><p>Maximum messages upper bound. Operator repartitions when the <em>msgProcessedPS</em> exceeds this bound. <em>msgProcessedPS</em> is the average number of messages processed per second by this operator.</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>byteRateUpperBound</p></td> |
| <td align="left"><p>Maximum bytes upper bound. Operator repartitions when the <em>bytesPS</em> exceeds this bound. <em>bytesPS</em> is the average number of bytes processed per second by this operator.</p> |
| <p></p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>offsetManager</p></td> |
| <td align="left"><p>This is an optional parameter that is useful when the application restarts or start at specific offsets (offsets are explained below)</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>repartitionInterval</p></td> |
| <td align="left"><p>Interval specified in milliseconds. This value specifies the minimum time required between two repartition actions. Default Value = 30 Seconds</p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>repartitionCheckInterval</p></td> |
| <td align="left"><p>Interval specified in milliseconds. This value specifies the minimum interval between two offset updates. Default Value = 5 Seconds</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>initialPartitionCount</p></td> |
| <td align="left"><p>When the ONE_TO_MANY partition strategy is enabled, this value indicates the number of Kafka input operator instances. Default Value = 1</p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>consumer</p></td> |
| <td align="left"><p>This is an instance of com.datatorrent.contrib.kafka.KafkaConsumer. Default Value = Instance of SimpleKafkaConsumer.</p></td> |
| </tr> |
| </tbody> |
| </table></p> |
| <h4 id="abstract-methods">Abstract Methods</h4> |
| <p>void emitTuple(Message message): Abstract method that emits tuples |
| extracted from Kafka message.</p> |
| <h3 id="kafkaconsumer">KafkaConsumer</h3> |
| <p>This is an abstract implementation of Kafka consumer. It sends the fetch |
| requests to the leading brokers of Kafka partitions. For each request, |
| it receives the set of messages and stores them into the buffer which is |
| ArrayBlockingQueue. SimpleKafkaConsumer which extends |
| KafkaConsumer and serves the functionality of Simple Consumer API and |
| HighLevelKafkaConsumer which extends KafkaConsumer and serves the |
| functionality of High Level Consumer API.</p> |
| <h3 id="pre-requisites">Pre-requisites</h3> |
| <p>This operator referred the Kafka Consumer API of version |
| 0.8.1.1. So, this operator will work with any 0.8.x and 0.7.x version of Apache Kafka.</p> |
| <h4 id="configuration-parameters_1">Configuration Parameters</h4> |
| <table> |
| <col width="15%" /> |
| <col width="15%" /> |
| <col width="15%" /> |
| <col width="55%" /> |
| <tbody> |
| <tr class="odd"> |
| <td align="left"><p>Parameter</p></td> |
| <td align="left"><p>Type</p></td> |
| <td align="left"><p>Default</p></td> |
| <td align="left"><p>Description</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>zookeeper</p></td> |
| <td align="left"><p>String</p></td> |
| <td align="left"><p></p></td> |
| <td align="left"><p>Specifies the zookeeper quorum of Kafka clusters that you want to consume messages from. zookeeper is a string in the form of hostname1:port1,hostname2:port2,hostname3:port3 where hostname1,hostname2,hostname3 are hosts and port1,port2,port3 are ports of zookeeper server. If the topic name is the same across the Kafka clusters and want to consume data from these clusters, then configure the zookeeper as follows: c1::hs1:p1,hs2:p2,hs3:p3;c2::hs4:p4,hs5:p5,c3::hs6:p6</p> |
| <p>where</p> |
| <p>c1,c2,c3 indicates the cluster names, hs1,hs2,hs3,hs4,hs5,hs6 are zookeeper hosts and p1,p2,p3,p4,p5,p6 are corresponding ports. Here, cluster name is optional in case of single cluster</p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>cacheSize</p></td> |
| <td align="left"><p>int</p></td> |
| <td align="left"><p>1024</p></td> |
| <td align="left"><p>Maximum of buffered messages hold in memory.</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>topic</p></td> |
| <td align="left"><p>String</p></td> |
| <td align="left"><p>default_topic</p></td> |
| <td align="left"><p>Indicates the name of the topic.</p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>initialOffset</p></td> |
| <td align="left"><p>String</p></td> |
| <td align="left"><p>latest</p></td> |
| <td align="left"><p>Indicates the type of offset i.e, “earliest or latest”. If initialOffset is “latest”, then the operator consumes messages from latest point of Kafka queue. If initialOffset is “earliest”, then the operator consumes messages starting from message queue. This can be overridden by OffsetManager.</p></td> |
| </tr> |
| </tbody> |
| </table> |
| |
| <h4 id="abstract-methods_1">Abstract Methods</h4> |
| <ol> |
| <li>void commitOffset(): Commit the offsets at checkpoint.</li> |
| <li>Map <KafkaPartition, Long> getCurrentOffsets(): Return the current |
| offset status.</li> |
| <li>resetPartitionsAndOffset(Set <KafkaPartition> partitionIds, |
| Map <KafkaPartition, Long> startOffset): Reset the partitions with |
| parittionIds and offsets with startOffset.</li> |
| </ol> |
| <h4 id="configuration-parameters-for-simplekafkaconsumer">Configuration Parameters for SimpleKafkaConsumer</h4> |
| <table> |
| <col width="25%" /> |
| <col width="15%" /> |
| <col width="15%" /> |
| <col width="45%" /> |
| <tbody> |
| <tr class="odd"> |
| <td align="left"><p>Parameter</p></td> |
| <td align="left"><p>Type</p></td> |
| <td align="left"><p>Default</p></td> |
| <td align="left"><p>Description</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>bufferSize</p></td> |
| <td align="left"><p>int</p></td> |
| <td align="left"><p>1 MB</p></td> |
| <td align="left"><p>Specifies the maximum total size of messages for each fetch request.</p></td> |
| </tr> |
| <tr class="odd"> |
| <td align="left"><p>metadataRefreshInterval</p></td> |
| <td align="left"><p>int</p></td> |
| <td align="left"><p>30 Seconds</p></td> |
| <td align="left"><p>Interval in between refresh the metadata change(broker change) in milliseconds. Enabling metadata refresh guarantees an automatic reconnect when a new broker is elected as the host. A value of -1 disables this feature.</p></td> |
| </tr> |
| <tr class="even"> |
| <td align="left"><p>metadataRefreshRetryLimit</p></td> |
| <td align="left"><p>int</p></td> |
| <td align="left"><p>-1</p></td> |
| <td align="left"><p>Specifies the maximum brokers' metadata refresh retry limit. -1 means unlimited retry.</p></td> |
| </tr> |
| </tbody> |
| </table> |
| |
| <h3 id="offsetmanager">OffsetManager</h3> |
| <p>This is an interface for offset management and is useful when consuming data |
| from specified offsets. Updates the offsets for all the Kafka partitions |
| periodically. Below is the code snippet: </p> |
| <pre><code class="java">public interface OffsetManager |
| { |
| public Map<KafkaPartition, Long> loadInitialOffsets(); |
| public void updateOffsets(Map<KafkaPartition, Long> offsetsOfPartitions); |
| } |
| </code></pre> |
| |
| <h4 id="abstract-methods_2">Abstract Methods</h4> |
| <p>Map <KafkaPartition, Long> loadInitialOffsets(): Specifies the initial offset for consuming messages; called at the activation stage.</p> |
| <p>updateOffsets(Map <KafkaPartition, Long> offsetsOfPartitions): This |
| method is called at every repartitionCheckInterval to update offsets.</p> |
| <h3 id="partitioning">Partitioning</h3> |
| <p>The logical instance of the KafkaInputOperator acts as the Partitioner |
| as well as a StatsListener. This is because the |
| AbstractKafkaInputOperator implements both the |
| com.datatorrent.api.Partitioner and com.datatorrent.api.StatsListener |
| interfaces and provides an implementation of definePartitions(...) and |
| processStats(...) which makes it auto-scalable.</p> |
| <h4 id="response-processstatsbatchedoperatorstats-stats">Response processStats(BatchedOperatorStats stats)</h4> |
| <p>The application master invokes this method on the logical instance with |
| the stats (tuplesProcessedPS, bytesPS, etc.) of each partition. |
| Re-partitioning happens based on whether any new Kafka partitions added for |
| the topic or bytesPS and msgPS cross their respective upper bounds.</p> |
| <h4 id="definepartitions">DefinePartitions</h4> |
| <p>Based on the repartitionRequired field of the Response object which is |
| returned by processStats(...) method, the application master invokes |
| definePartitions(...) on the logical instance which is also the |
| partitioner instance. Dynamic partition can be disabled by setting the |
| parameter repartitionInterval value to a negative value.</p> |
| <h3 id="abstractsingleportkafkainputoperator">AbstractSinglePortKafkaInputOperator</h3> |
| <p>This class extends AbstractKafkaInputOperator and having single output |
| port, will emit the messages through this port.</p> |
| <h4 id="ports">Ports</h4> |
| <p>outputPort <T>: Tuples extracted from Kafka messages are emitted through |
| this port.</p> |
| <h4 id="abstract-methods_3">Abstract Methods</h4> |
| <p>T getTuple(Message msg) : Converts the Kafka message to tuple.</p> |
| <h3 id="concrete-classes">Concrete Classes</h3> |
| <ol> |
| <li> |
| <p>KafkaSinglePortStringInputOperator : |
| This class extends AbstractSinglePortKafkaInputOperator and getTuple() method extracts string from Kafka message.</p> |
| </li> |
| <li> |
| <p>KafkaSinglePortByteArrayInputOperator: |
| This class extends AbstractSinglePortKafkaInputOperator and getTuple() method extracts byte array from Kafka message.</p> |
| </li> |
| </ol> |
| <h3 id="application-example">Application Example</h3> |
| <p>This section builds an Apex application using Kafka input operator. |
| Below is the code snippet:</p> |
| <pre><code class="java">@ApplicationAnnotation(name = "KafkaApp") |
| public class ExampleKafkaApplication implements StreamingApplication |
| { |
| @Override |
| public void populateDAG(DAG dag, Configuration entries) |
| { |
| KafkaSinglePortByteArrayInputOperator input = dag.addOperator("MessageReader", new KafkaSinglePortByteArrayInputOperator()); |
| |
| ConsoleOutputOperator output = dag.addOperator("Output", new ConsoleOutputOperator()); |
| |
| dag.addStream("MessageData", input.outputPort, output.input); |
| } |
| } |
| </code></pre> |
| |
| <p>Below is the configuration for “test” Kafka topic name and |
| “localhost:2181” is the zookeeper forum:</p> |
| <pre><code class="xml"><property> |
| <name>dt.operator.MessageReader.prop.topic</name> |
| <value>test</value> |
| </property> |
| |
| <property> |
| <name>dt.operator.KafkaInputOperator.prop.zookeeper</nam> |
| <value>localhost:2181</value> |
| </property> |
| </code></pre> |
| |
| </div> |
| </div> |
| <footer> |
| |
| <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation"> |
| |
| <a href="../file_splitter/" class="btn btn-neutral float-right" title="File Splitter">Next <span class="icon icon-circle-arrow-right"></span></a> |
| |
| |
| <a href="../.." class="btn btn-neutral" title="Apache Apex Malhar"><span class="icon icon-circle-arrow-left"></span> Previous</a> |
| |
| </div> |
| |
| |
| <hr/> |
| |
| <div role="contentinfo"> |
| <!-- Copyright etc --> |
| |
| </div> |
| |
| Built with <a href="http://www.mkdocs.org">MkDocs</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. |
| </footer> |
| |
| </div> |
| </div> |
| |
| </section> |
| |
| </div> |
| |
| <div class="rst-versions" role="note" style="cursor: pointer"> |
| <span class="rst-current-version" data-toggle="rst-current-version"> |
| |
| |
| <span><a href="../.." style="color: #fcfcfc;">« Previous</a></span> |
| |
| |
| <span style="margin-left: 15px"><a href="../file_splitter/" style="color: #fcfcfc">Next »</a></span> |
| |
| </span> |
| </div> |
| |
| </body> |
| </html> |
