Rechercher dans le manuel MySQL

8.2.4 Optimizing Performance Schema Queries

Table des matières

Applications that monitor databases may make frequent use of Performance Schema tables. To write queries for these tables most efficiently, take advantage of their indexes. For example, include a WHERE clause that restricts retrieved rows based on comparison to specific values in an indexed column.

Most Performance Schema tables have indexes. Tables that do not are those that normally contain few rows or are unlikely to be queried frequently. Performance Schema indexes give the optimizer access to execution plans other than full table scans. These indexes also improve performance for related objects, such as sys schema views that use those tables.

To see whether a given Performance Schema table has indexes and what they are, use SHOW INDEX or SHOW CREATE TABLE:

  1. mysql> SHOW INDEX FROM performance_schema.accounts\G
  2. *************************** 1. row ***************************
  3.         Table: accounts
  4.    Non_unique: 0
  5.      Key_name: ACCOUNT
  6.  Seq_in_index: 1
  7.   Column_name: USER
  8.     Collation: NULL
  9.   Cardinality: NULL
  10.      Sub_part: NULL
  11.        Packed: NULL
  12.          Null: YES
  13.    Index_type: HASH
  14.       Comment:
  15. Index_comment:
  16.       Visible: YES
  17. *************************** 2. row ***************************
  18.         Table: accounts
  19.    Non_unique: 0
  20.      Key_name: ACCOUNT
  21.  Seq_in_index: 2
  22.   Column_name: HOST
  23.     Collation: NULL
  24.   Cardinality: NULL
  25.      Sub_part: NULL
  26.        Packed: NULL
  27.          Null: YES
  28.    Index_type: HASH
  29.       Comment:
  30. Index_comment:
  31.       Visible: YES
  32.  
  33. mysql> SHOW CREATE TABLE performance_schema.rwlock_instances\G
  34. *************************** 1. row ***************************
  35.        Table: rwlock_instances
  36. Create Table: CREATE TABLE `rwlock_instances` (
  37.   `NAME` varchar(128) NOT NULL,
  38.   `OBJECT_INSTANCE_BEGIN` bigint(20) unsigned NOT NULL,
  39.   `WRITE_LOCKED_BY_THREAD_ID` bigint(20) unsigned DEFAULT NULL,
  40.   `READ_LOCKED_BY_COUNT` int(10) unsigned NOT NULL,
  41.   PRIMARY KEY (`OBJECT_INSTANCE_BEGIN`),
  42.   KEY `NAME` (`NAME`),
  43.   KEY `WRITE_LOCKED_BY_THREAD_ID` (`WRITE_LOCKED_BY_THREAD_ID`)
  44. ) ENGINE=PERFORMANCE_SCHEMA DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci

To see the execution plan for a Performance Schema query and whether it uses any indexes, use EXPLAIN:

  1. mysql> EXPLAIN SELECT * FROM performance_schema.accounts
  2.        WHERE (USER,HOST) = ('root','localhost')\G
  3. *************************** 1. row ***************************
  4.            id: 1
  5.   select_type: SIMPLE
  6.         table: accounts
  7.    partitions: NULL
  8.          type: const
  9. possible_keys: ACCOUNT
  10.           key: ACCOUNT
  11.       key_len: 278
  12.           ref: const,const
  13.          rows: 1
  14.      filtered: 100.00
  15.         Extra: NULL

The EXPLAIN output indicates that the optimizer uses the accounts table ACCOUNT index that comprises the USER and HOST columns.

Performance Schema indexes are virtual: They are a construct of the Performance Schema storage engine and use no memory or disk storage. The Performance Schema reports index information to the optimizer so that it can construct efficient execution plans. The Performance Schema in turn uses optimizer information about what to look for (for example, a particular key value), so that it can perform efficient lookups without building actual index structures. This implementation provides two important benefits:

  • It entirely avoids the maintenance cost normally incurred for tables that undergo frequent updates.

  • It reduces at an early stage of query execution the amount of data retrieved. For conditions on the indexed columns, the Performance Schema efficiently returns only table rows that satisfy the query conditions. Without an index, the Performance Schema would return all rows in the table, requiring that the optimizer later evaluate the conditions against each row to produce the final result.

Performance Schema indexes are predefined and cannot be dropped, added, or altered.

Performance Schema indexes are similar to hash indexes. For example:

  • They are used only for equality comparisons that use the = or <=> operators.

  • They are unordered. If a query result must have specific row ordering characteristics, include an ORDER BY clause.

For additional information about hash indexes, see Section 8.3.9, “Comparison of B-Tree and Hash Indexes”.


Rechercher dans le manuel MySQL

Traduction non disponible

Le manuel MySQL n'est pas encore traduit en français sur l'infobrol. Seule la version anglaise est disponible pour l'instant.

Document créé le 26/06/2006, dernière modification le 26/10/2018
Source du document imprimé : https://www.gaudry.be/mysql-rf-performance-schema-optimization.html

L'infobrol est un site personnel dont le contenu n'engage que moi. Le texte est mis à disposition sous licence CreativeCommons(BY-NC-SA). Plus d'info sur les conditions d'utilisation et sur l'auteur.

Références

  1. Consulter le document html Langue du document :en Manuel MySQL : https://dev.mysql.com/

Ces références et liens indiquent des documents consultés lors de la rédaction de cette page, ou qui peuvent apporter un complément d'information, mais les auteurs de ces sources ne peuvent être tenus responsables du contenu de cette page.
L'auteur de ce site est seul responsable de la manière dont sont présentés ici les différents concepts, et des libertés qui sont prises avec les ouvrages de référence. N'oubliez pas que vous devez croiser les informations de sources multiples afin de diminuer les risques d'erreurs.

Table des matières Haut