如何在MySQL中不使用子查询查询特定记录（千万级大表）

Got it, let's organize your database table details clearly with proper formatting, and throw in a couple quick tips since you're dealing with massive 100M-row tables—those can get tricky if you don't plan ahead!

• uph表：存储订单详情，包含约1亿条记录
• urs表：存储客户详情，包含约1亿条记录

Below is the output of the DESC uph; command:

mysql> desc uph;
+------------+--------------+------+-----+---------+----------------+
| Field      | Type         | Null | Key | Default | Extra          |
+------------+--------------+------+-----+---------+----------------+
| id         | int(11)      | NO   | PRI | NULL    | auto_increment |
| uid        | int(11)      | NO   |     | NULL    |                |
| order_from | varchar(255) | NO   |     | ...     |                |
+------------+--------------+------+-----+---------+----------------+

Since you're working with such large datasets, here are a few actionable tips to avoid performance bottlenecks:

• Add a regular index on uph.uid—this will drastically speed up joins between the order table (uph) and customer table (urs)
• Double-check that the corresponding customer ID field in urs (likely id) is set as the primary key or has a unique index—this ensures efficient lookups during joins
• If you frequently filter queries by order_from, consider a prefix index for this field (e.g., CREATE INDEX idx_order_from ON uph(order_from(10));) instead of a full-length index, to save storage and improve query speed

内容的提问来源于stack exchange，提问作者Arti Singh