EXPLAIN
功能
显示输入查询语句的逻辑或物理执行计划。关于如何分析查询计划，请参考 分析 Query Plan。
提示
该操作不需要权限。
语法
EXPLAIN [ LOGICAL | VERBOSE | COSTS ] <query>
提示
执行计划详细程度：EXPLAIN LOGICAL < EXPLAIN < EXPLAIN VERBOSE < EXPLAIN COSTS。
自 v3.3.5 起，默认的执行计划详细级别改为 COSTS。您可以使用动态 FE 参数 query_detail_explain_level 配置 EXPLAIN 返回的信息详细级别。
通常情况下使用 EXPLAIN 即可，EXPLAIN VERBOSE 和 EXPLAIN COSTS 会打印大量冗余信息，主要用于调试计划。
参数说明
| 参数 | 说明 |
|---|---|
| LOGICAL | 显示简要的逻辑执行计划。 |
| VERBOSE | 显示详细的逻辑执行计划，包括数据类型，nullable 信息，优化策略等。 |
| COSTS | 显示详细的逻辑执行计划，包括统计信息。 |
| query | 需要查看执行计划的查询语句。 |
返回
返回输入查询语句的执行计划。
| 返回 | 说明 |
|---|---|
| avgRowSize | 扫描数据行的平均大小。 |
| cardinality | 扫描表的数据总行数。 |
| colocate | 是否为 Colocate 表。 |
| numNodes | 待扫描的节点数。 |
| rollup | 使用的同步物化视图。 |
| preaggregation | 预聚合的状态。 |
| predicates | 查询中使用的谓词。 |
| column statistics | 列统计信息，依次为MIN,MAX, NULL个数，平均大小，基数 |
示例
EXPLAIN
MySQL tpcds> explain select
sr_customer_sk as ctr_customer_sk,
sr_store_sk as ctr_store_sk,
sum(SR_RETURN_AMT) as ctr_total_return
from
store_returns,
date_dim
where
sr_returned_date_sk = d_date_sk and d_year = 2000
group by
sr_customer_sk,
sr_store_sk
+-----------------------------------------------------------------------------+
| Explain String |
+-----------------------------------------------------------------------------+
| PLAN FRAGMENT 0 |
| OUTPUT EXPRS:5: sr_customer_sk | 9: sr_store_sk | 49: sum |
| PARTITION: UNPARTITIONED |
| |
| RESULT SINK |
| |
| 9:EXCHANGE |
| |
| PLAN FRAGMENT 1 |
| OUTPUT EXPRS: |
| PARTITION: HASH_PARTITIONED: 5: sr_customer_sk, 9: sr_store_sk |
| |
| STREAM DATA SINK |
| EXCHANGE ID: 09 |
| UNPARTITIONED |
| |
| 8:AGGREGATE (merge finalize) |
| | output: sum(49: sum) |
| | group by: 5: sr_customer_sk, 9: sr_store_sk |
| | |
| 7:EXCHANGE |
| |
| PLAN FRAGMENT 2 |
| OUTPUT EXPRS: |
| PARTITION: RANDOM |
| |
| STREAM DATA SINK |
| EXCHANGE ID: 07 |
| HASH_PARTITIONED: 5: sr_customer_sk, 9: sr_store_sk |
| |
| 6:AGGREGATE (update serialize) |
| | STREAMING |
| | output: sum(12: sr_return_amt) |
| | group by: 5: sr_customer_sk, 9: sr_store_sk |
| | |
| 5:Project |
| | <slot 5> : 5: sr_customer_sk |
| | <slot 9> : 9: sr_store_sk |
| | <slot 12> : 12: sr_return_amt |
| | |
| 4:HASH JOIN |
| | join op: INNER JOIN (BROADCAST) |
| | colocate: false, reason: |
| | equal join conjunct: 3: sr_returned_date_sk = 21: d_date_sk |
| | |
| |----3:EXCHANGE |
| | |
| 0:OlapScanNode |
| TABLE: store_returns |
| PREAGGREGATION: ON |
| PREDICATES: 3: sr_returned_date_sk IS NOT NULL |
| partitions=1/1 |
| rollup: store_returns |
| tabletRatio=10/10 |
| tabletList=10241,10243,10245,10247,10249,10251,10253,10255,10257,10259 |
| cardinality=277502 |
| avgRowSize=20.0 |
| |
| PLAN FRAGMENT 3 |
| OUTPUT EXPRS: |
| PARTITION: RANDOM |
| |
| STREAM DATA SINK |
| EXCHANGE ID: 03 |
| UNPARTITIONED |
| |
| 2:Project |
| | <slot 21> : 21: d_date_sk |
| | |
| 1:OlapScanNode |
| TABLE: date_dim |
| PREAGGREGATION: ON |
| PREDICATES: 27: d_year = 2000 |
| partitions=1/1 |
| rollup: date_dim |
| tabletRatio=5/5 |
| tabletList=11543,11545,11547,11549,11551 |
| cardinality=362 |
| avgRowSize=8.0 |
+-----------------------------------------------------------------------------+
79 rows in set
EXPLAIN LOGICAL
MySQL tpcds> explain logical select
sr_customer_sk as ctr_customer_sk,
sr_store_sk as ctr_store_sk,
sum(SR_RETURN_AMT) as ctr_total_return
from
store_returns,
date_dim
where
sr_returned_date_sk = d_date_sk and d_year = 2000
group by
sr_customer_sk,
sr_store_sk
+---------------------------------------------------------------------------------------------------------------------------------+
| Explain String |
+---------------------------------------------------------------------------------------------------------------------------------+
| - Output => [5:sr_customer_sk, 9:sr_store_sk, 49:sum] |
| - AGGREGATE(GLOBAL) [5:sr_customer_sk, 9:sr_store_sk] |
| Estimates: {row: 49728, cpu: 795660.58, memory: 795660.58, network: 0.00, cost: 8405375.45} |
| 49:sum := sum(49:sum) |
| - EXCHANGE(SHUFFLE) [5, 9] |
| Estimates: {row: 49728, cpu: 79566.06, memory: 0.00, network: 0.00, cost: 6416223.99} |
| - AGGREGATE(LOCAL) [5:sr_customer_sk, 9:sr_store_sk] |
| Estimates: {row: 49728, cpu: 119349.09, memory: 79566.06, network: 0.00, cost: 6376440.96} |
| 49:sum := sum(12:sr_return_amt) |
| - HASH/INNER JOIN [3:sr_returned_date_sk = 21:d_date_sk] => [5:sr_customer_sk, 9:sr_store_sk, 12:sr_return_amt] |
| Estimates: {row: 49728, cpu: 6744977.39, memory: 1446.51, network: 0.00, cost: 6157634.30} |
| - SCAN [store_returns] => [3:sr_returned_date_sk, 5:sr_customer_sk, 9:sr_store_sk, 12:sr_return_amt] |
| Estimates: {row: 277502, cpu: 5550040.00, memory: 0.00, network: 0.00, cost: 2775020.00} |
| partitionRatio: 1/1, tabletRatio: 10/10 |
| predicate: 3:sr_returned_date_sk IS NOT NULL |
| - EXCHANGE(BROADCAST) |
| Estimates: {row: 361, cpu: 1446.51, memory: 1446.51, network: 1446.51, cost: 7232.57} |
| - SCAN [date_dim] => [21:d_date_sk] |
| Estimates: {row: 361, cpu: 2893.03, memory: 0.00, network: 0.00, cost: 1446.51} |
| partitionRatio: 1/1, tabletRatio: 5/5 |
| predicate: 27:d_year = 2000 |
+---------------------------------------------------------------------------------------------------------------------------------+
21 rows in set
EXPLAIN VERBOSE
MySQL tpcds> explain verbose select
sr_customer_sk as ctr_customer_sk,
sr_store_sk as ctr_store_sk,
sum(SR_RETURN_AMT) as ctr_total_return
from
store_returns,
date_dim
where
sr_returned_date_sk = d_date_sk and d_year = 2000
group by
sr_customer_sk,
sr_store_sk
+---------------------------------------------------------------------------------------------------------------------------------------------------------+
| Explain String |
+---------------------------------------------------------------------------------------------------------------------------------------------------------+
| RESOURCE GROUP: default_wg |
| |
| PLAN FRAGMENT 0(F04) |
| Fragment Cost: 0.0 |
| Output Exprs:5: sr_customer_sk | 9: sr_store_sk | 49: sum |
| Input Partition: UNPARTITIONED |
| RESULT SINK |
| |
| 9:EXCHANGE |
| cardinality: 49729 |
| |
| PLAN FRAGMENT 1(F03) |
| Fragment Cost: 2028934.4876271961 |
| |
| Input Partition: HASH_PARTITIONED: 5: sr_customer_sk, 9: sr_store_sk |
| OutPut Partition: UNPARTITIONED |
| OutPut Exchange Id: 09 |
| |
| 8:AGGREGATE (merge finalize) |
| | aggregate: sum[([49: sum, DECIMAL128(38,2), true]); args: DECIMAL64; result: DECIMAL128(38,2); args nullable: true; result nullable: true] |
| | group by: [5: sr_customer_sk, INT, true], [9: sr_store_sk, INT, true] |
| | cardinality: 49729 |
| | |
| 7:EXCHANGE |
| distribution type: SHUFFLE |
| partition exprs: [5: sr_customer_sk, INT, true], [9: sr_store_sk, INT, true] |
| cardinality: 49729 |
| |
| PLAN FRAGMENT 2(F00) |
| Fragment Cost: 6376440.959353934 |
| |
| Input Partition: RANDOM |
| OutPut Partition: HASH_PARTITIONED: 5: sr_customer_sk, 9: sr_store_sk |
| OutPut Exchange Id: 07 |
| |
| 6:AGGREGATE (update serialize) |
| | STREAMING |
| | aggregate: sum[([12: sr_return_amt, DECIMAL64(7,2), true]); args: DECIMAL64; result: DECIMAL128(38,2); args nullable: true; result nullable: true] |
| | group by: [5: sr_customer_sk, INT, true], [9: sr_store_sk, INT, true] |
| | cardinality: 49729 |
| | |
| 5:Project |
| | output columns: |
| | 5 <-> [5: sr_customer_sk, INT, true] |
| | 9 <-> [9: sr_store_sk, INT, true] |
| | 12 <-> [12: sr_return_amt, DECIMAL64(7,2), true] |
| | cardinality: 49729 |
| | |
| 4:HASH JOIN |
| | join op: INNER JOIN (BROADCAST) |
| | equal join conjunct: [3: sr_returned_date_sk, INT, true] = [21: d_date_sk, INT, false] |
| | build runtime filters: |
| | - filter_id = 0, build_expr = (21: d_date_sk), remote = false |
| | output columns: 5, 9, 12 |
| | can local shuffle: true |
| | cardinality: 49729 |
| | |
| |----3:EXCHANGE |
| | distribution type: BROADCAST |
| | cardinality: 362 |
| | |
| 0:OlapScanNode |
| table: store_returns, rollup: store_returns |
| preAggregation: on |
| Predicates: 3: sr_returned_date_sk IS NOT NULL |
| partitionsRatio=1/1, tabletsRatio=10/10 |
| tabletList=10241,10243,10245,10247,10249,10251,10253,10255,10257,10259 |
| actualRows=287514, avgRowSize=20.0 |
| cardinality: 277502 |
| probe runtime filters: |
| - filter_id = 0, probe_expr = (3: sr_returned_date_sk) |
| |
| PLAN FRAGMENT 3(F01) |
| Fragment Cost: 1446.5148514851485 |
| |
| Input Partition: RANDOM |
| OutPut Partition: UNPARTITIONED |
| OutPut Exchange Id: 03 |
| |
| 2:Project |
| | output columns: |
| | 21 <-> [21: d_date_sk, INT, false] |
| | cardinality: 362 |
| | |
| 1:OlapScanNode |
| table: date_dim, rollup: date_dim |
| preAggregation: on |
| Predicates: [27: d_year, INT, false] = 2000 |
| partitionsRatio=1/1, tabletsRatio=5/5 |
| tabletList=11543,11545,11547,11549,11551 |
| actualRows=73049, avgRowSize=8.0 |
| cardinality: 362 |
+---------------------------------------------------------------------------------------------------------------------------------------------------------+
EXPLAIN COSTS
+---------------------------------------------------------------------------------------------------------------------------------------------------------+
| Explain String |
+---------------------------------------------------------------------------------------------------------------------------------------------------------+
| PLAN FRAGMENT 0(F04) |
| Output Exprs:5: sr_customer_sk | 9: sr_store_sk | 49: sum |
| Input Partition: UNPARTITIONED |
| RESULT SINK |
| |
| 9:EXCHANGE |
| cardinality: 49729 |
| |
| PLAN FRAGMENT 1(F03) |
| |
| Input Partition: HASH_PARTITIONED: 5: sr_customer_sk, 9: sr_store_sk |
| OutPut Partition: UNPARTITIONED |
| OutPut Exchange Id: 09 |
| |
| 8:AGGREGATE (merge finalize) |
| | aggregate: sum[([49: sum, DECIMAL128(38,2), true]); args: DECIMAL64; result: DECIMAL128(38,2); args nullable: true; result nullable: true] |
| | group by: [5: sr_customer_sk, INT, true], [9: sr_store_sk, INT, true] |
| | cardinality: 49729 |
| | column statistics: |
| | * sr_customer_sk-->[1.0, 100000.0, 2.010867271218169E-5, 4.0, 49728.786461450894] ESTIMATE |
| | * sr_store_sk-->[1.0, 10.0, 0.14285714285714285, 4.0, 6.0] ESTIMATE |
| | * sum-->[0.0, 16917.12, 0.034878301578357924, 8.0, 49728.786461450894] ESTIMATE |
| | |
| 7:EXCHANGE |
| distribution type: SHUFFLE |
| partition exprs: [5: sr_customer_sk, INT, true], [9: sr_store_sk, INT, true] |
| cardinality: 49729 |
| |
| PLAN FRAGMENT 2(F00) |
| |
| Input Partition: RANDOM |
| OutPut Partition: HASH_PARTITIONED: 5: sr_customer_sk, 9: sr_store_sk |
| OutPut Exchange Id: 07 |
| |
| 6:AGGREGATE (update serialize) |
| | STREAMING |
| | aggregate: sum[([12: sr_return_amt, DECIMAL64(7,2), true]); args: DECIMAL64; result: DECIMAL128(38,2); args nullable: true; result nullable: true] |
| | group by: [5: sr_customer_sk, INT, true], [9: sr_store_sk, INT, true] |
| | cardinality: 49729 |
| | column statistics: |
| | * sr_customer_sk-->[1.0, 100000.0, 2.010867271218169E-5, 4.0, 49728.786461450894] ESTIMATE |
| | * sr_store_sk-->[1.0, 10.0, 0.14285714285714285, 4.0, 6.0] ESTIMATE |
| | * sum-->[0.0, 16917.12, 0.034878301578357924, 8.0, 49728.786461450894] ESTIMATE |
| | |
| 5:Project |
| | output columns: |
| | 5 <-> [5: sr_customer_sk, INT, true] |
| | 9 <-> [9: sr_store_sk, INT, true] |
| | 12 <-> [12: sr_return_amt, DECIMAL64(7,2), true] |
| | cardinality: 49729 |
| | column statistics: |
| | * sr_customer_sk-->[1.0, 100000.0, 0.034836564480338346, 4.0, 49728.786461450894] ESTIMATE |
| | * sr_store_sk-->[1.0, 10.0, 0.034996556689413386, 4.0, 6.0] ESTIMATE |
| | * sr_return_amt-->[0.0, 16917.12, 0.034878301578357924, 8.0, 49728.786461450894] ESTIMATE |
| | |
| 4:HASH JOIN |
| | join op: INNER JOIN (BROADCAST) |
| | equal join conjunct: [3: sr_returned_date_sk, INT, true] = [21: d_date_sk, INT, false] |
| | build runtime filters: |
| | - filter_id = 0, build_expr = (21: d_date_sk), remote = false |
| | output columns: 5, 9, 12 |
| | can local shuffle: true |
| | cardinality: 49729 |
| | column statistics: |
| | * sr_returned_date_sk-->[2450820.0, 2452822.0, 0.0, 4.0, 361.6287128712871] ESTIMATE |
| | * sr_customer_sk-->[1.0, 100000.0, 0.034836564480338346, 4.0, 49728.786461450894] ESTIMATE |
| | * sr_store_sk-->[1.0, 10.0, 0.034996556689413386, 4.0, 6.0] ESTIMATE |
| | * sr_return_amt-->[0.0, 16917.12, 0.034878301578357924, 8.0, 49728.786461450894] ESTIMATE |
| | * d_date_sk-->[2450820.0, 2452822.0, 0.0, 4.0, 361.6287128712871] ESTIMATE |
| | |
| |----3:EXCHANGE |
| | distribution type: BROADCAST |
| | cardinality: 362 |
| | |
| 0:OlapScanNode |
| table: store_returns, rollup: store_returns |
| preAggregation: on |
| Predicates: 3: sr_returned_date_sk IS NOT NULL |
| partitionsRatio=1/1, tabletsRatio=10/10 |
| tabletList=10241,10243,10245,10247,10249,10251,10253,10255,10257,10259 |
| actualRows=287514, avgRowSize=20.0 |
| cardinality: 277502 |
| probe runtime filters: |
| - filter_id = 0, probe_expr = (3: sr_returned_date_sk) |
| column statistics: |
| * sr_returned_date_sk-->[2450820.0, 2452822.0, 0.0, 4.0, 2018.0] ESTIMATE |
| * sr_customer_sk-->[1.0, 100000.0, 0.034836564480338346, 4.0, 86524.0] ESTIMATE |
| * sr_store_sk-->[1.0, 10.0, 0.034996556689413386, 4.0, 6.0] ESTIMATE |
| * sr_return_amt-->[0.0, 16917.12, 0.034878301578357924, 8.0, 113968.0] ESTIMATE |
| |
| PLAN FRAGMENT 3(F01) |
| |
| Input Partition: RANDOM |
| OutPut Partition: UNPARTITIONED |
| OutPut Exchange Id: 03 |
| |
| 2:Project |
| | output columns: |
| | 21 <-> [21: d_date_sk, INT, false] |
| | cardinality: 362 |
| | column statistics: |
| | * d_date_sk-->[2415022.0, 2488070.0, 0.0, 4.0, 361.6287128712871] ESTIMATE |
| | |
| 1:OlapScanNode |
| table: date_dim, rollup: date_dim |
| preAggregation: on |
| Predicates: [27: d_year, INT, false] = 2000 |
| partitionsRatio=1/1, tabletsRatio=5/5 |
| tabletList=11543,11545,11547,11549,11551 |
| actualRows=73049, avgRowSize=8.0 |
| cardinality: 362 |
| column statistics: |
| * d_date_sk-->[2415022.0, 2488070.0, 0.0, 4.0, 361.6287128712871] ESTIMATE |
| * d_year-->[2000.0, 2000.0, 0.0, 4.0, 202.0] ESTIMATE |
+---------------------------------------------------------------------------------------------------------------------------------------------------------+
114 rows in set