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

Burp Suite · Job board · Hiring: remote burpsuite-jobs engineer (full-time, long-term)

575
BSr/burpsuite-jobs·posted by bob_chen·just nowTutorial

Hiring: remote burpsuite-jobs engineer (full-time, long-term)

Most burpsuite-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.

One last trap: in container environments remember to adjust the memory-related parameters in step. Otherwise the host limit and the process expectation disagree, and the symptom is intermittent, unreproducible failure.

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.

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 burpsuite-jobs itself but our upstream connection reuse — the load test traffic was too clean and hid the long-tail requests.

83 comments

83 comments

M
Zzhou_yi·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.

379
Bbob_chen·2 days 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.

374
Rran_bo·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.

350
WwinterOP·2 days ago

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

266
Ddev_zhouOP·3 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.

234
Zzhu_zong·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.

277
Sswoole_lee·28 minutes ago

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

241
Hhuang_ke·2 days ago

One counter-example: below burpsuite-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.

228
Lli_ming·2 days ago

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

219
Aalice_dev·2 days ago

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

169
Bbob_chen·2 days ago

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

55
Kkite·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.

250
Wwinter·just now

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.

5
Lli_ming·2 days agoeditedLevel 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.

177
Cchen_dev·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.

51
Rran_bo·2 hours 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.

217
Cchen_dev·28 minutes ago

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

81
Kkernel_panic·2 days ago

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

240
Ddev_zhou·2 days 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.

303
Mmike_xu·5 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.

55
Aalice_dev·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.

178
Wwinter·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.

156
Lli_ming·1 hour ago

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

141
Mmike_xu·just now

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

214
Wwinter·2 days agoedited

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

3
Rran_bo·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.

100
Kkernel_panicOP·2 days agoedited

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

34
Ddev_zhouOP·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.

34
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.

11
Ddev_zhou·28 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.

75
Hhuang_ke·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.

67
Oops_wang·yesterday

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

61
Ttang_haoOP·1 hour ago

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

440
Aalice_dev·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.

180
Zzhu_zong·28 minutes ago

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

53
Lli_mingOP·2 hours ago

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

57
Kkite·1 hour ago

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

133
Hhuang_ke·yesterday

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.

40
Hhuang_ke·2 days ago

One counter-example: below burpsuite-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.

154
Ddev_zhou·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.

1
Kkernel_panic·2 days ago

One counter-example: below burpsuite-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.

1
Zzhu_zong·2 days ago

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

150
Wwinter·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.

34
Nnikic·2 days ago

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

33
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.

32
Aalice_dev·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.

14
Zzhou_yi·2 days ago

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

6
Aalice_dev·12 minutes agoedited

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

5
Zzhu_zongOP·28 minutes ago

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

131
Zzhu_zong·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.

500
Zzhou_yi·2 days agoedited

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

122
Sswoole_lee·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.

8
Mmike_xu·yesterdayedited

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

290
Nnikic·2 days agoLevel 6

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.

244
Wwinter·just nowLevel 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.

90
Wwinter·2 days agoLevel 6

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.

8
Nnikic·2 days ago

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

253
Wwinter·yesterdayLevel 6

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

458
Cchen_dev·2 days agoeditedLevel 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.

132
Llinlin·2 days agoLevel 6

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

8
Ttang_hao·2 days ago

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

175
Aalice_dev·2 days agoedited

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

122
Hhuang_ke·2 days agoLevel 6

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

139
Zzhou_yi·2 days ago

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

1
Aalice_dev·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.

206
Rrase·2 days ago

One counter-example: below burpsuite-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.

60
NnikicOP·2 days ago

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

162
Ddev_zhouOP·2 days 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.

68
Sslow_query·28 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.

3
Oops_wang·28 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.

275
Kkernel_panicMod·just now

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

105
Cchen_dev·12 minutes ago

One counter-example: below burpsuite-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.

33
Cchen_dev·2 days agoedited

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

3
Wwinter·2 days ago

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

390
Sswoole_lee·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.

3
Kkite·3 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.

2
Rrase·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.

2
Mmike_xu·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.

53
Ddev_zhou·2 days ago

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

9
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.

1
Wwinter·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.

80
Wwinter·just nowedited

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.

120
Rrase·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.

203

This is the post detail page /en/c/burpsuite-jobs/post/p13. 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 →