[NDMII-3236] Update devicecheck profile when ProfileProvider has changed #32185
Conversation
The .Metrics, .MetricTags, etc. fields on CheckConfig are storing all the same attributes as a profile; it is easier if we just create a Profile from the CheckConfig and pass that around. This also removes the caching of those fields from the CheckConfig; a later commit caches this generated profile in the DeviceCheck.
Uncompressed package size comparisonComparison with ancestor Diff per package
Decision |
Test changes on VMUse this command from test-infra-definitions to manually test this PR changes on a VM: inv aws.create-vm --pipeline-id=51169033 --os-family=ubuntuNote: This applies to commit d43ce0b |
Regression DetectorRegression Detector ResultsMetrics dashboard Baseline: 786bce8 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gate_idle_all_features | memory utilization | +2.22 | [+2.11, +2.33] | 1 | Logs bounds checks dashboard |
| ➖ | otel_to_otel_logs | ingress throughput | +1.02 | [+0.35, +1.69] | 1 | Logs |
| ➖ | file_tree | memory utilization | +0.65 | [+0.54, +0.77] | 1 | Logs |
| ➖ | uds_dogstatsd_to_api_cpu | % cpu utilization | +0.31 | [-0.42, +1.05] | 1 | Logs |
| ➖ | quality_gate_logs | % cpu utilization | +0.11 | [-2.84, +3.06] | 1 | Logs |
| ➖ | file_to_blackhole_1000ms_latency_linear_load | egress throughput | +0.05 | [-0.41, +0.51] | 1 | Logs |
| ➖ | file_to_blackhole_500ms_latency | egress throughput | +0.04 | [-0.73, +0.81] | 1 | Logs |
| ➖ | file_to_blackhole_300ms_latency | egress throughput | +0.04 | [-0.59, +0.67] | 1 | Logs |
| ➖ | file_to_blackhole_0ms_latency_http2 | egress throughput | +0.03 | [-0.82, +0.87] | 1 | Logs |
| ➖ | file_to_blackhole_0ms_latency_http1 | egress throughput | +0.01 | [-0.81, +0.83] | 1 | Logs |
| ➖ | tcp_dd_logs_filter_exclude | ingress throughput | -0.00 | [-0.01, +0.01] | 1 | Logs |
| ➖ | uds_dogstatsd_to_api | ingress throughput | -0.00 | [-0.10, +0.09] | 1 | Logs |
| ➖ | file_to_blackhole_0ms_latency | egress throughput | -0.02 | [-0.81, +0.76] | 1 | Logs |
| ➖ | file_to_blackhole_100ms_latency | egress throughput | -0.03 | [-0.78, +0.72] | 1 | Logs |
| ➖ | tcp_syslog_to_blackhole | ingress throughput | -0.17 | [-0.23, -0.12] | 1 | Logs |
| ➖ | file_to_blackhole_1000ms_latency | egress throughput | -0.29 | [-1.08, +0.51] | 1 | Logs |
| ➖ | quality_gate_idle | memory utilization | -0.49 | [-0.53, -0.44] | 1 | Logs bounds checks dashboard |
Bounds Checks: ❌ Failed
| perf | experiment | bounds_check_name | replicates_passed | links |
|---|---|---|---|---|
| ❌ | file_to_blackhole_0ms_latency_http2 | lost_bytes | 9/10 | |
| ✅ | file_to_blackhole_0ms_latency | lost_bytes | 10/10 | |
| ✅ | file_to_blackhole_0ms_latency | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_0ms_latency_http1 | lost_bytes | 10/10 | |
| ✅ | file_to_blackhole_0ms_latency_http1 | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_0ms_latency_http2 | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_1000ms_latency | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_1000ms_latency_linear_load | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_100ms_latency | lost_bytes | 10/10 | |
| ✅ | file_to_blackhole_100ms_latency | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_300ms_latency | lost_bytes | 10/10 | |
| ✅ | file_to_blackhole_300ms_latency | memory_usage | 10/10 | |
| ✅ | file_to_blackhole_500ms_latency | lost_bytes | 10/10 | |
| ✅ | file_to_blackhole_500ms_latency | memory_usage | 10/10 | |
| ✅ | quality_gate_idle | memory_usage | 10/10 | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | memory_usage | 10/10 | bounds checks dashboard |
| ✅ | quality_gate_logs | lost_bytes | 10/10 | |
| ✅ | quality_gate_logs | 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".
CI Pass/Fail Decision
✅ Passed. All Quality Gates passed.
- quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
|
Closing this to reopen based on a different branch (I'm learning how to use |
What does this PR do?
This PR adds a method to the ProfileProvider that indicates when the profiles were last updated, and updates the device check to cache the timestamp of the last time it gathered the metrics, tags, etc. from the matching profile, so that it can regenerate them if the ProfileProvider receives new profiles.
This also pulls the cached values into the DeviceCheck, instead of storing them on the CheckConfig, and since the values being cached are a set of metrics, a set of SNMP tags, a set of static tags, and a Metadata object, we now store them in a Profile instead of as separate fields.
Motivation
https://datadoghq.atlassian.net/browse/NDMII-3236
Describe how you validated your changes
Unit tests pass; I haven't QA'd it locally yet but will do so on Monday 12/16.
Additional Notes
Although this adds support for the set of profiles changing, no ProfileProvider exists yet that actually does change them over time, so the behavior of the agent should be unchanged, EXCEPT in one very specific case: If an integration is configured with an explicit profile name, and that profile doesn't exist, the current behavior is to error out; after this PR, the integration will log a warning but will not error out, and any metrics or tags explicitly present in the initConfig will be monitored even though the profile could not be found. If the profile becomes available later (currently impossible), the check will begin working again.