[APR-205] dogstatsd: optimize multi-value distribution decoding

[tobz]

Context

Currently, when decoding multi-value metric payloads, we use an approach that wraps an iterator over the raw values, and then for each value returned from the iterator, we create the corresponding metric.

However, this is suboptimal for distributions because we can add multiple values to a distribution, which is in fact all that will end up happening when these metrics are aggregated.

Solution

We've updated ValueIter<'a> to take this optimized approach when dealing with distributions, instead just building one sketch, adding all of the values it pulls out of the value iterator (well, a new value iterator just for floats, FloatIter<'a>), and then returning a single distribution metric.

[pr-commenter]

Regression Detector (DogStatsD)

Regression Detector Results

Run ID: cdf053a4-5912-483f-913c-755181fd3f7c

Baseline: 7.52.0
Comparison: 7.52.1

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

No significant changes in experiment optimization goals

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

There were no significant changes in experiment optimization goals at this confidence level and effect size tolerance.

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	links
➖	dsd_uds_100mb_3k_contexts_distributions_only	memory utilization	+0.40	[+0.19, +0.60]
➖	dsd_uds_1mb_3k_contexts	ingress throughput	+0.05	[-0.02, +0.12]
➖	dsd_uds_100mb_3k_contexts	ingress throughput	+0.01	[-0.00, +0.02]
➖	dsd_uds_1mb_50k_contexts_memlimit	ingress throughput	+0.00	[-0.03, +0.03]
➖	dsd_uds_500mb_3k_contexts	ingress throughput	+0.00	[-0.01, +0.01]
➖	dsd_uds_100mb_250k_contexts	ingress throughput	-0.00	[-0.09, +0.09]
➖	dsd_uds_1mb_50k_contexts	ingress throughput	-0.02	[-0.04, +0.01]
➖	dsd_uds_10mb_3k_contexts	ingress throughput	-0.03	[-0.06, +0.00]
➖	dsd_uds_512kb_3k_contexts	ingress throughput	-0.03	[-0.07, +0.00]

Explanation

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:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. 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.

3. Its configuration does not mark it "erratic".

[pr-commenter]

Regression Detector (Saluki)

Regression Detector Results

Run ID: 482821be-c32c-4cf3-b268-07dd293b15c7

Baseline: 464a24a
Comparison: 10ab5f6

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

No significant changes in experiment optimization goals

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

There were no significant changes in experiment optimization goals at this confidence level and effect size tolerance.

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	links
➖	dsd_uds_500mb_3k_contexts	ingress throughput	+1.46	[+1.36, +1.55]
➖	dsd_uds_100mb_3k_contexts_distributions_only	memory utilization	+0.12	[-0.03, +0.26]
➖	dsd_uds_100mb_250k_contexts	ingress throughput	+0.04	[-0.23, +0.31]
➖	dsd_uds_512kb_3k_contexts	ingress throughput	+0.03	[-0.13, +0.18]
➖	dsd_uds_100mb_3k_contexts	ingress throughput	+0.01	[-0.00, +0.02]
➖	dsd_uds_50mb_10k_contexts_no_inlining_no_allocs	ingress throughput	+0.00	[-0.05, +0.05]
➖	dsd_uds_1mb_50k_contexts	ingress throughput	-0.00	[-0.02, +0.02]
➖	dsd_uds_10mb_3k_contexts	ingress throughput	-0.00	[-0.09, +0.09]
➖	dsd_uds_50mb_10k_contexts_no_inlining	ingress throughput	-0.00	[-0.02, +0.02]
➖	dsd_uds_1mb_3k_contexts	ingress throughput	-0.05	[-0.22, +0.12]
➖	dsd_uds_1mb_50k_contexts_memlimit	ingress throughput	-2.34	[-5.56, +0.88]

Explanation

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:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. 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.

3. Its configuration does not mark it "erratic".

[pr-commenter]

Regression Detector Links

Experiment Result Links

experiment	link(s)
dsd_uds_100mb_250k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_100mb_3k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_100mb_3k_contexts_distributions_only	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_10mb_3k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_1mb_3k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_1mb_50k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_1mb_50k_contexts_memlimit	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_500mb_3k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_512kb_3k_contexts	[Profiling (ADP)] [Profiling (DSD)] [SMP Dashboard]
dsd_uds_50mb_10k_contexts_no_inlining (ADP only)	[Profiling (ADP)] [SMP Dashboard]
dsd_uds_50mb_10k_contexts_no_inlining_no_allocs (ADP only)	[Profiling (ADP)] [SMP Dashboard]