Support AWS Kinesis Data Streams as a Source

data-prepper-plugins/kinesis-source/build.gradle

@@ -0,0 +1,42 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+plugins {
+    id 'java'
+}
+
+dependencies {
+    implementation project(':data-prepper-api')
+    implementation project(':data-prepper-plugins:common')
+    implementation project(path: ':data-prepper-plugins:buffer-common')
+    implementation project(path: ':data-prepper-plugins:aws-plugin-api')
+    implementation 'com.fasterxml.jackson.core:jackson-core'
+    implementation 'io.micrometer:micrometer-core'
+    implementation 'software.amazon.kinesis:amazon-kinesis-client:2.6.0'
+    compileOnly 'org.projectlombok:lombok:1.18.20'
+    annotationProcessor 'org.projectlombok:lombok:1.18.20'
+
+    testImplementation 'com.fasterxml.jackson.dataformat:jackson-dataformat-yaml'
+    testImplementation 'com.fasterxml.jackson.datatype:jackson-datatype-jsr310'
+    testImplementation project(':data-prepper-test-common')
+    testImplementation project(':data-prepper-test-event')
+    testImplementation project(':data-prepper-core')
+    testImplementation project(':data-prepper-plugin-framework')
+    testImplementation project(':data-prepper-pipeline-parser')
+    testImplementation 'com.fasterxml.jackson.dataformat:jackson-dataformat-yaml'
+    testImplementation project(':data-prepper-plugins:parse-json-processor')
+    testImplementation project(':data-prepper-plugins:newline-codecs')
+}
+
+jacocoTestCoverageVerification {
+    dependsOn jacocoTestReport
+    violationRules {
+        rule { //in addition to core projects rule
+            limit {
+                minimum = 1.0
+            }
+        }
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/extension/KinesisLeaseConfig.java

@@ -0,0 +1,20 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.extension;
+
+import com.fasterxml.jackson.annotation.JsonProperty;
+import lombok.Getter;
+
+@Getter
+public class KinesisLeaseConfig {
+    @JsonProperty("lease_coordination")
+    private KinesisLeaseCoordinationTableConfig leaseCoordinationTable;
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/extension/KinesisLeaseConfigExtension.java

@@ -0,0 +1,31 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.extension;
+
+import org.opensearch.dataprepper.model.annotations.DataPrepperExtensionPlugin;
+import org.opensearch.dataprepper.model.annotations.DataPrepperPluginConstructor;
+import org.opensearch.dataprepper.model.plugin.ExtensionPlugin;
+import org.opensearch.dataprepper.model.plugin.ExtensionPoints;
+
+@DataPrepperExtensionPlugin(modelType = KinesisLeaseConfig.class, rootKeyJsonPath = "/kinesis", allowInPipelineConfigurations = true)
+public class KinesisLeaseConfigExtension implements ExtensionPlugin {
+
+    private KinesisLeaseConfigSupplier kinesisLeaseConfigSupplier;
+    @DataPrepperPluginConstructor
+    public KinesisLeaseConfigExtension(final KinesisLeaseConfig kinesisLeaseConfig) {
+        this.kinesisLeaseConfigSupplier = new KinesisLeaseConfigSupplier(kinesisLeaseConfig);
+    }
+
+    @Override
+    public void apply(final ExtensionPoints extensionPoints) {
+        extensionPoints.addExtensionProvider(new KinesisLeaseConfigProvider(this.kinesisLeaseConfigSupplier));
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/extension/KinesisLeaseConfigProvider.java

@@ -0,0 +1,33 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.extension;
+
+import org.opensearch.dataprepper.model.plugin.ExtensionProvider;
+
+import java.util.Optional;
+
+class KinesisLeaseConfigProvider implements ExtensionProvider<KinesisLeaseConfigSupplier> {
+    private final KinesisLeaseConfigSupplier kinesisLeaseConfigSupplier;
+
+    public KinesisLeaseConfigProvider(final KinesisLeaseConfigSupplier kinesisLeaseConfigSupplier) {
+        this.kinesisLeaseConfigSupplier = kinesisLeaseConfigSupplier;
+    }
+
+    @Override
+    public Optional<KinesisLeaseConfigSupplier> provideInstance(Context context) {
+        return Optional.of(this.kinesisLeaseConfigSupplier);
+    }
+
+    @Override
+    public Class<KinesisLeaseConfigSupplier> supportedClass() {
+        return KinesisLeaseConfigSupplier.class;
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/extension/KinesisLeaseConfigSupplier.java

@@ -0,0 +1,26 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.extension;
+
+import java.util.Optional;
+
+public class KinesisLeaseConfigSupplier {
+
+    private KinesisLeaseConfig kinesisLeaseConfig;
+
+    public KinesisLeaseConfigSupplier(final KinesisLeaseConfig kinesisLeaseConfig) {
+        this.kinesisLeaseConfig = kinesisLeaseConfig;
+    }
+
+    public Optional<KinesisLeaseConfig> getKinesisExtensionLeaseConfig() {
+        return Optional.ofNullable(kinesisLeaseConfig);
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/extension/KinesisLeaseCoordinationTableConfig.java

@@ -0,0 +1,32 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.extension;
+
+import com.fasterxml.jackson.annotation.JsonProperty;
+import lombok.Getter;
+import lombok.NonNull;
+import software.amazon.awssdk.regions.Region;
+
+@Getter
+public class KinesisLeaseCoordinationTableConfig {
+
+    @JsonProperty("table_name")
+    @NonNull
+    private String tableName;
+
+    @JsonProperty("region")
+    @NonNull
+    private String region;
+
+    public Region getAwsRegion() {
+        return Region.of(region);
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/source/HostNameWorkerIdentifierGenerator.java

@@ -0,0 +1,39 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.source;
+
+import java.net.InetAddress;
+import java.net.UnknownHostException;
+
+/**
+ * Generate a unique ID to represent a consumer application instance.
+ */
+public class HostNameWorkerIdentifierGenerator implements WorkerIdentifierGenerator {
+
+    private static final String hostName;
+
+    static {
+        try {
+            hostName = InetAddress.getLocalHost().getHostName();
+        } catch (final UnknownHostException e) {
+            throw new RuntimeException(e);
+        }
+    }
+
+
+    /**
+     * @return Default to use host name.
+     */
+    @Override
+    public String generate() {
+        return hostName;
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/source/KinesisClientFactory.java

@@ -0,0 +1,61 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.source;
+
+import org.opensearch.dataprepper.aws.api.AwsCredentialsOptions;
+import org.opensearch.dataprepper.aws.api.AwsCredentialsSupplier;
+import org.opensearch.dataprepper.plugins.kinesis.source.configuration.AwsAuthenticationConfig;
+import software.amazon.awssdk.auth.credentials.AwsCredentialsProvider;
+import software.amazon.awssdk.regions.Region;
+import software.amazon.awssdk.services.cloudwatch.CloudWatchAsyncClient;
+import software.amazon.awssdk.services.dynamodb.DynamoDbAsyncClient;
+import software.amazon.awssdk.services.kinesis.KinesisAsyncClient;
+import software.amazon.kinesis.common.KinesisClientUtil;
+
+public class KinesisClientFactory {
+    private final AwsCredentialsProvider awsCredentialsProvider;
+    private final AwsCredentialsProvider defaultCredentialsProvider;
+    private final AwsAuthenticationConfig awsAuthenticationConfig;
+
+    public KinesisClientFactory(final AwsCredentialsSupplier awsCredentialsSupplier,
+                                final AwsAuthenticationConfig awsAuthenticationConfig) {
+        awsCredentialsProvider = awsCredentialsSupplier.getProvider(AwsCredentialsOptions.builder()
+                .withRegion(awsAuthenticationConfig.getAwsRegion())
+                .withStsRoleArn(awsAuthenticationConfig.getAwsStsRoleArn())
+                .withStsExternalId(awsAuthenticationConfig.getAwsStsExternalId())
+                .withStsHeaderOverrides(awsAuthenticationConfig.getAwsStsHeaderOverrides())
+                .build());
+        defaultCredentialsProvider = awsCredentialsSupplier.getProvider(AwsCredentialsOptions.defaultOptions());
+        this.awsAuthenticationConfig = awsAuthenticationConfig;
+    }
+
+    public DynamoDbAsyncClient buildDynamoDBClient(Region region) {
+        return DynamoDbAsyncClient.builder()
+                .credentialsProvider(defaultCredentialsProvider)
+                .region(region)
+                .build();
+    }
+
+    public KinesisAsyncClient buildKinesisAsyncClient(Region region) {
+        return KinesisClientUtil.createKinesisAsyncClient(
+                KinesisAsyncClient.builder()
+                    .credentialsProvider(awsCredentialsProvider)
+                        .region(region)
+        );
+    }
+
+    public CloudWatchAsyncClient buildCloudWatchAsyncClient(Region region) {
+        return CloudWatchAsyncClient.builder()
+                .credentialsProvider(defaultCredentialsProvider)
+                .region(region)
+                .build();
+    }
+}

data-prepper-plugins/kinesis-source/src/main/java/org/opensearch/dataprepper/plugins/kinesis/source/KinesisMultiStreamTracker.java

@@ -0,0 +1,88 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * The OpenSearch Contributors require contributions made to
+ * this file be licensed under the Apache-2.0 license or a
+ * compatible open source license.
+ *
+ */
+
+package org.opensearch.dataprepper.plugins.kinesis.source;
+
+import com.amazonaws.arn.Arn;
+import org.opensearch.dataprepper.plugins.kinesis.source.configuration.KinesisSourceConfig;
+import org.opensearch.dataprepper.plugins.kinesis.source.configuration.KinesisStreamConfig;
+import software.amazon.awssdk.services.kinesis.KinesisAsyncClient;
+import software.amazon.awssdk.services.kinesis.model.DescribeStreamRequest;
+import software.amazon.awssdk.services.kinesis.model.DescribeStreamResponse;
+import software.amazon.awssdk.services.kinesis.model.StreamDescription;
+import software.amazon.kinesis.common.InitialPositionInStreamExtended;
+import software.amazon.kinesis.common.StreamConfig;
+import software.amazon.kinesis.common.StreamIdentifier;
+import software.amazon.kinesis.processor.FormerStreamsLeasesDeletionStrategy;
+import software.amazon.kinesis.processor.MultiStreamTracker;
+
+import java.time.Duration;
+import java.util.ArrayList;
+import java.util.List;
+
+
+public class KinesisMultiStreamTracker implements MultiStreamTracker {
+    private static final String COLON = ":";
+
+    private final KinesisAsyncClient kinesisClient;
+    private final KinesisSourceConfig sourceConfig;
+    private final String applicationName;
+
+    public KinesisMultiStreamTracker(KinesisAsyncClient kinesisClient, final KinesisSourceConfig sourceConfig, final String applicationName) {
+        this.kinesisClient = kinesisClient;
+        this.sourceConfig = sourceConfig;
+        this.applicationName = applicationName;
+    }
+
+    @Override
+    public List<StreamConfig> streamConfigList()  {
+        List<StreamConfig> streamConfigList = new ArrayList<>();
+        for (KinesisStreamConfig kinesisStreamConfig : sourceConfig.getStreams()) {
+            StreamConfig streamConfig = getStreamConfig(kinesisStreamConfig);
+            streamConfigList.add(streamConfig);
+        }
+        return streamConfigList;
+    }
+
+    private StreamConfig getStreamConfig(KinesisStreamConfig kinesisStreamConfig) {
+        StreamIdentifier sourceStreamIdentifier = getStreamIdentifier(kinesisStreamConfig);
+        return new StreamConfig(sourceStreamIdentifier,
+                InitialPositionInStreamExtended.newInitialPosition(kinesisStreamConfig.getInitialPosition()));
+    }
+
+    private StreamIdentifier getStreamIdentifier(KinesisStreamConfig kinesisStreamConfig) {
+        DescribeStreamRequest describeStreamRequest = DescribeStreamRequest.builder()
+                .streamName(kinesisStreamConfig.getName())
+                .build();
+        DescribeStreamResponse describeStreamResponse = kinesisClient.describeStream(describeStreamRequest).join();
+        String streamIdentifierString = getStreamIdentifierString(describeStreamResponse.streamDescription());
+        return StreamIdentifier.multiStreamInstance(streamIdentifierString);
+    }
+
+    private String getStreamIdentifierString(StreamDescription streamDescription) {
+        String accountId = Arn.fromString(streamDescription.streamARN()).getAccountId();
+        long creationEpochSecond = streamDescription.streamCreationTimestamp().getEpochSecond();
+        return String.join(COLON, accountId, streamDescription.streamName(), String.valueOf(creationEpochSecond));
+    }
+
+    /**
+     * Setting the deletion policy as autodetect and release shard lease with a wait time of 10 sec
+     */
+    @Override
+    public FormerStreamsLeasesDeletionStrategy formerStreamsLeasesDeletionStrategy() {
+        return new FormerStreamsLeasesDeletionStrategy.AutoDetectionAndDeferredDeletionStrategy() {
+            @Override
+            public Duration waitPeriodToDeleteFormerStreams() {
+                return Duration.ofSeconds(10);
+            }
+        };
+
+    }
+}