[Exporter.Geneva] Opt in resources #3552

mattsains · 2025-11-20T18:26:22Z

This is a re-submission of #3367 , after @cijothomas recommend that it should be re-introduced.

I did change the name of the config option to WithResourceAttributes to match the Rust option in the user events log exporter

I also changed the logic when this field is not provided. The default is now to include no resource attributes except for mapped fields (eg., service.name -> Part A cloud extension)

Changes

Allows users of the Geneva exporter to specify which resource attributes they want sent to Geneva. This prevents excessive trace data from being exported if there are many resource attributes which do not need to be stored.

Merge requirement checklist

CONTRIBUTING guidelines followed (license requirements, nullable enabled, static analysis, etc.)
Unit tests added/updated
Appropriate CHANGELOG.md files updated for non-trivial changes
Changes in public API reviewed (if applicable)

cijothomas · 2025-11-20T18:31:42Z

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs

+    /// WantedResourceAttributes specifies which resource attribute fields should be sent to Geneva.
+    ///
+    /// Any resource attributes not in WantedResourceAttributes are ignored.
+    /// If WantedResourceAttributes is not provided, all resource attributes will be sent to Geneva.


all will be sent or none will be sent?

In my opinion, it should be none, so the behavior is back-compatible. To get resource attributes, users must explicitly opt-in.

if a customer really wants all resource attributes, how might they accomplish that? By the way, there is a feature flag which is default off that controls this feature

I've made the change you requested

do you think the special resource attributes, service.name and service.instanceId, which are mapped to part A cloud extension, should also be affected by WithResourceAttributes being set to null?

By the way, there is a feature flag which is default off that controls this feature

I am not sure if we need to feature flag this at all. The WithResourceAttribute is sufficient on its own right?

if a customer really wants all resource attributes, how might they accomplish that?

By listing each attribute. Not ideal, but I also don't think this will be the common case with GenevaExporter. If there is strong need for such a capability, we can introduce a boolean 'with_all_resource`/similar that adds all attributes effectively ignore WithResourceAttributes. Alternatively, "*" can be used to mean - get all attributes - we do such thing elsewhere in GenevaExporter.

I can remove the feature flag in a future PR

feature flag has been removed, and the behaviour of the attribute is that no resource attributes are sent if it is null

codecov · 2025-11-20T18:38:04Z

❌ 1 Tests Failed:

Tests completed	Failed	Passed	Skipped
528	1	527	13

View the top 1 failed test(s) by shortest run time

OpenTelemetry.Exporter.Geneva.Tests.GenevaTraceExporterTests::GenevaTraceExporter_constructor_Invalid_Input

Stack Traces | 0.01s run time

at OpenTelemetry.Exporter.Geneva.Tests.GenevaTraceExporterTests.GenevaTraceExporter_constructor_Invalid_Input() in .../_/test/OpenTelemetry.Exporter.Geneva.Tests/GenevaTraceExporterTests.cs:line 61

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

cijothomas · 2025-11-25T23:48:57Z

src/OpenTelemetry.Exporter.Geneva/Internal/MsgPack/MsgPackTraceExporter.cs

                }
            }

+            if (!isDedicatedField && (this.WithResourceAttributes == null || !this.WithResourceAttributes.Contains(entry.Key)))


shouldn't we short-circuit this entire method if this.WithResourceAttributes == null ?

I decided that service.name and service.instanceId fields should still be used even if they're not explicitly mentioned in WithResourceAttributes, because they have a special mapping into part A cloud extension. So I don't believe the code should skip this section if it is one of those two resource attributes

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs

cijothomas · 2025-11-25T23:51:07Z

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs

+    /// Any resource attributes not in WithResourceAttributes are ignored.
+    /// If WithResourceAttributes is not provided, no resource attributes will be sent to Geneva.
+    /// </summary>
+    public IEnumerable<string>? WithResourceAttributes { get; set; }


nit: is this still controlled by a feature flag in connection string?

please refer to this previous conversation: #3552 (comment)

rajkumar-rangaraj · 2025-11-26T17:39:26Z

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs

+    /// Any resource attributes not in WithResourceAttributes are ignored.
+    /// If WithResourceAttributes is not provided, no resource attributes will be sent to Geneva.
+    /// </summary>
+    public IEnumerable<string>? WithResourceAttributes { get; set; }


Should this be --> public IEnumerable? WithResourceAttributes { get; init; }

Also, AllowedResourceAttributes or ResourceAttributeFilter naming might fit better here. No strong preference will leave it to you.

sure - would you like me to do the same to PrepopulatedFields?

github-actions · 2025-12-04T03:36:47Z

This PR was marked stale due to lack of activity. It will be closed in 7 days.

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs

src/OpenTelemetry.Exporter.Geneva/.publicApi/PublicAPI.Shipped.txt

src/OpenTelemetry.Exporter.Geneva/Internal/MsgPack/MsgPackTraceExporter.cs

rajkumar-rangaraj · 2025-12-09T20:30:08Z

Consider expanding the test coverage for these scenarios

Reserved field name in ResourceFieldNames throws
Invalid type in resource attribute
service.name auto-mapping with ResourceFieldNames
Empty ResourceFieldNames = [] vs null

…utes is not set

mattsains · 2025-12-10T17:50:21Z

Consider expanding the test coverage for these scenarios

* Reserved field name in `ResourceFieldNames` throws

* Invalid type in resource attribute

* `service.name` auto-mapping with `ResourceFieldNames`

* Empty `ResourceFieldNames = []` vs null

Except for service.name mapping, which was already tested, I've added all of these tests in the new revision

github-actions bot added the comp:exporter.geneva Things related to OpenTelemetry.Exporter.Geneva label Nov 20, 2025

cijothomas reviewed Nov 20, 2025

View reviewed changes

mattsains marked this pull request as ready for review November 20, 2025 18:57

mattsains requested a review from a team as a code owner November 20, 2025 18:57

github-actions bot assigned rajkumar-rangaraj and xiang17 Nov 20, 2025

github-actions bot requested review from rajkumar-rangaraj and xiang17 November 20, 2025 19:04

Kielek changed the title ~~Opt in resources~~ [Exporter.Geneva] Opt in resources Nov 21, 2025

cijothomas reviewed Nov 25, 2025

View reviewed changes

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs Outdated Show resolved Hide resolved

cijothomas reviewed Nov 25, 2025

View reviewed changes

rajkumar-rangaraj reviewed Nov 26, 2025

View reviewed changes

github-actions bot added the Stale label Dec 4, 2025

mattsains force-pushed the opt-in-resources branch from 90ce62b to 5385b5e Compare December 4, 2025 22:04

github-actions bot removed the Stale label Dec 5, 2025

rajkumar-rangaraj reviewed Dec 9, 2025

View reviewed changes

src/OpenTelemetry.Exporter.Geneva/GenevaExporterOptions.cs Show resolved Hide resolved

rajkumar-rangaraj reviewed Dec 9, 2025

View reviewed changes

src/OpenTelemetry.Exporter.Geneva/.publicApi/PublicAPI.Shipped.txt Outdated Show resolved Hide resolved

rajkumar-rangaraj reviewed Dec 9, 2025

View reviewed changes

src/OpenTelemetry.Exporter.Geneva/Internal/MsgPack/MsgPackTraceExporter.cs Show resolved Hide resolved

rajkumar-rangaraj reviewed Dec 9, 2025

View reviewed changes

src/OpenTelemetry.Exporter.Geneva/Internal/MsgPack/MsgPackTraceExporter.cs Show resolved Hide resolved

rajkumar-rangaraj added the keep-open Prevents issues and pull requests being closed as stale label Dec 9, 2025

mattsains added 6 commits December 10, 2025 09:01

add WantedResourceAttributes option

f46d5d8

PR self-reference

7640eec

rename to WithResourceAttributes

99c9f25

update PR number

49dfb0c

seriously...

45cbc31

insane

5d4b68a

mattsains added 9 commits December 10, 2025 09:01

change behaviour to send no resource attributes if WithResourceAttrib…

19e4645

…utes is not set

lint

775af9b

always allow mapped resource attributes, and fix tests

10e636b

wip

0b2ad42

ready for PR I think

ba949b0

change set to init

22c266c

fix version

1668119

fix readme and test

368289b

PR feedback changes

9ac9337

mattsains force-pushed the opt-in-resources branch from 5255e5d to 9ac9337 Compare December 10, 2025 17:49

[Exporter.Geneva] Opt in resources #3552

Are you sure you want to change the base?

[Exporter.Geneva] Opt in resources #3552

Uh oh!

Conversation

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

Uh oh!

Changes

Merge requirement checklist

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

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

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mattsains Dec 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

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

Uh oh!

❌ 1 Tests Failed:

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Dec 4, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

rajkumar-rangaraj commented Dec 9, 2025

Uh oh!

mattsains commented Dec 10, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

mattsains commented Nov 20, 2025 •

edited

Loading

mattsains Nov 20, 2025 •

edited

Loading

mattsains Dec 8, 2025 •

edited

Loading

codecov bot commented Nov 20, 2025 •

edited

Loading