kishoreg commented on a change in pull request #4397: Adding Support for Kafka 2.0 Consumer URL: https://github.com/apache/incubator-pinot/pull/4397#discussion_r306662939
########## File path: pinot-connectors/pinot-connector-kafka-2.0/src/main/java/org/apache/pinot/core/realtime/impl/kafka2/KafkaStreamLevelConsumer.java ########## @@ -0,0 +1,180 @@ +/** + * 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.pinot.core.realtime.impl.kafka2; + +import com.yammer.metrics.core.Meter; +import java.time.Duration; +import java.util.Arrays; +import java.util.HashMap; +import java.util.Iterator; +import java.util.Map; +import org.apache.kafka.clients.consumer.ConsumerRecord; +import org.apache.kafka.clients.consumer.ConsumerRecords; +import org.apache.kafka.clients.consumer.KafkaConsumer; +import org.apache.kafka.clients.consumer.OffsetAndMetadata; +import org.apache.kafka.common.TopicPartition; +import org.apache.kafka.common.utils.Bytes; +import org.apache.pinot.common.data.Schema; +import org.apache.pinot.common.metadata.instance.InstanceZKMetadata; +import org.apache.pinot.common.metrics.ServerMeter; +import org.apache.pinot.common.metrics.ServerMetrics; +import org.apache.pinot.core.data.GenericRow; +import org.apache.pinot.core.realtime.stream.StreamConfig; +import org.apache.pinot.core.realtime.stream.StreamDecoderProvider; +import org.apache.pinot.core.realtime.stream.StreamLevelConsumer; +import org.apache.pinot.core.realtime.stream.StreamMessageDecoder; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + + +/** + * An implementation of a {@link StreamLevelConsumer} which consumes from the kafka stream + */ +public class KafkaStreamLevelConsumer implements StreamLevelConsumer { + + private StreamMessageDecoder _messageDecoder; + private Logger INSTANCE_LOGGER; + + private String _clientId; + private String _tableAndStreamName; + + private StreamConfig _streamConfig; + private KafkaHighLevelStreamConfig _kafkaHighLevelStreamConfig; + + private KafkaConsumer<Bytes, Bytes> consumer; + private ConsumerRecords<Bytes, Bytes> consumerRecords; + private Iterator<ConsumerRecord<Bytes, Bytes>> kafkaIterator; + private Map<Integer, Long> consumerOffsets = new HashMap<>(); // tracking current consumed records offsets. + + private long lastLogTime = 0; + private long lastCount = 0; + private long currentCount = 0L; + + private ServerMetrics _serverMetrics; + private Meter tableAndStreamRowsConsumed = null; + private Meter tableRowsConsumed = null; + + public KafkaStreamLevelConsumer(String clientId, String tableName, StreamConfig streamConfig, Schema schema, + InstanceZKMetadata instanceZKMetadata, ServerMetrics serverMetrics) { + _clientId = clientId; + _streamConfig = streamConfig; + _kafkaHighLevelStreamConfig = new KafkaHighLevelStreamConfig(streamConfig, tableName, instanceZKMetadata); + _serverMetrics = serverMetrics; + + _messageDecoder = StreamDecoderProvider.create(streamConfig, schema); + + _tableAndStreamName = tableName + "-" + streamConfig.getTopicName(); + INSTANCE_LOGGER = LoggerFactory + .getLogger(KafkaStreamLevelConsumer.class.getName() + "_" + tableName + "_" + streamConfig.getTopicName()); + INSTANCE_LOGGER.info("KafkaStreamLevelConsumer: streamConfig : {}", _streamConfig); + } + + @Override + public void start() + throws Exception { + consumer = KafkaStreamLevelConsumerManager.acquireKafkaConsumerForConfig(_kafkaHighLevelStreamConfig); + } + + private void updateKafkaIterator() { + consumerRecords = consumer.poll(Duration.ofMillis(_streamConfig.getFetchTimeoutMillis())); + kafkaIterator = consumerRecords.iterator(); + } + + private void resetOffsets() { + for (int partition : consumerOffsets.keySet()) { + long offsetToSeek = consumerOffsets.get(partition); + consumer.seek(new TopicPartition(_streamConfig.getTopicName(), partition), offsetToSeek); + } + } + + @Override + public GenericRow next(GenericRow destination) { + if (kafkaIterator == null || !kafkaIterator.hasNext()) { + updateKafkaIterator(); + } + if (kafkaIterator.hasNext()) { + try { + final ConsumerRecord<Bytes, Bytes> record = kafkaIterator.next(); + updateOffsets(record.partition(), record.offset()); + destination = _messageDecoder.decode(record.value().get(), destination); + tableAndStreamRowsConsumed = _serverMetrics + .addMeteredTableValue(_tableAndStreamName, ServerMeter.REALTIME_ROWS_CONSUMED, 1L, + tableAndStreamRowsConsumed); + tableRowsConsumed = + _serverMetrics.addMeteredGlobalValue(ServerMeter.REALTIME_ROWS_CONSUMED, 1L, tableRowsConsumed); + + ++currentCount; + + final long now = System.currentTimeMillis(); + // Log every minute or 100k events + if (now - lastLogTime > 60000 || currentCount - lastCount >= 100000) { + if (lastCount == 0) { + INSTANCE_LOGGER.info("Consumed {} events from kafka stream {}", currentCount, _streamConfig.getTopicName()); + } else { + INSTANCE_LOGGER.info("Consumed {} events from kafka stream {} (rate:{}/s)", currentCount - lastCount, + _streamConfig.getTopicName(), (float) (currentCount - lastCount) * 1000 / (now - lastLogTime)); + } + lastCount = currentCount; + lastLogTime = now; + } + return destination; + } catch (Exception e) { + INSTANCE_LOGGER.warn("Caught exception while consuming events", e); + _serverMetrics.addMeteredTableValue(_tableAndStreamName, ServerMeter.REALTIME_CONSUMPTION_EXCEPTIONS, 1L); + _serverMetrics.addMeteredGlobalValue(ServerMeter.REALTIME_CONSUMPTION_EXCEPTIONS, 1L); + throw e; Review comment: should we throw an exception or retry with exponentially back-off? ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
