An error occurred when querying empty partitions in the Hive external table.

[Astralidea]

Steps to reproduce the behavior (Required)

1. create table test_part_table(
   word string,
   num bigint
   )partitioned by(dt string)
   STORED AS ORC;
   2.CREATE EXTERNAL TABLE test (
   word varchar(256) NULL COMMENT "",
   num bigint(11) NULL COMMENT "",
   dt varchar(256) NULL COMMENT ""
   ) ENGINE=HIVE
   COMMENT "PARTITION BY (dt)"
   PROPERTIES (
   "database" = "lxy",
   "table" = "test_part_table",
   "resource" = "hive0",
   "hive.metastore.uris" = "thrift://xxx"
   );

insert into test_part_table partition(dt) select '1' as word, 2 as num, '20190808' as dt;
alter table test_part_table add if not exists partition(dt='20190807') location '20190807';
currently if you run
hadoop dfs -ls hdfs://172.26.92.154:10000/user/hive/warehouse/lxy.db/test_part_table
it will show only one directory although you already create 20190807 partition in hive meta
hdfs://xxx:10000/user/hive/warehouse/lxy.db/test_part_table/20190808

restart starrocks FE
then run
select * from test where dt='20190807';

Expected behavior (Required)

Empty set

Real behavior (Required)

in cli
ERROR 1064 (HY000): get table lxy.test_part_table partition meta info failed.
in fe log
2021-10-15 18:14:18,145 WARN (starrocks-mysql-nio-pool-0|83) [StmtExecutor.execute():454] execute Exception, sql select * from test where dt='20190807'
com.starrocks.sql.common.StarRocksPlannerException: get table lxy.test_part_table partition meta info failed.
at com.starrocks.sql.optimizer.statistics.StatisticsCalculator.getTableRowCount(StatisticsCalculator.java:444) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.statistics.StatisticsCalculator.computeHiveScanNode(StatisticsCalculator.java:220) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.statistics.StatisticsCalculator.visitLogicalHiveScan(StatisticsCalculator.java:208) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.statistics.StatisticsCalculator.visitLogicalHiveScan(StatisticsCalculator.java:111) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.operator.logical.LogicalHiveScanOperator.accept(LogicalHiveScanOperator.java:105) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.statistics.StatisticsCalculator.estimatorStats(StatisticsCalculator.java:130) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.task.DeriveStatsTask.execute(DeriveStatsTask.java:90) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.task.SeriallyTaskScheduler.executeTasks(SeriallyTaskScheduler.java:36) ~[starrocks-fe.jar:?]
at com.starrocks.sql.optimizer.Optimizer.optimize(Optimizer.java:150) ~[starrocks-fe.jar:?]
at com.starrocks.sql.StatementPlanner.createQueryPlan(StatementPlanner.java:75) ~[starrocks-fe.jar:?]
at com.starrocks.sql.StatementPlanner.plan(StatementPlanner.java:46) ~[starrocks-fe.jar:?]
at com.starrocks.qe.StmtExecutor.execute(StmtExecutor.java:299) [starrocks-fe.jar:?]
at com.starrocks.qe.ConnectProcessor.handleQuery(ConnectProcessor.java:248) [starrocks-fe.jar:?]
at com.starrocks.qe.ConnectProcessor.dispatch(ConnectProcessor.java:397) [starrocks-fe.jar:?]
at com.starrocks.qe.ConnectProcessor.processOnce(ConnectProcessor.java:633) [starrocks-fe.jar:?]
at com.starrocks.mysql.nio.ReadListener.lambda$handleEvent$0(ReadListener.java:54) [starrocks-fe.jar:?]
at com.starrocks.mysql.nio.ReadListener$$Lambda$51/1870280360.run(Unknown Source) [starrocks-fe.jar:?]
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) [?:1.8.0_41]
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) [?:1.8.0_41]
at java.lang.Thread.run(Thread.java:745) [?:1.8.0_41]

StarRocks version (Required)

• You can get the StarRocks version by executing SQL select current_version()
  1.19