TUTurborepo · Job board
Community homeHot discussions Latest posts Best of Knowledge base Resources About

Turborepo · Job board · This turborepo-jobs bug hid for three years before anyone reported it

67
TUr/turborepo-jobs·posted by swoole_lee·3 days agoJobs

This turborepo-jobs bug hid for three years before anyone reported it

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

What genuinely surprised me was the tail. The average looked great while P99 jumped by an order of magnitude past some threshold. The cause was not turborepo-jobs itself but our upstream connection reuse — the load test traffic was too clean and hid the long-tail requests.

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

114 votes total

32 comments

32 comments

M
Ttang_hao·2 days ago

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

409
Aalice_devOP·2 days ago

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

300
Rran_bo·2 days 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.

323
Mmike_xu·2 days ago

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

312
Mmike_xu·28 minutes agoedited

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

310
Hhuang_ke·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.

287
Ttang_hao·just nowedited

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

1
Zzhou_yi·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.

512
Aalice_devOP·12 minutes ago

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

277
Rrase·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.

356
Ttang_hao·2 days 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.

106
Ddev_zhou·2 days ago

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

104
Ttang_hao·2 days ago

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

374
Wwinter·28 minutes agoedited

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

86
Llinlin·3 minutes agoedited

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

1
Kkernel_panic·2 days 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.

17
Rrase·5 hours ago

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

278
Kkernel_panic·12 minutes ago

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

178
Zzhu_zongMod·2 days 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.

115
Kkite·yesterday

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.

415
Hhuang_ke·just now

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.

40
Lli_ming·yesterday

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.

6
Kkernel_panic·5 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.

6
Ttang_haoMod·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.

115
Nnikic·12 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.

59
Bbob_chen·just nowedited

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.

49
Rran_bo·28 minutes ago

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

9
Bbob_chenMod·5 hours 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.

108
Zzhou_yi·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
Lli_ming·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.

15
Oops_wang·2 days ago

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

14
Zzhou_yi·3 minutes agoedited

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

1

This is the post detail page /en/c/turborepo-jobs/post/p7. 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 →