feat(openfeature): add Datadog OpenFeature provider for server-side feature flag evaluation #4056

eliottness · 2025-10-20T16:30:02Z

What does this PR do?

Implements a complete OpenFeature-compatible feature flag provider for Go that integrates with Datadog Remote Config
Supports all OpenFeature flag types: boolean, string, integer, float, and JSON objects
Provides advanced targeting capabilities with attribute-based conditions and traffic sharding

Motivation

This PR adds the Datadog OpenFeature provider to dd-trace-go, enabling Go applications to evaluate feature flags using the OpenFeature standard
interface.

Testing Instructions

Set environment variable: export DD_EXPERIMENTAL_FLAGGING_PROVIDER_ENABLED=true
Run unit tests: go test ./openfeature/...
Run integration tests: go test -v ./openfeature/...
Test with Remote Config by starting tracer and creating provider

System Tests run and DataDog/system-tests#5580

Benchmarks

goos: linux
goarch: amd64
pkg: github.com/DataDog/dd-trace-go/v2/openfeature
cpu: 12th Gen Intel(R) Core(TM) i9-12900H
BenchmarkEvaluation
BenchmarkEvaluation-20                    	               4850703	       245.4 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingContextSize
BenchmarkEvaluationWithVaryingContextSize/1field
BenchmarkEvaluationWithVaryingContextSize/1field-20         	 4679434	       246.4 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingContextSize/5fields
BenchmarkEvaluationWithVaryingContextSize/5fields-20        	 4783665	       243.7 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingContextSize/10fields
BenchmarkEvaluationWithVaryingContextSize/10fields-20       	 4668084	       266.6 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingContextSize/20fields
BenchmarkEvaluationWithVaryingContextSize/20fields-20       	 4098861	       261.5 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingFlagCounts
BenchmarkEvaluationWithVaryingFlagCounts/5flags
BenchmarkEvaluationWithVaryingFlagCounts/5flags-20          	 4383056	       301.1 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingFlagCounts/10flags
BenchmarkEvaluationWithVaryingFlagCounts/10flags-20         	 4741010	       250.4 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingFlagCounts/50flags
BenchmarkEvaluationWithVaryingFlagCounts/50flags-20         	 4491292	       246.4 ns/op	      48 B/op	       3 allocs/op
BenchmarkEvaluationWithVaryingFlagCounts/100flags
BenchmarkEvaluationWithVaryingFlagCounts/100flags-20        	 4890760	       246.2 ns/op	      48 B/op	       3 allocs/op
BenchmarkConcurrentEvaluations
BenchmarkConcurrentEvaluations-20                           	11141725	       107.9 ns/op	      48 B/op	       3 allocs/op
PASS

pr-commenter · 2025-10-20T16:40:54Z

Benchmarks

Benchmark execution time: 2025-10-27 13:21:22

Comparing candidate commit f9ae5bf in PR branch eliottness/ffe with baseline commit 9225b43 in branch main.

Found 0 performance improvements and 1 performance regressions! Performance is the same for 5 metrics, 0 unstable metrics.

scenario:BenchmarkStartSpanConfig/scenario_WithStartSpanConfig-24

🟥 execution_time [+84.363ns; +99.037ns] or [+2.406%; +2.824%]

leoromanovsky · 2025-10-20T17:10:42Z

Benchmarks

Benchmark execution time: 2025-10-20 17:04:18

Comparing candidate commit bdc04fe in PR branch eliottness/ffe with baseline commit 585513f in branch main.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 3 metrics, 0 unstable metrics.

Automatied benchmarks is really interested - will we be able to wire in the flag evaluations into it to see if changes cause regressions too? I took a pass at doing this in the eppo repository for memory allocations, which are a customer sensitivity - we'd like to operate with a light touch in our customers' environments: https://github.com/Eppo-exp/golang-sdk/blob/main/.github/workflows/memory-profile-compare.yml

leoromanovsky

Very strong foundation!

go.mod

openfeature/evaluator.go

openfeature/provider.go

openfeature/remoteconfig.go

datadog-datadog-prod-us1 · 2025-10-21T09:17:41Z

⚠️ Tests

⚠️ Warnings

❄️ 1 New flaky test detected

TestReportHealthMetricsAtInterval from github.com/DataDog/dd-trace-go/v2/ddtrace/tracer (Datadog)

Failed

=== RUN   TestReportHealthMetricsAtInterval
    metrics_test.go:64: 
        	Error Trace:	/home/runner/work/dd-trace-go/dd-trace-go/ddtrace/tracer/metrics_test.go:64
        	Error:      	Condition never satisfied
        	Test:       	TestReportHealthMetricsAtInterval
--- FAIL: TestReportHealthMetricsAtInterval (5.03s)

ℹ️ Info

🧪 All tests passed

_{This comment will be updated automatically if new data arrives.

🔗 Commit SHA: f9ae5bf | Docs | Was this helpful? Give us feedback!}

eliottness · 2025-10-22T17:14:02Z

Automatied benchmarks is really interested - will we be able to wire in the flag evaluations into it to see if changes cause regressions too? I took a pass at doing this in the eppo repository for memory allocations, which are a customer sensitivity - we'd like to operate with a light touch in our customers' environments: https://github.com/Eppo-exp/golang-sdk/blob/main/.github/workflows/memory-profile-compare.yml

@leoromanovsky I added benchmarks and registered them into the continuous benchmarking platform. This will take a while for them to show up because it needs to gather a lot of data points by running on main

darccio · 2025-10-23T09:34:38Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

ℹ️ About Codex in GitHub

Your team has set up Codex to 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 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

openfeature/provider.go

darccio · 2025-10-23T10:45:06Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

ℹ️ About Codex in GitHub

Your team has set up Codex to 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 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2025-10-23T10:49:05Z

openfeature/provider.go

+// Init initializes the provider. For the Datadog provider,
+// this is waiting for the first configuration to be loaded.
+func (p *DatadogProvider) Init(openfeature.EvaluationContext) error {
+	p.mu.Lock()
+	defer p.mu.Unlock()
+	for p.configuration == nil {
+		p.configChange.Wait()
+	}


Avoid blocking forever when no configuration arrives

The provider’s Init method waits on the condition variable until p.configuration becomes non‑nil and never returns otherwise. If Remote Config is unavailable, misconfigured, or simply has no flags for this environment, this wait loop never exits and the OpenFeature client never finishes initialization. In practice this can hang application startup indefinitely even though evaluations could proceed with default values. Consider adding a timeout or returning immediately when Remote Config is unavailable so the provider can be registered and errors can be reported per evaluation instead of deadlocking initialization.

Useful? React with 👍 / 👎.

@leoromanovsky Is this a known issue of the openfeature.StateHandler interface not giving us a way to timeout ? I am against putting a constant timeout to this waiting time because remoteconfig api contract has no guaranteed init time

Not known - when you encounter something like that which you feel we could improve OF upstream with could you please add a ticket to (https://datadoghq.atlassian.net/browse/FFL-1147); we will follow up with pull requests.

openfeature/evaluator.go

openfeature/doc.go

leoromanovsky · 2025-10-24T03:47:02Z

openfeature/evaluator.go

+	// Special handling for "id" attribute: if not explicitly provided, use targeting key
+	if condition.Attribute == "id" && !exists {
+		if targetingKey, ok := context[of.TargetingKey].(string); ok {
+			attributeValue = targetingKey
+			exists = true
+		}
+	}


no special handling is desired here; callers must provide the targeting key

I added this to make one of the system-tests pass 🤔 @khanayan123 What do you think ?

Julio-Guerra · 2025-10-24T09:19:43Z

openfeature/provider.go

+// Returns an error if the default configuration of the Remote Config client is NOT working
+// In this case, please call tracer.Start before creating the provider.
+func NewDatadogProvider() (*DatadogProvider, error) {
+	if !internal.BoolEnv(ffeProductEnvVar, false) {


@leoromanovsky, are we sure about this activation env var? Please think of the customer implications and experience. IMHO, this is useful for emergency disablement for ops teams (ie. without touching the source code), and that can be achieved in other ways too.

openfeature/doc.go

openfeature/provider.go

Signed-off-by: Eliott Bouhana <[email protected]>

eliottness · 2025-10-27T13:08:22Z

/merge

dd-devflow-routing-codex · 2025-10-27T13:08:26Z

View all feedbacks in Devflow UI.

2025-10-27 13:08:26 UTC ℹ️ Start processing command /merge

2025-10-27 13:08:35 UTC ℹ️ MergeQueue: waiting for PR to be ready

This merge request is not mergeable according to GitHub. Common reasons include pending required checks, missing approvals, or merge conflicts — but it could also be blocked by other repository rules or settings.
It will be added to the queue as soon as checks pass and/or get approvals.
Note: if you pushed new commits since the last approval, you may need additional approval.
You can remove it from the waiting list with /remove command.

2025-10-27 14:10:21 UTC ℹ️ MergeQueue: merge request added to the queue

The expected merge time in main is approximately 18m (p90).

2025-10-27 14:26:30 UTC ℹ️ MergeQueue: This merge request was merged

Signed-off-by: Eliott Bouhana <[email protected]>

eliottness changed the title ~~eat(openfeature): add Datadog OpenFeature provider for server-side feature flag evaluation~~ feat(openfeature): add Datadog OpenFeature provider for server-side feature flag evaluation Oct 20, 2025

leoromanovsky reviewed Oct 21, 2025

View reviewed changes

go.mod Show resolved Hide resolved

openfeature/evaluator.go Outdated Show resolved Hide resolved

openfeature/evaluator.go Outdated Show resolved Hide resolved

openfeature/provider.go Show resolved Hide resolved

openfeature/remoteconfig.go Outdated Show resolved Hide resolved

eliottness force-pushed the eliottness/ffe branch from ff5303e to 32d50dc Compare October 22, 2025 13:55

eliottness added enhancement quick change/addition that does not need full team approval AI Assisted AI/LLM assistance used in this PR (partially or fully) labels Oct 22, 2025

eliottness force-pushed the eliottness/ffe branch from 4906534 to 71cd401 Compare October 22, 2025 17:00

github-actions bot added the apm:ecosystem contrib/* related feature requests or bugs label Oct 22, 2025

eliottness requested a review from leoromanovsky October 22, 2025 17:14

eliottness marked this pull request as ready for review October 22, 2025 17:14

eliottness requested review from a team as code owners October 22, 2025 17:14

darccio approved these changes Oct 23, 2025

View reviewed changes

chatgpt-codex-connector bot reviewed Oct 23, 2025

View reviewed changes

openfeature/provider.go Show resolved Hide resolved

chatgpt-codex-connector bot reviewed Oct 23, 2025

View reviewed changes

leoromanovsky approved these changes Oct 24, 2025

View reviewed changes

Julio-Guerra reviewed Oct 24, 2025

View reviewed changes

genesor approved these changes Oct 24, 2025

View reviewed changes

openfeature/doc.go Show resolved Hide resolved

openfeature/provider.go Show resolved Hide resolved

eliottness added 5 commits October 27, 2025 14:08

squashed rebase because of go.mod conflits

dabe130

Signed-off-by: Eliott Bouhana <[email protected]>

add benchmarks

0ada717

Signed-off-by: Eliott Bouhana <[email protected]>

restore rc

e9dde60

Signed-off-by: Eliott Bouhana <[email protected]>

remove from gitlab-ci

5b14308

Signed-off-by: Eliott Bouhana <[email protected]>

rework broadcast mechanism

0fd9b5f

Signed-off-by: Eliott Bouhana <[email protected]>

eliottness added 3 commits October 27, 2025 14:08

apply @codex and @leoromanovsky comments

91e0358

Signed-off-by: Eliott Bouhana <[email protected]>

remove me debug log

167a5ff

Signed-off-by: Eliott Bouhana <[email protected]>

apply @genesor suggestions

c9e1b6c

Signed-off-by: Eliott Bouhana <[email protected]>

eliottness force-pushed the eliottness/ffe branch from 98754ac to c9e1b6c Compare October 27, 2025 13:08

dd-devflow bot added the mergequeue-status: waiting label Oct 27, 2025

generate and put back nil check

f9ae5bf

Signed-off-by: Eliott Bouhana <[email protected]>

dd-devflow bot added mergequeue-status: queued mergequeue-status: in_progress and removed mergequeue-status: waiting mergequeue-status: queued labels Oct 27, 2025

dd-mergequeue bot merged commit 99623ba into main Oct 27, 2025
291 of 295 checks passed

dd-mergequeue bot deleted the eliottness/ffe branch October 27, 2025 14:26

dd-devflow bot added mergequeue-status: done and removed mergequeue-status: in_progress labels Oct 27, 2025

feat(openfeature): add Datadog OpenFeature provider for server-side feature flag evaluation #4056

feat(openfeature): add Datadog OpenFeature provider for server-side feature flag evaluation #4056

Uh oh!

Conversation

eliottness commented Oct 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What does this PR do?

Motivation

Testing Instructions

Benchmarks

Uh oh!

pr-commenter bot commented Oct 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmarks

scenario:BenchmarkStartSpanConfig/scenario_WithStartSpanConfig-24

Uh oh!

leoromanovsky commented Oct 20, 2025

Benchmarks

Uh oh!

leoromanovsky left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

datadog-datadog-prod-us1 bot commented Oct 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⚠️ Warnings

ℹ️ Info

Uh oh!

eliottness commented Oct 22, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

darccio commented Oct 23, 2025

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

darccio commented Oct 23, 2025

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector bot Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

eliottness Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

leoromanovsky Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

leoromanovsky Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

eliottness Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

Julio-Guerra Oct 24, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

eliottness commented Oct 27, 2025

Uh oh!

dd-devflow-routing-codex bot commented Oct 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

eliottness commented Oct 20, 2025 •

edited

Loading

pr-commenter bot commented Oct 20, 2025 •

edited

Loading

datadog-datadog-prod-us1 bot commented Oct 21, 2025 •

edited

Loading

eliottness commented Oct 22, 2025 •

edited

Loading

dd-devflow-routing-codex bot commented Oct 27, 2025 •

edited

Loading