知识卡片
扩展目标要有单机基线
可扩展性是在增加资源时提高吞吐并维持可接受延迟,不是无限扩容。先测单节点能力,再按可预见负载和重写时间窗口定目标。多份无状态服务加共享数据库能覆盖很多需求,数据库成为共同瓶颈时再用缓存或后台处理缓解。不要因“像大型平台”而提前引入分布式复杂度,也不能把书内TPS示例当自己系统的实测容量。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- 高可用与扩展 · 查看原书依据
Before scaling, you need to understand the limits of a single node and your performance goals. At a minimum, you need to do a PoC to understand the single-node performance. Sometimes, it is possible to tweak the single node to achieve your performance goals. If so, you should use that option because it will save you a lot of complexity in development and operation.
参考来源
- 位置:《Software Architecture and Decision-Making》第09章《高可用与扩展》(来源项目:book-sources;源文件:extracted/Software Architecture and Decision-Making Leveraging Leadership, Technology, and Product Management to Build Great Products… (Srinath Perera)/_epub-src/ch09.xhtml)
- 结论依据:所引连续原文给出本卡机制或取舍;条件与示例结合本节已读上下文解释。
- 原始内容:Before scaling, you need to understand the limits of a single node and your performance goals. At a minimum, you need to do a PoC to understand the single-node performance. Sometimes, it is possible to tweak the single node to achieve your performance goals. If so, you should use that option because it will save you a lot of complexity in development and operation.