PawSQL优化案例分析：TPC-H Query 9性能提升1195.14%

PawSQL

发布于 2024-09-26 16:23:57

710

发布于 2024-09-26 16:23:57

引言

PawSQL对TPC-H基准的Query 9进行的自动性能优化后，Query9的性能提升了1195.14%。本文将根据PawSQL的优化过程，探讨其优化 SQL 性能的有效策略与实际效果。通过对比优化前后的查询结构、索引设计及执行计划，阐述如何实现显著的性能提升。

本文的案例可在线查看：https://pawsql.com/statement/1837384704875499522

TPC-H Query 9介绍：TPC-H 的第9个查询统计每个国家每年所有被订购零件的总利润。其查询特性如下：

包含分组、排序、聚集操作
存在多表连接和子查询
使用了全匹配的LIKE 操作符

1️⃣ 查询重写优化

原始查询：‍

SELECT profit.nation, profit.o_year, SUM(profit.amount) AS sum_profit
FROM (
    SELECT nation.n_name AS nation, EXTRACT(YEAR FROM orders.o_orderdate) AS o_year,
           lineitem.l_extendedprice * (1 - lineitem.l_discount) - partsupp.ps_supplycost * lineitem.l_quantity AS amount
    FROM part, supplier, lineitem, partsupp, orders, nation
    WHERE supplier.s_suppkey = lineitem.l_suppkey
      AND partsupp.ps_suppkey = lineitem.l_suppkey
      AND partsupp.ps_partkey = lineitem.l_partkey
      AND part.p_partkey = lineitem.l_partkey
      AND orders.o_orderkey = lineitem.l_orderkey
      AND supplier.s_nationkey = nation.n_nationkey
      AND part.p_name LIKE '%dim%'
) AS profit
GROUP BY profit.nation, profit.o_year
ORDER BY profit.nation, profit.o_year DESC;

优化后的查询：‍

SELECT profit_nation.n_name AS nation, EXTRACT(YEAR FROM profit_orders.o_orderdate) AS o_year,
       SUM(profit_lineitem.l_extendedprice * (1 - profit_lineitem.l_discount) - profit_partsupp.ps_supplycost * profit_lineitem.l_quantity) AS sum_profit
FROM part AS profit_part, supplier AS profit_supplier,
     lineitem AS profit_lineitem, partsupp AS profit_partsupp,
     orders AS profit_orders, nation AS profit_nation
WHERE profit_supplier.s_suppkey = profit_lineitem.l_suppkey
  AND profit_partsupp.ps_suppkey = profit_lineitem.l_suppkey
  AND profit_partsupp.ps_partkey = profit_lineitem.l_partkey
  AND profit_part.p_partkey = profit_lineitem.l_partkey
  AND profit_orders.o_orderkey = profit_lineitem.l_orderkey
  AND profit_supplier.s_nationkey = profit_nation.n_nationkey
  AND profit_part.p_name LIKE '%dim%'
GROUP BY profit_nation.n_name, o_year
ORDER BY profit_nation.n_name, o_year DESC;

重写优化要点：

消除子查询：将子查询的逻辑提升到主查询中，简化结构。
表别名重命名：提高查询的可读性与可维护性。
保持计算逻辑：确保查询的业务逻辑未发生变化。

2️⃣ 🔍索引优化策略

PawSQL 提出的索引优化方案：

CREATE INDEX PAWSQL_IDX0485218972 ON tpch.lineitem(L_PARTKEY, L_SUPPKEY, L_SHIPDATE);
CREATE INDEX PAWSQL_IDX0214365528 ON tpch.supplier(S_NATIONKEY, S_SUPPKEY, S_NAME, S_ADDRESS);
CREATE INDEX PAWSQL_IDX0327029402 ON tpch.part(P_NAME);