Files
computer-playbook/scripts/github/update
kevinveenbirkenbachandClaude Opus 5 623ecc0248 feat(update): retune the preferred collection source from measured latency
Which of the two sources is faster is not a constant. galaxy.ansible.com
degraded from 207s to 687s over one day and then served 502s outright,
while the pinned git fallback carried the same install in 31s; a week
later the order may be the other way round. The daily job times both and
swaps the default only when the preferred one's median reaches three
times the other's over twelve samples.

A failed install is recorded as None and ranks worse than any duration,
so an outage counts instead of dropping out of the statistic. Below a
full window nothing changes, which keeps a single bad day from flipping
the default.

History lives in an artifact rather than the Actions cache: a schedule
run resolves to the default branch and has read-only cache access there,
so it could not carry the window forward.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-16 20:59:45 +02:00
..

Update Pipeline 🔄

Shell helpers that drive the scheduled 🔄 Update versions workflow.

Scope 📋

This directory contains the helpers invoked by cron-update.yml to bump pinned upstream versions and open a pull request once per day. The helpers are thin wrappers: the image-version and repository-ref bumpers delegate to Python CLIs under cli/contributing/ that scan roles/*/meta/services.yml and rewrite version pins in place; the PR-open helper commits, force-pushes the update branch, and creates or refreshes the PR via gh. Helpers MUST NOT be called from workflows other than 🔄 Update versions; the CI_ENABLE_AUTO_UPDATES gate is enforced once, at the workflow's job level, and bypassing it from another caller would create PRs even when auto-updates are intentionally disabled.

Each shim MUST resolve the repository root relative to its own location so it can run from any working directory, MUST forward arguments to the underlying CLI so --dry-run and similar flags keep working from the shell, and MUST exit non-zero on CLI failure so the workflow surfaces the error.

For the workflow catalog see workflows.md. For the repository variable that gates the workflow see configuration.md. For the App-token secrets that the PR-opening step requires see secrets.md.