awelless commented on code in PR #10077:
URL: https://github.com/apache/nifi/pull/10077#discussion_r2301286885


##########
nifi-extension-bundles/nifi-aws-bundle/nifi-aws-kinesis/src/main/java/org/apache/nifi/processors/aws/kinesis/ConsumeKinesis.java:
##########
@@ -0,0 +1,607 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.nifi.processors.aws.kinesis;
+
+import jakarta.annotation.Nullable;
+import org.apache.nifi.annotation.behavior.InputRequirement;
+import org.apache.nifi.annotation.behavior.SystemResource;
+import org.apache.nifi.annotation.behavior.SystemResourceConsideration;
+import org.apache.nifi.annotation.behavior.WritesAttribute;
+import org.apache.nifi.annotation.behavior.WritesAttributes;
+import org.apache.nifi.annotation.configuration.DefaultSettings;
+import org.apache.nifi.annotation.documentation.CapabilityDescription;
+import org.apache.nifi.annotation.documentation.Tags;
+import org.apache.nifi.annotation.lifecycle.OnScheduled;
+import org.apache.nifi.annotation.lifecycle.OnStopped;
+import org.apache.nifi.components.DescribedValue;
+import org.apache.nifi.components.PropertyDescriptor;
+import org.apache.nifi.flowfile.FlowFile;
+import org.apache.nifi.processor.AbstractProcessor;
+import org.apache.nifi.processor.DataUnit;
+import org.apache.nifi.processor.ProcessContext;
+import org.apache.nifi.processor.ProcessSession;
+import org.apache.nifi.processor.Relationship;
+import org.apache.nifi.processor.exception.ProcessException;
+import org.apache.nifi.processor.util.StandardValidators;
+import 
org.apache.nifi.processors.aws.credentials.provider.service.AWSCredentialsProviderService;
+import 
org.apache.nifi.processors.aws.kinesis.ReaderRecordProcessor.ProcessingResult;
+import org.apache.nifi.processors.aws.kinesis.RecordBuffer.ShardBufferId;
+import org.apache.nifi.processors.aws.kinesis.RecordBuffer.ShardBufferLease;
+import org.apache.nifi.processors.aws.region.RegionUtilV2;
+import org.apache.nifi.proxy.ProxyConfiguration;
+import org.apache.nifi.proxy.ProxyConfigurationService;
+import org.apache.nifi.proxy.ProxySpec;
+import org.apache.nifi.serialization.RecordReaderFactory;
+import org.apache.nifi.serialization.RecordSetWriterFactory;
+import software.amazon.awssdk.auth.credentials.AwsCredentialsProvider;
+import software.amazon.awssdk.http.Protocol;
+import software.amazon.awssdk.http.async.SdkAsyncHttpClient;
+import software.amazon.awssdk.http.nio.netty.Http2Configuration;
+import software.amazon.awssdk.http.nio.netty.NettyNioAsyncHttpClient;
+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.awssdk.services.kinesis.KinesisAsyncClientBuilder;
+import software.amazon.kinesis.common.ConfigsBuilder;
+import software.amazon.kinesis.common.InitialPositionInStream;
+import software.amazon.kinesis.common.InitialPositionInStreamExtended;
+import software.amazon.kinesis.coordinator.Scheduler;
+import software.amazon.kinesis.lifecycle.events.InitializationInput;
+import software.amazon.kinesis.lifecycle.events.LeaseLostInput;
+import software.amazon.kinesis.lifecycle.events.ProcessRecordsInput;
+import software.amazon.kinesis.lifecycle.events.ShardEndedInput;
+import software.amazon.kinesis.lifecycle.events.ShutdownRequestedInput;
+import software.amazon.kinesis.processor.ShardRecordProcessor;
+import software.amazon.kinesis.processor.ShardRecordProcessorFactory;
+import software.amazon.kinesis.processor.SingleStreamTracker;
+import software.amazon.kinesis.retrieval.KinesisClientRecord;
+
+import java.net.URI;
+import java.nio.channels.Channels;
+import java.time.Duration;
+import java.time.Instant;
+import java.util.Date;
+import java.util.List;
+import java.util.Optional;
+import java.util.Set;
+import java.util.UUID;
+
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.APPROXIMATE_ARRIVAL_TIMESTAMP;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.MIME_TYPE;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.PARTITION_KEY;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.RECORD_COUNT;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.RECORD_ERROR_MESSAGE;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.SEQUENCE_NUMBER;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.SHARD_ID;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.STREAM_NAME;
+import static 
org.apache.nifi.processors.aws.kinesis.ConsumeKinesisAttributes.SUB_SEQUENCE_NUMBER;
+
+@InputRequirement(InputRequirement.Requirement.INPUT_FORBIDDEN)
+@Tags({"amazon", "aws", "kinesis", "consume", "stream", "record"})
+@CapabilityDescription("""
+        Consumes data from the specified AWS Kinesis stream and outputs a 
FlowFile for every processed Record (raw) \
+        or a FlowFile for a batch of processed records if a Record Reader and 
Record Writer are configured.
+        AWS Kinesis Client Library can take a few minutes on the first start 
and several seconds on subsequent start \
+        to initialise before starting to fetch data.
+        Uses DynamoDB for check pointing and coordination, and CloudWatch 
(optional) for metrics.
+        Ensure that the credentials provided have access to DynamoDB and 
CloudWatch (optional) along with Kinesis.""")
+@WritesAttributes({
+        @WritesAttribute(attribute = STREAM_NAME,
+                description = "The name of the Kinesis Stream from which all 
Kinesis Records in the Flow File were read"),
+        @WritesAttribute(attribute = SHARD_ID,
+                description = "Shard ID from which all Kinesis Records in the 
Flow File were read"),
+        @WritesAttribute(attribute = PARTITION_KEY,
+                description = "Partition key of the last Kinesis Record in the 
Flow File"),
+        @WritesAttribute(attribute = SEQUENCE_NUMBER,
+                description = "A Sequence Number of the last Kinesis Record in 
the Flow File"),
+        @WritesAttribute(attribute = SUB_SEQUENCE_NUMBER,
+                description = "A SubSequence Number of the last Kinesis Record 
in the Flow File. Generated by KPL when aggregating records into a single 
Kinesis Record"),
+        @WritesAttribute(attribute = APPROXIMATE_ARRIVAL_TIMESTAMP,
+                description = "Approximate arrival timestamp of the last 
Kinesis Record in the Flow File"),
+        @WritesAttribute(attribute = MIME_TYPE,
+                description = "Sets the mime.type attribute to the MIME Type 
specified by the Record Writer (if configured)"),
+        @WritesAttribute(attribute = RECORD_COUNT,
+                description = "Number of records written to the FlowFiles by 
the Record Writer (if configured)"),
+        @WritesAttribute(attribute = RECORD_ERROR_MESSAGE,
+                description = "This attribute provides on failure the error 
message encountered by the Record Reader or Record Writer (if configured)")
+})
+@DefaultSettings(yieldDuration = "100 millis")
+@SystemResourceConsideration(resource = SystemResource.CPU, description = """
+        Kinesis Client Library is used to create a Worker thread for 
consumption of Kinesis Records.
+        The Worker is initialised and started when this Processor has been 
scheduled. It runs continually, spawning Kinesis Record Processors as required \
+        to fetch Kinesis Records. The Worker Thread (and any child Record 
Processor threads) are not controlled by the normal NiFi scheduler as part of 
the \
+        Concurrent Thread pool and are not released until this processor is 
stopped.""")
+@SystemResourceConsideration(resource = SystemResource.NETWORK, description = 
"""
+        Kinesis Client Library will continually poll for new Records, \
+        requesting up to a maximum number of Records/bytes per call. This can 
result in sustained network usage.""")
+@SystemResourceConsideration(resource = SystemResource.MEMORY, description = 
"""
+        ConsumeKinesis buffers Kinesis Records in memory until they can be 
processed.
+        The maximum size of the buffer is controlled by the 'Max Bytes to 
Buffer' property.
+        In addition, Kinesis Client Library may cache some amount of data for 
each shard when the processor's buffer is full.""")
+public class ConsumeKinesis extends AbstractProcessor {
+
+    private static final Duration HTTP_CLIENTS_CONNECTION_TIMEOUT = 
Duration.ofSeconds(30);
+    private static final Duration HTTP_CLIENTS_READ_TIMEOUT = 
Duration.ofMinutes(3);
+
+    /**
+     * Best balance between throughput and CPU usage by KCL.
+     */
+    private static final int KINESIS_HTTP_CLIENT_CONCURRENCY_PER_TASK = 16;
+    private static final int KINESIS_HTTP_CLIENT_WINDOW_SIZE = 512 * 1024; // 
512 KiB
+    private static final Duration KINESIS_HTTP_HEALTH_CHECK_PERIOD = 
Duration.ofMinutes(1);
+
+    static final PropertyDescriptor KINESIS_STREAM_NAME = new 
PropertyDescriptor.Builder()
+            .name("Amazon Kinesis Stream Name")
+            .description("The name of the Kinesis stream to consume from.")
+            .required(true)
+            .addValidator(StandardValidators.NON_EMPTY_VALIDATOR)
+            .build();
+
+    static final PropertyDescriptor APPLICATION_NAME = new 
PropertyDescriptor.Builder()
+            .name("Application Name")
+            .description("The name of the Kinesis application. This is used 
for DynamoDB table naming and worker coordination.")
+            .required(true)
+            .addValidator(StandardValidators.NON_EMPTY_VALIDATOR)
+            .build();
+
+    static final PropertyDescriptor AWS_CREDENTIALS_PROVIDER_SERVICE = new 
PropertyDescriptor.Builder()
+            .name("AWS Credentials Provider service")
+            .displayName("AWS Credentials Provider Service")
+            .description("The Controller Service that is used to obtain AWS 
credentials provider.")
+            .required(true)
+            .identifiesControllerService(AWSCredentialsProviderService.class)
+            .build();
+
+    static final PropertyDescriptor REGION = new PropertyDescriptor.Builder()
+            .name("Region")
+            .description("AWS Region in which the Kinesis stream is located.")
+            .required(true)
+            .allowableValues(RegionUtilV2.getAvailableRegions())
+            
.defaultValue(RegionUtilV2.createAllowableValue(Region.US_WEST_2).getValue())
+            .build();
+
+    static final PropertyDescriptor RECORD_READER = new 
PropertyDescriptor.Builder()
+            .name("Record Reader")
+            .description("""
+                    The Record Reader to use for parsing the data received 
from Kinesis.
+                    If not set, the records are written one per FlowFile 
without parsing.""")
+            .required(false)
+            .identifiesControllerService(RecordReaderFactory.class)
+            .build();
+
+    static final PropertyDescriptor RECORD_WRITER = new 
PropertyDescriptor.Builder()
+            .name("Record Writer")
+            .description("The Record Writer to use for serializing records.")
+            .required(true)
+            .dependsOn(RECORD_READER)
+            .identifiesControllerService(RecordSetWriterFactory.class)
+            .build();
+
+    static final PropertyDescriptor INITIAL_STREAM_POSITION = new 
PropertyDescriptor.Builder()
+            .name("Initial Stream Position")
+            .description("The position in the stream where the processor 
should start reading.")
+            .required(true)
+            .allowableValues(InitialPosition.class)
+            .defaultValue(InitialPosition.TRIM_HORIZON)
+            .build();
+
+    static final PropertyDescriptor STREAM_POSITION_TIMESTAMP = new 
PropertyDescriptor.Builder()
+            .name("Stream Position Timestamp")
+            .description("Timestamp position in stream from which to start 
reading Kinesis Records. The timestamp must be in ISO 8601 format.")
+            .addValidator(StandardValidators.ISO8601_INSTANT_VALIDATOR)
+            .dependsOn(INITIAL_STREAM_POSITION, InitialPosition.AT_TIMESTAMP)
+            .required(true)
+            .build();
+
+    static final PropertyDescriptor MAX_BYTES_TO_BUFFER = new 
PropertyDescriptor.Builder()
+            .name("Max Bytes to Buffer")
+            .description("""
+                    The maximum size of Kinesis Records that can be buffered 
in memory before being processed by NiFi.
+                    If the buffer size exceeds the limit, the KCL will stop 
consuming new records until free space is available.
+
+                    Using a larger value may improve throughput, but will do 
so at the expense of using additional heap.
+                    Using a smaller value may back off the Kinesis Client 
Library (KCL) from consuming records if the buffer is full, which may result in 
lower throughput.
+                    """)
+            .required(true)
+            .addValidator(StandardValidators.DATA_SIZE_VALIDATOR)
+            .defaultValue("100 MB")
+            .build();
+
+    static final PropertyDescriptor CHECKPOINT_INTERVAL = new 
PropertyDescriptor.Builder()
+            .name("Checkpoint Interval")
+            .description("""
+                    Interval between checkpointing consumed Kinesis records. 
To checkpoint records on each NiFi session commit, set this value to 0 seconds.
+
+                    Checkpointing too frequently may result in performance 
degradation and higher DynamoDB costs.
+                    Checkpointing too rarely may result in duplicated records 
whenever a Shard lease is lost or NiFi server restarts.

Review Comment:
   Not really. Even when records are checkpointed on each NiFi commit, it's 
possible NiFi shuts down after committing the session, but right before 
performing the checkpoint. In that case we will reprocess the same records on 
retry.
   
   As for other streaming processors, downstream flow should still handle 
duplicates properly.
   
   Lowering the checkpoint interval makes sense when downstream processing is 
very expensive or duplicate detection is limited in time, so we want to 
decrease the number of duplicates as much as possible.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to