VUVue · Remote
Community homeHot discussions Latest posts Best of Knowledge base Resources About

Vue · Remote · Two weeks with vue-remote: what feels right and what drives me up the wall

312
VUr/vue-remote·posted by mike_xu·yesterdayTooling

Two weeks with vue-remote: what feels right and what drives me up the wall

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

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.

Image placeholder · object storage in production
POST /api/uploads → CDN origin pull
36 comments

36 comments

M
Oops_wang·5 hours agoedited

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

520
Rrase·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.

487
Ttang_hao·12 minutes agoedited

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

499
Lli_ming·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.

430
Hhuang_ke·yesterdayedited

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

93
Wwinter·12 minutes agoedited

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

180
Sswoole_lee·2 days ago

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

4
Llinlin·12 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.

482
Llinlin·1 hour 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.

93
Rran_boMod·2 days 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.

518
Oops_wang·2 days agoedited

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

364
Hhuang_ke·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.

336
Zzhu_zong·12 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.

258
Mmike_xuMod·just nowedited

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

251
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.

1
Kkernel_panic·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.

131
WwinterOP·28 minutes ago

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

39
Zzhou_yiOP·2 days ago

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

33
Kkite·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.

375
Lli_ming·2 hours ago

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

78
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.

44
Hhuang_ke·2 days ago

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

38
Mmike_xu·28 minutes ago

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

1
Aalice_devMod·3 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.

33
Zzhou_yi·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.

27
Rrase·2 days agoedited

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

18
Nnikic·28 minutes ago

One counter-example: below vue-remote 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
Bbob_chen·3 minutes ago

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

1
Sslow_query·3 minutes ago

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

6
Nnikic·1 hour ago

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

16
Mmike_xuOP·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.

211
Ddev_zhou·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.

268
Zzhou_yiOP·1 hour 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.

21
Aalice_dev·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.

2
Oops_wang·1 hour ago

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

301
Zzhou_yi·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.

201

This is the post detail page /en/c/vue-remote/post/p12. 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 →