Hi! I'm the author and the person who was on a road trip when I decided to do this experiment back in July. I'm super excited to finally get the blog post out, and even better, alongside a post describing the root cause process that went into the project. You can read about Tailscale's story here: https://tailscale.com/blog/sqlite-wal-reset-bug
I'm only sad that I didn't put SQLite under test earlier in the year, or I would have found this issue right away. If you look at the workload[1], you can see how simple it is. Exactly the same kind of workload we write every day to help our customers find bugs.
If you have any questions about our process or how debugging with Antithesis works, please let me know! Thanks for reading!
How difficult would it have been to isolate that problem if you didn't already know the SQLite subsystem it was in? This feels, to someone relatively ignorant of the SQLite / Tailscale / Antithesis architectures as a "hindsight is 20/20" kind of thing, but I'm open to learning more.
Great question! The general approach we take with transactional systems like this is to put reachability statements throughout the complex stateful machinery and then stress-test them in Antithesis.
The workload I put in place does exactly that[1]. It runs a write workload from multiple processes concurrently on the same SQLite database to cause writes to build up in the WAL, and runs checkpoints concurrently. This exercises the portion of the WAL code that, from a trivial read-through, is most likely to contain bugs (and turns out, did!).
Said differently, this is exactly the approach we take with all stateful transactional systems. I am only sad that I didn't do this experiment months earlier, as it would have saved Tailscale and the SQLite team a lot of time.
you asserted that Antithesis would have found it right away, but it seems like from your write-up that you first set up instrumentation on the source, then told it to reproduce the WAL bug.
i don't feel like i gained a strong appreciation for what the actual Antithesis capabilities are or why you had confidence this was anything the Antithesis tools would have helped find? i want to know more about what kind of analysis or capabilities Antithesis has and how it could have helped find the issue. what's clear is that Antithesis could confirm an already found issue. and even there I don't feel like I gained any connection or knowledge or understanding of what the tooling was doing or how it would help. i'd really like more technical meat to this post to understand the value proposition.
Nice article, but could you get someone to fix the blog layout? Having the author card pinned to the window rather than being able to scroll past it is distracting.
as a long time lurker who usually enjoys antithesis posts, this was a pretty existentially sad read. we all know that company blog posts here always tend to be ads at their core, but historically afaicr antithesis generally toes the line well of keeping the "thing we sell is really good" as a side dish to the "i (author) did very interesting deep thinking and/or hard work to accomplish something cool" entree.
this one falls very flat on that front imo. the "antithesis finds rare bugs" is the overwhelming majority of the actual content here, where as the usual meat of the blog, the hard work, the deep thinking, really distills down to "typed this into claude mobile, typed that into claude mobile". yes, its a statement to the power of the technologies of today that's all it takes to find this bug, but man it's a sad signal for those of us looking for compelling technical reads instead of b2b sales pitches.
I am okay with companies shilling their tools. Ultimately I'd want my employer to buy good tools that make our lives easier specifically wrt debugging which I genuinely hate. However this post could have been a better walk through about what Claude found, what was the exact root cause and how it can be fixed instead of the generic Anthithesis found it in 15m.
yeah im fine with shilling that has actual interesting substance beyond the ad. imo this is an ad + AI handwaving and nothing else.
this is a pretty wide trend for blog content nowadays so im not suprised overall, i just commented because previous antithesis blogs were better than this to that criteria. just take a look at the other older write ups on their site compared to this one and you’ll see a very clear difference in depth and effort of content.
I understand the author feels he missed an opportunity to catch a bug in Sqlite - but this still feels like catching something in hindsight. Just knowing the bug the is in Sqlite's concurrency engine is a huge win.
But reading over tailscale's blogpost, I'm not sure how I would connect the dots to "this issue I'm seeing in production" to "load this up in antithesis". It seemed half the battle was figuring out if the bug was in Sqlite, Tailscale, or Linux, and in that situation if you gave Antithesis all the context, would it reliably find the bug?
I went clicking through to see if I could find the prompt they fed the AI to locate the issue / write the test suite.
I couldn't find it, so its unclear if the prompt was completely "make a test suite" or was lead towards finding it in the first place, which wouldn't be a fair test.
The closest I mention of the prompt I could find was:
> Then I asked it to write a simple workload which exercised the WAL insert and checkpoint code. Notably, this is a completely generic workload.
Hey, author here! Good question. The prompt was to use the Antithesis skills[1] to write a workload that exercises SQLite's insertion and checkpoint machinery across multiple processes using a specific version of SQLite (3.51.2). The agent who did the work was aware of the bug and knew I wanted to see if Antithesis could reproduce it. However, I specifically designed the workload the same way we approach testing other WAL checkpoint-style systems: by putting them under stress. The workload itself does not know about the bug.
My goal, which I believe I accomplished, was to determine if Antithesis would have found this bug if we had put the SQLite WAL under test.
Their founder guy says[1] they found some other bugs while doing this that they will report soon, so if that is true, seems more plausibly like a pretty generic thing.
Looking forward to seeing the other bugs they found.
I feel like we’re in a weird P=NP age, where after someone finds an issue, someone points out a cheaper way to find the same issue, which inevitably involves knowing the issue is there in the first place and it’s approximate location.
We wanted to publish this ASAP since people are talking about this bug today. In a follow-up, we will show how our automated causality analysis could have saved Tailscale and SQLite from 6 months of root cause analysis. Stay tuned!
Carl (author) here. This is a poor reading of the blog post. Notably, I built a general-purpose workload that simply exercises the write/checkpoint machinery. The exact same workload we use to break all kinds of transactional systems. This generic workload found the bug without any special knowledge of the bug, due to how Antithesis works.
I think the issue is that you knew there was a bug to be found in the write and checkpoint interactions, which then determined which workload you built.
My understanding is that this workload is very uncommon: the Tailscale blog says they used a custom unusual configuration to have many checkpoints like this. So without knowing the bug, it seems unlikely one would build this workload and then find the bug. If that makes sense.
Edit: just want to say that you being able to repro it is awesome, but that the overall claim seems a little overstated to me.
Hi! I'm the author and the person who was on a road trip when I decided to do this experiment back in July. I'm super excited to finally get the blog post out, and even better, alongside a post describing the root cause process that went into the project. You can read about Tailscale's story here: https://tailscale.com/blog/sqlite-wal-reset-bug
I'm only sad that I didn't put SQLite under test earlier in the year, or I would have found this issue right away. If you look at the workload[1], you can see how simple it is. Exactly the same kind of workload we write every day to help our customers find bugs.
If you have any questions about our process or how debugging with Antithesis works, please let me know! Thanks for reading!
[1]: https://github.com/antithesishq/sqlite/blob/3.51.2-instrumen...
How difficult would it have been to isolate that problem if you didn't already know the SQLite subsystem it was in? This feels, to someone relatively ignorant of the SQLite / Tailscale / Antithesis architectures as a "hindsight is 20/20" kind of thing, but I'm open to learning more.
Great question! The general approach we take with transactional systems like this is to put reachability statements throughout the complex stateful machinery and then stress-test them in Antithesis.
The workload I put in place does exactly that[1]. It runs a write workload from multiple processes concurrently on the same SQLite database to cause writes to build up in the WAL, and runs checkpoints concurrently. This exercises the portion of the WAL code that, from a trivial read-through, is most likely to contain bugs (and turns out, did!).
Said differently, this is exactly the approach we take with all stateful transactional systems. I am only sad that I didn't do this experiment months earlier, as it would have saved Tailscale and the SQLite team a lot of time.
[1]: https://github.com/antithesishq/sqlite/blob/3.51.2-instrumen...
10 replies →
So you got information where the bug is, and now you brag that your AI friend found it quickly :joy:
Go find bugs that haven't been yet found, and then brag about them :joy:
you asserted that Antithesis would have found it right away, but it seems like from your write-up that you first set up instrumentation on the source, then told it to reproduce the WAL bug.
i don't feel like i gained a strong appreciation for what the actual Antithesis capabilities are or why you had confidence this was anything the Antithesis tools would have helped find? i want to know more about what kind of analysis or capabilities Antithesis has and how it could have helped find the issue. what's clear is that Antithesis could confirm an already found issue. and even there I don't feel like I gained any connection or knowledge or understanding of what the tooling was doing or how it would help. i'd really like more technical meat to this post to understand the value proposition.
Nice article, but could you get someone to fix the blog layout? Having the author card pinned to the window rather than being able to scroll past it is distracting.
> I'm only sad that I didn't put SQLite under test earlier in the year, or I would have found this issue right away
Unless you guys routinely add assertions to libraries I don't see how you would have caught this?
I mean for sure databases should guard against lost writes but how would you have known to add that particular assert?
Typo in the post: "SQL 3.51.2" should presumably be "SQLite 3.51.2".
as a long time lurker who usually enjoys antithesis posts, this was a pretty existentially sad read. we all know that company blog posts here always tend to be ads at their core, but historically afaicr antithesis generally toes the line well of keeping the "thing we sell is really good" as a side dish to the "i (author) did very interesting deep thinking and/or hard work to accomplish something cool" entree.
this one falls very flat on that front imo. the "antithesis finds rare bugs" is the overwhelming majority of the actual content here, where as the usual meat of the blog, the hard work, the deep thinking, really distills down to "typed this into claude mobile, typed that into claude mobile". yes, its a statement to the power of the technologies of today that's all it takes to find this bug, but man it's a sad signal for those of us looking for compelling technical reads instead of b2b sales pitches.
I am okay with companies shilling their tools. Ultimately I'd want my employer to buy good tools that make our lives easier specifically wrt debugging which I genuinely hate. However this post could have been a better walk through about what Claude found, what was the exact root cause and how it can be fixed instead of the generic Anthithesis found it in 15m.
yeah im fine with shilling that has actual interesting substance beyond the ad. imo this is an ad + AI handwaving and nothing else.
this is a pretty wide trend for blog content nowadays so im not suprised overall, i just commented because previous antithesis blogs were better than this to that criteria. just take a look at the other older write ups on their site compared to this one and you’ll see a very clear difference in depth and effort of content.
I understand the author feels he missed an opportunity to catch a bug in Sqlite - but this still feels like catching something in hindsight. Just knowing the bug the is in Sqlite's concurrency engine is a huge win.
But reading over tailscale's blogpost, I'm not sure how I would connect the dots to "this issue I'm seeing in production" to "load this up in antithesis". It seemed half the battle was figuring out if the bug was in Sqlite, Tailscale, or Linux, and in that situation if you gave Antithesis all the context, would it reliably find the bug?
Maybe it wouldn't, but being able to cheaply test SQLite within a few minutes is still very valuable, even if just to rule it out.
I went clicking through to see if I could find the prompt they fed the AI to locate the issue / write the test suite.
I couldn't find it, so its unclear if the prompt was completely "make a test suite" or was lead towards finding it in the first place, which wouldn't be a fair test.
The closest I mention of the prompt I could find was:
> Then I asked it to write a simple workload which exercised the WAL insert and checkpoint code. Notably, this is a completely generic workload.
With a skeptical lens, unclear.
Hey, author here! Good question. The prompt was to use the Antithesis skills[1] to write a workload that exercises SQLite's insertion and checkpoint machinery across multiple processes using a specific version of SQLite (3.51.2). The agent who did the work was aware of the bug and knew I wanted to see if Antithesis could reproduce it. However, I specifically designed the workload the same way we approach testing other WAL checkpoint-style systems: by putting them under stress. The workload itself does not know about the bug.
My goal, which I believe I accomplished, was to determine if Antithesis would have found this bug if we had put the SQLite WAL under test.
[1]: https://github.com/antithesishq/antithesis-skills
Thanks for clarifying! It would be really interesting if Antithesis finds the bug when:
1. The specific bug isnt mentioned 2. (If youre game) a model with a knowledge-cutoff date before the report is used
3 replies →
Their founder guy says[1] they found some other bugs while doing this that they will report soon, so if that is true, seems more plausibly like a pretty generic thing.
Looking forward to seeing the other bugs they found.
1. https://news.ycombinator.com/item?id=49278351
looks like the bug fixes are in individual commits here: https://github.com/sqlite/sqlite/compare/master...antithesis...
I'd be interested in seeing what the SQLite people have to say about them if/when they are submitted upstream
3 replies →
I feel like we’re in a weird P=NP age, where after someone finds an issue, someone points out a cheaper way to find the same issue, which inevitably involves knowing the issue is there in the first place and it’s approximate location.
We wanted to publish this ASAP since people are talking about this bug today. In a follow-up, we will show how our automated causality analysis could have saved Tailscale and SQLite from 6 months of root cause analysis. Stay tuned!
related: Tailscale Traces Database Corruption to 16y/o SQLite WAL-Reset Bug (tailscale.com) https://news.ycombinator.com/item?id=49272832
[flagged]
Carl (author) here. This is a poor reading of the blog post. Notably, I built a general-purpose workload that simply exercises the write/checkpoint machinery. The exact same workload we use to break all kinds of transactional systems. This generic workload found the bug without any special knowledge of the bug, due to how Antithesis works.
I think the issue is that you knew there was a bug to be found in the write and checkpoint interactions, which then determined which workload you built.
My understanding is that this workload is very uncommon: the Tailscale blog says they used a custom unusual configuration to have many checkpoints like this. So without knowing the bug, it seems unlikely one would build this workload and then find the bug. If that makes sense.
Edit: just want to say that you being able to repro it is awesome, but that the overall claim seems a little overstated to me.
3 replies →
I read your ad.
It states you replicated the bug once the SQLite team fixed it, and published it.
Not sure what’s difficult about replicating behaviour when it’s spelled out for you.
2 replies →
Okay, so use the same methodology to find a new, unknown bug? That'd be actually impressive.
As-is, this is just 20/20 hindsight with concerns about leading the AI on through the prompt hand-waved away. Come on.
As the post (and linked repo) pretty clearly indicate, zero info about the existence of the bug went into this work.
Oh btw, we also found some other ones… Stay tuned!
So you found this before SQLite published the fix?
Related: finding the same bug with TLA+ https://news.ycombinator.com/item?id=48730953
Omg the one thing I hate more than ai art is ai pixelart where the pixels are all different sizes
I love hearing about this product, I hope one day I'll get to use anything like it.
[flagged]
[dead]