Linux7 下Hadoop集群用户管理方案之五安装Hadoop集群遇到的坑

关注

原创最新推荐文章于 2026-08-09 09:39:59 发布 8k 阅读 AI 写同款· GEO 优化 ›

Hadoop 配置 Kerberos 认证_hadoop kerberos认证 rw-r–r–. 1 root root 1298 10月 5 17:14 bd_ca_cert。-rw-r–r–. 1 root root 1834 10月 5 17:14 bd_ca_key。阅读详情

自己在安装时候遇到的一部分的坑。

密码都不能用了。。

难道是因为。。改了kdc.conf?? 从ip改成端口了

PS。这个还没有解决。。Kerberos创建用户，创建的用户密码过期之后不可用的问题。Kerberos还没有完全吃透，还没解决。

/opt/cm-5.9.0/share/cmf/bin/gen_credentials.sh failed with exit code 1 and output of <<
+ export PATH=/usr/kerberos/bin:/usr/kerberos/sbin:/usr/lib/mit/sbin:/usr/sbin:/usr/lib/mit/bin:/usr/bin:/sbin:/usr/sbin:/bin:/usr/bin
+ PATH=/usr/kerberos/bin:/usr/kerberos/sbin:/usr/lib/mit/sbin:/usr/sbin:/usr/lib/mit/bin:/usr/bin:/sbin:/usr/sbin:/bin:/usr/bin
+ CMF_REALM=HADOOP.COM
+ KEYTAB_OUT=/tmp/cmf1986556804032057570.keytab
+ PRINC=yarn/slave2@HADOOP.COM
+ MAX_RENEW_LIFE=432000
+ KADMIN='kadmin -k -t /tmp/cmf817148388354671948.keytab -p cloudera-scm/admin@HADOOP.COM -r HADOOP.COM'
+ RENEW_ARG=
+ '[' 432000 -gt 0 ']'
+ RENEW_ARG='-maxrenewlife "432000 sec"'
+ '[' -z /tmp/krb59017674432558873467.conf ']'
+ echo 'Using custom config path '\''/tmp/krb59017674432558873467.conf'\'', contents below:'
+ cat /tmp/krb59017674432558873467.conf
+ kadmin -k -t /tmp/cmf817148388354671948.keytab -p cloudera-scm/admin@HADOOP.COM -r HADOOP.COM -q 'addprinc -maxrenewlife "432000 sec" -randkey yarn/slave2@HADOOP.COM'
kadmin: Incorrect password while initializing kadmin interface

>>

就是因为密码不行了。。

怎么弄一个永久的密码呢

kinit: Cannot determine realm for host (principal host/6accda1c7f91@)

Failed on local exception: java.io.IOException: java.lang.IllegalArgumentException: Server has invalid Kerberos prin

cipal: hdfs/master@HADOOP.COM; Host Details : local host is: "6accda1c7f91/172.19.0.2"; destination host is: "10.20.66.212":8020;

解决方案：需要在/etc/hosts 中配置 10.20.66.212 master

   /opt/cm-5.9.0/share/cmf/bin/gen_credentials.sh failed with exit code 1 and output of <<
  
 + export PATH=/usr/kerberos/bin:/usr/kerberos/sbin:/usr/lib/mit/sbin:/usr/sbin:/usr/lib/mit/bin:/usr/bin:/sbin:/usr/sbin:/bin:/usr/bin
  
 + PATH=/usr/kerberos/bin:/usr/kerberos/sbin:/usr/lib/mit/sbin:/usr/sbin:/usr/lib/mit/bin:/usr/bin:/sbin:/usr/sbin:/bin:/usr/bin
  
 + CMF_REALM=HADOOP.COM
  
 + KEYTAB_OUT=/tmp/cmf2885609246198664122.keytab
  
 + PRINC=yarn/slave2@HADOOP.COM
  
 + MAX_RENEW_LIFE=432000
  
 + KADMIN='kadmin -k -t /tmp/cmf8365314439897372734.keytab -p cloudera-scm/admin@HADOOP.COM -r 
  HADOOP.COM'
  
 + RENEW_ARG=
  
 + '[' 432000 -gt 0 ']'
  
 + RENEW_ARG='-maxrenewlife "432000 sec"'
  
 + '[' -z /tmp/krb54503148931985644829.conf ']'
  
 + echo 'Using custom config path '\''/tmp/krb54503148931985644829.conf'\'', contents below:'
  
 + cat /tmp/krb54503148931985644829.conf
  
 + kadmin -k -t /tmp/cmf8365314439897372734.keytab -p cloudera-scm/admin@HADOOP.COM -r 
  HADOOP.COM -q 'addprinc -maxrenewlife "432000 sec" -randkey yarn/slave2@HADOOP.COM'
  
 kadmin: Cannot contact any KDC for realm '
  HADOOP.COM' while initializing kadmin interface
  
   >>

kinit: Cannot contact any KDC for realm 'HADOOP.COM' while getting initial credentials

解决方案 1：查看/etc/krb5.conf

2: 查看/etc/hosts

:查看Kerberos服务有没有启动

启动Agent报错

.tail -200f /opt/cm-5.9.0/log/cloudera-scm-agent/cloudera-scm-agent.log

[25/Apr/2017 13:15:31 +0000] 15953 MainThread agent INFO CM server guid: 8dd2c1ca-ef6c-434f-93bf-267c0fd763c7

[25/Apr/2017 13:15:31 +0000] 15953 MainThread agent INFO Using parcels directory from server provided value: /opt/cloudera/parcels

[25/Apr/2017 13:15:31 +0000] 15953 MainThread parcel INFO Agent does create users/groups and apply file permissions

[25/Apr/2017 13:15:31 +0000] 15953 MainThread parcel_cache INFO Using /opt/cloudera/parcel-cache for parcel cache

[25/Apr/2017 13:15:31 +0000] 15953 MainThread agent ERROR Caught unexpected exception in main loop.

Traceback (most recent call last):

File "/opt/cm-5.9.0/lib64/cmf/agent/build/env/lib/python2.7/site-packages/cmf-5.9.0-py2.7.egg/cmf/agent.py", line 758, in start

self._init_after_first_heartbeat_response(resp_data)

File "/opt/cm-5.9.0/lib64/cmf/agent/build/env/lib/python2.7/site-packages/cmf-5.9.0-py2.7.egg/cmf/agent.py", line 938, in _init_after_first_heartbeat_response

self.client_configs.load()

File "/opt/cm-5.9.0/lib64/cmf/agent/build/env/lib/python2.7/site-packages/cmf-5.9.0-py2.7.egg/cmf/client_configs.py", line 682, in load

new_deployed.update(self._lookup_alternatives(fname))

File "/opt/cm-5.9.0/lib64/cmf/agent/build/env/lib/python2.7/site-packages/cmf-5.9.0-py2.7.egg/cmf/client_configs.py", line 432, in _lookup_alternatives

return self._parse_alternatives(alt_name, out)

File "/opt/cm-5.9.0/lib64/cmf/agent/build/env/lib/python2.7/site-packages/cmf-5.9.0-py2.7.egg/cmf/client_configs.py", line 444, in _parse_alternatives

path, _, _, priority_str = line.rstrip().split(" ")

ValueError: too many values to unpack

解决方案：

http://community.cloudera.com/t5/Cloudera-Manager-Installation/Freeze-while-activating-the-CDH-Parcels/td-p/47771

好像是JDK的版本问题？

https://community.cloudera.com/t5/Cloudera-Manager-Installation/Freeze-while-activating-the-CDH-Parcels/m-p/47805#M8795

需要删除OpenJDK。使用Oracle JDK。

注意source

另外。。注意 export JAVA_HOME ……犯了低级错误

无法使用pstree命令

#On Mac OS
brew install pstree
#On Fedora/Red Hat/CentOS
yum install psmisc #using psmisc package for pstree
#On Ubuntu/Debian APT
apt-get install psmisc

又一个问题。

   Failed to format NameNode

/usr/lib/hadoop-hdfs/bin/hdfs: 没有那个文件或目录

一直听在添加节点的页面。。显示delete CDH

难道因为没有用户？？

useradd --system --home=/opt/cm- 5.6 . 0 /run/cloudera-scm-server --shell=/bin/ false --comment "Cloudera SCM User" cloudera-scm

chown cloudera-scm:cloudera-scm /opt/cloudera/parcels

mkdir -p /opt/cloudera/parcel-repo

chown cloudera-scm:cloudera-scm /opt/cloudera/parcel-repo

mkdir -p /opt/cloudera/parcels

chown cloudera-scm:cloudera-scm /opt/cloudera/parcels

解决方案：

果然是没有用户，还有就是免密登陆

另外就是把agent节点上 parcel-repo 目录删掉。

把parcles目录中的残留数据清空

启动Hive的时候 message:Version information not found in metastore

Parcel not distribute d but have active state ACTIVATING

可以尝试把 /opt/cloudera/parcels 下的残留文件删除

Failed to format namenode

在每个节点可以尝试创建cloudera-scm用户另外设置对应的权限。

新的一个Hadoop的机器。访问总是跟我说

Server has invalid Kerberos principal: hdfs/edu-bigdata-01.novalocal@HADOOP.COM; Host Details : local host is: "shihys-MacBook-Pro.local/10.20.5.172"; destination host is: "10.10.106.156":8020;

想起来没有设置 /etc/hosts

设置上之后。。

又出新错。

Failed on local exception: java.io.IOException: Couldn't setup connection for hdfs@HADOOP.COM to edu-bigdata-01.novalocal/10.10.106.156:8020; Host Details : local host is: "shihys-MacBook-Pro.local/10.20.5.174"; destination host is: "edu-bigdata-01.novalocal":8020;

[Caused by GSSException: No valid credentials provided (Mechanism level: Failed to find any Kerberos tgt)]

kinit: krb5_get_init_creds: time skew (43202) larger than max (300)

时间不对。相差太多

我去了。。。竟然是因为Kerberos 时间跟本地时间相差太多。。

终于找到原因了。。晕死了。。

主要需要看一下时区。。

EDU CST。。。

date -U 看UTC时间。。最准了。

收录于