无法读取hive外部表

版本:2.5.2
HIVE版本:2.1.1
想知道创建外部表或catalog时,能否指定hive版本?

starrocks上创建了hive外部表,可以用show database / show tables 看到库和表,explain select * from xxx limit 1也能看到执行计划,但一执行sql语句就卡到超时;

我怀疑是不是hive客户端不兼容的问题?
P.S:在doris上指定了hive version就能成功读取数据 :joy:

#这是 show tables 命令,hadoop-cmf-hive-HIVEMETASTORE里的日志
2023-04-24 10:10:48,337 INFO  org.apache.hadoop.hive.metastore.HiveMetaStore: [pool-8-thread-198]: 182: source:10.207.xx.xx get_all_tables: db=test
2023-04-24 10:10:48,338 INFO  org.apache.hadoop.hive.metastore.HiveMetaStore.audit: [pool-8-thread-198]: ugi=root	ip=10.207.xx.xx	cmd=source:10.207.xx.xx get_all_tables: db=test	
#这是 explain select * from product limit 1 的hive metadata日志输出
2023-04-24 10:10:59,731 INFO  org.apache.hadoop.hive.metastore.HiveMetaStore: [pool-8-thread-198]: 182: source:10.207.xx.xx get_partition_names : db=test tbl=product
2023-04-24 10:10:59,731 INFO  org.apache.hadoop.hive.metastore.HiveMetaStore.audit: [pool-8-thread-198]: ugi=root	ip=10.207.xx.xx	cmd=source:10.207.xx.xx get_partition_names : db=test tbl=product	


#这是 select * from product limit 1 的hive metadata日志输出,卡了30分钟查询超时退出,hive日志端无报错
2023-04-24 10:11:12,903 INFO  org.apache.hadoop.hive.metastore.HiveMetaStore: [pool-8-thread-198]: 182: source:10.207.xx.xx get_partition_names : db=test tbl=product
2023-04-24 10:11:12,903 INFO  org.apache.hadoop.hive.metastore.HiveMetaStore.audit: [pool-8-thread-198]: ugi=root	ip=10.207.xx.xx	cmd=source:10.207.xx.xx get_partition_names : db=test tbl=product

能发下查询期间的be.info和fe.log日志么,我们看下,估计是hive链接信息没有配置正确,导致元数据获取超时了

日志已私信
P.S:看上去像驱动问题?

升级到2.5.5解决。 :joy:
P.S:建议官方在release note里提一下
PP.S:试了下3.0,也没问题

是不是:Starrocks-2.5.4查询hive catalog卡住不动

1赞