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

Turborepo · Job board · Those easily-missed type details in turborepo-jobs

94
TUr/turborepo-jobs·posted by slow_query·5 minutes agoOpen sourcePinned

Those easily-missed type details in turborepo-jobs

Most turborepo-jobs articles stop at "how to use it" and never cover "when not to use it". This is an attempt at the second half.

We also fixed monitoring along the way: replaced average-based alerts with percentiles and split them per endpoint. False alerts dropped by about seventy percent and the on-call rotation visibly cheered up.

Performance44%
Maintainability28%
Ecosystem and community17%
Hiring difficulty11%

160 votes total

49 comments

49 comments

M
Zzhu_zong·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.

515
Zzhu_zong·3 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".

118
Zzhou_yi·3 minutes ago

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.

183
Oops_wang·yesterday

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.

190
Mmike_xu·2 days ago

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

504
Kkernel_panicMod·12 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.

383
Zzhu_zong·3 minutes ago

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

369
KkiteMod·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.

152
Sswoole_lee·2 days 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.

376
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.

284
Mmike_xu·3 minutes 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.

266
Llinlin·2 hours agoedited

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

84
Lli_ming·just now

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.

292
Kkite·3 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.

121
Aalice_dev·12 minutes ago

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

219
Sslow_queryMod·2 days ago

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

188
Nnikic·2 days agoedited

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.

171
Ttang_haoOP·28 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.

90
Sswoole_lee·2 hours 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.

376
Mmike_xu·just now

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.

306
Mmike_xuOP·28 minutes ago

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

2
Oops_wang·yesterday

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

338
Zzhou_yiOP·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.

234
Rran_bo·2 days ago

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

403
Bbob_chen·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.

18
Kkernel_panic·2 days ago

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

473
Rran_bo·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.

144
Hhuang_ke·3 minutes 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.

119
Wwinter·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.

457
Llinlin·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.

466
WwinterOPMod·5 hours 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".

348
Sslow_query·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.

314
Ttang_haoMod·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.

5
Rrase·2 days agoedited

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

473
Zzhu_zong·2 days agoedited

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

1
Zzhou_yi·3 minutes agoedited

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.

12
Lli_ming·3 minutes 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.

326
Zzhu_zongOP·28 minutes 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.

339
Llinlin·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".

2
Hhuang_ke·1 hour 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.

94
Oops_wang·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.

93
Oops_wang·2 days ago

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

225
Kkite·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.

36
Oops_wang·3 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.

367
Llinlin·2 days agoedited

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

58
Zzhou_yi·3 minutes ago

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

1
Wwinter·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.

15
Oops_wang·12 minutes ago

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

1
Wwinter·2 days ago

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.

105

This is the post detail page /en/c/turborepo-jobs/post/p0. 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 →