The sub-millisecond writes with data in S3 is false and impossible. If you look at the benchmark the fsync is not timed, so this is just the latency of either the network or in kernel file operations depending on the mount settings
I hate it when databases celebrate their performance without synchronous flushing. You should be clear about data loss window (which should be zero for committed transactions by default!) and the flushing interval to persistent storage.
I'm okay if you batch writes, I'm okay if you offer a low-latency mode with less durability, but by being unclear about this it just feels like a scam.
Yeah in this case the footnote to the write latency specifically says “at rest in S3”, which is what caused me to go look at the source. To be clear I have no problem with the ZeroFS of only flushing on fsync.
I am very excited for object storage first systems like this to leverage low latency zonal storage for write ahead logs to keep the disaggregated storage but greatly reduce write latency. That ends up being more expensive, but is likely a good tradeoff in lots of cases I have seen
ZeroFS aims to be a POSIX filesystem, the semantics here are the standard ones (ext4, xfs behave the same): write() is buffered (that's the batching) and "committed" maps to fsync(), which returns only once data is durable.
Nothing wrong with that, but you should remove the “at rest in S3” footnote from the write latency on the frontpage of the website, because that is not what is measured
Read/write operations in object storage are _far more_ expensive than stored bytes. I'm always afraid of anything that abstracts over S3/GCS access specifically for that reason.
Since you are harnessing the sorcery of AI, have it write really good benchmarks, run tests and comparisons on competitive products, (and publish them), look up common pitfalls, often requested features, run security analysis.
Also with marketing texts, write your self first and then you can ask AI to hone it or give you feedback. AI slopped marketing text is visible from miles and really, really puts people off. Even if the product itself would be fine, there is some much slop slushing around in the pipes at the moment.
I really like this project and want to see it succeed! Don't let naysayers wear you down.
If one of your goals is to get others to adopt the software, I recommend you redo the marketing page and readme from scratch. Delete them without looking at them again, then hand write the content for them. Once you have the content, you call tell an LLM to format it into a nice landing page, but strictly keep your wording without changes.
How does this compare to JuiceFS or SeaweedFS in terms of metadata latency? The LSM tree approach is interesting but compaction pauses on a remote-backed store seem like they could be painful.
Thanks for being explicit, AI written marketing site. Wouldn't have been able to figure that out! Every currently maintained and reasonably popular open source project either runs CI in public or makes the tests extremely easy to run.
> These are asciinema recordings of real terminal sessions, rendered as text rather than video. Playback caps idle pauses at two seconds and changes nothing else.
Thanks? This sounds like it's the LLM's response to the prompter, not something you should display on the page itself...
I feel bad for actually liking that part now. Capping pauses at 2 seconds would show you where it hung 2+ seconds without wasting your time. Smart I thought.
I believe the first version of this required the metadata to be stored on the ZeroFS server, making HA kinda hard.
This has changed now that if I stop the server and create a new instance with the same configuration file it'll pickup the existing metadata from the bucket?
We have done loads of research into using object storage wherever we can (given how cheap it is compared to SSDs), and so far it seems like making your application object store-aware is a far surer bet than abstracting S3 behind the file system. The behavior is just too different.
I'm more interested in applications that cleverly use object storage, e.g. AutoMQ, which is quite compatible with Kafka APIs but needs no HDDs.
The sub-millisecond writes with data in S3 is false and impossible. If you look at the benchmark the fsync is not timed, so this is just the latency of either the network or in kernel file operations depending on the mount settings
I hate it when databases celebrate their performance without synchronous flushing. You should be clear about data loss window (which should be zero for committed transactions by default!) and the flushing interval to persistent storage.
I'm okay if you batch writes, I'm okay if you offer a low-latency mode with less durability, but by being unclear about this it just feels like a scam.
Yeah in this case the footnote to the write latency specifically says “at rest in S3”, which is what caused me to go look at the source. To be clear I have no problem with the ZeroFS of only flushing on fsync.
I am very excited for object storage first systems like this to leverage low latency zonal storage for write ahead logs to keep the disaggregated storage but greatly reduce write latency. That ends up being more expensive, but is likely a good tradeoff in lots of cases I have seen
ZeroFS aims to be a POSIX filesystem, the semantics here are the standard ones (ext4, xfs behave the same): write() is buffered (that's the batching) and "committed" maps to fsync(), which returns only once data is durable.
Nothing wrong with that, but you should remove the “at rest in S3” footnote from the write latency on the frontpage of the website, because that is not what is measured
I don’t get why they went for NFSv3, v4 is quite old and I can’t think of any reason why you would choose v3 over v4.
From the docs:
> ZeroFS fetches object data in 128 KiB parts
Read/write operations in object storage are _far more_ expensive than stored bytes. I'm always afraid of anything that abstracts over S3/GCS access specifically for that reason.
One of the reasons why ZeroFS seems interesting is they use SlateDB under the hood, which optimizes the requests that hit S3 behind the scenes.
Especially that the “one fetch” is who knows how many reads and retries under the hood.
I’ve seen things like this before; your key differentiator needs to be efficiency and safety compared to other options.
OP this is the best advice here.
Since you are harnessing the sorcery of AI, have it write really good benchmarks, run tests and comparisons on competitive products, (and publish them), look up common pitfalls, often requested features, run security analysis.
Also with marketing texts, write your self first and then you can ask AI to hone it or give you feedback. AI slopped marketing text is visible from miles and really, really puts people off. Even if the product itself would be fine, there is some much slop slushing around in the pipes at the moment.
I really like this project and want to see it succeed! Don't let naysayers wear you down.
Thank you, appreciated!
Entrusting data storage to a vibe coded filesystem seems imprudent.
Is it? :)
Ask me anything!
Here is my unsolicited advice:
If one of your goals is to get others to adopt the software, I recommend you redo the marketing page and readme from scratch. Delete them without looking at them again, then hand write the content for them. Once you have the content, you call tell an LLM to format it into a nice landing page, but strictly keep your wording without changes.
That's fair advice, thanks.
FYI it looks like some of your comments are getting auto-flagged by the HN moderation system and marked as dead
How does this compare to JuiceFS or SeaweedFS in terms of metadata latency? The LSM tree approach is interesting but compaction pauses on a remote-backed store seem like they could be painful.
> The test suites run in public CI.
> Each card links to the CI pipeline.
Thanks for being explicit, AI written marketing site. Wouldn't have been able to figure that out! Every currently maintained and reasonably popular open source project either runs CI in public or makes the tests extremely easy to run.
I got the same vibe from
> These are asciinema recordings of real terminal sessions, rendered as text rather than video. Playback caps idle pauses at two seconds and changes nothing else.
Thanks? This sounds like it's the LLM's response to the prompter, not something you should display on the page itself...
I feel bad for actually liking that part now. Capping pauses at 2 seconds would show you where it hung 2+ seconds without wasting your time. Smart I thought.
Thank you for the feedback, the idea behind this was to say "We make claims that are backed by workflows you can verify". I'll improve the phrasing.
I believe the first version of this required the metadata to be stored on the ZeroFS server, making HA kinda hard.
This has changed now that if I stop the server and create a new instance with the same configuration file it'll pickup the existing metadata from the bucket?
> I believe the first version of this required the metadata to be stored on the ZeroFS server, making HA kinda hard.
Metadata has always been in the bucket itself.
For HA, there's now a "replicated mode" if you want automatic failover:
https://www.zerofs.net/docs/high-availability
See also: JuiceFS, S3FS, and quite a few others.
We have done loads of research into using object storage wherever we can (given how cheap it is compared to SSDs), and so far it seems like making your application object store-aware is a far surer bet than abstracting S3 behind the file system. The behavior is just too different.
I'm more interested in applications that cleverly use object storage, e.g. AutoMQ, which is quite compatible with Kafka APIs but needs no HDDs.
s3fs doesn't provide posix semantics. It's good enough™ for some uses, but not comparable to what this one is ostensibly providing.
Seems purely vibecoded
wonder when we get agents good enough that we can't say vibecode any more and have to say 'code'.
there was slop with ai jesus but now gpt image is just a photo with hidden watermark
Why does this landing page load js from merklemap.com?
Both projects have the same author.
Just a self hosted plausible instance :)
Under the hood, S3's storage nodes are also built on a log-structured file system: https://cdn.amazon.science/77/5e/4a7c238f4ce890efdc325df8326...
(Not posix compliant because it doesn't need to be.)