feat(router): Subgraph Timeout Configuration #541

ardatan · 2025-10-31T10:55:39Z

Implementation of Timeout in #317
Ref ROUTER-110
Ref ROUTER-151
This also adds subgraphs and all options to traffic_shaping as in Apollo Router. So subgraph specific configuration can be done with subgraphs;

Apollo Router -> https://www.apollographql.com/docs/graphos/routing/performance/traffic-shaping#configuration

traffic_shaping:
   all:
      request_timeout: 5s
   subgraphs:
       products:
           request_timeout:
                expression: |
                   if (.request.operation.kind == "mutation") {
                        "15s"
                    } else {
                        "5s"
                    }

Documentation -> graphql-hive/console#7214

gemini-code-assist · 2025-10-31T10:56:05Z

Summary of Changes

Hello @ardatan, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly improves the router's resilience and configurability by adding a robust timeout feature for subgraph requests. It allows operators to define timeouts either as fixed durations or dynamically via VRL expressions, applicable globally or on a per-subgraph basis. This enhancement is integrated into a refactored traffic shaping configuration, providing greater control over request execution behavior.

Highlights

Subgraph Timeout Implementation: Introduced a timeout mechanism for subgraph requests, allowing for both fixed duration and dynamic VRL expression-based timeouts.
Flexible Traffic Shaping Configuration: Enhanced the traffic shaping configuration to support global settings (all) and specific overrides for individual subgraphs (subgraphs), mirroring Apollo Router's approach.
VRL Expression Integration: Centralized VRL expression handling into a new Expression primitive, simplifying its use across various configurations like subgraph URLs, headers, and now, timeouts.
Error Handling: Added specific error types for timeout failures and issues with VRL expression resolution for timeouts.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces a valuable subgraph timeout feature, enhancing the router's resilience. The implementation is well-structured, using tokio::timeout and providing flexibility with dynamic timeouts through VRL expressions. The configuration refactoring to support per-subgraph settings is also a welcome improvement.

My review focuses on several key areas to further refine the implementation:

Performance: I've identified a high-severity issue where new HTTP clients and connection pools are created unnecessarily. I've also pointed out a couple of medium-severity opportunities to reduce allocations on hot paths, aligning with the repository's performance-first ethos.
Correctness & Readability: I've suggested improvements to the timeout expression evaluation to provide clearer, more accurate error messages and handle edge cases more robustly. Additionally, I've noted some documentation comments that could be clarified for better user understanding.

The proposed changes aim to enhance performance, improve error handling, and increase the overall clarity and maintainability of the new feature.

lib/executor/src/executors/map.rs

lib/executor/src/executors/http.rs

lib/router-config/src/traffic_shaping.rs

github-actions · 2025-10-31T11:01:33Z

✅ `k6-benchmark` results

     ✓ response code was 200
     ✓ no graphql errors
     ✓ valid response structure

     █ setup

     checks.........................: 100.00% ✓ 197886      ✗ 0    
     data_received..................: 5.8 GB  193 MB/s
     data_sent......................: 77 MB   2.6 MB/s
     http_req_blocked...............: avg=3.65µs   min=661ns   med=1.68µs  max=17.14ms  p(90)=2.39µs  p(95)=2.79µs  
     http_req_connecting............: avg=887ns    min=0s      med=0s      max=2.54ms   p(90)=0s      p(95)=0s      
     http_req_duration..............: avg=22.24ms  min=2.34ms  med=21.3ms  max=114.75ms p(90)=30.18ms p(95)=33.36ms 
       { expected_response:true }...: avg=22.24ms  min=2.34ms  med=21.3ms  max=114.75ms p(90)=30.18ms p(95)=33.36ms 
     http_req_failed................: 0.00%   ✓ 0           ✗ 65982
     http_req_receiving.............: avg=138.07µs min=24.63µs med=39.61µs max=28.1ms   p(90)=87.5µs  p(95)=402.49µs
     http_req_sending...............: avg=24.51µs  min=5.63µs  med=10.86µs max=21.46ms  p(90)=15.89µs p(95)=28.06µs 
     http_req_tls_handshaking.......: avg=0s       min=0s      med=0s      max=0s       p(90)=0s      p(95)=0s      
     http_req_waiting...............: avg=22.08ms  min=2.28ms  med=21.16ms max=97.52ms  p(90)=29.92ms p(95)=33.06ms 
     http_reqs......................: 65982   2194.589839/s
     iteration_duration.............: avg=22.73ms  min=7.02ms  med=21.65ms max=249.48ms p(90)=30.6ms  p(95)=33.82ms 
     iterations.....................: 65962   2193.92463/s
     vus............................: 50      min=50        max=50 
     vus_max........................: 50      min=50        max=50

github-actions · 2025-10-31T11:01:35Z

🐋 This PR was built and pushed to the following Docker images:

Image Names: ghcr.io/graphql-hive/router

Platforms: linux/amd64,linux/arm64

Image Tags: ghcr.io/graphql-hive/router:pr-541 ghcr.io/graphql-hive/router:sha-cc9fb63

Docker metadata

{
"buildx.build.ref": "builder-129e9f65-ee08-4627-875e-3da8e8b8be74/builder-129e9f65-ee08-4627-875e-3da8e8b8be740/r7r547s2pph1jq3p3xzs6727d",
"containerimage.descriptor": {
  "mediaType": "application/vnd.oci.image.index.v1+json",
  "digest": "sha256:023f5b5f8266ea36afabf41292c8d3d982fd4f48b5e3bee2d8a209242929f49c",
  "size": 1609
},
"containerimage.digest": "sha256:023f5b5f8266ea36afabf41292c8d3d982fd4f48b5e3bee2d8a209242929f49c",
"image.name": "ghcr.io/graphql-hive/router:pr-541,ghcr.io/graphql-hive/router:sha-cc9fb63"
}

The previous implementation had expression handling scattered across multiple locations: - `duration_or_prog.rs` handled duration expressions - `utils/expression.rs` had generic compilation logic - Type-specific conversions were ad-hoc This refactor centralizes all expression handling into a module with: 1. Generic traits for extensibility - `FromVrlValue`: Provides a way to convert VRL values to specific types - `CompileExpression`: Compiles VRL expression strings into executable programs - `ExecutableProgram`: Executes compiled VRL programs with proper context setup 2. Type-safe ValueOrProgram enum - A generic enum that can represent either a static value or a computed expression - Eliminates duplication (previously had DurationOrProgram only) - Enables easy extension for new types (StringOrProgram, HeaderValueOrProgram) 3. Proper error handling - Dedicated error types for compilation and execution failures - Type-specific conversion errors (DurationConversionError, StringConversionError, etc.)

ardatan changed the base branch from main to primitive-expression October 31, 2025 10:55

gemini-code-assist bot reviewed Oct 31, 2025

View reviewed changes

lib/executor/src/executors/map.rs Outdated Show resolved Hide resolved

lib/executor/src/executors/http.rs Outdated Show resolved Hide resolved

lib/executor/src/executors/http.rs Outdated Show resolved Hide resolved

lib/router-config/src/traffic_shaping.rs Outdated Show resolved Hide resolved

ardatan changed the title ~~Subgraph timeout~~ feat(router): Subgraph Timeout Configuration Oct 31, 2025

ardatan force-pushed the subgraph_timeout branch 3 times, most recently from 0a61e05 to e319e11 Compare October 31, 2025 15:10

ardatan force-pushed the primitive-expression branch 2 times, most recently from 128b10a to 766bd0e Compare November 3, 2025 11:30

ardatan force-pushed the subgraph_timeout branch 4 times, most recently from 98ceed4 to 7cc1030 Compare November 3, 2025 13:31

ardatan mentioned this pull request Nov 4, 2025

docs(router): subgraph timeout configuration graphql-hive/console#7214

Open

ardatan force-pushed the primitive-expression branch from 046cc07 to 8650809 Compare November 5, 2025 23:25

Base automatically changed from primitive-expression to main November 21, 2025 15:38

kamilkisiela force-pushed the subgraph_timeout branch from d069f91 to 48c0b6b Compare November 25, 2025 09:14

ardatan and others added 8 commits November 25, 2025 10:17

feat(router): Subgraph Timeout Configuration

40f60ab

I guess?

09e1181

Make sure Duration in config is treated as String

2fe644d

e2e tests, fix config and adjust naming

a0aa887

Compile timeout expressions during config load

0e5bf38

Use a link to the vector.dev documentation for the VRL error messages

a91a23e

syntax

c776475

Introduce .default value as a fallback

7d062f6

kamilkisiela force-pushed the subgraph_timeout branch from 48c0b6b to 7d062f6 Compare November 26, 2025 12:38

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat(router): Subgraph Timeout Configuration #541

feat(router): Subgraph Timeout Configuration #541

ardatan commented Oct 31, 2025 •

edited

Loading

Uh oh!

gemini-code-assist bot commented Oct 31, 2025

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions bot commented Oct 31, 2025 •

edited

Loading

Uh oh!

github-actions bot commented Oct 31, 2025 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

feat(router): Subgraph Timeout Configuration #541

Are you sure you want to change the base?

feat(router): Subgraph Timeout Configuration #541

Conversation

ardatan commented Oct 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gemini-code-assist bot commented Oct 31, 2025

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions bot commented Oct 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ k6-benchmark results

Uh oh!

github-actions bot commented Oct 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

ardatan commented Oct 31, 2025 •

edited

Loading

github-actions bot commented Oct 31, 2025 •

edited

Loading

✅ `k6-benchmark` results

github-actions bot commented Oct 31, 2025 •

edited

Loading