Blog/Papers I loved - November 2023

It just occurred to me that I've been reading some (scientific) papers lately, mostly due to my current job that's way more challenging and interesting than my previous one (albeit, way more intense as well).

So I decided to keep track of the papers I've read (any why I liked them).

The papers in this pages aren't all from November 2023, but this is the first page in what could become a series so here we go...

For the time being, I'm mainly interested in keeping track of them, I'll come back one day and write more on each of those papers.

A Gossip-Style Failure Detection Service

"A Gossip-Style Failure Detection Service" is an interesting paper describing how to implement failure detection in a distributed system.

It's interesting because I've seen (a variation of) it implemented at work.

The DynamoDB paper

This paper (Amazon DynamoDB: A Scalable, Predictably Performant, and Fully Managed NoSQL Database Service) is nice because it describes a real-world very large system and (as a user) helps clear (and motivate) some of the terminology in DynamoDB.

sendmsg copy avoidance with MSG_ZEROCOPY

The "sendmsg copy avoidance with MSG_ZEROCOPY" is great not just for the topic itself, but because it sheds a lot of light on the amount of issues that needs to be accounted for when implementing something like this.

I particularly liked it because there were some issues I didn't even know existed, and now I have more known unknowns and less unknown unknowns.

The paper was presented at NetDevConf 2.1 in 2017 by Willem de Bruijn from Google.

In the page, besides the paper itself, there is also a link to the presentation recording and the slides from the recording.