Commit 95629c3

committed

Correct simple query to find unused indexes

Ensure that unique indices are not removed, even if they are not or only rarely used.

1 parent 2d6f69f commit 95629c3Copy full SHA for 95629c3

File tree

-19

lines changed

-19

lines changed

Lines changed: 3 additions & 1 deletion

Original file line number	Diff line number	Diff line change
`@@ -59,7 +59,9 @@ A basic query is simple:`
`59`	`59`	```sql
`60`	`60`	`select *`
`61`	`61`	`from pg_stat_user_indexes`
`62`		`-where idx_scan = 0;`
	`62`	`+where`
	`63`	`+ idx_scan = 0`
	`64`	`+ and not indisunique;`
`63`	`65`	```
`64`	`66`
`65`	`67`	`How to understand how long ago the stats were last reset:`

Lines changed: 0 additions & 18 deletions

Original file line number	Diff line number	Diff line change
`@@ -7,24 +7,6 @@ Originally from: [tweet](https://twitter.com/samokhvalov/status/1734467240832201`
`7`	`7`	`> I post a new PostgreSQL "howto" article every day. Join me in this`
`8`	`8`	`> journey – [subscribe](https://twitter.com/samokhvalov/), provide feedback, share!`
`9`	`9`
`10`		`----------`
`11`		`-🛠️ Correction: In [Day 75: How to find unused indexes](0075_how_to_find_unused_indexes.md), the provided "simple"`
`12`		`-version of the query finding unused indexes lacked an important filter to exclude unique indexes. Of course, unique`
`13`		`-indexes must be excluded from consideration – even if such an index is not used, it has an important role. The corrected`
`14`		`-version:`
`15`		`-`
`16`		-```sql
`17`		`-select *`
`18`		`-from pg_stat_user_indexes`
`19`		`-where`
`20`		`- idx_scan = 0`
`21`		`- and not indisunique;`
`22`		-```
`23`		`-`
`24`		`-The "advanced" version of the query provided in the post includes this filter.`
`25`		`-`
`26`		`----------`
`27`		`-`
`28`	`10`	`Let's discuss today how to find and clean up redundant indexes.`
`29`	`11`
`30`	`12`	`## What is a redundant index?`

Comments

(0)