flink-user-zh mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Becket Qin <becket....@gmail.com>
Subject Re: Re: Re: [Blink]sql client kafka sink 失败
Date Mon, 25 Feb 2019 06:34:54 GMT
@Kurt,

这个是符合预期的。为了防止和用户code中可能的Kafka依赖发生冲突。

On Mon, Feb 25, 2019 at 10:28 AM Kurt Young <ykt836@gmail.com> wrote:

> kafka的包看路径是shade过的,这是符合预期的吗? @Becket
>
> Best,
> Kurt
>
>
> On Mon, Feb 25, 2019 at 9:56 AM 张洪涛 <hongtao12310@163.com> wrote:
>
> >
> >
> > sql-client.sh 的启动参数首先在classpath里面会包含kafka相关的jar
 另外会有--jar
> > 包含所有connector的jar
> >
> >
> > 这些jars在sql-client提交job时候会上传到cluster的blob store 但是很奇怪为啥找不到
> >
> >
> >  00:00:06 /usr/lib/jvm/java-1.8.0-openjdk/bin/java
> > -Dlog.file=/bigdata/flink-1.5.1/log/flink-root-sql-client-gpu06.log
> > -Dlog4j.configuration=file:/bigdata/flink-1.5.1/conf/log4j-cli.properties
> > -Dlogback.configurationFile=file:/bigdata/flink-1.5.1/conf/logback.xml
> > -classpath
> >
> /bigdata/flink-1.5.1/lib/flink-python_2.11-1.5.1.jar:/bigdata/flink-1.5.1/lib/flink-shaded-hadoop2-uber-1.5.1.jar:/bigdata/flink-1.5.1/lib/log4j-1.2.17.jar:/bigdata/flink-1.5.1/lib/slf4j-log4j12-1.7.7.jar:/bigdata/flink-1.5.1/lib/flink-dist_2.11-1.5.1.jar::/bigdata/hadoop-2.7.5/etc/hadoop::/bigdata/flink-1.5.1/opt/connectors/kafka011/flink-connector-kafka-0.11_2.11-1.5.1-sql-jar.jar:/bigdata/flink-1.5.1/opt/connectors/kafka010/flink-connector-kafka-0.10_2.11-1.5.1-sql-jar.jar:/bigdata/flink-1.5.1/opt/connectors/kafka09/flink-connector-kafka-0.9_2.11-1.5.1-sql-jar.jar:/bigdata/flink-1.5.1/opt/connectors/kafka08/flink-connector-kafka-0.8_2.11-1.5.1.jar:/bigdata/flink-1.5.1/opt/connectors/flink-hbase_2.11-1.5.1.jar:/bigdata/flink-1.5.1/opt/connectors/flink-connector-hadoop-compatibility_2.11-1.5.1.jar:/bigdata/flink-1.5.1/opt/connectors/flink-connector-hive_2.11-1.5.1.jar:/bigdata/flink-1.5.1/opt/sql-client/datanucleus-api-jdo-4.2.4.jar:/bigdata/flink-1.5.1/opt/sql-client/javax.jdo-3.2.0-m3.jar:/bigdata/flink-1.5.1/opt/sql-client/datanucleus-core-4.1.17.jar:/bigdata/flink-1.5.1/opt/sql-client/datanucleus-rdbms-4.1.19.jar:/bigdata/flink-1.5.1/opt/sql-client/flink-sql-client-1.5.1.jar
> > org.apache.flink.table.client.SqlClient embedded -d
> > conf/sql-client-defaults.yaml --jar
> >
> /bigdata/flink-1.5.1/opt/connectors/kafka011/flink-connector-kafka-0.11_2.11-1.5.1-sql-jar.jar
> > --jar
> >
> /bigdata/flink-1.5.1/opt/connectors/kafka010/flink-connector-kafka-0.10_2.11-1.5.1-sql-jar.jar
> > --jar
> >
> /bigdata/flink-1.5.1/opt/connectors/kafka09/flink-connector-kafka-0.9_2.11-1.5.1-sql-jar.jar
> > --jar
> >
> /bigdata/flink-1.5.1/opt/connectors/kafka08/flink-connector-kafka-0.8_2.11-1.5.1.jar
> > --jar /bigdata/flink-1.5.1/opt/connectors/flink-hbase_2.11-1.5.1.jar
> --jar
> >
> /bigdata/flink-1.5.1/opt/connectors/flink-connector-hadoop-compatibility_2.11-1.5.1.jar
> > --jar
> > /bigdata/flink-1.5.1/opt/connectors/flink-connector-hive_2.11-1.5.1.jar
> > --jar /bigdata/flink-1.5.1/opt/sql-client/flink-sql-client-1.5.1.jar
> >
> >
> >
> >
> >
> >
> > 在 2019-02-22 19:32:18,"Becket Qin" <becket.qin@gmail.com> 写道:
> > >能不能看一下运行sql-client.sh的运行参数。具体做法是:
> > >
> > >运行sql-client.sh
> > >ps | grep sql-client
> > >
> > >查看一下其中是不是有这个 flink-connector-kafka-0.11 的 jar.
> > >
> > >Jiangjie (Becket) Qin
> > >
> > >On Fri, Feb 22, 2019 at 6:54 PM 张洪涛 <hongtao12310@163.com> wrote:
> > >
> > >>
> > >>
> > >> 是包含这个类的
> > >>
> > >>
> > >> jar -tf flink-connector-kafka-0.11_2.11-*.jar | grep Crc32C
> > >>
> > >>
> >
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/Crc32C$1.class
> > >>
> > >>
> >
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/Crc32C$ChecksumFactory.class
> > >>
> > >>
> >
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/Crc32C$Java9ChecksumFactory.class
> > >>
> > >>
> >
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/Crc32C$PureJavaChecksumFactory.class
> > >>
> >
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/Crc32C.class
> > >>
> > >>
> >
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/PureJavaCrc32C.class
> > >>
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> 在 2019-02-22 18:03:18,"Zhenghua Gao" <docete@gmail.com> 写道:
> > >> >能否看一下对应的包里是否有这个类, 方法如下(假设你的blink安装包在
/tmp/blink):
> > >> >
> > >> >cd /tmp/blink/opt/connectors/kafka011
> > >> >jar -tf flink-connector-kafka-0.11_2.11-*.jar | grep Crc32C
> > >> >
> > >> >On Fri, Feb 22, 2019 at 2:56 PM 张洪涛 <hongtao12310@163.com>
wrote:
> > >> >
> > >> >>
> > >> >>
> > >> >> 大家好!
> > >> >>
> > >> >>
> > >> >> 我正在测试Blink sql client kafka sink connector ,但是发现写入失败,以下是我的步骤
> > >> >>
> > >> >>
> > >> >> 环境配置
> > >> >> blink standalone 模式
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >> 1. 配置environment 启动sql client
> > >> >>
> > >> >>
> > >> >> 2. 创建kafka sink table
> > >> >> CREATETABLEkafka_sink(
> > >> >>    messageKeyVARBINARY,
> > >> >>    messageValueVARBINARY,
> > >> >>    PRIMARYKEY(messageKey))
> > >> >> with(
> > >> >>    type='KAFKA011',
> > >> >>    topic='sink-topic',
> > >> >>    `bootstrap.servers`='172.19.0.108:9092',
> > >> >>    retries='3'
> > >> >> );
> > >> >>
> > >> >>
> > >> >> 3. 创建查询语句并执行
> > >> >> INSERT INTO kafka_sink
> > >> >> SELECT CAST('123' AS VARBINARY) AS key,
> > >> >> CAST(CONCAT_WS(',', 'HELLO', 'WORLD') AS VARBINARY) AS msg;
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >> 错误日志(from task executor log)
> > >> >>
> > >> >>
> > >> >> 主要是找不到kafka common package下面的一个类, 但是启动sql
client 时候已经把kafka
> connector
> > >> >> 相关的jar包包括在内 在提交job时候 也会把这些jars
和
> jobgraph一并上传到cluster,理论上这些class都会被加载
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >> 2019-02-22 14:37:18,356 ERROR
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.common.utils.KafkaThread
> > >> >> - Uncaught exception in kafka-producer-network-thread | producer-1:
> > >> >> java.lang.NoClassDefFoundError:
> > >> >>
> org/apache/flink/kafka011/shaded/org/apache/kafka/common/utils/Crc32C
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.common.record.DefaultRecordBatch.writeHeader(DefaultRecordBatch.java:468)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.common.record.MemoryRecordsBuilder.writeDefaultBatchHeader(MemoryRecordsBuilder.java:339)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.common.record.MemoryRecordsBuilder.close(MemoryRecordsBuilder.java:293)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.clients.producer.internals.ProducerBatch.close(ProducerBatch.java:391)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.clients.producer.internals.RecordAccumulator.drain(RecordAccumulator.java:485)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.clients.producer.internals.Sender.sendProducerData(Sender.java:254)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.clients.producer.internals.Sender.run(Sender.java:223)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.kafka011.shaded.org.apache.kafka.clients.producer.internals.Sender.run(Sender.java:162)
> > >> >>         at java.lang.Thread.run(Thread.java:748)
> > >> >> Caused by: java.lang.ClassNotFoundException:
> > >> >>
> org.apache.flink.kafka011.shaded.org.apache.kafka.common.utils.Crc32C
> > >> >>         at
> java.net.URLClassLoader.findClass(URLClassLoader.java:381)
> > >> >>         at java.lang.ClassLoader.loadClass(ClassLoader.java:424)
> > >> >>         at
> > >> >>
> > >>
> >
> org.apache.flink.runtime.execution.librarycache.FlinkUserCodeClassLoaders$ChildFirstClassLoader.loadClass(FlinkUserCodeClassLoaders.java:120)
> > >> >>         at java.lang.ClassLoader.loadClass(ClassLoader.java:357)
> > >> >>
> > >> >>
> > >> >>
> > >> >>
> > >> >> --
> > >> >>   Best Regards
> > >> >>   Hongtao
> > >> >>
> > >> >>
> > >> >
> > >> >--
> > >> >若批評無自由,則讚美無意義!
> > >>
> > >>
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> --
> > >>   Best Regards,
> > >>   HongTao
> > >>
> > >>
> >
> >
> >
> >
> >
> >
> >
> > --
> >   Best Regards,
> >   HongTao
> >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message