Conversation
github-actions
bot
added
area/io
tobz
added
the
type/chore
Regression Detector (DogStatsD)Regression Detector ResultsRun ID: 94c8407a-b4a0-4cbb-83cc-637b82ef1ae5 Baseline: 7.65.0-rc.9 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gates_idle_rss | memory utilization | +0.07 | [-0.06, +0.21] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts | ingress throughput | +0.00 | [-0.10, +0.11] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts_dualship | 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_40mb_12k_contexts_40_senders | 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_512kb_3k_contexts | ingress throughput | -0.00 | [-0.01, +0.01] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts | ingress throughput | -0.00 | [-0.00, +0.00] | 1 | |
| ➖ | dsd_uds_100mb_250k_contexts | ingress throughput | -0.00 | [-0.13, +0.12] | 1 | |
| ➖ | dsd_uds_10mb_3k_contexts | ingress throughput | -0.01 | [-0.04, +0.02] | 1 | |
| ➖ | dsd_uds_500mb_3k_contexts | ingress throughput | -0.23 | [-0.32, -0.14] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts_distributions_only | memory utilization | -0.66 | [-0.87, -0.45] | 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".
Regression Detector (Checks Agent Go)Regression Detector ResultsRun ID: 12edfb03-da10-4bb6-ab6a-25106333c609 Baseline: f61d1f4e054b884cb1894254ab2714b84b4684cb Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gates_idle_rss | memory utilization | -0.59 | [-0.65, -0.53] | 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 (Checks Agent)Regression Detector ResultsRun ID: cd270469-db0b-4db6-b780-ec59d38232a8 Baseline: 7ee652b Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gates_idle_rss | memory utilization | +0.28 | [+0.27, +0.30] | 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".
tobz
changed the title
RequestBuilder<E, O>
Regression Detector (Saluki)Regression Detector ResultsRun ID: 85577f83-55d3-475a-aa4a-2fb627d7c658 Baseline: 7ee652b Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gates_idle_rss | memory utilization | +1.03 | [+0.94, +1.11] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts_distributions_only | memory utilization | +0.43 | [+0.27, +0.59] | 1 | |
| ➖ | dsd_uds_50mb_10k_contexts_no_inlining_no_allocs | ingress throughput | +0.02 | [-0.08, +0.12] | 1 | |
| ➖ | dsd_uds_100mb_3k_contexts | ingress throughput | +0.02 | [-0.05, +0.08] | 1 | |
| ➖ | dsd_uds_100mb_250k_contexts | ingress throughput | +0.01 | [-0.04, +0.07] | 1 | |
| ➖ | dsd_uds_50mb_10k_contexts_no_inlining | ingress throughput | +0.01 | [-0.09, +0.10] | 1 | |
| ➖ | dsd_uds_40mb_12k_contexts_40_senders | ingress throughput | +0.00 | [-0.03, +0.04] | 1 | |
| ➖ | dsd_uds_1mb_3k_contexts | 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_512kb_3k_contexts | ingress throughput | -0.00 | [-0.01, +0.01] | 1 | |
| ➖ | dsd_uds_1mb_50k_contexts | ingress throughput | -0.00 | [-0.02, +0.02] | 1 | |
| ➖ | dsd_uds_10mb_3k_contexts | ingress throughput | -0.03 | [-0.06, +0.01] | 1 | |
| ➖ | dsd_uds_500mb_3k_contexts | ingress throughput | -1.16 | [-1.27, -1.04] | 1 | |
| ➖ | dsd_uds_1mb_50k_contexts_memlimit | ingress throughput | -3.01 | [-4.73, -1.28] | 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
|
tobz
force-pushed
the
tobz/refactor-dd-metrics-to-generic-request-builder
branch
from
e100ac1 to
0dd1134
Compare
tobz
force-pushed
the
tobz/delimited-payloads-dd-request-builder
branch
from
4b8ecd2 to
65f8fc1
Compare
tobz
force-pushed
the
tobz/refactor-dd-metrics-to-generic-request-builder
branch
from
0dd1134 to
66bdf4b
Compare
tobz
force-pushed
the
tobz/delimited-payloads-dd-request-builder
branch
from
65f8fc1 to
e609a13
Compare
tobz
changed the base branch from
tobz/refactor-dd-metrics-to-generic-request-builder
to
graphite-base/660
tobz
force-pushed
the
tobz/delimited-payloads-dd-request-builder
branch
from
e609a13 to
397c069
Compare
…tBuilder<E, O>` (#660) ## Summary This PR adds support for delimited payloads to `RequestBuilder<E, O>`. In the case of the Datadog Metrics destination, individual inputs are self-delimiting by virtue of being elements in a repeated message field in the Protocol Buffers payload. This means that while the request itself is a single message with N inputs bundled into it, and not N messages, we still don't have to do anything special to wrap those inputs or separate them: just write them individually, concatenated, and the payload is valid. However, for other destinations, like Datadog Events/Service Checks, the payloads are JSON arrays. This means that the payload has a prefix and suffix (`[` and `]`) and each input needs to be separated by a comma. Naturally, you get this when encoding an array of values all at once, but we want to build our requests incrementally, input by input. This PR adds support for this model by allowing `EndpointEncoder` to specify a prefix, suffix, and input separator (referred to collectively as "delimiters") that must be used when building a payload. `RequestBuilder<E, O>` uses these to generate a valid payload, whether we've written a single input or 10,000 inputs. While these changes are simple, most of this PR revolves around ensuring that our size limiting logic and request splitting logic works correctly in delimited mode as well as non-delimited mode. We've done a decent amount of refactoring to try and share more of the encode/flush logic between the regular encode/flush functions, and their usage in the request splitting codepath. ## Change Type - [ ] Bug fix - [ ] New feature - [x] Non-functional (chore, refactoring, docs) - [ ] Performance ## How did you test this PR? Unit tests and correctness test. ## References AGTMETRICS-184 ad9cec3
2 participants
Summary
This PR adds support for delimited payloads to
RequestBuilder<E, O>.In the case of the Datadog Metrics destination, individual inputs are self-delimiting by virtue of being elements in a repeated message field in the Protocol Buffers payload. This means that while the request itself is a single message with N inputs bundled into it, and not N messages, we still don't have to do anything special to wrap those inputs or separate them: just write them individually, concatenated, and the payload is valid.
However, for other destinations, like Datadog Events/Service Checks, the payloads are JSON arrays. This means that the payload has a prefix and suffix (
[and]) and each input needs to be separated by a comma. Naturally, you get this when encoding an array of values all at once, but we want to build our requests incrementally, input by input.This PR adds support for this model by allowing
EndpointEncoderto specify a prefix, suffix, and input separator (referred to collectively as "delimiters") that must be used when building a payload.RequestBuilder<E, O>uses these to generate a valid payload, whether we've written a single input or 10,000 inputs.While these changes are simple, most of this PR revolves around ensuring that our size limiting logic and request splitting logic works correctly in delimited mode as well as non-delimited mode. We've done a decent amount of refactoring to try and share more of the encode/flush logic between the regular encode/flush functions, and their usage in the request splitting codepath.
Change Type
How did you test this PR?
Unit tests and correctness test.
References
AGTMETRICS-184