You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Jiangfei Liu (Jira)" <ji...@apache.org> on 2022/06/14 04:32:00 UTC
[jira] (FLINK-27828) FlinkKafkaProducer VS KafkaSink
[ https://issues.apache.org/jira/browse/FLINK-27828 ]
Jiangfei Liu deleted comment on FLINK-27828:
--------------------------------------
was (Author: JIRAUSER290004):
parallelism:3
checkpoint config:
CheckpointConfig checkpointConfig = env.getCheckpointConfig();
env.enableCheckpointing(CHECKPOINT_INTERVAL);
checkpointConfig.setCheckpointingMode(CHECKPOINT_MODE);
checkpointConfig.setCheckpointTimeout(CHECKPOINT_TIMEOUT);
checkpointConfig.setTolerableCheckpointFailureNumber(CHECKPOINT_FAILURE_NUMBER);
env.setRestartStrategy(RESTART_STRATEGY_CONFIGURATION);
checkpointConfig.setMaxConcurrentCheckpoints(CHECKPOINT_MAX_CONCURRENT);
checkpointConfig.setMinPauseBetweenCheckpoints(CHECKPOINT_MIN_PAUSE_BETWEEN);
checkpointConfig.setExternalizedCheckpointCleanup(CHECKPOINT_EXTERNALIZED_CLEANUP);
checkpointConfig.setCheckpointStorage(new FileSystemCheckpointStorage(HDFS_BASE + CHECKPOINT_BASE_PATH + path));
System.setProperty("HADOOP_USER_NAME", HADOOP_USER_NAME);
> FlinkKafkaProducer VS KafkaSink
> -------------------------------
>
> Key: FLINK-27828
> URL: https://issues.apache.org/jira/browse/FLINK-27828
> Project: Flink
> Issue Type: Bug
> Components: API / DataStream
> Affects Versions: 1.14.3
> Reporter: Jiangfei Liu
> Priority: Major
> Attachments: Snipaste_2022-05-25_19-52-11.png
>
>
> sorry,my english is bad.
> in flink1.14.3,write 10000 data to kafka.
> when use FlinkKafkaProducer,completed 7s
> when use KafkaSink,completed 1m40s
> why KafkaSink is low speed?
--
This message was sent by Atlassian Jira
(v8.20.7#820007)