Linux7 下Hadoop集群用户管理方案之五 安装Hadoop集群遇到的坑
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/[email protected] + MAX_RENEW_LIFE=432000 + KADMIN='kadmin -k -t /tmp/cmf817148388354671948.keytab -p cloudera-scm/[email protected] -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/[email protected] -r HADOOP.COM -q 'addprinc -maxrenewlife "432000 sec" -randkey yarn/[email protected]' 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/[email protected]; 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/[email protected]
+ MAX_RENEW_LIFE=432000
+ KADMIN='kadmin -k -t /tmp/cmf8365314439897372734.keytab -p cloudera-scm/[email protected] -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/[email protected] -r HADOOP.COM -q 'addprinc -maxrenewlife "432000 sec" -randkey yarn/[email protected]'
kadmin: Cannot contact any KDC for realm ' HADOOP.COM' while initializing kadmin interface
+ 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/[email protected]
+ MAX_RENEW_LIFE=432000
+ KADMIN='kadmin -k -t /tmp/cmf8365314439897372734.keytab -p cloudera-scm/[email protected] -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/[email protected] -r HADOOP.COM -q 'addprinc -maxrenewlife "432000 sec" -randkey yarn/[email protected]'
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/[email protected]; 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 [email protected] 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。