I built Ayder — a single-binary, HTTP-native durable event log written in C. The wedge is simple: curl is the client (no JVM, no ZooKeeper, no thick client libs).
There’s a 2-minute demo that starts with an unclean SIGKILL, then restarts and verifies offsets + data are still there.
Numbers (3-node Raft, real network, sync-majority writes, 64B payload): ~50K msg/s sustained (wrk2 @ 50K req/s), client P99 ~3.46ms. Crash recovery after SIGKILL is ~40–50s with ~8M offsets.
Repo link has the video, benchmarks, and quick start. I’m looking for a few early design partners (any event ingestion/streaming workload).
I wish there was a standard protocol for consuming event logs, and that all the client side tooling for processing them didn't care what server was there.
I was part of making this:
https://github.com/vippsas/feedapi-spec
https://github.com/vippsas/feedapi-spec/blob/main/SPEC.md
I hope some day there will be a widespread standard that looks something like this.
An ecosystem building on Kafka clients libraries with various non-Kafka servers would work fine too, but we didn't figure out how to easily do that.
For benchmarks: I used real network (not loopback) and sync-majority writes in a 3-node Raft cluster. Happy to answer questions about tradeoffs vs Kafka / Redis Streams and what’s still missing.
VSR makes a lot of sense for their problem space: fixed schema, deterministic state machine, and a very tight control over replication + execution order.
Ayder has a different set of constraints: - append-only logs with streaming semantics - dynamic topics / partitions - external clients producing arbitrary payloads over HTTP
Raft here is a pragmatic choice: it’s well understood, easier to reason about for operators, and fits the “easy to try, easy to operate” goal of the system.
That said, I think VSR is a great example of what’s possible when you fully own the problem and can specialize aggressively. Definitely a project I’ve learned from.
I think there’s a similar philosophy around simplicity and operator experience. Where Ayder diverges is in durability and recovery semantics nsq intentionally trades some of that off to stay lightweight.
The goal here is to keep the “easy to run” feeling, but with stronger guarantees around crash recovery and replication.
Thank you for sharing this with us.
> Numbers are real, not marketing.
I'm not questioning the actual benchmarks or anything, but this README is substantially AI generated, yeah?
The benchmarks, logs, scripts, and recovery scenarios are all real and hand-run that’s the part I care most about being correct.
For the README text itself: I did iterate on wording and structure (including tooling), but the system, measurements, and tradeoffs are mine.
If any part reads unclear or misleading, I’m very open to tightening it up. Happy to clarify specifics.
When I read this type of prose it makes me feel like the author is more worried about trying to sell me something than just describing the project.
For instance, you don't need to tell me the numbers are "real". You just have to show me they're covering real-world use-cases, etc. LLMs love this sort of "telling not showing" where it's constantly saying "this is what I'm going to tell you, this is what I'm telling you, this is what I told you" structure. They do it within sections and then again at higher levels. They have, I think, been overindexed on "five-paragraph essays". They do it way more than most human writers do.
You're welcome to make your substantive points thoughtfully, of course.
Also if you disapprove, modding down is enough, you don't need to start a meta-discussion thread, which is itself a discouraged practice.
The point of these numbers is the envelope: 3-node consensus (Raft), real network (not loopback), and sync-majority writes (ACK after 2/3 replicas) plus the crash/recovery semantics (SIGKILL → restart → offsets/data still there).
If you have a quick Python setup that does majority-acked replication + fast crash recovery with similar measurements, I’d honestly love to compare apples-to-apples happy to share exact scripts/config and run the same test conditions.