feat(runners): warm pool support for Windows runners

[Brend-Smits]

Warm pool stack (review in order; each PR is based on the previous one):

1. feat(compute-providers): add EC2 standby primitives for warm pools #5491 feat(compute-providers): add EC2 standby primitives for warm pools
2. feat(pool): keep a warm pool of stopped instances #5492 feat(pool): keep a warm pool of stopped instances
3. feat(scale-up): start warm instances before launching cold runners #5493 feat(scale-up): start warm instances before launching cold runners
4. feat(scale-down): sweep stopped warm instances #5494 feat(scale-down): sweep stopped warm instances
5. feat(runners): boot modes for warm pool instances #5495 feat(runners): boot modes for warm pool instances
6. feat(runners): add warm_pool option to the root module #5496 feat(runners): add warm_pool option to the root module
7. feat(multi-runner): support warm_pool in runner configs #5497 feat(multi-runner): support warm_pool in runner configs
8. docs(warm-pool): add warm pool guide, ADR and examples #5498 docs(warm-pool): add warm pool guide, ADR and examples
9. fix(compute-providers): add warm pool boot modes to the EC2 template #5499 fix(compute-providers): add warm pool boot modes to the EC2 template
10. feat(runners): warm pool support for Windows runners #5500 feat(runners): warm pool support for Windows runners (this PR)
11. fix(scale-down): do not sweep warm instances that are being activated #5501 fix(scale-down): do not sweep warm instances that are being activated

Description

Windows runners get the same RUN, PRIME and WAIT boot modes as Linux.

• EC2Launch runs user data only on the first boot, so a priming instance copies the start logic between the ghr:start-runner markers out of its user data (read from instance metadata) to C:\ProgramData\ghr\start-runner.ps1 and registers a ghr-start-runner scheduled task that runs it as SYSTEM at startup.
• The priming shutdown is delayed by 60 seconds so EC2Launch records the user data as completed and does not run it again.
• An activated instance disables the task, so later reboots do not rerun the start logic, and logs the time from activation to runner start.
• Both Terraform stacks now accept warm pools for linux and windows runners and still reject osx. Docs and the ADR are updated.

Test Plan

• New pwsh harness modules/runners/templates/tests/start-runner.test.ps1 (37 checks: RUN, WAIT, PRIME, activated boot from the startup task, missing markers, and a check that both template copies match).
• terraform test for the warm pool tests in modules/runners and modules/compute-providers/aws/ec2 (Windows plans, macOS is rejected), modules/multi-runner and modules/runner-config.
• Sandbox deployment of examples/multi-runner with warm mode on the Windows Server 2022 config: the instance primed and stopped itself about 10 minutes after launch; a job later activated it, the startup task ran it in RUN mode, and the job started 95 s after dispatch (activation to runner start 64 s) instead of a full Windows cold boot.
• Same round: the existing Amazon Linux arm64, Ubuntu 24.04 and Ubuntu 22.04 (custom user data template) example configs primed and served jobs from their warm instance in 46-63 s; the x64 config with instance metadata tags disabled is rejected by validation. The v2 example configs (including Windows) plan with warm mode.

Related Issues

Supersedes #5204.

[github-actions]

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

Scanned Files

None

[Copilot]

Copilot review overview

🟡 Changes recommended

Boot-hook installation failures can currently be ignored, allowing unusable Windows instances to enter the warm pool.

Review effort: Balanced
Findings: 2

Open (2)

• Ignored hook setup failures leave standby instances unusable · New
• Ignored hook setup failures leave standby instances unusable · New

What changed in this PR

Adds Windows runner support to the existing EC2 warm-pool lifecycle.

Changes:

• Adds Windows RUN, PRIME, and WAIT boot modes using a startup scheduled task.
• Allows Windows warm pools while continuing to reject macOS.
• Adds PowerShell tests and updates documentation.

File	Description
variables.tf	Documents Windows support.
README.md	Updates root module documentation.
modules/​runners/​warm-pool.tf	Allows Windows runners.
modules/​runners/​variables.tf	Updates warm-pool description.
modules/​runners/​tests/​warm-pool.tftest.hcl	Tests Windows and macOS validation.
modules/​runners/​templates/​tests/​start-runner.test.ps1	Tests Windows boot modes.
modules/​runners/​templates/​start-runner.ps1	Implements the Windows boot hook.
modules/​runners/​README.md	Documents Windows support.
modules/​runner-config/​variables.orchestration-provider.tf	Updates provider documentation.
modules/​runner-config/​README.md	Regenerates provider documentation.
modules/​multi-runner/​variables.tf	Updates multi-runner documentation.
modules/​multi-runner/​README.md	Regenerates multi-runner documentation.
modules/​compute-providers/​aws/​ec2/​validations.tf	Allows Windows warm pools.
modules/​compute-providers/​aws/​ec2/​tests/​warm-pool.tftest.hcl	Tests provider validation.
modules/​compute-providers/​aws/​ec2/​templates/​start-runner.ps1	Adds the v2 Windows boot hook.
docs/​warm-pool.md	Documents Windows behavior.
docs/​configuration.md	Explains Windows scheduled-task boot hooks.
docs/​adr/​0004-warm-pool-standby.md	Records expanded OS support.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.