Fixes / 01

Undersized runner

Build and test dominate the job: the machine is the limit.

What the numbers show

build + test ≥ 5 min and ≥ 60% of the job

When most of a job is compiling and running tests, the job is CPU- and memory-bound. More cores only help if the build tool and test runner use them, so check parallelism flags too.

The change

Move the job to a bigger or faster runner
jobs:  build:-   runs-on: ubuntu-latest+   runs-on: graphide
Let the tools use every core
- run: cargo build --jobs $(nproc)- run: make -j"$(nproc)"- run: npx jest --maxWorkers=100%

Runner options

Options for this case. Judged on setup effort, undo path and sizing. Full comparison
SetupOne line: runs-on: graphide
UndoA pull request restores runs-on
SizingWarm machine sized to the job, up to 32 vCPU
SetupCreate a larger runner, change its label
UndoChange the label back
SizingLarger sizes on paid plans
SetupProvision and patch your own pool
UndoChange the label back, retire machines
SizingWhatever you provision
SetupCheck each vendor's docs
UndoCheck each vendor's docs
SizingCheck each vendor's docs

Verify the verdict stays the same

  1. 01Run the same commit on both runners.
  2. 02The job's conclusion matches on both.
  3. 03The test count and the build artifacts match.
  4. 04Paste both runs into /baseline and read the per-job difference.