比较不同数据库模式的性能
我试图通过更改数据库模式来升级现有系统的性能，因为我认为旧模式不太适合存储大量数据。
现在我不知道如何指出新模式的影响。我应该使用什么指标来比较这两个使用不同模式的系统的性能？
提前致谢..
I am trying to upgrade the performance of an existing system by changing the database schema, because I think the old schema is not so good for storing a large data.
Now I'm not sure how to point out the effects of the new schema. What metrics should I use to compare the performance of those two systems that using different schemas?
Thanks in advance..
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（2）
那么，哪些指标让您得出需要重新设计模式的结论？
数据库性能不佳的最常见原因是 SQL 编写不当。有时，这是因为数据库模式阻止了写入最佳查询，但通常是因为开发人员对 SQL 或数据库设计不够了解，无法编写良好的查询，如果调用得足够频繁，可能会导致一两个错误的查询。杀死整个系统。
Well, what metrics have lead you to the conclusion that you need to re-design the schema?
The most common cause of poor database performance is badly-written SQL. Occasionally that is because the database schema prevents the writing of optimal queries, but usually it is because the developer didn't understand SQL or the database design well enough to write decent queries. Shonky SQL can cause all sorts of problems; one or two bad queries, if they are called often enough, can kill an entire system.
仅更改架构是不够的。您需要检查您的索引策略。
逻辑读取的数量是确定查询工作负载的好方法。
在代表正常活动（以及不经常发生的活动，例如月末报告）的时间段内运行分析跟踪，捕获逻辑读取。
Changing the schema alone is not sufficient. You need to examine your indexing strategy.
The number of logical reads is a good way of determining query workload.
Run a profiling trace for a time period that represents normal activity (and activity that occurs infrequently, such as end of month reporting), capturing logical reads.