HCHigh Concurrency · Code review
Community homeHot discussions Latest posts Best of Knowledge base Resources About

High Concurrency · Code review · Why does high-concurrency-review use twice the memory of comparable tools?

205
HCr/high-concurrency-review·posted by kite·5 minutes agoHelp

Why does high-concurrency-review use twice the memory of comparable tools?

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

Worth noting: the official docs do cover this, just in a very inconspicuous spot. I only found it reading the source comments, where the author explains the reasoning — roughly "so that it degrades into predictable behaviour in extreme cases".

Self-host — full control44%
Managed service — less work28%
Hybrid: core self-hosted17%
Not decided yet11%

349 votes total

43 comments

43 comments

M
Ddev_zhou·2 days ago

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

474
Sswoole_lee·12 minutes ago

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

449
Ttang_hao·2 days ago

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

397
Kkite·2 hours 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.

323
Zzhu_zongMod·3 minutes 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.

321
Kkite·28 minutes ago

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

307
Sslow_query·1 hour 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.

404
Wwinter·12 minutes agoedited

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

273
Kkernel_panic·just now

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

105
Nnikic·2 hours 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.

475
LlinlinOP·5 hours 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.

358
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.

103
Rran_bo·1 hour 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.

53
Sswoole_lee·12 minutes agoLevel 6

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.

88
Hhuang_ke·2 days ago

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

81
Hhuang_ke·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.

393
Oops_wang·2 hours ago

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

238
Zzhou_yiMod·2 days ago

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

221
Nnikic·12 minutes ago

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

83
Ttang_hao·5 hours 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.

305
Rrase·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.

294
Mmike_xu·2 days agoedited

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.

231
Mmike_xu·2 days ago

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

88
Wwinter·2 hours 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.

83
Oops_wang·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.

277
Hhuang_ke·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.

67
Sswoole_lee·2 days ago

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

49
Rrase·12 minutes 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.

14
Zzhu_zong·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.

305
Ttang_hao·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.

12
Wwinter·2 hours ago

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

279
Llinlin·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.

63
Nnikic·12 minutes ago

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

109
Oops_wang·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.

28
Ddev_zhou·28 minutes ago

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

337
Kkernel_panic·12 minutes ago

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

10
Nnikic·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.

9
Bbob_chen·just now

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

1
Mmike_xu·2 days ago

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

276
Wwinter·2 days ago

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

49
Zzhou_yi·2 days ago

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

40
Sslow_query·2 days ago

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

1
Bbob_chen·5 hours ago

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

420

This is the post detail page /en/c/high-concurrency-review/post/p9. 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 →