编写查询的开发人员没有很好地了解SQL应用程序(包括单个查询和多用户系统)的快慢。这包括对以下方面的无知:

physical I/O minimization strategies, given that most queries' bottleneck is I/O not CPU perf impact of different kinds of physical storage access (e.g. lots of sequential I/O will be faster than lots of small random I/O, although less so if your physical storage is an SSD!) how to hand-tune a query if the DBMS produces a poor query plan how to diagnose poor database performance, how to "debug" a slow query, and how to read a query plan (or EXPLAIN, depending on your DBMS of choice) locking strategies to optimize throughput and avoid deadlocks in multi-user applications importance of batching and other tricks to handle processing of data sets table and index design to best balance space and performance (e.g. covering indexes, keeping indexes small where possible, reducing data types to minimum size needed, etc.)

没有使用With子句或适当的连接并依赖子查询。

反模式:

select 
 ...
from data
where RECORD.STATE IN (
          SELECT STATEID
            FROM STATE
           WHERE NAME IN
                    ('Published to test',
                     'Approved for public',
                     'Published to public',
                     'Archived'
                    ))

好: 我喜欢使用with子句使我的意图更易于阅读。

with valid_states as (
          SELECT STATEID
            FROM STATE
           WHERE NAME IN
                    ('Published to test',
                     'Approved for public',
                     'Published to public',
                     'Archived'
                    )
select  ... from data, valid_states
where data.state = valid_states.state

最好的:

select 
  ... 
from data join states using (state)
where 
states.state in  ('Published to test',
                     'Approved for public',
                     'Published to public',
                     'Archived'
                    )

我最担心的是450列的访问表，这些表是由总经理最好的朋友狗美容师的8岁儿子整理的，还有那个不可靠的查找表，它之所以存在，是因为有人不知道如何正确地规范化数据结构。

通常，这个查找表是这样的:

ID INT,
Name NVARCHAR(132),
IntValue1 INT,
IntValue2 INT,
CharValue1 NVARCHAR(255),
CharValue2 NVARCHAR(255),
Date1 DATETIME,
Date2 DATETIME

我已经记不清有多少客户的系统依赖于这种可恶的东西了。

对于存储时间值，应该只使用UTC时区。不应使用当地时间。

不必深入浅出:不使用准备好的语句。

Human readable password fields, egad. Self explanatory. Using LIKE against indexed columns, and I'm almost tempted to just say LIKE in general. Recycling SQL-generated PK values. Surprise nobody mentioned the god-table yet. Nothing says "organic" like 100 columns of bit flags, large strings and integers. Then there's the "I miss .ini files" pattern: storing CSVs, pipe delimited strings or other parse required data in large text fields. And for MS SQL server the use of cursors at all. There's a better way to do any given cursor task.

编辑是因为有太多了!