Kylin 在 CDH 中的安装、错误解决

Apache Kylin

中文名麒（shen）麟（shou）是Hadoop动物园的重要成员。Apache Kylin是一个开源的分布式分析引擎，最初由eBay开发贡献至开源社区。它提供Hadoop之上的SQL查询接口及多维分析（OLAP）能力以支持大规模数据，能够处理TB乃至PB级别的分析任务，能够在亚秒级查询巨大的Hive表，并支持高并发。

安装

环境

CDH 5.10.0
apache-kylin-1.6.0-cdh5.7-bin

官网建议 CDH 5.10 安装 Kylin 2.0 ，尝试后发现部分查询有问题，后又换成 1.6.0 版本

下载并解压到 /opt 目录

tar zvxf apache-kylin-1.5.2.1-cdh5.7-bin.tar -C /opt

mv /opt/apache-kylin-1.6.0-cdh5.7-bin /opt/kylin-1.6.0

赋予权限

kylin 运行用户必须有，hdfs,hive,hbase 操作权限
/opt/kylin-1.6.0 以下目录，当前用户必须有读写修改权限

环境变量 /etc/profile

添加以下变量

export KYLIN_HOME=/opt/kylin-1.6.0
export HCAT_HOME=/opt/cloudera/parcels/CDH/lib/hive-hcatalog

由于使用的是 CDH 集群，其它组件环境已经配置好，只需配以上变量即可

配置 kylin.properties

# 修改
kylin.hive.client=beeline
kylin.hive.beeline.params=-n hive --hiveconf hive.security.authorization.sqlstd.confwhitelist.append='mapreduce.job.*|dfs.*' -u 'jdbc:hive2://master:10000'

# 增加
kylin.job.jar=/opt/kylin-1.6.0/lib/kylin-job-1.6.0.jar
kylin.coprocessor.local.jar=/opt/kylin-1.6.0/lib/kylin-coprocessor-1.6.0.jar
kylin.job.yarn.app.rest.check.status.url=http://master:8088/ws/v1/cluster/apps/${job_id}?anonymous=true
# kylin.job.mr.lib.dir=/opt/cloudera/parcels/CDH/lib/sentry/lib

环境检查

/opt/kylin-1.6.0/bin 目录下

./find-hive-dependency.sh

./find-hbase-dependency.sh

./check-env.sh

导入测试数据

./sample.sh

启动 Kylin

./kylin.sh start

访问 web 界面

登录后台：http://xxxxx:7070/kylin
账号密码：ADMIN/KYLIN

问题解决

HDFS 权限

dfs.permissions 设置为 false
目录权限可使用，hdfs dfs -chmod -R 777 /

kylin 2.0 版本中查询问题

./kylin.sh org.apache.kylin.storage.hbase.util.DeployCoprocessorCLI /opt/kylin-1.6.0/lib/kylin-coprocessor-1.6.0.jar all

构建 Cube 出错

如图，一直卡在构建 cube 的第三步，相应 mapreduce 任务在 map 阶段出错

Error: java.lang.ClassNotFoundException: org.apache.hadoop.hive.serde2.typeinfo.TypeInfo at java.net.URLClassLoader$1.run(URLClassLoader.java:366) at java.net.URLClassLoader$1.run(URLClassLoader.java:355) at java.security.AccessController.doPrivileged(Native Method) at java.net.URLClassLoader.findClass(URLClassLoader.java:354) at java.lang.ClassLoader.loadClass(ClassLoader.java:425) at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:308) at java.lang.ClassLoader.loadClass(ClassLoader.java:358) at java.lang.Class.forName0(Native Method) at java.lang.Class.forName(Class.java:270) at org.apache.hadoop.conf.Configuration.getClassByNameOrNull(Configuration.java:2138) at org.apache.hadoop.conf.Configuration.getClassByName(Configuration.java:2103) at org.apache.hadoop.conf.Configuration.getClass(Configuration.java:2197) at org.apache.hadoop.mapreduce.task.JobContextImpl.getInputFormatClass(JobContextImpl.java:184) at org.apache.hadoop.mapred.MapTask.runNewMapper(MapTask.java:749) at org.apache.hadoop.mapred.MapTask.run(MapTask.java:341) at org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:164) at java.security.AccessController.doPrivileged(Native Method) at javax.security.auth.Subject.doAs(Subject.java:415) at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1796) at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:158)

google 一番后,发现如下解决方案

I can still confirm this issue with kylin version 2.0 on cloudera CDH 5.11. I tried the exports (setting them to the parcel directory), but the error 
still appeared.
Adding kylin.job.mr.lib.dir=/opt/cloudera/parcels/CDH/lib/sentry/lib to the kylin.properties helped though.
Thanks for the workaround and help


Since 2.0, another workaround other than the "kylin.engine.mr.lib-dir" is set the below env vars.
export HADOOP_CONF_DIR=/etc/hadoop/conf
export HIVE_LIB=/usr/lib/hive
export HIVE_CONF=/etc/hive/conf
export HCAT_HOME=/usr/lib/hive-hcatalog
Then "bin/find-hive-dependencies.sh" should pick up the right hive jars.

最后在 kylin.properties 中添加 kylin.job.mr.lib.dir=/opt/cloudera/parcels/CDH/lib/sentry/lib 解决