Skip to content

Use right-sized runner tiers for infra workflows#2023

Merged
mattgodbolt merged 1 commit intomainfrom
molty-runner-tiers
Mar 12, 2026
Merged

Use right-sized runner tiers for infra workflows#2023
mattgodbolt merged 1 commit intomainfrom
molty-runner-tiers

Conversation

@mattgodbolt-molty
Copy link
Copy Markdown
Contributor

Summary

Assigns runner size labels to four infra workflows based on their actual resource needs, now that ce-ci supports small (8 vCPU) and medium (16 vCPU) tiers alongside the existing large (32 vCPU) runners.

  • install.yml -> small: downloads pre-built tarballs from S3, ~2 min, no compilation
  • consolidate.yml -> medium: runs mksquashfs with zstd, genuinely multi-threaded
  • adhoc-command.yml -> size dropdown (default small): most ad-hoc commands are lightweight
  • bespoke-build.yaml -> size dropdown (default large): full compiler compilations need all cores

For the dropdown workflows, selecting large gives the original unlabeled behavior (any available runner). Selecting small or medium routes to the appropriately sized fleet.

Depends on compiler-explorer/ce-ci#18 -- the small and medium runner fleets must be deployed before this takes effect.

Assign runner size labels to match actual workload needs:
- install.yml: small (8 vCPU) -- just downloads pre-built tarballs
- consolidate.yml: medium (16 vCPU) -- mksquashfs is genuinely multi-threaded
- adhoc-command.yml: size dropdown (default small) -- most ad-hoc commands are lightweight
- bespoke-build.yaml: size dropdown (default large) -- full compiler compilations

🤖 Generated by LLM (Claude, via OpenClaw)
@mattgodbolt mattgodbolt marked this pull request as ready for review March 12, 2026 21:13
@mattgodbolt mattgodbolt merged commit 52762da into main Mar 12, 2026
1 check passed
@mattgodbolt mattgodbolt deleted the molty-runner-tiers branch March 12, 2026 21:14
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants