A GitHub Action that monitors websites' uptime, logs results, and creates issues when sites are down. All managed entirely within your GitHub repository, with no external servers or cloud services required.
- ✅ Monitor multiple websites concurrently
- 📊 Log response times and status codes
- 🚨 Automatically create GitHub issues for failures
- 🌍 Timezone-aware timestamps
- ⚙️ Configurable timeouts and success codes
- 📝 Commit logs back to repository
If you’re new to GitHub Actions or prefer a detailed, step-by-step guide, check out GETTING_STARTED guide. Otherwise, continue here for a quick setup.
name: Website Uptime Monitor
on:
schedule:
- cron: '0 */12 * * 1-5' # run automatically every 12 hours, Monday–Friday
workflow_dispatch:
permissions:
contents: write
issues: write
jobs:
uptime-check:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Monitor websites
uses: cd-roman/uptime-monitor@v1
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
sites: |
https://example.com
https://api.example.com/health
https://status.example.com
timezone: "Europe/Paris"| Input | Description | Required | Default |
|---|---|---|---|
sites |
Newline-separated URLs to monitor | Yes | - |
timezone |
IANA timezone for timestamps | No | UTC |
log-file |
Path to log file | No | uptime-monitor-results.log |
timeout |
Request timeout in seconds | No | 15 |
success-codes |
HTTP codes considered successful | No | 200,201,202,204 |
max-concurrent |
Max concurrent requests | No | 5 |
create-issue |
Create GitHub issue on failures | No | true |
| Output | Description |
|---|---|
failed-sites |
Number of sites that failed |
total-sites |
Total number of sites checked |
log-file-path |
Path to the generated log file |
permissions:
contents: write # To commit log files
issues: write # To create issues on failuresname: Website Uptime Monitor
on:
schedule:
- cron: '0 */12 * * 1-5' # run automatically every 12 hours, Monday–Friday
workflow_dispatch:
permissions:
contents: write # for committing logs
issues: write # for opening issues
jobs:
uptime-check:
runs-on: ubuntu-latest
timeout-minutes: 20 # cancel the job if it runs over 20 minutes
steps:
- name: Checkout repo
uses: actions/checkout@v4
- name: Monitor websites
uses: cd-roman/uptime-monitor@v1
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
# 1) List of sites, newline-separated
sites: |
https://example.com
https://api.example.com/health
https://status.example.com
# 2) IANA timezone for log timestamps
timezone: "America/New_York"
# 3) Path to write the log file (relative to repo root)
log-file: "logs/uptime-monitor.log"
# 4) Commit message when pushing the updated log
commit-message: "📊 Update uptime log"
# 5) Per-request timeout (seconds)
timeout: "15"
# 6) Which HTTP codes count as “success”
success-codes: "200,201,202,204,301,302"
# 7) Max number of concurrent requests
max-concurrent: "5"
# 8) Whether to auto-create an issue on non-success code
create-issue: "true"Ensure your workflow declares:
permissions:
contents: write
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}Verify log-file is a path within the repository and not excluded by .gitignore.
Add a timeout-minutes: under your job to cap total runtime:
jobs:
uptime-check:
timeout-minutes: 20Try lowering the timeout value (in seconds) or reducing max-concurrent to avoid long-running or overlapping fetch requests.
Set create-issue: "false" in your step’s with: block. The log will still record failures, but no issue will be opened.
The action treats a failed fetch() as status 0 and logs “Error: Timeout” or the exception message.
You can adjust timeout (e.g. 5 seconds) or disable issue creation for noisy endpoints.
Provide a comma-separated list without spaces, e.g. "200,204,301".
The action parses each code with parseInt; invalid entries are ignored.
The action opens issues using the GITHUB_TOKEN, so they’re authored by the GitHub Actions bot and you won’t be notified unless you’re subscribed to issue events or participate in the thread.
To receive emails for new issues, go to your repository, click Watch > Custom, select Issues, and then Apply. Once subscribed, you’ll get an email each time the action generates an issue.
By default, GitHub cancels jobs after 6 hours. Use timeout-minutes: X in your workflow to shorten this and avoid runaway bills.
Contributions are welcome! Refer to CONTRIBUTING.md for guidelines on getting started.
This project is licensed under the MIT License. See LICENSE for details.
If you have any questions, found a bug, want to request a feature, or need help with setup, please open an issue.
✨ Happy monitoring!