[only21mil]

Problem

ResourceTelemetry.test.ts and ProcessDiagnostics.test.ts fail whenever the test worker's PID equals one of the fixture's fixed PIDs. The code under test treats process.pid as the server, and the fixtures put process.pid next to hardcoded 4242, 5000 and 9000 in one process table. When the worker lands on one of those, two fixture processes share a PID, the later one replaces the earlier one, and the counts come out wrong. This matches the merge-CI failure in #15983: backend processStarts 2 instead of 3, and backend processCount 1 instead of 2.

Change

Each synthetic PID is now an offset from process.pid (SERVER_PID + 1, + 2 and so on), so no fixture PID can equal the worker PID. This covers the codex child, the short-lived child, Electron and the resource monitor in ResourceTelemetry.test.ts, and the agent child in ProcessDiagnostics.test.ts. Expected messages that print the PID now build it from the same constant. Test files only; no production code changes.

Scope and approval

Fixes #15983. The triage comment, posted by Grok on behalf of @juliusmarminge, traced the failure to this PID collision and noted that ProcessDiagnostics.test.ts can flake the same way. It listed two fixes: derive the synthetic PIDs as offsets from process.pid, or pass serverPid explicitly as Model.test.ts and ResourceTelemetryHistory.test.ts do. It left the choice to a maintainer. This PR takes the first option because it stays inside the tests. These suites test the services, and the services take no serverPid: ResourceTelemetry.make() reads process.pid directly (ResourceTelemetry.ts lines 175, 285 and 437), and so does ProcessDiagnostics.ts (lines 79 and 92). Passing it explicitly would mean adding an injection point to production code.

Verification

I reproduced the collision for real by running vitest in a fresh Linux PID namespace with ns_last_pid set so the worker gets the chosen PID. The run used the threads pool, so process.pid inside the tests is that PID:

sudo unshare --pid --fork --mount-proc bash -c '
  cd apps/server
  echo 4241 > /proc/sys/kernel/ns_last_pid
  setpriv --reuid=1000 --regid=1000 --init-groups node <vitest.mjs> run --pool=threads \
    src/resourceTelemetry/ResourceTelemetry.test.ts src/diagnostics/ProcessDiagnostics.test.ts
  exit $?'

A --require probe logged the PID of each worker thread to confirm it.

Worker PID	Before (main 05696c636e)	After
4242	5 failed, 2 passed	7 passed
5000	2 failed, 5 passed	7 passed
9000	2 failed, 5 passed	7 passed
31337 (control)	7 passed	7 passed

At PID 4242, main reproduces the processStarts assertion from #15983: ResourceTelemetry.test.ts:260 gets expected 2 to be 3 (line 269 at the CI revision). The test that held the processCount assertion is no longer on main. All three ProcessDiagnostics tests fail at that PID too.

Also ran:

• vp test run on both files under normal PIDs: 7 passed.
• pnpm run typecheck in apps/server on a fresh install: passes.
• vp lint and vp fmt --check on both files: clean.

Production fix, tests, PR description and images written with Claude Opus 5.5 in T3 Code Nightly.

🤖 Generated with Claude Opus 5.5 in T3 Code Nightly

[macroscopeapp]

Approvability

Verdict: Approved at a183588

Macroscope's review found this PR approvable — The changes are confined to ignored, test-only telemetry fixtures and assertions, addressing PID collisions without altering production behavior, defaults, or static-analysis configuration.

Notes:

• No code objects were reviewed. Approvability was decided on eligibility alone.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Path: .coderabbit.config.ts
• Review profile: CHILL
• Plan: Advanced
• Run ID: cc163734-aafc-453d-bb7d-95599f34e958

📥 Commits

Reviewing files that changed from the base of the PR and between ab6be13 and a183588.

📒 Files selected for processing: 2

• apps/server/src/diagnostics/ProcessDiagnostics.test.ts
• apps/server/src/resourceTelemetry/ResourceTelemetry.test.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 5 remain after this review.

📝 Walkthrough

Two server test suites now derive process fixture IDs from the test worker PID. Fixtures, requests, lookups, and expected results use these IDs instead of fixed values. The existing process projection, refusal, rejection, and telemetry assertions remain unchanged.

Changes: Process-ID test fixtures

Layer / File(s)	Summary
Resource telemetry process fixtures apps/server/src/resourceTelemetry/ResourceTelemetry.test.ts	The suite uses worker-relative IDs for process fixtures, lookups, and assertions. Process relationships, sample data, counts, and telemetry totals remain unchanged.
Process diagnostics fixtures and assertions apps/server/src/diagnostics/ProcessDiagnostics.test.ts	The suite uses worker-relative server and child IDs in fixtures, signal requests, and expected results. The tested projection, stale-identity refusal, refresh-failure refusal, and Electron-target rejection remain unchanged.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Suggested reviewers: saphid, juliusmarminge

Merge Risk: ⚪ Minimal · up to a1835

The test fixtures now use worker-relative process IDs without changing production behavior. No actionable merge risk remains.

Pre-merge checks: 4

Check name	Status	Explanation
Linked Issues check		Issue #15983 requires a repair for nondeterministic ResourceTelemetry process-count assertions. The PR derives synthetic fixture PIDs from process.pid with distinct offsets in `ResourceTelemetry.tes…
Out of Scope Changes check		The changes are limited to the two test files named in the PR scope. They update synthetic PID fixtures and related expectations. No production code or unrelated behavior changes are reported. The tes…
Title check		The title clearly and concisely describes the main change: preventing test fixture PIDs from colliding with the test worker PID.
Description check		The pull request description includes all required sections. It clearly explains the PID collision, the test-only fix, linked issue and approval context, focused verification results, and the agent mo…

Full details: Linked Issues check

Explanation

Issue #15983 requires a repair for nondeterministic ResourceTelemetry process-count assertions. The PR derives synthetic fixture PIDs from process.pid with distinct offsets in ResourceTelemetry.test.ts and ProcessDiagnostics.test.ts. The reported reproductions at worker PIDs 4242, 5000, and 9000 pass after the change. The change also updates PID-based expectations. This meets the linked issue's coding requirement.

Full details: Out of Scope Changes check

Explanation

The changes are limited to the two test files named in the PR scope. They update synthetic PID fixtures and related expectations. No production code or unrelated behavior changes are reported. The test changes directly support issue #15983.

Full details: Description check

Explanation

The pull request description includes all required sections. It clearly explains the PID collision, the test-only fix, linked issue and approval context, focused verification results, and the agent model and harness used.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

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