组合索引：最左优先原则

如：组合索引(col_a,col_b,col_c)

使用到索引的情况如下：

col_a = "some value"

col_a = "some value" and col_b = "some value"

col_a = "some value" and col_c = "some value"

col_a = "some value" and col_b = "some value" and col_c = "some value"

col_a = "some value" and col_c = "some value" and col_b = "some value"

col_b = "some value" and col_a = "some value" and col_c = "some value"

col_b = "some value" and col_c = "some value" and col_a = "some value"

col_c = "some value" and col_a = "some value" and col_b = "some value"

col_c = "some value" and col_b = "some value" and col_a = "some value"

未使用索引的情况

col_b = "some value"

col_c = "some value"

col_b = "some value" and col_c = "some value"

col_c = "some value" and col_b = "some value"

总结，组合索引从最左开始组合，只要包含第一列的查询都会用到该组合索引

为什么组合索引遵循最左优先原则？

索引方法是Btree，树状的，搜索时需要从根节点出发，上层节点对应靠左的值，所以有最左优先原则。