success
bench
queued <1s ran 17s total 17s
1 jobs
0 failed
17s duration
— tests
1 artifacts
bench
successbench
successactions/checkout@11d5960a326750d5838078e36cf38b85af677262 <1s
successmark job-start 2s
Run
rid="${GITHUB_RUN_ID:-0}-${GITHUB_RUN_ATTEMPT:-1}-$(git rev-parse --short HEAD)"
echo "BENCH_RUN_ID=$rid" >> "$GITHUB_ENV"
source "${GITHUB_WORKSPACE:-$PWD}/scripts/emit_timing.sh"
BENCH_RUN_ID="$rid" bench_mark job-start
successrust cold build 8s
Run
set -euo pipefail
source "${GITHUB_WORKSPACE:-$PWD}/scripts/emit_timing.sh"
bench_step rust-cold-build bash -ec \
'rm -rf workloads/rust-build/target && cargo build --manifest-path workloads/rust-build/Cargo.toml'
successrust warm build <1s
Run
set -euo pipefail
source "${GITHUB_WORKSPACE:-$PWD}/scripts/emit_timing.sh"
bench_step rust-warm-build cargo build --manifest-path workloads/rust-build/Cargo.toml
successrust test <1s
Run
set -euo pipefail
source "${GITHUB_WORKSPACE:-$PWD}/scripts/emit_timing.sh"
bench_step rust-test cargo test --manifest-path workloads/rust-build/Cargo.toml
successdocker build 3s
Run
set -euo pipefail
source "${GITHUB_WORKSPACE:-$PWD}/scripts/emit_timing.sh"
if command -v docker >/dev/null 2>&1; then
bench_step docker-build bash -ec \
'docker build -t ci-bench:docker workloads/docker-build && docker run --rm ci-bench:docker'
else
bench_skip docker-build "docker not available on runner"
fi
successprobe runner 3s
Run
bash workloads/probe/probe.sh
successshow results <1s
Run
cat "${GITHUB_WORKSPACE:-$PWD}/results.jsonl"successactions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 <1s
No test reports for this run.
Tell a job where its test runner writes its report and this tab fills in: how many tests passed, failed and were skipped, a bar of the result at a glance, then every failure first — with the assertion that failed and the output around it — over a per-suite list of cases and their durations. No scrolling a log for the word FAIL.
It is one reports: key on the job — a job key, not a step, so the report is collected whatever the job concluded. JUnit XML and go test -json are both understood, and most runners already emit one of them. reports: is a rickub extension and works only in .rickub/workflows/; in .github/workflows/ it is a parse error on purpose, because those files are also meant to run on github.com.