The physical execution plans of kyuubi and sparkthriftserver are different

[xcl1102]

kyuubi version is 1.7.0，The version of spark is 3.1.3。Under the premise that the spark-default.conf configuration is the same, kyuubi and sparkthriftserver are used to execute order by limit sql, but the physical execution is different.
The execution performance of kyuubi is worse than that of spark thrift server. Can you help me answer the question?

The statement to create the table is as follows：

CREATE TABLE `tpcds_1t_parquet_no_v4`.`store_sales` (
  `ss_sold_time_sk` BIGINT,
  `ss_item_sk` BIGINT,
  `ss_customer_sk` BIGINT,
  `ss_cdemo_sk` BIGINT,
  `ss_hdemo_sk` BIGINT,
  `ss_addr_sk` BIGINT,
  `ss_store_sk` BIGINT,
  `ss_promo_sk` BIGINT,
  `ss_ticket_number` BIGINT,
  `ss_quantity` INT,
  `ss_wholesale_cost` DECIMAL(7,2),
  `ss_list_price` DECIMAL(7,2),
  `ss_sales_price` DECIMAL(7,2),
  `ss_ext_discount_amt` DECIMAL(7,2),
  `ss_ext_sales_price` DECIMAL(7,2),
  `ss_ext_wholesale_cost` DECIMAL(7,2),
  `ss_ext_list_price` DECIMAL(7,2),
  `ss_ext_tax` DECIMAL(7,2),
  `ss_coupon_amt` DECIMAL(7,2),
  `ss_net_paid` DECIMAL(7,2),
  `ss_net_paid_inc_tax` DECIMAL(7,2),
  `ss_net_profit` DECIMAL(7,2),
  `ss_sold_date_sk` BIGINT)
USING parquet
OPTIONS (
  `parquet.compression` 'SNAPPY',
  `parquet.enable.dictionary` 'false')
PARTITIONED BY (ss_sold_date_sk);

The executed sql is as follows：

select ss_sold_time_sk, ss_item_sk, ss_customer_sk, ss_cdemo_sk, ss_hdemo_sk, ss_addr_sk, ss_store_sk, ss_promo_sk, ss_ticket_number, 
ss_quantity, ss_wholesale_cost,ss_list_price, ss_sales_price, ss_ext_discount_amt, ss_ext_sales_price, ss_ext_wholesale_cost, 
ss_ext_list_price, ss_ext_tax, ss_coupon_amt, ss_net_paid, ss_net_paid_inc_tax, ss_net_profit, ss_sold_date_sk from store_sales
where ss_sold_date_sk >= 2452500 and ss_sold_date_sk <= 2452600 and ss_sold_time_sk >= 30000 and ss_sold_time_sk <= 60000 order by ss_sold_time_sk DESC LIMIT 10000;

The execution plan is as follows：
kyuubi:

| == Physical Plan ==
CollectLimit 10000
+- *(2) Sort [ss_sold_time_sk#40L DESC NULLS LAST], true, 0
   +- Exchange rangepartitioning(ss_sold_time_sk#40L DESC NULLS LAST, 200), ENSURE_REQUIREMENTS, [id=#43]
      +- *(1) Filter ((isnotnull(ss_sold_time_sk#40L) AND (ss_sold_time_sk#40L >= 30000)) AND (ss_sold_time_sk#40L <= 60000))
         +- *(1) ColumnarToRow
            +- FileScan parquet tpcds_1t_parquet_no_v4.store_sales[ss_sold_time_sk#40L,ss_item_sk#41L,ss_customer_sk#42L,ss_cdemo_sk#43L,ss_hdemo_sk#44L,ss_addr_sk#45L,ss_store_sk#46L,ss_promo_sk#47L,ss_ticket_number#48L,ss_quantity#49,ss_wholesale_cost#50,ss_list_price#51,ss_sales_price#52,ss_ext_discount_amt#53,ss_ext_sales_price#54,ss_ext_wholesale_cost#55,ss_ext_list_price#56,ss_ext_tax#57,ss_coupon_amt#58,ss_net_paid#59,ss_net_paid_inc_tax#60,ss_net_profit#61,ss_sold_date_sk#62L] Batched: true, DataFilters: [isnotnull(ss_sold_time_sk#40L), (ss_sold_time_sk#40L >= 30000), (ss_sold_time_sk#40L <= 60000)], Format: Parquet, Location: InMemoryFileIndex[hdfs://HACluster/home/work/hive/warehouse/tpcds_1t_parquet_no_v4.db/store_sales..., PartitionFilters: [isnotnull(ss_sold_date_sk#62L), (ss_sold_date_sk#62L >= 2452500), (ss_sold_date_sk#62L <= 2452600)], PushedFilters: [IsNotNull(ss_sold_time_sk), GreaterThanOrEqual(ss_sold_time_sk,30000), LessThanOrEqual(ss_sold_t..., ReadSchema: struct<ss_sold_time_sk:bigint,ss_item_sk:bigint,ss_customer_sk:bigint,ss_cdemo_sk:bigint,ss_hdemo...

sparkthrifyserver:

| == Physical Plan ==
TakeOrderedAndProject(limit=10000, orderBy=[ss_sold_time_sk#102L DESC NULLS LAST], output=[ss_sold_time_sk#102L,ss_item_sk#103L,ss_customer_sk#104L,ss_cdemo_sk#105L,ss_hdemo_sk#106L,ss_addr_sk#107L,ss_store_sk#108L,ss_promo_sk#109L,ss_ticket_number#110L,ss_quantity#111,ss_wholesale_cost#112,ss_list_price#113,ss_sales_price#114,ss_ext_discount_amt#115,ss_ext_sales_price#116,ss_ext_wholesale_cost#117,ss_ext_list_price#118,ss_ext_tax#119,ss_coupon_amt#120,ss_net_paid#121,ss_net_paid_inc_tax#122,ss_net_profit#123,ss_sold_date_sk#124L])
+- *(1) Filter ((isnotnull(ss_sold_time_sk#102L) AND (ss_sold_time_sk#102L >= 30000)) AND (ss_sold_time_sk#102L <= 60000))
   +- *(1) ColumnarToRow
      +- FileScan parquet tpcds_1t_parquet_no_v4.store_sales[ss_sold_time_sk#102L,ss_item_sk#103L,ss_customer_sk#104L,ss_cdemo_sk#105L,ss_hdemo_sk#106L,ss_addr_sk#107L,ss_store_sk#108L,ss_promo_sk#109L,ss_ticket_number#110L,ss_quantity#111,ss_wholesale_cost#112,ss_list_price#113,ss_sales_price#114,ss_ext_discount_amt#115,ss_ext_sales_price#116,ss_ext_wholesale_cost#117,ss_ext_list_price#118,ss_ext_tax#119,ss_coupon_amt#120,ss_net_paid#121,ss_net_paid_inc_tax#122,ss_net_profit#123,ss_sold_date_sk#124L] Batched: true, DataFilters: [isnotnull(ss_sold_time_sk#102L), (ss_sold_time_sk#102L >= 30000), (ss_sold_time_sk#102L <= 60000)], Format: Parquet, Location: InMemoryFileIndex[hdfs://HACluster/home/work/hive/warehouse/tpcds_1t_parquet_no_v4.db/store_sales..., PartitionFilters: [isnotnull(ss_sold_date_sk#124L), (ss_sold_date_sk#124L >= 2452500), (ss_sold_date_sk#124L <= 245..., PushedFilters: [IsNotNull(ss_sold_time_sk), GreaterThanOrEqual(ss_sold_time_sk,30000), LessThanOrEqual(ss_sold_t..., ReadSchema: struct<ss_sold_time_sk:bigint,ss_item_sk:bigint,ss_customer_sk:bigint,ss_cdemo_sk:bigint,ss_hdemo...

[pan3793]

This is caused by #5005

[xcl1102]

thanks for your answer