Explain / Desc execution plan analysis

Simple Instructions

In MySQL, both explain and desc can generate information about SQL execution by the optimizer, so no matter which one is used, the result is the same, this article only usesdesc for description.

The following is a query statement and the corresponding SQL execution plan:

As can be seen from the figure, the SQL execution plan shows 12element information.

id

The first is id, which indicates the execution order of the operation table of the select statement.
The larger the value of id, the higher the priority.
When id is the same, it is generally executed in order from top to bottom, but it is determined by the optimizer.
When the id is different, there may be nested sub-queries, and the sub-queries have higher priority and are executed first.

select_type

The second is select_type, which distinguishes between different query types, including SIMPLE, PRIMARY,
SUBQUERY , DERIVED, UNION, UNION RESULT.

SIMPLE--Simplest query, excluding subqueries and intersection unions.PRIMARY-- contains subsections, the outermost is PRIMARY.SUBQUERY-- select or where contains a subqueryDERIVED-- Subquery after from ; If UNION is after from, the outermost layer is DERIVEDUNION-- the select statement after union is UNION

table

The third one is the table name.
may also be a temporary table, an alias for a table.

partitions

The fourth represents the partition information of the query table.

null if there is no partitionIf there is a partition, the partition status will be displayed

type

The fifth shows the type of query, reflecting the difference in performance.

The performance of SQL optimization from good to bad is as follows:system > const > eq_ref > ref > ref_or_null > index_merge > unique_subquery > index_subquery > range > index >ALL

1) system: There is only one row of records in the table, the amount of data is very small, and no disk IO is required

2) const: Hit the primary key or hit the unique index or hit a constant value when querying.The at most one matching line.

3) eq_ref: Same as const usage.The difference is that eq_ref is used for join table queries (composed of a row).

4) ref: Compared with eq_ref, a non-unique index can be used, and multiple rows of query results will be returned.

5) ref_or_null: Compared with ref, the query result returns a value containing null.

6) index_merge: Index merge optimization, use more than two indexes when querying.

7) unique_subquery: replace the IN subquery, which returns unique values.

8) index_subquery: Different from unique_subquery, it is used for non-unique indexes and can return duplicate values.

9) range: For the index field, query the rows of a given range.Generally, where is followed by bettween...and, <,>, <=, in and other conditional queries

10) index: traverse the index tree and perform a full table scan.

11) all: read from the hard disk, full table scan.

possible_keys

The sixth means that all indexes contained in this field will be listed, but these indexes will not necessarily go in the end.

key

The seventh represents the index actually used.

key_len

The eighth indicates the index length (in bytes) used by the query. The shorter the index, the better the performance.

Only the index length of the where condition column is calculated, and the index of the sorting and grouping statements is not counted.

ref

The ninth indicates the column that the index is used for.

When using constant equivalent query, display const,When the query is related, the related fields of the corresponding related table will be displayedIf the query condition uses an expression, function, or the condition column is internally implicitly converted, it may be displayed as funcOtherwise it shows null

rows

The tenth indicates that the system estimates the number of lines that should be read to find the desired result.
The smaller the value, the better the performance.

filtered

The eleventh indicates the percentage of record data in the table that meet the conditions.

Extra

The twelfth represents additional information that cannot be displayed in the first eleven columns.

1) Using index: Covering index is used.

2) Using where: The query does not go through the index, and the corresponding result set is matched through the where condition.

3) Using temporary: The results of the query are stored in a temporary table, usually sorted or grouped.

4) Using filesort: The sorting operation cannot be completed with an index. The ORDER BY field has no index and needs to be optimized.

5) Impossible where: The desired result could not be found due to an incorrect where statement.

6) No tables used: There is no from statement.