Hive

# 基于 Ambari 安装 Hive 高可用

# 方案一：安装时直接选择高可用

在 Hive 的安装过程中，第一种方案是直接在安装时选择高可用配置。此方案非常适合于从一开始就设置高可用模式。

提示

注: 对于这种配置方式，具体的安装流程不再赘述，请参考本站其他文章了解更多关于 Hive 安装的详细内容。

选择高可用安装模式

# 方案二：安装后通过添加服务实现高可用

如果你选择了安装简版的 Hive，而后又需要配置高可用，可以按照以下步骤进行：

# 1. 等待服务启动并遇到报错

由于 Ambari 在默认安装过程中没有初始化多节点的 Hive schema，可能会在启动时遇到错误。

启动报错

错误信息:

0: jdbc:mysql://hadoop1/hive> CREATE TABLE IF NOT EXISTS `COLUMNS_V2` ( `CD_ID`  
bigint(20) NOT NULL, `COMMENT` varchar(256) CHARACTER SET latin1 COLLATE latin1_ 
bin DEFAULT NULL, `COLUMN_NAME` varchar(767) CHARACTER SET latin1 COLLATE latin1 
_bin NOT NULL, `TYPE_NAME` MEDIUMTEXT DEFAULT NULL, `INTEGER_IDX` int(11) NOT NU 
LL, PRIMARY KEY (`CD_ID`,`COLUMN_NAME`), KEY `COLUMNS_V2_N49` (`CD_ID`), CONSTRA 
INT `COLUMNS_V2_FK1` FOREIGN KEY (`CD_ID`) REFERENCES `CDS` (`CD_ID`) ) ENGINE=I 
nnoDB DEFAULT CHARSET=latin1
No rows affected (0.01 seconds)
0: jdbc:mysql://hadoop1/hive> /*!40101 SET character_set_client = @saved_cs_clie 
nt */
No rows affected (0.001 seconds)
0: jdbc:mysql://hadoop1/hive> /*!40101 SET @saved_cs_client     = @@character_se 
t_client */
No rows affected (0.001 seconds)
0: jdbc:mysql://hadoop1/hive> /*!40101 SET character_set_client = utf8 */
No rows affected (0 seconds)
0: jdbc:mysql://hadoop1/hive> CREATE TABLE IF NOT EXISTS `DATABASE_PARAMS` ( `DB 
_ID` bigint(20) NOT NULL, `PARAM_KEY` varchar(180) CHARACTER SET latin1 COLLATE  
latin1_bin NOT NULL, `PARAM_VALUE` varchar(4000) CHARACTER SET latin1 COLLATE la 
tin1_bin DEFAULT NULL, PRIMARY KEY (`DB_ID`,`PARAM_KEY`), KEY `DATABASE_PARAMS_N 
49` (`DB_ID`), CONSTRAINT `DATABASE_PARAMS_FK1` FOREIGN KEY (`DB_ID`) REFERENCES 
 `DBS` (`DB_ID`) ) ENGINE=InnoDB DEFAULT CHARSET=latin1
No rows affected (0.01 seconds)
0: jdbc:mysql://hadoop1/hive> /*!40101 SET character_set_client = @saved_cs_clie 
nt */
No rows affected (0 seconds)
0: jdbc:mysql://hadoop1/hive> CREATE TABLE `CTLGS` ( `CTLG_ID` BIGINT PRIMARY KE 
Y, `NAME` VARCHAR(256), `DESC` VARCHAR(4000), `LOCATION_URI` VARCHAR(4000) NOT N 
ULL, UNIQUE KEY `UNIQUE_CATALOG` (`NAME`) ) ENGINE=InnoDB DEFAULT CHARSET=latin1 
Error: Table 'CTLGS' already exists (state=42S01,code=1050)
Closing: 0: jdbc:mysql://hadoop1/hive
org.apache.hadoop.hive.metastore.HiveMetaException: Schema initialization FAILED! Metastore state would be inconsistent !!
Underlying cause: java.io.IOException : Schema script failed, errorcode 2
org.apache.hadoop.hive.metastore.HiveMetaException: Schema initialization FAILED! Metastore state would be inconsistent !!
	at org.apache.hive.beeline.HiveSchemaTool.doInit(HiveSchemaTool.java:594)
	at org.apache.hive.beeline.HiveSchemaTool.doInit(HiveSchemaTool.java:567)
	at org.apache.hive.beeline.HiveSchemaTool.main(HiveSchemaTool.java:1517)
	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 org.apache.hadoop.util.RunJar.run(RunJar.java:323)
	at org.apache.hadoop.util.RunJar.main(RunJar.java:236)
Caused by: java.io.IOException: Schema script failed, errorcode 2
	at org.apache.hive.beeline.HiveSchemaTool.runBeeLine(HiveSchemaTool.java:1226)
	at org.apache.hive.beeline.HiveSchemaTool.runBeeLine(HiveSchemaTool.java:1204)
	at org.apache.hive.beeline.HiveSchemaTool.doInit(HiveSchemaTool.java:590)
	... 8 more
*** schemaTool failed ***