ECEdge Computing · Showcase
Community homeHot discussions Latest posts Best of Knowledge base Resources About

Edge Computing · Showcase · Hiring: remote edge-computing-show engineer (full-time, long-term)

351
ECr/edge-computing-show·posted by rase·yesterdayJobs

Hiring: remote edge-computing-show engineer (full-time, long-term)

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

Order of investigation, by return on effort: 1. Check downstream latency first — usually it is not your problem 2. Then pool hit rate and wait-queue length 3. Only then GC and allocation 4. Suspect the framework last

The first thing was to collapse the variables. We were changing config and upgrading the version at the same time, and afterwards nobody could say which change caused what. We rolled back to moving one variable at a time, re-ran three times, and only then did the curve settle. Tedious, but not skippable.

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.

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".

89 comments

89 comments

M
LlinlinMod·12 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.

512
Aalice_dev·28 minutes ago

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

156
Llinlin·12 minutes ago

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

75
Llinlin·28 minutes ago

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

90
Llinlin·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.

217
Ddev_zhou·2 days agoLevel 6

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.

58
Zzhou_yiMod·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.

142
NnikicOPMod·2 days 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
Ttang_hao·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.

25
Zzhu_zong·28 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.

1
Aalice_dev·1 hour ago

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

3
Lli_ming·2 days ago

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

271
Sslow_queryMod·2 days agoeditedLevel 6

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.

54
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.

8
Ddev_zhou·2 days agoLevel 6

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

1
Hhuang_keOP·2 days ago

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

444
Lli_mingOP·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.

433
Oops_wang·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.

490
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.

454
Lli_ming·just nowedited

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

435
Ttang_hao·3 minutes ago

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

401
Cchen_dev·12 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.

492
Wwinter·2 days ago

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

299
Bbob_chen·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.

344
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.

430
Ddev_zhou·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.

411
Ttang_hao·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.

362
Bbob_chen·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.

348
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.

293
Kkite·2 days ago

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

21
Sswoole_lee·2 days ago

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

242
Zzhu_zong·2 days ago

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

216
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.

10
Lli_mingMod·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.

194
Cchen_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.

191
Kkernel_panic·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.

175
Oops_wang·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.

50
Cchen_devOP·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.

99
Wwinter·2 days ago

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

425
Mmike_xu·2 days ago

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

325
Kkite·2 days ago

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

63
Zzhu_zongOP·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.

73
Wwinter·2 days ago

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

58
Mmike_xu·2 days ago

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

48
Mmike_xu·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.

383
Aalice_dev·2 days ago

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

84
Oops_wang·2 days ago

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

173
Llinlin·5 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.

118
Zzhou_yi·3 minutes ago

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

2
Sslow_queryOP·2 days ago

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

1
Rrase·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.

4
Sslow_query·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.

83
Ddev_zhou·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.

81
Bbob_chen·2 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.

45
Rrase·2 days ago

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

71
Mmike_xu·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.

42
Sslow_query·2 days ago

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

32
Kkernel_panic·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.

24
Mmike_xu·2 days ago

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

433
Zzhu_zongMod·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.

426
Ttang_hao·2 days ago

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

286
Kkite·just now

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

15
Cchen_dev·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.

189
Mmike_xu·2 days ago

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

469
Sslow_query·yesterday

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

12
Zzhou_yi·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.

9
Oops_wang·1 hour 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.

14
Wwinter·2 days ago

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

2
Kkernel_panicOP·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.

352
Sswoole_lee·2 days ago

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

4
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.

182
Bbob_chen·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.

124
Cchen_dev·2 days ago

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

40
Ddev_zhou·3 minutes ago

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

1
Wwinter·2 days ago

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

403
Llinlin·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.

351
Wwinter·3 minutes ago

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

333
Wwinter·28 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.

1
Sslow_query·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.

167
Kkite·just now

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

33
Hhuang_keMod·yesterday

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

25
Ttang_hao·28 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.

126
Bbob_chen·yesterday

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.

165
Sslow_query·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.

43
Zzhu_zong·2 days ago

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

21
Llinlin·12 minutes ago

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

2
Zzhu_zong·2 days ago

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

473
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.

76
Sswoole_lee·2 hours 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.

1

This is the post detail page /en/c/edge-computing-show/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 →