Env: HDP 3.1.5 Hive: hive_3_1_5_0_152 We have a hive table with 6k partitions, it is very normal when describing a single partition but super slow when describing the whole table, describing non-partitioned tables is normal. desc hive_desc_partition_tbl_01; -- cost 1k+ sec desc hive_desc_partition_tbl_02 partition(ns='NS1',slicetime=20200102);  -- cost 3 sec alter table hive_desc_partition_tbl_01 change column id id string comment 'change';  -- cost 3600 sec   The DDL of the table: create TABLE `hive_desc_partition_tbl_01` ( `id` bigint, `b` string, `c` string, `d` string, `e` string, `ns_name` string) COMMENT 'compress test' PARTITIONED BY ( `ns` string, `slicetime` string)   Any ideas to optimize? Thanks in advance! ... View more