chore(ci): try using SMP-provided links in regression experiment PR comments #645
Conversation
tobz
added
the
type/chore
Regression Detector (DogStatsD)Regression Detector ResultsMetrics dashboard Baseline: 7.65.0-rc.9 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | dsd_uds_500mb_3k_contexts | ingress throughput | +1.72 | [+1.62, +1.82] | 1 | Logs |
| ➖ | quality_gates_idle_rss | memory utilization | +1.30 | [+1.18, +1.41] | 1 | Logs |
| ➖ | dsd_uds_100mb_3k_contexts | ingress throughput | +0.00 | [-0.10, +0.10] | 1 | Logs |
| ➖ | dsd_uds_100mb_250k_contexts | ingress throughput | +0.00 | [-0.12, +0.13] | 1 | Logs |
| ➖ | dsd_uds_1mb_50k_contexts | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | Logs |
| ➖ | dsd_uds_1mb_3k_contexts | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | Logs |
| ➖ | dsd_uds_1mb_3k_contexts_dualship | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | Logs |
| ➖ | dsd_uds_1mb_50k_contexts_memlimit | ingress throughput | -0.00 | [-0.00, +0.00] | 1 | Logs |
| ➖ | dsd_uds_512kb_3k_contexts | ingress throughput | -0.00 | [-0.01, +0.01] | 1 | Logs |
| ➖ | dsd_uds_40mb_12k_contexts_40_senders | ingress throughput | -0.00 | [-0.00, +0.00] | 1 | Logs |
| ➖ | dsd_uds_10mb_3k_contexts | ingress throughput | -0.02 | [-0.08, +0.04] | 1 | Logs |
| ➖ | dsd_uds_100mb_3k_contexts_distributions_only | memory utilization | -0.17 | [-0.38, +0.04] | 1 | Logs |
Bounds Checks: ❌ Failed
| perf | experiment | bounds_check_name | replicates_passed | links |
|---|---|---|---|---|
| ❌ | quality_gates_idle_rss | memory_usage | 0/10 |
Explanation
Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%
Performance changes are noted in the perf column of each table:
- ✅ = significantly better comparison variant performance
- ❌ = significantly worse comparison variant performance
- ➖ = no significant change in performance
A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".
For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:
-
Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.
-
Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.
-
Its configuration does not mark it "erratic".
Regression Detector (Checks Agent Go)Regression Detector ResultsRun ID: e9b626ff-3949-408e-85bd-74eff29e4688 Baseline: f61d1f4e054b884cb1894254ab2714b84b4684cb Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gates_idle_rss | memory utilization | +0.37 | [+0.31, +0.44] | 1 |
Explanation
Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%
Performance changes are noted in the perf column of each table:
- ✅ = significantly better comparison variant performance
- ❌ = significantly worse comparison variant performance
- ➖ = no significant change in performance
A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".
For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:
-
Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.
-
Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.
-
Its configuration does not mark it "erratic".
Regression Detector (Saluki)Regression Detector ResultsMetrics dashboard Baseline: 9d02818 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | dsd_uds_1mb_50k_contexts_memlimit | ingress throughput | +2.77 | [+1.18, +4.36] | 1 | Logs |
| ➖ | dsd_uds_500mb_3k_contexts | ingress throughput | +1.03 | [+0.93, +1.13] | 1 | Logs |
| ➖ | dsd_uds_10mb_3k_contexts | ingress throughput | +0.02 | [+0.00, +0.04] | 1 | Logs |
| ➖ | dsd_uds_100mb_250k_contexts | ingress throughput | +0.01 | [-0.07, +0.09] | 1 | Logs |
| ➖ | dsd_uds_1mb_3k_contexts_dualship | ingress throughput | +0.01 | [-0.00, +0.02] | 1 | Logs |
| ➖ | dsd_uds_1mb_50k_contexts | ingress throughput | +0.01 | [-0.00, +0.02] | 1 | Logs |
| ➖ | dsd_uds_40mb_12k_contexts_40_senders | ingress throughput | +0.00 | [-0.03, +0.04] | 1 | Logs |
| ➖ | dsd_uds_50mb_10k_contexts_no_inlining | ingress throughput | +0.00 | [-0.10, +0.11] | 1 | Logs |
| ➖ | dsd_uds_100mb_3k_contexts | ingress throughput | +0.00 | [-0.05, +0.06] | 1 | Logs |
| ➖ | dsd_uds_512kb_3k_contexts | ingress throughput | -0.01 | [-0.02, +0.01] | 1 | Logs |
| ➖ | dsd_uds_1mb_3k_contexts | ingress throughput | -0.01 | [-0.03, +0.00] | 1 | Logs |
| ➖ | dsd_uds_50mb_10k_contexts_no_inlining_no_allocs | ingress throughput | -0.01 | [-0.12, +0.09] | 1 | Logs |
| ➖ | quality_gates_idle_rss | memory utilization | -0.06 | [-0.08, -0.05] | 1 | Logs |
| ➖ | dsd_uds_100mb_3k_contexts_distributions_only | memory utilization | -0.43 | [-0.59, -0.27] | 1 | Logs |
Bounds Checks: ✅ Passed
| perf | experiment | bounds_check_name | replicates_passed | links |
|---|---|---|---|---|
| ✅ | quality_gates_idle_rss | memory_usage | 10/10 |
Explanation
Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%
Performance changes are noted in the perf column of each table:
- ✅ = significantly better comparison variant performance
- ❌ = significantly worse comparison variant performance
- ➖ = no significant change in performance
A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".
For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:
-
Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.
-
Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.
-
Its configuration does not mark it "erratic".
Regression Detector (Checks Agent)Regression Detector ResultsRun ID: fd7bf412-16f8-46c6-aff3-273aaa2fca33 Baseline: 9d02818 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gates_idle_rss | memory utilization | +0.34 | [+0.32, +0.36] | 1 |
Bounds Checks: ✅ Passed
| perf | experiment | bounds_check_name | replicates_passed | links |
|---|---|---|---|---|
| ✅ | quality_gates_idle_rss | memory_usage | 10/10 |
Explanation
Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%
Performance changes are noted in the perf column of each table:
- ✅ = significantly better comparison variant performance
- ❌ = significantly worse comparison variant performance
- ➖ = no significant change in performance
A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".
For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:
-
Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.
-
Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.
-
Its configuration does not mark it "erratic".
Regression Detector LinksADP Experiment Result Links
Checks Agent Experiment Result Links
|
1 participant
Summary
As stated in the PR title.
It would be nice to get rid of our custom PR comment with all of the links since it's extra code to manage and would require additional work to always post links even if the regression experiment jobs "fail" due to a regression being detected.
Change Type
How did you test this PR?
N/A
References
AGTMETRICS-184