POPolars · Beginners
Community homeHot discussions Latest posts Best of Knowledge base Resources About

Polars · Beginners · Ten polars-newbie anti-patterns — how many are you guilty of?

460
POr/polars-newbie·posted by alice_dev·just nowPostmortem

Ten polars-newbie anti-patterns — how many are you guilty of?

It took me two weeks of on-and-off digging and plenty of wrong turns. Writing the process down as it happened so the next person spends less time.

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
55 comments

55 comments

M
Aalice_dev·28 minutes agoedited

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.

482
Zzhu_zongOP·5 hours ago

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

337
Rran_bo·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.

385
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.

375
Nnikic·2 days ago

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

369
Ddev_zhou·12 minutes agoedited

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.

357
Rrase·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.

305
Nnikic·2 days ago

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

58
Cchen_dev·yesterday

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

296
Ttang_hao·12 minutes agoedited

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

282
Sslow_queryMod·12 minutes ago

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

266
Kkernel_panic·3 minutes ago

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

76
Zzhou_yi·3 minutes agoedited

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

51
Zzhou_yi·3 minutes ago

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

371
Kkernel_panic·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.

221
Kkernel_panicOP·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.

149
Hhuang_ke·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.

105
Kkernel_panic·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.

278
Ttang_hao·12 minutes ago

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

104
Nnikic·3 minutes ago

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

44
Rrase·2 hours ago

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

96
Cchen_devOP·2 days ago

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

169
Rran_boOP·1 hour ago

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

38
Cchen_dev·just nowLevel 6

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.

492
Rrase·2 days agoLevel 6

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

4
Nnikic·2 days 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.

58
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.

21
Bbob_chen·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.

235
Rran_bo·2 hours ago

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

36
Rrase·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.

4
Ddev_zhouMod·2 days ago

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

458
Rrase·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.

14
Mmike_xuOP·2 days agoLevel 6

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

125
Sslow_query·2 days agoLevel 6

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.

11
Oops_wang·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.

13
Sslow_query·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.

8
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.

203
LlinlinMod·2 days ago

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

210
Sslow_query·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.

414
Kkite·28 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.

124
Rran_bo·3 minutes ago

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

108
Bbob_chen·2 days ago

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

101
Nnikic·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.

97
Rrase·5 hours ago

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

83
Oops_wang·yesterday

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

43
Sswoole_lee·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.

19
Zzhou_yi·yesterday

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

41
Mmike_xu·1 hour ago

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

31
Zzhu_zong·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.

26
Rran_bo·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.

12
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.

23
Zzhu_zong·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.

21
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.

9
Cchen_dev·2 days ago

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

3
Nnikic·2 days agoedited

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.

15

This is the post detail page /en/c/polars-newbie/post/p3. 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 →