enhancement: add more aggregator-specific telemetry #478
Conversation
github-actions
bot
added
area/core
Regression Detector (Saluki)Regression Detector ResultsRun ID: f4b1d1ad-a286-4f99-9b9a-0daca4c89f18 Baseline: 4e8add1 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | dsd_uds_500mb_3k_contexts | ingress throughput | +0.84 | [+0.71, +0.97] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts_distributions_only | memory utilization | +0.59 | [+0.48, +0.70] | 1 | |
| ➖ | quality_gates_idle_rss | memory utilization | +0.44 | [+0.41, +0.47] | 1 | |
| ➖ | dsd_uds_40mb_12k_contexts_40_senders | ingress throughput | +0.02 | [-0.01, +0.05] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts_dualship | ingress throughput | +0.01 | [-0.00, +0.02] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts | ingress throughput | +0.00 | [-0.00, +0.01] | 1 | |
| ➖ | dsd_uds_1mb_50k_contexts | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_512kb_3k_contexts | ingress throughput | -0.00 | [-0.01, +0.01] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts | ingress throughput | -0.00 | [-0.05, +0.04] | 1 | |
| ➖ | dsd_uds_50mb_10k_contexts_no_inlining | ingress throughput | -0.01 | [-0.08, +0.07] | 1 | |
| ➖ | dsd_uds_10mb_3k_contexts | ingress throughput | -0.01 | [-0.04, +0.02] | 1 | |
| ➖ | dsd_uds_50mb_10k_contexts_no_inlining_no_allocs | ingress throughput | -0.01 | [-0.07, +0.05] | 1 | |
| ➖ | dsd_uds_100mb_250k_contexts | ingress throughput | -0.01 | [-0.05, +0.03] | 1 | |
| ➖ | dsd_uds_1mb_50k_contexts_memlimit | ingress throughput | -0.43 | [-0.72, -0.15] | 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 LinksExperiment Result Links
|
Regression Detector (DogStatsD)Regression Detector ResultsRun ID: 4c6cc300-6ecc-4f1b-82ea-7a6a7c44a04e Baseline: 7.63.0-rc.2 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | dsd_uds_500mb_3k_contexts | ingress throughput | +2.54 | [+2.45, +2.64] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts_distributions_only | memory utilization | +1.28 | [+1.13, +1.44] | 1 | |
| ➖ | quality_gates_idle_rss | memory utilization | +0.36 | [+0.25, +0.46] | 1 | |
| ➖ | dsd_uds_40mb_12k_contexts_40_senders | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_1mb_50k_contexts_memlimit | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts_dualship | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_100mb_250k_contexts | ingress throughput | +0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_1mb_50k_contexts | ingress throughput | -0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts | ingress throughput | -0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_512kb_3k_contexts | ingress throughput | -0.00 | [-0.01, +0.01] | 1 | |
| ➖ | dsd_uds_10mb_3k_contexts | ingress throughput | -0.00 | [-0.02, +0.01] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts | ingress throughput | -0.02 | [-0.05, +0.02] | 1 |
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".
Summary
This PR adds a few new metrics for the aggregate transform, as well as some remapped ones:
aggregate_active_contexts_bytes_by_type(new in Saluki, tracks total context-related bytes per metric type)aggregator.dogstatsd_contexts_by_mtype(new remapped metric, derived fromaggregate_active_contexts_by_type)aggregator.dogstatsd_contexts_bytes_by_mtype(new remapped metric, derived fromaggregate_active_contexts_by_type)no_aggregation.processed(new remapped metric, derived fromaggregate_passthrough_metrics_total)no_aggregation.flush(new remapped metric, derived fromaggregate_passthrough_flushes_total)The bulk of this PR, however, is the actual addition of the logic to expose the size of a context. I believe the code comments and unit tests should be sufficient to explain what/why it's calculating things the way it is. I chose to do it the calculations rather than go full-blown trait-based approach just to keep things simpler for the moment.
Change Type
How did you test this PR?
I ran ADP locally, and ran one of our SMP experiments against it:
dsd_uds_10mb_3k_contexts. I checked the internal telemetry (http://localhost:5051/metrics) and observed all aforementioned metrics to be present (albeit sanitized for Prometheus), with the expected values.References
N/A