hdp 整合impala
访问量:1613
创建时间:2020-06-10
建议，我这里hdp为2.6.5,impala安装2.12版本的报错方法找不到，查找相关方法在高版本hadoop中存在，hdp2.6.5中没有，安装impala2.4,后正常运行。
配置yum源
[root@ecs-hdp-dev-ambari ~]# yum install nginx
[root@ecs-hdp-dev-ambari ~]# vim /etc/nginx/nginx.conf
root /var/www/html; #修改root目录
#配置cdh yum源
[root@ecs-hdp-dev-ambari html]# vim /etc/yum.repos.d/cdh.repo
[cloudera-cdh5]
# Packages for Cloudera's Distribution for Hadoop, Version 5, on RedHat or CentOS 7 x86_64
name=Cloudera's Distribution for Hadoop, Version 5
baseurl=https://archive.cloudera.com/cdh5/redhat/7/x86_64/cdh/5/
gpgkey =https://archive.cloudera.com/cdh5/redhat/7/x86_64/cdh/RPM-GPG-KEY-cloudera
gpgcheck = 1
#reposync过程较慢，可以使用screen同步
[root@ecs-hdp-dev-ambari html]# reposync cloudera-cdh5
[root@ecs-hdp-dev-ambari html]# createrepo cloudera-cdh5/
Spawning worker 0 with 31 pkgs
Spawning worker 1 with 31 pkgs
Spawning worker 2 with 31 pkgs
Spawning worker 3 with 31 pkgs
Workers Finished
Saving Primary metadata
Saving file lists metadata
Saving other metadata
Generating sqlite DBs
Sqlite DBs complete
#添加所有hdp集群机器的yum源（在所有机器上执行）
[root@name1 html]# vim /etc/yum.repos.d/cdh.repo
[cloudera-cdh5]
# Packages for Cloudera's Distribution for Hadoop, Version 5, on RedHat or CentOS 7 x86_64
name=Cloudera's Distribution for Hadoop, Version 5
baseurl=http://172.16.96.62/cloudera-cdh5/
gpgcheck = 0
enabled=1
[root@ecs-hdp-dev-ambari html]# systemctl start nginx
配置ambari
#查看集群版本
[root@name1 html]# hdp-select status hadoop-client
hadoop-client - 2.6.5.0-292
#进入对应ambari版本的hdp目录
[root@ecs-hdp-dev-ambari html]# cd /var/lib/ambari-server/resources/stacks/HDP/2.6/services
[root@ecs-hdp-dev-ambari services]# git clone https://github.com/cas-bigdatalab/ambari-impala-service.git IMPALA
Cloning into 'IMPALA'...
remote: Enumerating objects: 283, done.
remote: Total 283 (delta 0), reused 0 (delta 0), pack-reused 283
Receiving objects: 100% (283/283), 516.21 KiB | 56.00 KiB/s, done.
Resolving deltas: 100% (142/142), done.
#重启ambari服务
[root@ecs-hdp-dev-ambari services]# ambari-server restart
通过ambari控制台安装impala
在Install, Start and Test是各种key error，配置ssh无密码访问。后面就是next即可
处理启动故障 1
查看日志：
[root@name2 impala]# pwd
/var/log/impala
[root@name2 impala]# cat impala-state-store.log
Unable to find Java. JAVA_HOME should be set in /etc/default/bigtop-utils
Unable to find Java. JAVA_HOME should be set in /etc/default/bigtop-utils
```shell
[root@node3 ~]# vim /etc/default/bigtop-utils
export JAVA_HOME=/usr/jdk64/jdk1.8.0_112
处理启动故障 2
查看impala deamon节点日志：
[root@node2 impala]# tail -f impalad.ERROR
Log file created at: 2020/06/10 14:17:05
Running on machine: node2.hdp.work.com
Log line format: [IWEF]mmdd hh:mm:ss.uuuuuu threadid file:line] msg
E0610 14:17:05.607198 25912 logging.cc:121] stderr will be logged to this file.
log4j:WARN No appenders could be found for logger (org.apache.hadoop.metrics2.lib.MutableMetricsFactory).
log4j:WARN Please initialize the log4j system properly.
log4j:WARN See http://logging.apache.org/log4j/1.2/faq.html#noconfig for more info.
E0610 14:17:12.706774 25912 authentication.cc:176] SASL message (Kerberos (internal)): No worthy mechs found
E0610 14:17:15.709295 25912 authentication.cc:176] SASL message (Kerberos (internal)): No worthy mechs found
E0610 14:17:36.721410 25912 authentication.cc:176] SASL message (Kerberos (internal)): No worthy mechs found
E0610 14:17:39.721897 25912 authentication.cc:176] SASL message (Kerberos (internal)): No worthy mechs found
E0610 14:17:39.722074 25912 impalad-main.cc:92] Impalad services did not start correctly, exiting. Error: Couldn't open transport for name2.hdp.work.com:24000 (SASL(-4): no mechanism available: No worthy mechs found)
Statestore subscriber did not start up.
[root@node2 impala]# tail -f impalad.INFO
E0610 14:14:19.483218 15485 authentication.cc:176] SASL message (Kerberos (internal)): No worthy mechs found
I0610 14:14:19.483361 15485 thrift-client.cc:94] Unable to connect to name2.hdp.work.com:24000
I0610 14:14:19.483366 15485 thrift-client.cc:100] (Attempt 9 of 10)
E0610 14:14:22.491096 15485 authentication.cc:176] SASL message (Kerberos (internal)): No worthy mechs found
I0610 14:14:22.491261 15485 thrift-client.cc:94] Unable to connect to name2.hdp.work.com:24000
I0610 14:14:22.491269 15485 thrift-client.cc:100] (Attempt 10 of 10)
I0610 14:14:22.491289 15485 statestore-subscriber.cc:238] statestore registration unsuccessful: Couldn't open transport for name2.hdp.work.com:24000 (SASL(-4): no mechanism available: No worthy mechs found)
E0610 14:14:22.495311 15485 impalad-main.cc:92] Impalad services did not start correctly, exiting. Error: Couldn't open transport for name2.hdp.work.com:24000 (SASL(-4): no mechanism available: No worthy mechs found)
Statestore subscriber did not start up.
#安装依赖包，在ambari中重新启动对应进程
[root@node2 impala]# yum install cyrus-sasl-plain cyrus-sasl-devel cyrus-sasl-gssapi cyrus-sasl-md5
hue配置impala
修改hue的配置文件，我这里的hue版本是4.7.1,3版本应该也可以（docker安装可以进入docker修改）
[root@name1 hue-4.7.1]# pwd
/hadoop/hue/hue-4.7.1
[root@name1 hue-4.7.1]# vim desktop/conf/hue.ini
[impala]
server_host=name2.work.com
server_port=21050
#如果你的hadoop集群没有开启kerberos，下面2项目不用配置。
use_sasl=true
impala_principal=impala/_HOST@work.com
#配置说明server_host，你的imapala安装的机器，server_port，impala监听端口，impala_principal=impala/_HOST@xxx特定格式,xxx替换为你的kdc realms
在hue中简单测试impala是否正常：
故障处理3,hue中select时报错如下（2.6版本impala报错）：
AnalysisException: Failed to load metadata for table: 'customeraccounthisdetail' CAUSED BY: TableLoadingException: java.lang.NoClassDefFoundError: Could not initialize class com.cloudera.impala.catalog.HBaseTable CAUSED BY: ExecutionException: java.lang.NoClassDefFoundError: Could not initialize class com.cloudera.impala.catalog.HBaseTable CAUSED BY: NoClassDefFoundError: Could not initialize class com.cloudera.impala.catalog.HBaseTable
或者：
AnalysisException: java.lang.NoClassDefFoundError: org/apache/hadoop/hbase/HBaseConfiguration CAUSED BY: ExecutionException: java.lang.NoClassDefFoundError: org/apache/hadoop/hbase/HBaseConfiguration CAUSED BY: NoClassDefFoundError: org/apache/hadoop/hbase/HBaseConfiguration CAUSED BY: ClassNotFoundException: org.apache.hadoop.hbase.HBaseConfiguration CAUSED BY: TableLoadingException: java.lang.NoClassDefFoundError: org/apache/hadoop/hbase/HBaseConfiguration CAUSED BY: ExecutionException: java.lang.NoClassDefFoundError: org/apache/hadoop/hbase/HBaseConfiguration CAUSED BY: NoClassDefFoundError: org/apache/hadoop/hbase/HBaseConfiguration CAUSED BY: ClassNotFoundException: org.apache.hadoop.hbase.HBaseConfiguration
处理如下：（注意impala要配合cdh的hbase的jar包，将cdh的hbase的对应版本的jar包cp过来，创建软连接，下面这个是用hdp的hbase包，启动会报错，过程只作参考，用cdh的hbase的4个jar包）
[root@client1 ~]# cp /usr/hdp/2.6.5.0-292/hbase/lib/hbase-annotations-1.1.2.2.6.5.0-292.jar /usr/lib/impala/lib/
[root@client1 ~]# cp /usr/hdp/2.6.5.0-292/hbase/lib/hbase-client-1.1.2.2.6.5.0-292.jar /usr/lib/impala/lib/
[root@client1 ~]# cp /usr/hdp/2.6.5.0-292/hbase/lib/hbase-common-1.1.2.2.6.5.0-292.jar /usr/lib/impala/lib/
[root@client1 ~]# cp /usr/hdp/2.6.5.0-292/hbase/lib/hbase-protocol-1.1.2.2.6.5.0-292.jar /usr/lib/impala/lib/
[root@client1 ~]# cd !$
cd /usr/lib/impala/lib/
[root@client1 lib]# ll hbase-*
-rw-r--r-- 1 root root 20807 Jun 16 15:01 hbase-annotations-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 36 Jun 16 13:38 hbase-annotations.jar -> /usr/lib/hbase/hbase-annotations.jar
-rw-r--r-- 1 root root 1407871 Jun 16 15:01 hbase-client-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 31 Jun 16 13:38 hbase-client.jar -> /usr/lib/hbase/hbase-client.jar
-rw-r--r-- 1 root root 577452 Jun 16 15:01 hbase-common-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 31 Jun 16 13:38 hbase-common.jar -> /usr/lib/hbase/hbase-common.jar
-rw-r--r-- 1 root root 4988181 Jun 16 15:01 hbase-protocol-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 33 Jun 16 13:38 hbase-protocol.jar -> /usr/lib/hbase/hbase-protocol.jar
[root@client1 lib]# ln -s hbase-annotations-1.1.2.2.6.5.0-292.jar hbase-annotations.jar
ln: failed to create symbolic link ‘hbase-annotations.jar’: File exists
[root@client1 lib]# rm hbase-annotations.jar
rm: remove symbolic link ‘hbase-annotations.jar’? y
[root@client1 lib]# rm hbase-client.jar
rm: remove symbolic link ‘hbase-client.jar’? y
[root@client1 lib]# rm hbase-common.jar
rm: remove symbolic link ‘hbase-common.jar’? y
[root@client1 lib]# rm hbase-protocol.jar
rm: remove symbolic link ‘hbase-protocol.jar’? y
[root@client1 lib]# ln -s hbase-annotations-1.1.2.2.6.5.0-292.jar hbase-annotations.jar
[root@client1 lib]# ln -s hbase-client-1.1.2.2.6.5.0-292.jar hbase-client.jar
[root@client1 lib]# ln -s hbase-common-1.1.2.2.6.5.0-292.jar hbase-common.jar
[root@client1 lib]# ln -s hbase-protocol-1.1.2.2.6.5.0-292.jar hbase-protocol.jar
[root@client1 lib]# ll hbase-*
-rw-r--r-- 1 root root 20807 Jun 16 15:01 hbase-annotations-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 39 Jun 16 15:02 hbase-annotations.jar -> hbase-annotations-1.1.2.2.6.5.0-292.jar
-rw-r--r-- 1 root root 1407871 Jun 16 15:01 hbase-client-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 34 Jun 16 15:03 hbase-client.jar -> hbase-client-1.1.2.2.6.5.0-292.jar
-rw-r--r-- 1 root root 577452 Jun 16 15:01 hbase-common-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 34 Jun 16 15:03 hbase-common.jar -> hbase-common-1.1.2.2.6.5.0-292.jar
-rw-r--r-- 1 root root 4988181 Jun 16 15:01 hbase-protocol-1.1.2.2.6.5.0-292.jar
lrwxrwxrwx 1 root root 36 Jun 16 15:03 hbase-protocol.jar -> hbase-protocol-1.1.2.2.6.5.0-292.jar
[root@client1 lib]#
#重启impala
故障处理4,No usable temporary files: space could not be allocated on any tempor
处理方法一：
[root@hdpprdd02 ~]# ll /tmp/impala_scratch
ls: cannot access /tmp/impala_scratch: No such file or directory
[root@hdpprdd02 ~]# ll /hadoop01/tmp/impala_scratch
total 0
[root@hdpprdd02 ~]# mkdir -pv /tmp/impala_scratch
mkdir: created directory ‘/tmp/impala_scratch’
[root@hdpprdd02 ~]# chmod 777 /tmp/impala_scratch
处理方法二：
修改impala的配置文件/etc/default/impala（注意本人尝试在hdp的ambari中修改，但是不生效。）
故障处理5,Memory limit exceeded
处理方法：修改相关内存参数。
修改impala的配置文件/etc/default/impala（注意本人尝试在hdp的ambari中修改，但是不生效）
登陆评论:
使用GITHUB登陆