Fix runners monitoring workflow: improve reliability and reduce noise

[igorpecovnik]

• Remove push trigger (only run scheduled and manual)
• Change schedule from hourly to daily at 6 AM UTC
• Enable and fix concurrency group (runners-status, cancel-in-progress: true)
• Move NETBOX_TOKEN and NETBOX_API to env block for consistency
• Remove duplicate GH_TOKEN variable assignment
• Add set -euo pipefail for proper error handling
• Add error checking for GitHub API calls (gh and curl fallback)
• Add error checking for NetBox API call
• Add validation that received data is not empty
• Fix runner status logic: only show 🟢 for runners with "online" status
• Previously showed 🟢 for any runner with status or "github" label

[coderabbitai]

Walkthrough

This PR modifies the GitHub Actions workflow for monitoring runner status. The trigger schedule changes from push and hourly execution to daily at 06:00 UTC. Concurrency configuration is updated with a new group name and cancel-in-progress set to true. Environment variable handling is simplified with explicit NETBOX_TOKEN and NETBOX_API variables. Script robustness is enhanced with set -euo pipefail. Error handling is strengthened for GitHub Runners and NetBox API fetches, including validation of non-empty responses. Data rendering logic is updated to show green dots for online runners and red dots with indices for others.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• Display self hosted runners status #16: Implements a similar GitHub Actions workflow that fetches org runners and NetBox data to render per-runner status with overlapping feature logic.

Suggested reviewers

• armbianworker

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main changes: improving reliability and reducing noise in the runners monitoring workflow through error handling and logic fixes.
Description check	✅ Passed	The description is directly related to the changeset, providing a detailed list of specific improvements made to the workflow including schedule changes, error handling, and logic fixes.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch runners

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In @.github/workflows/monitoring-runners-status.yml:
- Around line 67-76: The curl fallback that runs the command with GH_TOKEN and
writes to "$runners_nd" only requests per_page=100 and doesn't implement
pagination, so it will miss runners when there are >100; update the fallback
(the curl invocation and its enclosing if/then block) to loop through pages
(e.g., increment a page parameter and append each page's .runners[] into
"$runners_nd" until an empty page or fewer than requested items is returned) or,
if you prefer a minimal change, add a clear comment/log message in that block
documenting the 100-item limit and recommending use of "gh api --paginate" (the
code locations to change are the curl command that calls
"https://api.github.com/orgs/armbian/actions/runners?per_page=100", the
surrounding if/then block that exits on failure, and the variable
"$runners_nd").

📜 Review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Disabled knowledge base sources:

• Jira integration is disabled by default for public repositories

You can enable these sources in your CodeRabbit configuration.

📥 Commits

Reviewing files that changed from the base of the PR and between 9ac3db0 and 60bc8b6.

📒 Files selected for processing (1)

• .github/workflows/monitoring-runners-status.yml

🔇 Additional comments (5)

.github/workflows/monitoring-runners-status.yml (5)

5-10: LGTM!

Schedule change to daily at 06:00 UTC and the concurrency configuration look good. The cancel-in-progress: true will prevent duplicate workflow runs from piling up.

---

44-47: LGTM!

Moving NETBOX_TOKEN and NETBOX_API to the env block improves consistency, and adding set -euo pipefail is a best practice for catching errors early in shell scripts.

---

78-81: LGTM!

Good defensive check to ensure the runners data file is not empty before proceeding.

---

93-105: LGTM!

Solid error handling for the NetBox API call with -fsS flags and explicit empty-response validation.

---

126-132: LGTM! This is the key fix.

The updated logic correctly checks both that the runner exists in the lookup (-n "$runner_status") and that its status is specifically "online" before showing the green indicator. This fixes the false-positive issue described in the PR objectives.

_{✏️ Tip: You can disable this entire section by setting review_details to false in your review settings.}