知识卡片
毒消息需要运维上下文
毒消息携带无法处理的坏数据,重复投递不会修复内容,只增加负载;重试耗尽后进入死信队列仍需要人工诊断、修正和重投能力。消息经过多条渠道与服务后,来源上下文容易丢失。失败流程实例能保留起点、路径和变量,帮助定位根因,但运维工具仍需明确处理责任。
这张卡片看不懂?
这张卡片的反馈已满 10 条,暂时无法继续提交。
已有反馈 · 0 / 10 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- 工作流引擎与集成挑战 · 查看原书依据
My favorite examples are so-called poisoned messages. Suppose your service receives new customer orders via messaging. There is a bug in the frontend that puts some broken data into that message, making it “poisoned.” Your service will throw exceptions when processing the message. There is no client you can hand this exception to, so the
参考来源
- 位置:《Practical Process Automation》第9章《工作流引擎与集成挑战》(来源项目:book-sources;源文件:extracted/Practical Process Automation (Bernd Ruecker)/_epub-src/OEBPS/ch09.xhtml)
- 结论依据:原文给出的机制、实例与边界支持本卡的解释。
- 原始内容:My favorite examples are so-called poisoned messages. Suppose your service receives new customer orders via messaging. There is a bug in the frontend that puts some broken data into that message, making it “poisoned.” Your service will throw exceptions when processing the message. There is no client you can hand this exception to, so the