Sql server SQL Server查询某些时间使用freetexttable花费太多时间-查找索引问题

以下是我的查询，它需要大约1分钟的时间，有时会在一秒钟内给出结果， 问题发生了，特别是我在一段时间前执行了查询，或者在查询中添加了一些新关键字。 看起来像是索引问题，当我执行执行计划时，RID查找成本为60%。源表有大约2-5个Lacs数据，每天将添加大约10000-20000行。请告诉我。谢谢

SELECT *
FROM   (SELECT Row_number() OVER (ORDER BY rank DESC, jobid DESC) AS rnum,
               *
        FROM   (SELECT rank,
                       joblistview.*
                FROM   joblistview,
                       FREETEXTTABLE(joblistview, jobtitle, 'seo manager') f
                WHERE  joblistview.jobid = f.[key]
                       AND CONTAINS(joblistview.joblocation, 'mumbai')
                UNION
                SELECT rank,
                       joblistview.*
                FROM   joblistview,
                       FREETEXTTABLE(joblistview, jobdescription, 'seo manager')
                       f
                WHERE  joblistview.jobid = f.[key]
                       AND CONTAINS(joblistview.joblocation, 'mumbai')
                UNION
                SELECT rank,
                       joblistview.*
                FROM   joblistview,
                       FREETEXTTABLE(joblistview, company_name, 'seo manager') f
                WHERE  joblistview.jobid = f.[key]
                       AND CONTAINS(joblistview.joblocation, 'mumbai')) AS xx)AS
       tt
WHERE  rnum BETWEEN 11 AND 20  

执行计划

---

您可以在一次搜索中指定多个列，这将消除使用联合进行多个查询的需要

SELECT *
FROM   (SELECT Row_number() OVER (ORDER BY rank DESC, jobid DESC) AS rnum,
               *
        FROM   (SELECT rank,
                       joblistview.*
                FROM   joblistview,
                       FREETEXTTABLE(joblistview, (jobtitle,jobdescription,company_name), 'seo manager') f
                WHERE  joblistview.jobid = f.[key]
                       AND CONTAINS(joblistview.joblocation, 'mumbai')
               ) AS xx
       ) AS tt
WHERE  rnum BETWEEN 11 AND 20  

---

是否尝试添加包含所有搜索列数据的计算列？然后将此计算列配置为持久化，并对其进行全文索引。然后你可以试试

SELECT *
FROM   (SELECT Row_number() OVER (ORDER BY rank DESC, jobid DESC) AS rnum,
               *
        FROM   (SELECT rank,
                       joblistview.*
                FROM   joblistview,
                       FREETEXTTABLE(joblistview, (<<<ComputedColumn>>>), 'seo manager') f
                WHERE  joblistview.jobid = f.[key]
                       AND CONTAINS(joblistview.joblocation, 'mumbai')
               ) AS xx
       ) AS tt
WHERE  rnum BETWEEN 11 AND 20 

---

您已将同一查询联合了三次-给出了什么？它们不是同一个查询，每个查询使用FREETEXTTABLE函数搜索不同的列jobtitle、jobdescription、company\u name。您好，谢谢您的回答。。我已经这样做了，但结果在与工会的相关性方面要好得多。有多列的偏离路线结果会很快出现。与工会的合作结果也不错，但问题是我认为这与指数有关。。每当有新的术语出现在搜索中，都会花费太多的时间。