试图了解如何提高此 Sql 查询的性能

Question

我有一个 SQL 查询，它试图查找某些县的所有社区。​​p>

当我使用 SQL Sentry Plan Explorer 来可视化查询时（IMO，比 MS SSMS 提供的工具好一点），它突出显示了一个执行速度非常慢的部分：-

完整计划

放大....

详情

SQL 脚本：

 -- Update which Neighbourhoods are in these Counties.
INSERT INTO @NeighbourhoodCounties (NeighbourhoodId, CountyId)
 SELECT SubQuery.NeighbourhoodId, SubQuery.CountyId
 FROM (
  SELECT e.LocationId AS NeighbourhoodId, b.LocationId AS CountyId, 
      c.OriginalBoundary.STArea() AS CountyArea,
      c.OriginalBoundary.STIntersection(d.OriginalBoundary).STArea() AS IntersectionArea
  FROM @CountyIds a
   INNER JOIN [dbo].[Counties] b ON a.Id = b.LocationId
   INNER JOIN [dbo].[GeographyBoundaries] c ON b.LocationId = c.LocationId
   INNER JOIN [dbo].[GeographyBoundaries] d ON c.OriginalBoundary.STIntersects(d.OriginalBoundary) = 1
   INNER JOIN [dbo].[Neighbourhoods] e ON d.LocationId = e.LocationId
   ) SubQuery
    WHERE (SubQuery.IntersectionArea / SubQuery.CountyArea) * 100 > 5 -- a Neighbourhood has to be 5% or more to be considered 'Inside'

谁能帮助解释这个查询？所有这些数字是什么意思？如何使用这些数字来帮助诊断和改进我的查询？

I tried to make an indexed view on the spatial table 但惨败。

谁能帮忙？

Answer 1

这很正常。在大多数情况下，您在任何地方都没有粗条。

但是，我确实看到您在 JOIN 中的列上有一个函数

ON c.OriginalBoundary.STIntersects(d.OriginalBoundary) = 1

这无济于事。计算列也无济于事

您还可以在 WHERE = non-sargable 中对列进行计算

(SubQuery.IntersectionArea / SubQuery.CountyArea) * 100

从表面上看，64.5% 的搜索可能是这些 JOIN 和围绕它们工作的优化器的结果