M
Mahesh Balakrishnan
maheshba.bitbucket.io · Personal blog · on HN since 2021
Mahesh Balakrishnan writes about distributed systems, replicated databases, consensus protocols like Paxos and Raft, and engineering leadership.
What HN saysCommenters engage with his arguments on consensus and consistency, citing his post that Raft is less modular and harder to understand than MultiPaxos.
15HN people shared it
3linked it in comments
19stories submitted
959story points
6 months agolast shared
2021people per year2026
Most-shared posts
- 1Things I learned from building a production database3 people · 556 points · first shared 2021 · 106 comments on HN
- 2What we talk about when we talk about System Design3 people · 4 points · first shared 2023 · discussion on HN
- 3The leadership myth in replicated databases (2023)2 people · 62 points · first shared 2023 · 3 comments on HN
- 4The Art of Managing Skunks2 people · 58 points · first shared 2025 · 10 comments on HN
- 5Jeeps, Ferraris, and Other Engineers2 people · 3 points · first shared 2025 · discussion on HN
- 6Paxos Made Abstract2 people · 3 points · first shared 2021 · discussion on HN
- 7Is Raft more modular than MultiPaxos?2 people · 3 points · first shared 2021 · discussion on HN
- 8Agent is a distributed system (and fails like one)2 people · 2 points · first shared 2026 · discussion on HN
- 9Three Laws of Software Complexity1 people · 268 points · first shared 2024 · 138 comments on HN
- 10Osr2024.pdf1 people · first shared 2024
What HN people wrote when they linked it
rubiquity · 2024-09-08
I don’t agree with much in this writing other than that eventual consistency is a bad choice. Distributed systems are hard but in 2024 there are enough known patterns and techniques to make them less icky. Systems built on total ordering are much more tractable than weaker protocols. Mahesh Balakrishnan’s recent paper[0] on the Shared Log abstraction is a great recipe, for example.
As an aside, I’ve never enjoyed the defeatist culture that permeated operations and distributed systems pop culture, which this post seems to re-enforce.
0 - https://maheshba.bitbucket.io/papers/osr2024.pdf
shikhar · 2024-02-08
I believe TigerBeetle are alluding to their integration of protocol-aware recovery [1], which is a worthy consideration for the log implementation. Yet another engineering concern which can be offloaded.
If the disaggregated log integrates some mechanism to support leadership "above" it [2], it can be functionally identical to a converged log. Efficiency-wise yes there will be some extra network messages – but networks are very high throughput [3] and fast (sub-millisecond within a cloud region) these days!
[1] https://www.usenix.org/conference/fast18/presentation/alagappan [2] https://maheshba.bitbucket.io/blog/2023/05/06/Leadership.html, also on HN yesterday [3] https://blog.enfabrica.net/the-next-step-in-high-performance-distribut…
senderista · 2024-09-27
Agree, Raft is less modular and therefore harder to understand than MultiPaxos:
https://maheshba.bitbucket.io/blog/2021/12/14/Modularity.html