问题描述
我在 Redshift 中每小时自动执行一系列 ~10 个查询(可能报告成功/失败).
I have series of ~10 queries to be executed every hour automatically in Redshift (maybe report success/failure).
大多数查询都是对我的表进行聚合.
Most queries are aggregation on my tables.
我尝试将 AWS Lambda 与 CloudWatch Events 结合使用，但 Lambda 函数最多只能存活 5 分钟，而我的查询最多需要 25 分钟.
I have tried using AWS Lambda with CloudWatch Events, but Lambda functions only survive for 5 minutes max and my queries can take up to 25 minutes.
推荐答案
AWS 不提供简单的分布式 cron 样式服务有点奇怪.它对很多事情都有用.有 SWF，但时间/调度方面由用户决定.您可以使用 Lambda/Cloudwatch 来触发 SWF 事件.获得合理的 cron 之类的活动需要大量开销.
It's kind of strange that AWS doesn't provide a simple distributed cron style service. It would be useful for so many things. There is SWF, but the timing/scheduling aspect is left up to the user. You could use Lambda/Cloudwatch to trigger SWF events. That's a lot of overhead to get reasonable cron like activity.
就像评论所说，最简单的方法是运行一个小实例并在那里托管 cron 作业.使用自动缩放组 1 以获得一些可靠性.一种类似但更复杂的方法是使用 elastic beanstalk.
Like the comment says the easiest way would be to run a small instance and host cron jobs there. Use an autoscale group of 1 for some reliability. A similar but more complicated approach is to use elastic beanstalk.
如果您真的想要冗余、可靠性、可见性等，可能值得查看 第三方解决方案，例如Airflow.根据您的偏好语言，还有许多其他语言.
If you really want redundancy, reliability, visibility, etc. it might be worth looking at a third party solution like Airflow. There are many others depending on your language of preference.
这是一个类似问题，提供更多信息.
Here's a similar question with more info.
这篇关于如何在 Amazon Redshift 上执行预定的 SQL 脚本?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！