ECS: 네트워크 분리를 사용하여 ECS에서 Hadoop 구성

Summary: 네트워크 분리를 사용하여 ECS에서 Hadoop을 구성합니다.

This article applies to This article does not apply to This article is not tied to any specific product. Not all product versions are identified in this article.
Check out other resources

Instructions

요약:     
ECS 클러스터가 관리 및 데이터 트래픽을 분할하기 위해 네트워크를 분리하도록 구성된 경우 Hadoop 노드가 트래픽을 '데이터' IP 주소로 보낼 가능성이 높습니다.

Hadoop에 대한 Kerberos 인증을 구성할 때 사용자는 'mgmt' 호스트 이름에 대한 보안 주체 keytab을 구성할 수 있습니다.

Active Directory(또는 LDAP)는 요청을 인증하려고 할 때 ECS 노드 IP 주소의 역방향 조회를 수행하여 호스트 이름을 확인합니다. 이렇게 하면 '데이터' IP가 전달되고 'mgmt' 호스트 이름에 대해 구성되었기 때문에 원하는 keytab을 사용하여 인증하지 못합니다.

지침:    
이 문제가 있는지 확인하려면 Hadoop 클라이언트 오류 및 ECS dataheadsvc를 검토하십시오.   

Hadoop 클라이언트 오류 예:      
20/10/12 15:57:27 DEBUG vipr.ClientConnection: Read StatRequestMessage[kind=STAT_REQUEST,namespace=ns-datalake,bucket=ns-datalake-s3-bucket-ifrs17,path=/,hdfsTrustedStatus=HDFS_USER_TRUSTED] from '10.20.46.71'
20/10/12 15:57:27 DEBUG vipr.ClientConnection: The Message request kind from the client is RESPONSE The registered Socket Address is: /10.20.46.71 and port is 9040
20/10/12 15:57:27 DEBUG vipr.ViPRFileSystemClientBase: Generic failure for request: User: hdfs@HDFS.COM (auth:KERBEROS), host: ns-datalake-s3-bucket-ifrs17.ns-datalake.federation1, namespace: ns-datalake, bucket: ns-datalake-s3-bucket-ifrs17
20/10/12 15:57:27 DEBUG vipr.ViPRFileSystemClientBase: Request message sent: StatRequestMessage[kind=STAT_REQUEST,namespace=ns-datalake,bucket=ns-datalake-s3-bucket-ifrs17,path=/,hdfsTrustedStatus=HDFS_USER_TRUSTED]
20/10/12 15:57:27 DEBUG vipr.ViPRFileSystemClientBase: Reply message received: ResponseMessage[kind=RESPONSE,status=ERROR(100)]
20/10/12 15:57:27 DEBUG util.ExceptionHelper: ExceptionHelper l_method checkResponse, l_errstr ERROR_INSUFFICIENT_PERMISSIONS
java.lang.Exception: Stack Trace
        at com.emc.hadoop.fs.vipr.util.ExceptionHelper.throwException(ExceptionHelper.java:110)
        at com.emc.hadoop.fs.vipr.ViPRFileSystemClientBase.checkResponse(ViPRFileSystemClientBase.java:988)
        at com.emc.hadoop.fs.vipr.ViPRFileSystemClientBase.sendFileStatusRequest(ViPRFileSystemClientBase.java:1008)
        at com.emc.hadoop.fs.vipr.ViPRFileSystemClientBase.stat(ViPRFileSystemClientBase.java:654)
        at com.emc.hadoop.fs.vipr.ViPRFileSystemBase.getFileStatus(ViPRFileSystemBase.java:313)
        at org.apache.hadoop.fs.Globber.getFileStatus(Globber.java:65)
        at org.apache.hadoop.fs.Globber.doGlob(Globber.java:294)
        at org.apache.hadoop.fs.Globber.glob(Globber.java:149)
        at org.apache.hadoop.fs.FileSystem.globStatus(FileSystem.java:2016)
        at org.apache.hadoop.fs.shell.PathData.expandAsGlob(PathData.java:353)
        at org.apache.hadoop.fs.shell.Command.expandArgument(Command.java:250)
        at org.apache.hadoop.fs.shell.Command.expandArguments(Command.java:233)
        at org.apache.hadoop.fs.shell.FsCommand.processRawArguments(FsCommand.java:104)
        at org.apache.hadoop.fs.shell.Command.run(Command.java:177)
        at org.apache.hadoop.fs.FsShell.run(FsShell.java:328)
        at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:76)
        at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:90)
        at org.apache.hadoop.fs.FsShell.main(FsShell.java:391)
ls: ViPRFS internal error (ERROR).
20/10/12 15:57:27 DEBUG util.ShutdownHookManager: Completed shutdown in 0.003 seconds; Timeouts: 0
20/10/12 15:57:27 DEBUG util.ShutdownHookManager: ShutdownHookManger completed shutdown.

 

ECS dataheadsvc:      
2020-10-12T08:51:17,178 [pool-103-thread-1] ERROR  TokenAuthenticator.java (line 100) Exception in authenticating user
javax.security.auth.login.LoginException: Client not found in Kerberos database (6)
        at com.sun.security.auth.module.Krb5LoginModule.attemptAuthentication(Krb5LoginModule.java:804)
        at com.sun.security.auth.module.Krb5LoginModule.login(Krb5LoginModule.java:617)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:498)
        at javax.security.auth.login.LoginContext.invoke(LoginContext.java:755)
        at javax.security.auth.login.LoginContext.access$000(LoginContext.java:195)
        at javax.security.auth.login.LoginContext$4.run(LoginContext.java:682)
        at javax.security.auth.login.LoginContext$4.run(LoginContext.java:680)
        at java.security.AccessController.doPrivileged(Native Method)
        at javax.security.auth.login.LoginContext.invokePriv(LoginContext.java:680)
        at javax.security.auth.login.LoginContext.login(LoginContext.java:587)
        at com.emc.vipr.hdfs.auth.TokenAuthenticator.authenticateClientPrincipal(TokenAuthenticator.java:81)
        at com.emc.vipr.hdfs.fs.RequestProcessor.getPrincipal(RequestProcessor.java:2200)
        at com.emc.vipr.hdfs.fs.RequestProcessor.accept(RequestProcessor.java:195)
        at com.emc.vipr.hdfs.net.ConnectionManager$RequestThread.run(ConnectionManager.java:141)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748)
Caused by: KrbException: Client not found in Kerberos database (6)
        at sun.security.krb5.KrbAsRep.<init>(KrbAsRep.java:76)
        at sun.security.krb5.KrbAsReqBuilder.send(KrbAsReqBuilder.java:316)
        at sun.security.krb5.KrbAsReqBuilder.action(KrbAsReqBuilder.java:361)
        at com.sun.security.auth.module.Krb5LoginModule.attemptAuthentication(Krb5LoginModule.java:776)
        ... 19 more
Caused by: KrbException: Identifier doesn't match expected value (906)
        at sun.security.krb5.internal.KDCRep.init(KDCRep.java:140)
        at sun.security.krb5.internal.ASRep.init(ASRep.java:64)
        at sun.security.krb5.internal.ASRep.<init>(ASRep.java:59)
        at sun.security.krb5.KrbAsRep.<init>(KrbAsRep.java:60)
        ... 22 more
2020-10-12T08:51:17,217 [pool-103-thread-1] ERROR  RequestProcessor.java (line 2110) Request from 10.20.136.117 : ERROR : StatRequestMessage[kind=STAT_REQUEST,namespace=ns-datalake,bucket=ns-datalake-s3-bucket-ifrs17,path=/,hdfsTrustedStatus=HDFS_USER_TRUSTED]

 

해결 방법:     
해결 방법은 Active Directory 호스트에서 setspn을 사용하여 'mgmt' 기반 보안 주체에 대한 별칭을 생성하여 'data' 기반 사용자를 포함하는 것입니다.

예:
ECS에는 두 개의 IP 주소가 있습니다. 데이터의 경우 10.10.10.1, 관리의 경우 10.10.10.2입니다.
데이터 IP에 대한 DNS 조회가 ecsdata1.corp.com 이고 관리 IP에 대한 DNS 조회가 ecsmgmt1.corp.com 입니다. 
노드의 호스트 이름이 'ecsmgmt1.corp.com'입니다.
  
setspn -U -S vipr/ecsmgmt1.corp.com ecsdata1.corp.com
setspn -U -S vipr/ecsmgmt2.corp.com ecsdata2.corp.com
setspn -U -S vipr/ecsmgmt3.corp.com ecsdata3.corp.com
setspn -U -S vipr/ecsmgmt4.corp.com ecsdata4.corp.com
setspn -U -S vipr/ecsmgmt5.corp.com ecsdata5.corp.com
setspn -U -S vipr/ecsmgmt6.corp.com ecsdata6.corp.com

 

setspn -Q 를 사용 하 여 올바른지 확인 합니다.      
setspn -Q vipr/ecsmgmt1.corp.com

Checking domain DC=atradiusnet,DC=com
CN=vipr/gbcdfecsdata1.atradiusnet.com,OU=HDP_DEV,OU=Dedicated,OU=GroupWide,DC=atradiusnet,DC=com
        vipr/ecsmgmt1.corp.com
        vipr/ecsdata1.corp.com


이제 Hadoop 노드에서 보낸 요청이 'ecsdata1.corp.com' IP 주소를 사용하더라도 Active Directory가 'vipr/ecsmgmt1.corp.com@CORP.COM' 보안 주체를 인증합니다. 

Affected Products

ECS Appliance

Products

ECS Appliance, ECS Appliance Hardware Series
Article Properties
Article Number: 000180919
Article Type: How To
Last Modified: 02 Jul 2025
Version:  5
Find answers to your questions from other Dell users
Support Services
Check if your device is covered by Support Services.