Micah Lerner writes summaries of academic and industry computer science papers, mostly on distributed systems and infrastructure.
The author describes it as a way to learn and share research. Commenters call the posts well presented and link them when related papers come up.
Very nicely presented. The terminology here paints a easy to understand picture. There's an ideal form, & a solidication/realization of that intent. Annealing is a great word for the realization of an intent.
Semi sort of related, I notice a little passing similarity in Delos, Facebook's Log-Structured Protocol, comprised of two primary verbs that are sent around: apply, to apply an intent to the underlying platform, and sync, for the platform to report newer state up to the app.
Delos is not focused on this infrastructure topic, but it describes an intent-based system, where eventual resolution is expected. I could imagine using Delos or something like it as a way to actuate infrastructure in a similar manner to this paper. I also really enjoy a lot of the neat composability capabilities of Delos.
Kind of hard finding what to cite for Delos. There's dozens of submissions here on HN, but a lot of them are abstract only access, none have gotten much traction.
https://www.usenix.org/conference/osdi20/presentation/balakrishnan
https://www.micahlerner.com/2021/11/23/log-structured-protocols-in-delos.html
https://engineering.fb.com/2019/06/06/data-center-engineering/delos/
I'm surprised that they published the more recent paper discussed in part 2 [0] since, IIRC, TAO has supported atomic transactions since at least when I was there a few years ago, but better late than never I suppose. I wonder if that means there's still hope that someone at the company will write a blog post about frameworks like:
- EntSchema/EntPattern
- EntQL
- EntAction
- PrivacyPolicy
- Ent Deletion framework
- Ent consistency checkers/fixers (I forget the name of the framework/tool)
It's sad to think that people's only glimpse into these tools is through https://entgo.io which is a really crude approximation of the real thing due to Go's lack of generics and expressiveness. Imagine if such amazing tools could be used at actually ethical companies.
[0] https://www.micahlerner.com/2021/10/23/ramp-tao-layering-atomic-transactions-on-facebooks-online-tao-data-store.html
"Specifically, Homa aims to replace TCP, which was designed in the era before modern data center environments existed. Consequently, TCP doesn’t take into account the unique properties of data center networks (like high-speed, high-reliability, and low-latency). Furthermore, the nature of RPC traffic is different - RPC communication in a data center often involve enormous amounts of small messages and communication between many different machines."[0]
0: https://www.micahlerner.com/2021/08/15/a-linux-kernel-implementation-of-the-homa-transport-protocol.html
https://www.micahlerner.com
I write about new and foundational academic CS research - writing is a way for me to learn and share with others along the way.
Some of my most popular writing is:
- FoundationDB: A Distributed Unbundled Transactional Key Value Store (https://news.ycombinator.com/item?id=28740497)
- Monarch: Google’s Planet-Scale In-Memory Time Series Database (https://news.ycombinator.com/item?id=31379383)
- Ray: A Distributed Framework for Emerging AI Applications (https://news.ycombinator.com/item?id=27730807)
In case someone finds it interesting, kernel bypass also exposes hardware capabilities which can be used for optimization. A recent research attempts at leveraging NICs to accelerate data serialization https://www.sigops.org/s/conferences/hotos/2021/papers/hotos21-s10-raghavan.pdf
Source: https://www.micahlerner.com/2021/07/07/breakfast-of-champions-towards-zero-copy-serialization-with-nic-scatter-gather.html
Just last week, the Meta paper about their Defcon system landed on HN front page.
https://www.micahlerner.com/2023/07/23/defcon-preventing-overload-with-graceful-feature-degradation.html