Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

OFC: An Opportunistic Caching System for FaaS Platforms #10

Open
linxuyalun opened this issue Apr 27, 2021 · 3 comments
Open

OFC: An Opportunistic Caching System for FaaS Platforms #10

linxuyalun opened this issue Apr 27, 2021 · 3 comments
Labels
done Finish reading serverless Papers about serverless

Comments

@linxuyalun
Copy link
Owner

EuroSys ’21
https://doi.org/10.1145/3447786.3456239

@linxuyalun linxuyalun added todo todo serverless Papers about serverless labels Apr 27, 2021
@linxuyalun
Copy link
Owner Author

OFC 这篇文章从整体的 idea 而言在于利用用户定义的内存总是 over provide 的特点,通过 ML 的手段给函数部署真正需要的内存大小,然后多出来的部分来作为 cache 减少中间函数的传输代价。

@linxuyalun
Copy link
Owner Author

三个 challenge:

  1. 怎么知道每个函数真正要用的内存是多少?
  2. 怎么保证本地 cache 系统的垂直扩展能力?
  3. 怎么保证 cache 系统的 transparent,efficient 和 reliable?

@linxuyalun
Copy link
Owner Author

  1. 使用 ML 的方法去做预测;
  2. 第二个 challenge 的难点在于当一个节点内存不够时能够 scaling down。论文采用两种策略,业务上来讲,尽可能驱逐已经没用的 cache,从实现来说,提出一个算法用来进行 object 迁移;
  3. 这个挑战本质上也是提出一些策略去解决,首先只有当一个数据可能改善执行时间时才会缓存,然后 cache 存储的时候采用异步存储以及一些其他常见的技术手段。主要也是通过 ML 和 cache 的思路。

@linxuyalun linxuyalun added done Finish reading and removed todo todo labels May 1, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
done Finish reading serverless Papers about serverless
Projects
None yet
Development

No branches or pull requests

1 participant