feat(fleet): Add per-extension registry overrides for BYOC#47984
feat(fleet): Add per-extension registry overrides for BYOC#47984BaptisteFoy wants to merge 2 commits intomainfrom
Conversation
BaptisteFoy
added
changelog/no-changelog
Files inventory check summaryFile checks results against ancestor b8036ed6: Results for datadog-agent_7.78.0~devel.git.675.a8370ca.pipeline.103323611-1_amd64.deb:No change detected |
Static quality checks✅ Please find below the results from static quality gates Successful checksInfo
21 successful checks with minimal change (< 2 KiB)
On-wire sizes (compressed)
|
Regression DetectorRegression Detector ResultsMetrics dashboard Baseline: b8036ed Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ❌ | docker_containers_cpu | % cpu utilization | +7.12 | [+3.98, +10.25] | 1 | Logs |
Fine details of change detection per experiment
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ❌ | docker_containers_cpu | % cpu utilization | +7.12 | [+3.98, +10.25] | 1 | Logs |
| ➖ | quality_gate_logs | % cpu utilization | +2.69 | [+1.07, +4.30] | 1 | Logs bounds checks dashboard |
| ➖ | otlp_ingest_metrics | memory utilization | +0.38 | [+0.23, +0.54] | 1 | Logs |
| ➖ | ddot_logs | memory utilization | +0.37 | [+0.32, +0.43] | 1 | Logs |
| ➖ | ddot_metrics | memory utilization | +0.12 | [-0.05, +0.30] | 1 | Logs |
| ➖ | file_tree | memory utilization | +0.12 | [+0.06, +0.17] | 1 | Logs |
| ➖ | quality_gate_metrics_logs | memory utilization | +0.10 | [-0.14, +0.33] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_idle | memory utilization | +0.09 | [+0.04, +0.13] | 1 | Logs bounds checks dashboard |
| ➖ | file_to_blackhole_1000ms_latency | egress throughput | +0.02 | [-0.39, +0.43] | 1 | Logs |
| ➖ | tcp_dd_logs_filter_exclude | ingress throughput | +0.00 | [-0.10, +0.10] | 1 | Logs |
| ➖ | file_to_blackhole_100ms_latency | egress throughput | +0.00 | [-0.09, +0.09] | 1 | Logs |
| ➖ | uds_dogstatsd_to_api | ingress throughput | -0.00 | [-0.20, +0.19] | 1 | Logs |
| ➖ | file_to_blackhole_500ms_latency | egress throughput | -0.00 | [-0.39, +0.39] | 1 | Logs |
| ➖ | uds_dogstatsd_to_api_v3 | ingress throughput | -0.01 | [-0.21, +0.19] | 1 | Logs |
| ➖ | file_to_blackhole_0ms_latency | egress throughput | -0.03 | [-0.50, +0.44] | 1 | Logs |
| ➖ | tcp_syslog_to_blackhole | ingress throughput | -0.06 | [-0.18, +0.06] | 1 | Logs |
| ➖ | quality_gate_idle_all_features | memory utilization | -0.11 | [-0.15, -0.07] | 1 | Logs bounds checks dashboard |
| ➖ | docker_containers_memory | memory utilization | -0.22 | [-0.29, -0.15] | 1 | Logs |
| ➖ | ddot_metrics_sum_cumulativetodelta_exporter | memory utilization | -0.25 | [-0.47, -0.02] | 1 | Logs |
| ➖ | uds_dogstatsd_20mb_12k_contexts_20_senders | memory utilization | -0.30 | [-0.36, -0.24] | 1 | Logs |
| ➖ | ddot_metrics_sum_delta | memory utilization | -0.35 | [-0.52, -0.18] | 1 | Logs |
| ➖ | otlp_ingest_logs | memory utilization | -0.38 | [-0.47, -0.29] | 1 | Logs |
| ➖ | ddot_metrics_sum_cumulative | memory utilization | -0.50 | [-0.64, -0.36] | 1 | Logs |
Bounds Checks: ✅ Passed
| perf | experiment | bounds_check_name | replicates_passed | observed_value | links |
|---|---|---|---|---|---|
| ✅ | docker_containers_cpu | simple_check_run | 10/10 | 565 ≥ 26 | |
| ✅ | docker_containers_memory | memory_usage | 10/10 | 270.87MiB ≤ 370MiB | |
| ✅ | docker_containers_memory | simple_check_run | 10/10 | 698 ≥ 26 | |
| ✅ | file_to_blackhole_0ms_latency | memory_usage | 10/10 | 0.19GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_0ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | file_to_blackhole_1000ms_latency | memory_usage | 10/10 | 0.23GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_1000ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | file_to_blackhole_100ms_latency | memory_usage | 10/10 | 0.19GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_100ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | file_to_blackhole_500ms_latency | memory_usage | 10/10 | 0.22GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_500ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | quality_gate_idle | intake_connections | 10/10 | 3 = 3 | bounds checks dashboard |
| ✅ | quality_gate_idle | memory_usage | 10/10 | 173.29MiB ≤ 175MiB | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | intake_connections | 10/10 | 2 ≤ 3 | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | memory_usage | 10/10 | 488.53MiB ≤ 550MiB | bounds checks dashboard |
| ✅ | quality_gate_logs | intake_connections | 10/10 | 3 ≤ 6 | bounds checks dashboard |
| ✅ | quality_gate_logs | memory_usage | 10/10 | 202.09MiB ≤ 220MiB | bounds checks dashboard |
| ✅ | quality_gate_logs | missed_bytes | 10/10 | 0B = 0B | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | cpu_usage | 10/10 | 349.77 ≤ 2000 | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | intake_connections | 10/10 | 4 ≤ 6 | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | memory_usage | 10/10 | 417.08MiB ≤ 475MiB | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | missed_bytes | 10/10 | 0B = 0B | bounds checks dashboard |
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_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 5d973eb874
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
| envCopy := *d.env | ||
| if url != "" { | ||
| envCopy.RegistryOverride = url |
There was a problem hiding this comment.
WithRegistryOverride clones the full Env and only updates the global registry fields, but leaves RegistryOverrideByImage/auth-by-image maps intact; getRefAndKeychain gives those image-scoped entries precedence, so a set DD_INSTALLER_REGISTRY_URL_AGENT_PACKAGE can silently override the per-extension URL passed here. In that setup, extension-specific BYOC registry overrides are ignored and downloads still go to the package-scoped registry, so this needs to clear or rewrite image-scoped override maps on the copied env.
Useful? React with 👍 / 👎.
| span.SetTag("package_name", pkg.Name) | ||
| span.SetTag("package_version", pkg.Version) |
There was a problem hiding this comment.
❓: Not sure how the tags work, but won't they be overwrite each time and only keep the last ?
| // We expect download errors (no real registry), not a panic or grouping error. | ||
| assert.Error(t, err) |
There was a problem hiding this comment.
❓: Maybe assert for a "could not download package" pattern more reliable ?
| // Expect download failure (no real registry), not a grouping error | ||
| assert.Error(t, err) |
| err := installSingle(ctx, pkg, extension, isExperiment, hooks) | ||
| for _, group := range groups { | ||
| // Download package metadata once per distinct registry | ||
| pkg, err := group.downloader.Download(ctx, url) |
There was a problem hiding this comment.
❓: Not sure if it's the same url, but shouldn't we use the override one if any ?
2 participants
What does this PR do?
Adds support for per-extension registry overrides so each OCI extension (e.g.
ddot) can be downloaded from a different registry than the default one. This is needed for BYOC (Bring Your Own Cloud) scenarios where extensions are hosted in customer-controlled registries.Changes:
oci/download.go: NewDownloader.WithRegistryOverride()method that creates a new Downloader with overridden registry settings while sharing the same HTTP clientextensions/extensions.go: NewExtensionRegistrytype.Install()andRestore()now accept anoverridesparameter that groups extensions by their effective registry, downloading the package once per distinct registrydatadog_agent_extensions.go: NewextensionRegistryConfigstruct andExtensionsfield oninstallerRegistryConfig.setRegistryConfig()now returns parsed per-extension overrides frominstaller.registry.extensions.<pkg>.<ext>installer.go: UpdatedInstallExtensions()andRestoreExtensions()to passniloverrides (daemon path)Config format:
Motivation
BYOC customers need to host extensions in their own registries. Currently all extensions are downloaded from a single registry. This change allows per-extension registry overrides so each extension can come from a different registry.
Describe how you validated your changes
go build ./pkg/fleet/installer/...compilesgo test ./pkg/fleet/installer/oci/...— new tests forWithRegistryOverride(full + partial override, shared client, env isolation)go test ./pkg/fleet/installer/packages/...— new tests for YAML config parsing with extension overridesgo test ./pkg/fleet/installer/packages/extensions/...— new tests for Install grouping by registry and nil overrides backwards compatAdditional Notes
niloverrides = no change in behavior (full backwards compatibility)installer.go) passesnilsince it doesn't parse YAML config