OWOWASP · Open source
Community homeHot discussions Latest posts Best of Knowledge base Resources About

OWASP · Open source · Open-sourced a rate-limiting middleware for owasp-opensource with token bucket and sliding window

415
OWr/owasp-opensource·posted by swoole_lee·2 days agoJobs

Open-sourced a rate-limiting middleware for owasp-opensource with token bucket and sliding window

Some background first. Our setup is owasp-opensource plus three downstream services, seven figures of daily requests, peaking around nine in the evening.

On trade-offs, my view is this: if nobody on the team owns this area long-term, do not introduce a second mechanism. With two coexistence you first have to work out which one is even in play when things break, and that costs far more than the performance you saved.

Documentation first44%
Source code first28%
Just ask someone17%
Run a demo and learn by error11%

706 votes total

44 comments

44 comments

M
Cchen_dev·2 days agoedited

We have run this in production for two years without hitting it. That said, we never reached this scale, so our experience is not really evidence here.

466
Lli_ming·1 hour ago

This matches what we see in production. We only hit it past 3k QPS; the earlier load tests showed nothing — the test traffic was too clean, with no long-tail requests.

361
Ddev_zhou·yesterday

This matches what we see in production. We only hit it past 3k QPS; the earlier load tests showed nothing — the test traffic was too clean, with no long-tail requests.

311
Zzhou_yi·2 days ago

Saved. I am reworking this area this week — this saves a lot of wrong turns.

178
Ddev_zhouOP·2 days ago

Can you give a minimal reproduction? I ran it locally for ten minutes and could not reproduce on macOS with the latest version.

271
Bbob_chen·2 days ago

I see point 3 differently. The trade-off depends on your read/write ratio: read-heavy with little writing means caching actually widens the inconsistency window.

147
Cchen_dev·just now

We have run this in production for two years without hitting it. That said, we never reached this scale, so our experience is not really evidence here.

11
Ttang_hao·28 minutes ago

Agreeing with the above. One addition: with this option enabled the GC count in your metrics doubles, so adjust the alert threshold at the same time or it will keep firing.

302
Ddev_zhou·28 minutes agoedited

Sharing our numbers, 8 cores 16GB, same scenario:

| Concurrency | P50 | P99 |
|---|---|---|
| 200 | 12ms | 88ms |
| 500 | 31ms | 340ms |

P99 clearly collapses at 500 concurrency, which lines up with your knee point.

154
Oops_wang·3 minutes ago

I see point 3 differently. The trade-off depends on your read/write ratio: read-heavy with little writing means caching actually widens the inconsistency window.

56
NnikicMod·2 hours ago

Can you give a minimal reproduction? I ran it locally for ten minutes and could not reproduce on macOS with the latest version.

509
Rrase·3 minutes ago

Has anyone run a controlled experiment? I did, reducing it to a single variable, and the difference was 4% — within noise. So I suspect the main cause is something else.

110
Bbob_chen·3 minutes ago

One counter-example: below owasp-opensource 7.4 the semantics of that code are different, so do not copy it verbatim. We got burned in staging and rolled back once.

13
Lli_ming·3 minutes ago

Has anyone run a controlled experiment? I did, reducing it to a single variable, and the difference was 4% — within noise. So I suspect the main cause is something else.

251
Ddev_zhou·2 hours ago

Sharing our numbers, 8 cores 16GB, same scenario:

| Concurrency | P50 | P99 |
|---|---|---|
| 200 | 12ms | 88ms |
| 500 | 31ms | 340ms |

P99 clearly collapses at 500 concurrency, which lines up with your knee point.

387
Llinlin·3 minutes ago

This is not a owasp-opensource problem, it is a usage problem. The docs say this API is not thread-safe and you must lock around it yourself.

175
Zzhou_yiOP·2 days ago

There is actually a simpler fix that needs no architecture change: move this check up to the gateway and the problem disappears. The cost is one extra lookup at the gateway.

1
Ttang_hao·5 hours ago

Thanks for sharing real numbers — far more useful than the articles that only cover concepts.

7
Sswoole_lee·just now

Agreeing with the above. One addition: with this option enabled the GC count in your metrics doubles, so adjust the alert threshold at the same time or it will keep firing.

26
Rrase·12 minutes ago

There is actually a simpler fix that needs no architecture change: move this check up to the gateway and the problem disappears. The cost is one extra lookup at the gateway.

270
Sslow_query·1 hour agoedited

I just read the owasp-opensource source — the author actually explains the reasoning in a comment, roughly "so that it degrades into predictable behaviour in extreme cases".

185
Lli_mingOP·2 days ago

One counter-example: below owasp-opensource 7.4 the semantics of that code are different, so do not copy it verbatim. We got burned in staging and rolled back once.

235
Rrase·2 days agoedited

Agreeing with the above. One addition: with this option enabled the GC count in your metrics doubles, so adjust the alert threshold at the same time or it will keep firing.

10
Kkite·2 days agoedited

Saved. I am reworking this area this week — this saves a lot of wrong turns.

166
Rrase·2 days ago

Thanks for sharing real numbers — far more useful than the articles that only cover concepts.

144
Lli_mingMod·2 days ago

A question: what changes in a container with a 512Mi memory limit? That is how we run it in production.

142
Mmike_xu·2 hours ago

Has anyone run a controlled experiment? I did, reducing it to a single variable, and the difference was 4% — within noise. So I suspect the main cause is something else.

98
Lli_ming·2 days ago

I see point 3 differently. The trade-off depends on your read/write ratio: read-heavy with little writing means caching actually widens the inconsistency window.

48
Rran_bo·3 minutes agoedited

I just read the owasp-opensource source — the author actually explains the reasoning in a comment, roughly "so that it degrades into predictable behaviour in extreme cases".

28
Aalice_dev·5 hours agoedited

Worth learning from this debugging approach. We went straight at the logs and took a much longer route.

19
Kkernel_panicMod·5 hours ago

Can you give a minimal reproduction? I ran it locally for ten minutes and could not reproduce on macOS with the latest version.

297
KkiteOPMod·2 days ago

One counter-example: below owasp-opensource 7.4 the semantics of that code are different, so do not copy it verbatim. We got burned in staging and rolled back once.

155
Lli_ming·2 hours ago

This is not a owasp-opensource problem, it is a usage problem. The docs say this API is not thread-safe and you must lock around it yourself.

473
Wwinter·2 days ago

Sharing our numbers, 8 cores 16GB, same scenario:

| Concurrency | P50 | P99 |
|---|---|---|
| 200 | 12ms | 88ms |
| 500 | 31ms | 340ms |

P99 clearly collapses at 500 concurrency, which lines up with your knee point.

38
Rrase·yesterday

This matches what we see in production. We only hit it past 3k QPS; the earlier load tests showed nothing — the test traffic was too clean, with no long-tail requests.

125
Bbob_chen·2 days ago

A question: what changes in a container with a 512Mi memory limit? That is how we run it in production.

9
Zzhu_zong·2 days ago

This is not a owasp-opensource problem, it is a usage problem. The docs say this API is not thread-safe and you must lock around it yourself.

161
Sslow_query·1 hour ago

Worth learning from this debugging approach. We went straight at the logs and took a much longer route.

1
Mmike_xu·2 days ago

There is actually a simpler fix that needs no architecture change: move this check up to the gateway and the problem disappears. The cost is one extra lookup at the gateway.

24
Sswoole_lee·2 days ago

We have run this in production for two years without hitting it. That said, we never reached this scale, so our experience is not really evidence here.

1
Nnikic·2 days agoedited

Thanks for sharing real numbers — far more useful than the articles that only cover concepts.

7
Hhuang_keMod·2 days agoedited

A question: what changes in a container with a 512Mi memory limit? That is how we run it in production.

48
Oops_wang·2 days ago

Saved. I am reworking this area this week — this saves a lot of wrong turns.

5
Zzhu_zong·2 days ago

Worth learning from this debugging approach. We went straight at the logs and took a much longer route.

5

This is the post detail page /en/c/owasp-opensource/post/p6. Posts and comments are generated deterministically from a seeded PRNG, so the same post always renders the same content and the link can be shared, reloaded and indexed. In production this page reads MySQL for the post, Redis for hot-post caching, and fetches the whole comment tree in a single query on the path column.

See the database schema →