Help Center>
Data Ingestion Service>
User Guide>
Using DIS>
Using Kafka Adapter to Upload and Download Data>
Consumption Modes
Updated on 2024-03-05 GMT+08:00
Consumption Modes
Similar to Kafka, dis kafka adapter supports three consumption modes.
assign Mode
Users specify the partitions to be consumed by the consumer instance are manually. In this case, the group management mechanism is not used. That is, when the number of consumers in the group changes or the stream scaling is performed, the partitions will not be reallocated. A code example is provided as follows:
package com.huaweicloud.dis.demo.adapter;
import com.huaweicloud.dis.DISConfig;
import com.huaweicloud.dis.adapter.kafka.clients.consumer.*;
import com.huaweicloud.dis.adapter.kafka.common.PartitionInfo;
import com.huaweicloud.dis.adapter.kafka.common.TopicPartition;
import com.huaweicloud.dis.adapter.kafka.common.serialization.StringDeserializer;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.ArrayList;
import java.util.List;
import java.util.Map;
import java.util.Properties;
public class DISKafkaConsumerAssignDemo
{
private static final Logger LOGGER = LoggerFactory.getLogger(DISKafkaConsumerAssignDemo.class);
public static void main(String[] args)
{
// There will be security risks if the AK/SK used for authentication is directly written into code. Encrypt the AK/SK in the configuration file or environment variables for storage;
// In this example, the AK and SK stored in the environment variables are used for identity authentication. Before running this example, configure environment variables HUAWEICLOUD_SDK_AK and HUAWEICLOUD_SDK_SK in the local environment.
String ak = System.getenv("HUAWEICLOUD_SDK_AK");
String sk = System.getenv("HUAWEICLOUD_SDK_SK");
// YOU ProjectId
String projectId = "YOU_PROJECT_ID";
// YOU DIS Stream
String streamName = "YOU_STREAM_NAME";
//Consumption group ID, which is used to record the offset.
String groupId = "YOU_GROUP_ID";
// DIS region
String region = "your region";
Properties props = new Properties();
props.setProperty(DISConfig.PROPERTY_AK, ak);
props.setProperty(DISConfig.PROPERTY_SK, sk);
props.setProperty(DISConfig.PROPERTY_PROJECT_ID, projectId);
props.setProperty(DISConfig.PROPERTY_REGION_ID, region);
props.setProperty(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class.getName());
props.setProperty(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class.getName());
props.setProperty(ConsumerConfig.GROUP_ID_CONFIG, groupId);
props.setProperty(ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG, "false");
props.setProperty(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, OffsetResetStrategy.LATEST.name());
//By default, the domain name is automatically used for access instead of configuring an endpoint. If an endpoint is required, remove the following comments and set the endpoint:
// props.setProperty(DISConfig.PROPERTY_ENDPOINT, "https://dis-${region}.myhuaweicloud.com");
Consumer<String, String> consumer = new DISKafkaConsumer<>(props);
List<TopicPartition> topicPartitions = new ArrayList<>();
for (PartitionInfo partitionInfo : consumer.partitionsFor(streamName))
{
topicPartitions.add(new TopicPartition(partitionInfo.topic(), partitionInfo.partition()));
}
//Use the assign mode to specify the partition to be consumed.
consumer.assign(topicPartitions);
while (true)
{
try
{
ConsumerRecords<String, String> records = consumer.poll(Long.MAX_VALUE);
if (!records.isEmpty())
{
for (TopicPartition partition : records.partitions())
{
List<ConsumerRecord<String, String>> partitionRecords = records.records(partition);
for (ConsumerRecord<String, String> record : partitionRecords)
{
LOGGER.info("Value [{}], Partition [{}], Offset [{}], Key [{}]",
record.value(), record.partition(), record.offset(), record.key());
}
}
//Submit the current offset asynchronously after data processing is complete or submit commitSync synchronously.
consumer.commitAsync(new OffsetCommitCallback()
{
@Override
public void onComplete(Map<TopicPartition, OffsetAndMetadata> map, Exception e)
{
if (e == null)
{
LOGGER.debug("Success to commit offset [{}]", map);
}
else
{
LOGGER.error("Failed to commit offset [{}]", e.getMessage(), e);
}
}
});
}
}
catch (Exception e)
{
LOGGER.info(e.getMessage(), e);
}
}
}
}
After running the preceding program, if the data is sent to the stream, the following log is generated:
09:36:45.071 INFO c.h.d.a.k.c.DISKafkaConsumer - create DISKafkaConsumer successfully 09:36:49.842 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[sync]. 0], Partition [0], Offset [134], Key [769066] 09:36:49.963 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[sync]. 1], Partition [0], Offset [135], Key [700005] 09:36:50.145 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[sync]. 2], Partition [0], Offset [136], Key [338044] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[sync]. 3], Partition [0], Offset [137], Key [537495] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[sync]. 4], Partition [0], Offset [138], Key [980016] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[async]. 0], Partition [0], Offset [139], Key [182772] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[async]. 1], Partition [0], Offset [140], Key [830271] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[async]. 2], Partition [0], Offset [141], Key [927054] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[async]. 3], Partition [0], Offset [142], Key [268122] 09:36:51.093 INFO c.h.d.d.a.DISKafkaConsumerAssignDemo - Value [Hello world[async]. 4], Partition [0], Offset [143], Key [992787]
subscribe Mode
Users only need to specify stream names without specifying specific partitions. The server will trigger the group management mechanism based on the number of consumers or stream scaling changes to automatically allocate partitions to each consumer. This ensures that a partition is consumed by only one consumer.
A code example is provided as follows:
package com.huaweicloud.dis.demo.adapter;
import com.huaweicloud.dis.DISConfig;
import com.huaweicloud.dis.adapter.kafka.clients.consumer.*;
import com.huaweicloud.dis.adapter.kafka.common.TopicPartition;
import com.huaweicloud.dis.adapter.kafka.common.serialization.StringDeserializer;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.Collections;
import java.util.List;
import java.util.Map;
import java.util.Properties;
public class DISKafkaConsumerSubscribeDemo
{
private static final Logger LOGGER = LoggerFactory.getLogger(DISKafkaConsumerSubscribeDemo.class);
public static void main(String[] args)
{
// There will be security risks if the AK/SK used for authentication is directly written into code. Encrypt the AK/SK in the configuration file or environment variables for storage;
// In this example, the AK and SK stored in the environment variables are used for identity authentication. Before running this example, configure environment variables HUAWEICLOUD_SDK_AK and HUAWEICLOUD_SDK_SK in the local environment.
String ak = System.getenv("HUAWEICLOUD_SDK_AK");
String sk = System.getenv("HUAWEICLOUD_SDK_SK");
// YOU ProjectId
String projectId = "YOU_PROJECT_ID";
// YOU DIS Stream
String streamName = "YOU_STREAM_NAME";
//Consumption group ID, which is used to record the offset.
String groupId = "YOU_GROUP_ID";
// DIS region
String region = "your region";
Properties props = new Properties();
props.setProperty(DISConfig.PROPERTY_AK, ak);
props.setProperty(DISConfig.PROPERTY_SK, sk);
props.setProperty(DISConfig.PROPERTY_PROJECT_ID, projectId);
props.setProperty(DISConfig.PROPERTY_REGION_ID, region);
props.setProperty(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class.getName());
props.setProperty(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class.getName());
props.setProperty(ConsumerConfig.GROUP_ID_CONFIG, groupId);
props.setProperty(ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG, "false");
props.setProperty(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, OffsetResetStrategy.LATEST.name());
//By default, the domain name is automatically used for access instead of configuring an endpoint. If an endpoint is required, remove the following comments and set the endpoint:
// props.setProperty(DISConfig.PROPERTY_ENDPOINT, "https://dis-${region}.myhuaweicloud.com");
Consumer<String, String> consumer = new DISKafkaConsumer<>(props);
//When using the subscribe mode, specify the name of the stream to be consumed.
consumer.subscribe(Collections.singleton(streamName));
while (true)
{
try
{
ConsumerRecords<String, String> records = consumer.poll(Long.MAX_VALUE);
if (!records.isEmpty())
{
for (TopicPartition partition : records.partitions())
{
List<ConsumerRecord<String, String>> partitionRecords = records.records(partition);
for (ConsumerRecord<String, String> record : partitionRecords)
{
LOGGER.info("Value [{}], Partition [{}], Offset [{}], Key [{}]",
record.value(), record.partition(), record.offset(), record.key());
}
}
//Submit the current offset asynchronously after data processing is complete or submit commitSync synchronously.
consumer.commitAsync(new OffsetCommitCallback()
{
@Override
public void onComplete(Map<TopicPartition, OffsetAndMetadata> map, Exception e)
{
if (e == null)
{
LOGGER.debug("Success to commit offset [{}]", map);
}
else
{
LOGGER.error("Failed to commit offset [{}]", e.getMessage(), e);
}
}
});
}
}
catch (Exception e)
{
LOGGER.info(e.getMessage(), e);
}
}
}
}
When the program runs, it sends a heartbeat request (Heartbeat) every 10 seconds and then sends a (JoinGroup) request to join the consumer group. The server starts to allocate partitions to consumers in the consumer group. This process takes about 20s. After the operation is complete, the consumers send synchronization requests (SyncGroup) to obtain allocation results. If the information about heartbeat {"state": "STABLE"} is recorded in the log, it indicates that allocation to the entire consumer group has been completed, data is ready to be consumed.
The key logs in this process are described as follows:
- Heartbeat {"state":"JOINING"}
Heartbeat: indicates a heartbeat request. The heartbeat request is initiated every 10 seconds and is used to keep a connection with the server. If the server does not receive the heartbeat message within 1 minute, it considers that the consumer is offline and the partitions will be reallocated to the consumers in the consumer group. If the heartbeat result is JOINING, the consumer needs to join the consumer group again. If the heartbeat result is STABLE, the consumer group is stable.
- JoinGroup
If the heartbeat result is not STABLE, the consumer sends a joinGroup request to notify the server that it needs to join the consumer group. After receiving the join request from the client, the server will reallocate partitions for the consumer group. In this case, syncDelayedTimeMs is returned, indicating the allocation duration. After allocation is completed, the client sends a synchronization request (SyncGroup) to obtain the allocation result.
- SyncGroup
A SyncGroup request is used to obtain the allocation result. The returned assignment contains the stream name and partition to be consumed.
Run the sample program. After allocation is completed, send data to the stream. The complete log is as follows:
09:42:37.296 INFO c.h.d.a.k.c.DISKafkaConsumer - create DISKafkaConsumer successfully
09:42:37.354 INFO c.h.d.a.k.c.Coordinator - Heartbeat {"state":"JOINING"}
09:42:37.363 INFO c.h.d.a.k.c.Coordinator - joinGroupRequest {"groupId":"ding","clientId":"consumer-c2d43144-0823-4eea-aaa8-7af95c536144","interestedStream":["liuhao12"]}
09:42:37.406 INFO c.h.d.a.k.c.Coordinator - joinGroupResponse {"state":"OK","syncDelayedTimeMs":21000}
09:42:58.408 INFO c.h.d.a.k.c.Coordinator - syncGroup {"groupId":"ding","clientId":"consumer-c2d43144-0823-4eea-aaa8-7af95c536144","generation":-1}
09:42:58.451 INFO c.h.d.a.k.c.Coordinator - syncGroup {"state":"OK","generation":33,"assignment":{"dis-test":[0]}}
09:42:58.488 INFO c.h.d.a.k.c.Coordinator - Heartbeat {"state":"STABLE"}
09:43:08.960 INFO c.h.d.a.k.c.Coordinator - Heartbeat {"state":"STABLE"}
09:46:56.227 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[sync]. 0], Partition [0], Offset [144], Key [799704]
09:46:56.327 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[sync]. 1], Partition [0], Offset [145], Key [683757]
09:46:56.449 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[sync]. 2], Partition [0], Offset [146], Key [439062]
09:46:56.535 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[sync]. 3], Partition [0], Offset [147], Key [374939]
09:46:56.654 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[sync]. 4], Partition [0], Offset [148], Key [321528]
09:46:56.749 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[async]. 0], Partition [0], Offset [149], Key [964841]
09:46:56.749 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[async]. 1], Partition [0], Offset [150], Key [520262]
09:46:56.749 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[async]. 2], Partition [0], Offset [151], Key [619119]
09:46:56.749 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[async]. 3], Partition [0], Offset [152], Key [257094]
09:46:56.749 INFO c.h.d.d.a.DISKafkaConsumerSubscribeDemo - Value [Hello world[async]. 4], Partition [0], Offset [153], Key [310331]
subscribePattern Mode
Users specify a wildcard instead of specifying a stream name. For example, stream.* indicates that stream1, stream2, stream_123, and so on are consumed. An existing, added, or deleted stream can be consumed by a consumer group as long as it matches the wildcard.
A code example is provided as follows:
package com.huaweicloud.dis.demo.adapter;
import com.huaweicloud.dis.DISConfig;
import com.huaweicloud.dis.adapter.kafka.clients.consumer.*;
import com.huaweicloud.dis.adapter.kafka.common.TopicPartition;
import com.huaweicloud.dis.adapter.kafka.common.serialization.StringDeserializer;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.Collection;
import java.util.List;
import java.util.Map;
import java.util.Properties;
import java.util.regex.Pattern;
public class DISKafkaConsumerSubscribePatternDemo
{
private static final Logger LOGGER = LoggerFactory.getLogger(DISKafkaConsumerSubscribePatternDemo.class);
public static void main(String[] args)
{
// There will be security risks if the AK/SK used for authentication is directly written into code. Encrypt the AK/SK in the configuration file or environment variables for storage;
// In this example, the AK and SK stored in the environment variables are used for identity authentication. Before running this example, configure environment variables HUAWEICLOUD_SDK_AK and HUAWEICLOUD_SDK_SK in the local environment.
String ak = System.getenv("HUAWEICLOUD_SDK_AK");
String sk = System.getenv("HUAWEICLOUD_SDK_SK");
// YOU ProjectId
String projectId = "YOU_PROJECT_ID";
// YOU DIS Stream
String streamName = "YOU_STREAM_NAME";
//Consumption group ID, which is used to record the offset.
String groupId = "YOU_GROUP_ID";
// DIS region
String region = "your region";
Properties props = new Properties();
props.setProperty(DISConfig.PROPERTY_AK, ak);
props.setProperty(DISConfig.PROPERTY_SK, sk);
props.setProperty(DISConfig.PROPERTY_PROJECT_ID, projectId);
props.setProperty(DISConfig.PROPERTY_REGION_ID, region);
props.setProperty(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class.getName());
props.setProperty(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class.getName());
props.setProperty(ConsumerConfig.GROUP_ID_CONFIG, groupId);
props.setProperty(ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG, "false");
props.setProperty(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, OffsetResetStrategy.LATEST.name());
//By default, the domain name is automatically used for access instead of configuring an endpoint. If an endpoint is required, remove the following comments and set the endpoint:
// props.setProperty(DISConfig.PROPERTY_ENDPOINT, "https://dis-${region}.myhuaweicloud.com");
Consumer<String, String> consumer = new DISKafkaConsumer<>(props);
//With the subscribePattern mode, you only need to specify a wildcard.
consumer.subscribe(Pattern.compile(streamNamePattern), new ConsumerRebalanceListener()
{
@Override
public void onPartitionsRevoked(Collection<TopicPartition> collection)
{
LOGGER.info("onPartitionsRevoked [{}]", collection);
}
@Override
public void onPartitionsAssigned(Collection<TopicPartition> collection)
{
LOGGER.info("onPartitionsAssigned [{}]", collection);
}
});
while (true)
{
try
{
ConsumerRecords<String, String> records = consumer.poll(Long.MAX_VALUE);
if (!records.isEmpty())
{
for (TopicPartition partition : records.partitions())
{
List<ConsumerRecord<String, String>> partitionRecords = records.records(partition);
for (ConsumerRecord<String, String> record : partitionRecords)
{
LOGGER.info("Value [{}], Partition [{}], Offset [{}], Key [{}]",
record.value(), record.partition(), record.offset(), record.key());
}
}
//Submit the current offset asynchronously after data processing is complete or submit commitSync synchronously.
consumer.commitAsync(new OffsetCommitCallback()
{
@Override
public void onComplete(Map<TopicPartition, OffsetAndMetadata> map, Exception e)
{
if (e == null)
{
LOGGER.debug("Success to commit offset [{}]", map);
}
else
{
LOGGER.error("Failed to commit offset [{}]", e.getMessage(), e);
}
}
});
}
}
catch (Exception e)
{
LOGGER.info(e.getMessage(), e);
}
}
}
}
After the program runs, wait about 20s until the allocation is completed. The data can be consumed. The following is an example code execution log:
10:10:36.420 INFO c.h.d.a.k.c.DISKafkaConsumer - create DISKafkaConsumer successfully
10:10:36.481 INFO c.h.d.a.k.c.Coordinator - Heartbeat {"state":"JOINING"}
10:10:36.486 INFO c.h.d.a.k.c.Coordinator - joinGroupRequest {"groupId":"ding","clientId":"consumer-cad967ba-70ab-4e02-b184-f60b95fe3256","streamPattern":"stream.*"}
10:10:36.697 INFO c.h.d.a.k.c.Coordinator - joinGroupResponse {"state":"OK","subscription":["stream_hello","stream_world"],"syncDelayedTimeMs":21000}
10:10:57.699 INFO c.h.d.a.k.c.Coordinator - syncGroup {"groupId":"ding","clientId":"consumer-cad967ba-70ab-4e02-b184-f60b95fe3256","generation":-1}
10:10:57.746 INFO c.h.d.a.k.c.Coordinator - syncGroup {"state":"OK","generation":34,"assignment":{"stream_hello":[0],"stream_world":[0]}}
10:10:57.770 INFO c.h.d.d.a.DISKafkaConsumerSubscribePatternDemo - onPartitionsAssigned [[stream_hello-0, stream_world-0]]
10:10:57.770 INFO c.h.d.a.k.c.Coordinator - Heartbeat {"state":"STABLE"}
10:11:08.466 INFO c.h.d.a.k.c.Coordinator - Heartbeat {"state":"STABLE"}
10:11:09.992 INFO c.h.d.d.a.DISKafkaConsumerSubscribePatternDemo - Value [Hello world[sync]. 0], Partition [0], Offset [154], Key [181881]
10:11:09.993 INFO c.h.d.d.a.DISKafkaConsumerSubscribePatternDemo - Value [Hello world[sync]. 1], Partition [0], Offset [155], Key [483023]
10:11:09.993 INFO c.h.d.d.a.DISKafkaConsumerSubscribePatternDemo - Value [Hello world[sync]. 2], Partition [0], Offset [156], Key [32453]
10:11:10.093 INFO c.h.d.d.a.DISKafkaConsumerSubscribePatternDemo - Value [Hello world[sync]. 3], Partition [0], Offset [157], Key [111948]
10:11:10.180 INFO c.h.d.d.a.DISKafkaConsumerSubscribePatternDemo - Value [Hello world[sync]. 4], Partition [0], Offset [158], Key [822860]
Parent topic: Using Kafka Adapter to Upload and Download Data
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
The system is busy. Please try again later.
For any further questions, feel free to contact us through the chatbot.
Chatbot
