[MySQL] index invalidation and index optimization

BTree How index works

   If the index is index（a,b,c）, Will sort first a, If the same a, Reordering b, If the same b, Reordering c; If at this point in the middle b The query conditions are b>1, This is a range value range,mysql Cannot take advantage of the index that follows a pair of c Partial search , namely range Invalid index after type query field

Index failure

1. If the index column has more than one column , The leftmost prefix rule should be observed in the query criteria , That is, the query starts at the top left of the index and does not skip the columns in the index ; If this rule is not observed , The index will fail

2. Do anything on the index column （ Calculation , function , Automatic or manual type conversion ）, It will cause index invalidation and turn to full table scan

3. A storage index cannot use the column to the right of the range condition in the index

4. Try to use overlay index , Reduce select * , This will only query the index fields , Improved efficiency using index

5. Use is not equal to ！= or <> It will lead to index invalidation and full table scanning

6.like Use wildcards % start （%abc…）mysql The index will fail, resulting in a full table scan

   But in the actual demand , There must be % The opening query , To meet this need , We can use overlay indexes , So it's solved ;

7. The query criteria include is null,is not null You can't use indexes

8. String index is invalid without single quotation marks （ Because the system automatically converts data types by default ）

9.or The index fails when connecting

【 Index failure summary 】

Index optimization

• Single key index , Try to choose the current query Better filtering index
• Composite index , At present query The fields with the best filterability in the index field order can be placed as high as possible
• In combined query , Try to include the current query in where Index of more fields in clause
• Combination query join when ： Reduce NestedLoop cycles , follow “ Always drive large result sets with small results ”; First, optimize the inner loop ; Guarantee join Statement on the driven table join The condition field has been indexed , The best situation is join All tables connected by are indexed ; When there is no guarantee that the driven meter join If the condition field is indexed and the memory resource is sufficient , Don't skimp JoinBuffer Set up

【 Index optimization summary 】