fix(router): drain only in-flight requests of muxes the graph server tears down

[endigma]

Summary by CodeRabbit

• Bug Fixes
  • Improved shutdown and reload behavior so in-flight requests are tracked correctly during graph swaps.
  • Ensured the system waits only for requests owned by the server being shut down, while allowing reused request handlers to continue across reloads.
  • Added coverage for a shutdown scenario involving blocked requests, reloads, and configuration/version consistency.

[github-actions]

Router-nonroot image scan passed

✅ No security vulnerabilities found in image:

ghcr.io/wundergraph/cosmo/router:sha-9a44b1f4b0ff4a325fb94afe95f0d070e8b1df4a-nonroot

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: f47103da-2ec5-46b7-a860-2b109800924c

📥 Commits

Reviewing files that changed from the base of the PR and between 6c2dc0f and 0f2d926.

📒 Files selected for processing (3)

• router-tests/lifecycle/shutdown_test.go
• router/core/graph_server.go
• router/core/graph_server_test.go

💤 Files with no reviewable changes (1)

• router/core/graph_server_test.go

---

Walkthrough

This change moves in-flight request tracking from the graphServer level to individual graphMux instances, adding a per-mux atomic counter and an inFlightOwnedRequests helper that skips reused muxes during shutdown waits. Tests are updated and a new hot-reload shutdown test with a blocking middleware module is added.

Changes

Mux-owned in-flight tracking and shutdown wait

Layer / File(s)	Summary
Per-mux counter field router/core/graph_server.go	Removes the server-level inFlightRequests field and its initialization, adding a per-graphMux atomic.Int64 counter documented as owned by the mux, not drained when reused.
Request counting and shutdown wait logic router/core/graph_server.go	GraphQL request handling now increments/decrements the mux-level counter; a new inFlightOwnedRequests() sums counters across non-reused muxes; shutdown wait now checks this aggregate instead of the removed server-level counter.
Unit test literal updates router/core/graph_server_test.go	Removes the now-obsolete inFlightRequests field initialization from graphServer literals in existing shutdown subtests.
Blocking middleware and hot-reload shutdown test router-tests/lifecycle/shutdown_test.go	Adds a header-driven blocking middleware module and TestGraphServerShutdown, verifying that shutdown waits on the old server's torn-down mux request while a reused mux continues serving a blocked request after reload.

Estimated code review effort: 4 (Complex) | ~45 minutes

Possibly related PRs

• wundergraph/cosmo#2838: Both PRs adjust shutdown/swap behavior in graph_server.go to special-case reused graphMux instances and add related swap tests.
• wundergraph/cosmo#3035: Both PRs modify reused-mux reload/shutdown lifecycle semantics and corresponding TestGraphServerShutdown coverage in the same file.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and accurately summarizes the main router draining behavior change.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

Warning

There were issues while running some tools. Please review the errors and either fix the tool's configuration or disable the tool if it's a critical failure.

🔧 golangci-lint (2.12.2)

level=error msg="[linters_context] typechecking error: pattern ./...: directory prefix . does not contain main module or its selected dependencies"

---

_{Comment @coderabbitai help to get the list of available commands.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 61.89%. Comparing base (31d1baf) to head (314d6f4).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3064      +/-   ##
==========================================
+ Coverage   61.85%   61.89%   +0.03%     
==========================================
  Files         261      261              
  Lines       30649    30657       +8     
==========================================
+ Hits        18958    18974      +16     
+ Misses      10179    10173       -6     
+ Partials     1512     1510       -2     

Files with missing lines	Coverage Δ
router/core/graph_server.go	85.51% <100.00%> (+0.07%)	⬆️

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.