[
https://issues.apache.org/jira/browse/SPARK-31551?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yuqi Wang updated SPARK-31551:
------------------------------
Description:
See current *createSparkUser*:
[https://github.com/apache/spark/blob/263f04db865920d9c10251517b00a1b477b58ff1/core/src/main/scala/org/apache/spark/deploy/SparkHadoopUtil.scala#L66-L76]
{code:java}
def createSparkUser(): UserGroupInformation = {
val user = Utils.getCurrentUserName()
logDebug("creating UGI for user: " + user)
val ugi = UserGroupInformation.createRemoteUser(user)
transferCredentials(UserGroupInformation.getCurrentUser(), ugi)
ugi
}
def transferCredentials(source: UserGroupInformation, dest:
UserGroupInformation): Unit = {
dest.addCredentials(source.getCredentials())
}
def getCurrentUserName(): String = {
Option(System.getenv("SPARK_USER"))
.getOrElse(UserGroupInformation.getCurrentUser().getShortUserName())
}
{code}
The *transferCredentials* func can only transfer Hadoop creds such as
Delegation Tokens.
However, other creds stored in UGI.subject.getPrivateCredentials, will be lost
here, such as:
# Non-Hadoop creds:
Such as, Kafka creds,
# Newly supported or 3rd party supported Hadoop creds:
Such as to support OAuth/JWT token authn on Hadoop, we need to store the
OAuth/JWT token into UGI.subject.getPrivateCredentials. However, these tokens
are not supposed to be managed by Hadoop Credentials (currently it is only for
Hadoop secret keys and delegation tokens)
Another issue is that the *getCurrentUserName* only returns the
getShortUserName of the user, which may lost the user's fully qualified user
name that need to be passed to PRC server (such as YARN, HDFS, Kafka). We
should better use the *getUserName* to get fully qualified user name in our
client side. This is aligned to
*[HADOOP_PROXY_USER|https://github.com/apache/hadoop/blob/30ef8d0f1a1463931fe581a46c739dad4c8260e4/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/UserGroupInformation.java#L720]*
was:
Current createRemoteUser:
[https://github.com/apache/spark/blob/263f04db865920d9c10251517b00a1b477b58ff1/core/src/main/scala/org/apache/spark/deploy/SparkHadoopUtil.scala#L66-L76]
{code:java}
def createSparkUser(): UserGroupInformation = {
val user = Utils.getCurrentUserName()
logDebug("creating UGI for user: " + user)
val ugi = UserGroupInformation.createRemoteUser(user)
transferCredentials(UserGroupInformation.getCurrentUser(), ugi)
ugi
}
def transferCredentials(source: UserGroupInformation, dest:
UserGroupInformation): Unit = {
dest.addCredentials(source.getCredentials())
}
def getCurrentUserName(): String = {
Option(System.getenv("SPARK_USER"))
.getOrElse(UserGroupInformation.getCurrentUser().getShortUserName())
}
{code}
The transferCredentials func can only transfer Hadoop creds such as Delegation
Tokens.
However, other creds stored in UGI.subject.getPrivateCredentials, will be lost
here, such as:
1. Non-Hadoop creds:
Such as, Kafka creds,
https://github.com/apache/kafka/blob/f3c8bff311b0e4c4d0e316ac949fe4491f9b107f/clients/src/main/java/org/apache/kafka/common/security/oauthbearer/OAuthBearerLoginModule.java#L395
2. Customized Hadoop creds:
Such as support OAuth/JWT token authn on Hadoop, we need to store the
OAuth/JWT token into UGI.subject.getPrivateCredentials. However, these tokens
are not supposed to be managed by Hadoop Credentials (currently only for Hadoop
secret keys and delegation Tokens)
Another issue is that the getCurrentUserName only returns the getShortUserName
of the user, which may lost the user's fully qualified user name that need to
be passed to PRC server (such as YARN, HDFS, Kafka). We should use getUserName
to get fully qualified user name in our client side.
> createSparkUser lost user's non-Hadoop credentials and fully qualified user
> name
> --------------------------------------------------------------------------------
>
> Key: SPARK-31551
> URL: https://issues.apache.org/jira/browse/SPARK-31551
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 2.4.4, 2.4.5
> Reporter: Yuqi Wang
> Priority: Major
>
> See current *createSparkUser*:
> [https://github.com/apache/spark/blob/263f04db865920d9c10251517b00a1b477b58ff1/core/src/main/scala/org/apache/spark/deploy/SparkHadoopUtil.scala#L66-L76]
> {code:java}
> def createSparkUser(): UserGroupInformation = {
> val user = Utils.getCurrentUserName()
> logDebug("creating UGI for user: " + user)
> val ugi = UserGroupInformation.createRemoteUser(user)
> transferCredentials(UserGroupInformation.getCurrentUser(), ugi)
> ugi
> }
> def transferCredentials(source: UserGroupInformation, dest:
> UserGroupInformation): Unit = {
> dest.addCredentials(source.getCredentials())
> }
> def getCurrentUserName(): String = {
> Option(System.getenv("SPARK_USER"))
> .getOrElse(UserGroupInformation.getCurrentUser().getShortUserName())
> }
> {code}
> The *transferCredentials* func can only transfer Hadoop creds such as
> Delegation Tokens.
> However, other creds stored in UGI.subject.getPrivateCredentials, will be
> lost here, such as:
> # Non-Hadoop creds:
> Such as, Kafka creds,
> # Newly supported or 3rd party supported Hadoop creds:
> Such as to support OAuth/JWT token authn on Hadoop, we need to store the
> OAuth/JWT token into UGI.subject.getPrivateCredentials. However, these tokens
> are not supposed to be managed by Hadoop Credentials (currently it is only
> for Hadoop secret keys and delegation tokens)
> Another issue is that the *getCurrentUserName* only returns the
> getShortUserName of the user, which may lost the user's fully qualified user
> name that need to be passed to PRC server (such as YARN, HDFS, Kafka). We
> should better use the *getUserName* to get fully qualified user name in our
> client side. This is aligned to
> *[HADOOP_PROXY_USER|https://github.com/apache/hadoop/blob/30ef8d0f1a1463931fe581a46c739dad4c8260e4/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/UserGroupInformation.java#L720]*
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
