Find non-function based indexes (both unique and non-unique) that duplicate each other because their first column is identical and the operator class that is used in case of the first column is identical. Include unique indexes that support a constraint (primary key, unique, exclude), i.e., these indexes have been automatically created due to the constraint declaration.
Notes
Find non-function based indexes (both unique and non-unique) that have identical first column but the operator class that is used in case of the first column is different. Include unique indexes that support a constraint (primary key, unique, exclude), i.e., these indexes have been automatically created due to the constraint declaration. Exclude pairs of indexes where both support a constraint.
Type
Problem detection (Each row in the result could represent a flaw in the design)
Drop the redundant indexes. Do not drop indexes that suport constraints.
Data Source
system catalog only
SQL Query
WITH indexes AS (SELECT
n.nspname AS sch,
c2.relname AS tbl,
c.relname AS indx,
array_agg(a.attname ORDER BY a.attnum) AS clm,
i.indisunique,
indclass[0] AS operator_class,
EXISTS (SELECT * FROM pg_catalog.pg_constraint AS o WHERE o.conindid=i.indexrelid) AS is_constraint
FROM
pg_catalog.pg_class AS c INNER JOIN pg_catalog.pg_index AS i ON i.indexrelid = c.oid
INNER JOIN pg_catalog.pg_class AS c2 ON i.indrelid = c2.oid
INNER JOIN pg_catalog.pg_namespace AS n ON c.relnamespace = n.oid
INNER JOIN pg_catalog.pg_authid AS u ON n.nspowner = u.oid
INNER JOIN pg_catalog.pg_attribute AS a ON a.attrelid = c.oid
WHERE c.relkind = 'i'
AND i.indexprs IS NULL
AND a.attnum>=1
AND a.attisdropped = FALSE
AND (n.nspname = 'public'
OR u.rolname <> 'postgres')
GROUP BY n.nspname, c2.relname, c.relname, i.indisunique, indclass[0], is_constraint),
index_with_type AS (SELECT sch, tbl, indx, clm, operator_class, is_constraint,
CASE WHEN is_constraint=TRUE THEN 'UNIQUE CONSTRAINT INDEX'
WHEN indisunique=TRUE AND is_constraint=FALSE THEN 'UNIQUE NON-CONSTRAINT INDEX'
ELSE 'NON-UNIQUE INDEX' END AS index_type
FROM indexes)
SELECT A.sch AS table_schema, A.tbl AS table_name, A.index_type AS idx1_index_type, A.indx AS idx1, A.clm AS idx1_columns,
B.clm AS idx2_columns, B.index_type AS idx2_index_type, B.indx AS idx2
FROM index_with_type AS A, index_with_type AS B
WHERE A.sch=B.sch AND A.tbl=B.tbl AND A.indx>B.indx AND A.clm[1]=B.clm[1] AND A.operator_class=B.operator_class
AND NOT (a.is_constraint=TRUE AND b.is_constraint=TRUE)
ORDER BY A.sch, A.tbl, A.indx, B.indx;
SQL statements that help generate fixes for the identified problem.
SQL Query to Generate Fix
Description
WITH indexes AS (SELECT
n.nspname AS sch,
c2.relname AS tbl,
c.relname AS indx,
array_agg(a.attname ORDER BY a.attnum) AS clm,
i.indisunique,
indclass[0] AS operator_class,
EXISTS (SELECT * FROM pg_catalog.pg_constraint AS o WHERE o.conindid=i.indexrelid) AS is_constraint
FROM
pg_catalog.pg_class AS c INNER JOIN pg_catalog.pg_index AS i ON i.indexrelid = c.oid
INNER JOIN pg_catalog.pg_class AS c2 ON i.indrelid = c2.oid
INNER JOIN pg_catalog.pg_namespace AS n ON c.relnamespace = n.oid
INNER JOIN pg_catalog.pg_authid AS u ON n.nspowner = u.oid
INNER JOIN pg_catalog.pg_attribute AS a ON a.attrelid = c.oid
WHERE c.relkind = 'i'
AND i.indexprs IS NULL
AND a.attnum>=1
AND a.attisdropped = FALSE
AND (n.nspname = 'public'
OR u.rolname <> 'postgres')
GROUP BY n.nspname, c2.relname, c.relname, i.indisunique, indclass[0], is_constraint),
index_with_type AS (SELECT sch, tbl, indx, clm, operator_class, is_constraint,
CASE WHEN is_constraint=TRUE THEN 'UNIQUE CONSTRAINT INDEX'
WHEN indisunique=TRUE AND is_constraint=FALSE THEN 'UNIQUE NON-CONSTRAINT INDEX'
ELSE 'NON-UNIQUE INDEX' END AS index_type
FROM indexes),
duplicates AS (SELECT A.sch AS table_schema, A.tbl AS table_name, A.index_type AS idx1_index_type, A.indx AS idx1, A.clm AS idx1_columns,
B.clm AS idx2_columns, B.index_type AS idx2_index_type, B.indx AS idx2
FROM index_with_type AS A, index_with_type AS B
WHERE A.sch=B.sch AND A.tbl=B.tbl AND A.indx>B.indx AND A.clm[1]=B.clm[1] AND A.operator_class=B.operator_class
AND NOT (a.is_constraint=TRUE AND b.is_constraint=TRUE))
SELECT format('DROP INDEX %1$I.%2$I;', table_schema, idx1) AS statements
FROM duplicates
WHERE idx1_index_type<>'UNIQUE CONSTRAINT INDEX'
UNION SELECT format('DROP INDEX %1$I.%2$I;', table_schema, idx2) AS statements
FROM duplicates
WHERE idx2_index_type<>'UNIQUE CONSTRAINT INDEX'
ORDER BY statements;
Drop the index. Choose, which index to drop and which index should stay in place.
Collections
This query belongs to the following collections:
Name
Description
Find problems automatically
Queries, that results point to problems in the database. Each query in the collection produces an initial assessment. However, a human reviewer has the final say as to whether there is a problem or not .
Categories
This query is classified under the following categories:
Name
Description
Duplication of implementation elements
Queries of this catergory provide information about the duplication of the database objects.
Performance
Queries of this category provide information about indexes in a database.
Further reading and related materials:
Reference
Smell "Index abuse": Sharma, T., Fragkoulis, M., Rizou, S., Bruntink, M. and Spinellis, D.: Smelly relations: measuring and understanding database schema quality. In: Proceedings of the 40th International Conference on Software Engineering: Software Engineering in Practice, pp. 55-64. ACM, (2018).