gpt4 book ai didi

apache-spark - kerberos 环境的 Oozie Spark 操作失败

转载 作者:行者123 更新时间:2023-12-03 14:45:47 34 4
gpt4 key购买 nike

我正在通过 oozie spark 操作运行一个 spark 作业。 spark 作业使用 hivecontext 来执行一些要求。集群配置了 kerberos。当我使用 spark-submit 表单控制台提交作业时,它运行成功。但是当我从 oozie 运行作业时,出现以下错误。

18/03/18 03:34:16 INFO metastore: Trying to connect to metastore with URI thrift://localhost.local:9083
18/03/18 03:34:16 ERROR TSaslTransport: SASL negotiation failure
javax.security.sasl.SaslException: GSS initiate failed [Caused by GSSException: No valid credentials provided (Mechanism level: Failed to find any Kerberos tgt)]
at com.sun.security.sasl.gsskerb.GssKrb5Client.evaluateChallenge(GssKrb5Client.java:211)
at org.apache.thrift.transport.TSaslClientTransport.handleSaslStartMessage(TSaslClientTransport.java:94)

工作流.xml

<workflow-app xmlns="uri:oozie:workflow:0.5" name="workflow">
<start to="analysis" />
<!-- Bash script to do the spark-submit. The version numbers of these actions are magic. -->
<action name="Analysis">
<spark xmlns="uri:oozie:spark-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>${master}</master>
<name>Analysis</name>
<class>com.demo.analyzer</class>
<jar>${appLib}</jar>
<spark-opts>--jars ${sparkLib} --files ${config},${hivesite} --num-executors ${NoOfExecutors} --executor-cores ${ExecutorCores} --executor-memory ${ExecutorMemory} --driver-memory ${driverMemory}</spark-opts>
</spark>
<ok to="sendEmail" />
<error to="fail" />
</action>
<action name="sendEmail">
<email xmlns="uri:oozie:email-action:0.1">
<to>${emailToAddress}</to>
<subject>Output of workflow ${wf:id()}</subject>
<body>Results from line count: ${wf:actionData('shellAction')['NumberOfLines']}</body>
</email>
<ok to="end" />
<error to="end" />
</action>
<!-- You wish you'd ever get Oozie errors. -->
<kill name="fail">
<message>Bash action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
<end name="end" />
</workflow-app>

我是否需要在 workflow.xml 中配置任何与 Kerberos 相关的内容?我在这里遗漏了什么吗?

感谢任何帮助。

提前致谢。

最佳答案

您需要在 oozie 工作流中为 thrift uri 添加 hcat 凭据。这将使使用 Kerberos 的 thrift URI 成功验证 Metastore。

在 oozie 工作流中的 credentials 标签下方添加。

<credentials>
<credential name="credhive" type="hcat">
<property>
<name>hcat.metastore.uri</name>
<value>${thrift_uri}</value>
</property>
<property>
<name>hcat.metastore.principal</name>
<value>${principal}</value>
</property>
</credential>
</credentials>

并向 spark 操作提供凭证,如下所示:

<action name="Analysis" cred="credhive">
<spark xmlns="uri:oozie:spark-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<master>${master}</master>
<name>Analysis</name>
<class>com.demo.analyzer</class>
<jar>${appLib}</jar>
<spark-opts>--jars ${sparkLib} --files ${config},${hivesite} --num-executors ${NoOfExecutors} --executor-cores ${ExecutorCores} --executor-memory ${ExecutorMemory} --driver-memory ${driverMemory}</spark-opts>
</spark>
<ok to="sendEmail" />
<error to="fail" />
</action>

thrift_uriprincipal 可以在hive-site.xml 中找到。 thrift_uri 将在 hive-site.xml 属性中设置:

<property>
<name>hive.metastore.uris</name>
<value>thrift://xxxxxx:9083</value>
</property>

此外,主体将在 hive-site.xml 属性中设置:

 <property>
<name>hive.metastore.kerberos.principal</name>
<value>hive/_HOST@domain.COM</value>
</property>

关于apache-spark - kerberos 环境的 Oozie Spark 操作失败,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/49345746/

34 4 0
Copyright 2021 - 2024 cfsdn All Rights Reserved 蜀ICP备2022000587号
广告合作:1813099741@qq.com 6ren.com