Skip to content

About

No description or website provided.

Topics

Resources

Code of conduct

Security policy

Stars

5 stars

Watchers

0 watching

Forks

Repository files navigation

PagerDuty On-Call Scheduler

A Python CLI tool for analyzing PagerDuty on-call schedules and identifying team members who exceed configurable on-call day limits per month.

Features

  • Fetch on-call schedules by PagerDuty team
  • Timezone-aware per-user calculations - each user's on-call time is calculated in their local timezone
  • After-hours only counting - only counts days where on-call coverage extends past standard workday hours (configurable, default: 5 PM)
  • Calculate on-call days per user for a specified month
  • Identify users exceeding configurable daily limits
  • Beautiful color-coded terminal output using Rich
  • JSON output support for automation and scripting
  • Configurable via environment variables or CLI arguments

Installation

Prerequisites

  • Python 3.12 or higher
  • uv (fast Python package installer)
  • PagerDuty API key

Install with uv

# Clone the repository
git clone https://github.com/getsentry/on-call-scheduler.git
cd on-call-scheduler

# Install dependencies using uv
uv sync

# The tool is now ready to use
uv run oncall-scheduler --help

Install in development mode

# Install with development dependencies
uv sync --extra dev

# Run tests
uv run pytest

Configuration

PagerDuty API Key

  1. Log in to your PagerDuty account
  2. Navigate to Integrations → API Access Keys (General API token) or My Profile → User Settings (Personal API token).
  3. Create a new API key with read permissions
  4. Copy the API key for use in configuration

Environment Variables

Create a .env file in the project root (see .env.example):

# Required
PAGERDUTY_API_KEY=your_api_key_here

# Optional (can also be specified via CLI)
PAGERDUTY_TEAM_IDS=TEAM1,TEAM2,TEAM3
ONCALL_MAX_DAYS=10

# Optional: Sentry error tracking
SENTRY_DSN=https://your-sentry-dsn@sentry.io/project-id
SENTRY_ENVIRONMENT=production

Configuration Priority

  1. CLI arguments (highest priority)
  2. Environment variables
  3. .env file
  4. Default values (lowest priority)

Sentry Error Tracking

The application includes built-in Sentry integration for error tracking and monitoring:

  1. Create a Sentry Project: Sign up at sentry.io and create a new project
  2. Get your DSN: Navigate to Settings → Projects → [Your Project] → Client Keys (DSN)
  3. Configure: Add the SENTRY_DSN to your .env file
  4. Optional: Set SENTRY_ENVIRONMENT to differentiate between environments (e.g., "production", "staging", "development")

When configured, the application will automatically:

  • Capture and report unhandled exceptions
  • Track performance metrics
  • Provide detailed error context and stack traces
  • Associate errors with specific releases

To disable Sentry, simply omit the SENTRY_DSN environment variable.

How It Works: Timezone-Aware After-Hours Counting

The scheduler uses timezone-aware calculations to only count days where users have on-call coverage extending past standard workday hours. This means:

Key Concepts

  1. User Timezones: Each user is assigned a timezone (from PagerDuty API, or configured via USER_TIMEZONES, or defaults to schedule timezone)

  2. Workday End Hour: Configurable via WORKDAY_END_HOUR (default: 17 for 5 PM)

  3. After-Hours Coverage: A day is only counted if the user's on-call shift extends past the workday end hour on that specific day in their local timezone

Examples

Scenario 1: Full Day Coverage

  • User timezone: America/New_York (EST/EDT)
  • Shift: Jan 1, 12:00 AM - Jan 2, 12:00 AM EST
  • Result: Jan 1 is counted (coverage extends past 5 PM EST)

Scenario 2: Business Hours Only

  • User timezone: America/New_York
  • Shift: Jan 1, 9:00 AM - Jan 1, 5:00 PM EST
  • Result: Jan 1 is NOT counted (coverage ends at 5 PM, doesn't extend past it)

Scenario 3: Evening Shift

  • User timezone: America/New_York
  • Shift: Jan 1, 5:00 PM - Jan 2, 9:00 AM EST
  • Result: Both Jan 1 and Jan 2 are counted (coverage extends past 5 PM on Jan 1, and past 5 PM on Jan 2)

Scenario 4: Night Shift

  • User timezone: Europe/London
  • Shift: Jan 1, 10:00 PM - Jan 2, 6:00 AM GMT
  • Result: Jan 1 is counted (shift extends past 5 PM), Jan 2 is NOT counted (shift ends at 6 AM, before 5 PM)

This approach ensures that only after-hours on-call burden is measured, not business-hours coverage.

Usage

Basic Usage

Check on-call schedules for the current month:

uv run oncall-scheduler check --team PXXXXXX

Specify Multiple Teams

uv run oncall-scheduler check --team TEAM1 --team TEAM2 --team TEAM3

Custom Month and Year

uv run oncall-scheduler check --team PXXXXXX --month 2 --year 2026

Custom Day Limit

uv run oncall-scheduler check --team PXXXXXX --max-days 15

Analyze Multiple Months

Analyze the next 3 months starting from the current month:

uv run oncall-scheduler check --team PXXXXXX --months 3

Analyze 6 months starting from a specific month:

uv run oncall-scheduler check --team PXXXXXX --month 1 --year 2026 --months 6

Verbose Output

Show all users including those under the limit:

uv run oncall-scheduler check --team PXXXXXX --verbose

JSON Output

Output results as JSON for scripting:

uv run oncall-scheduler check --team PXXXXXX --output json

Using with jq

# Get names of users over the limit
uv run oncall-scheduler check --team PXXXXXX --output json | jq '.over_limit[].user.name'

# Count users over limit
uv run oncall-scheduler check --team PXXXXXX --output json | jq '.over_limit | length'

Output Examples

Table Output (Default)

PagerDuty On-Call Analysis - January 2026
Max Days Allowed: 10

┏━━━━━━━━━━━━━━━━━━┳━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━┓
┃ User             ┃ Days ┃ Schedule          ┃ Dates               ┃ Status ┃
┡━━━━━━━━━━━━━━━━━━╇━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━┩
│ Alice Smith      │ 14   │ Primary On-Call   │ Jan 01-07, 15-21    │ Over   │  (red)
│ Bob Jones        │ 12   │ Secondary On-Call │ Jan 05-10, 20-25    │ Over   │  (red)
│ Carol White      │ 10   │ Primary On-Call   │ Jan 08-12, 22-26    │ At     │  (yellow)
└──────────────────┴──────┴───────────────────┴─────────────────────┴────────┘

Users under limit: 5 (use --verbose to see details)

Summary: 2 over limit, 1 at limit, 5 under limit

JSON Output

{
  "month": 1,
  "year": 2026,
  "max_days": 10,
  "over_limit": [
    {
      "user": {
        "id": "PXXXXXX",
        "name": "Alice Smith",
        "email": "alice@example.com",
        "html_url": "https://example.pagerduty.com/users/PXXXXXX"
      },
      "schedule_name": "Primary On-Call",
      "total_days": 14,
      "scheduled_dates": ["2026-01-01", "2026-01-02", "..."]
    }
  ],
  "at_limit": [...],
  "under_limit": [...],
  "summary": {
    "over_limit_count": 2,
    "at_limit_count": 1,
    "under_limit_count": 5
  }
}

Automation

Cron Job

Run daily at 9 AM:

0 9 * * * cd /path/to/on-call-scheduler && /path/to/uv run oncall-scheduler check >> /var/log/oncall-check.log 2>&1

Integration with Slack/Email

Use the JSON output to integrate with notification systems:

#!/bin/bash
RESULT=$(uv run oncall-scheduler check --team PXXXXXX --output json)
OVER_COUNT=$(echo "$RESULT" | jq '.summary.over_limit_count')

if [ "$OVER_COUNT" -gt 0 ]; then
    # Send alert to Slack, email, etc.
    echo "$RESULT" | jq '.over_limit[]' | curl -X POST -H 'Content-Type: application/json' \
        -d @- https://hooks.slack.com/services/YOUR/WEBHOOK/URL
fi

Development

Project Structure

src/
└── oncall_scheduler/
    ├── __init__.py
    ├── __main__.py          # Entry point
    ├── cli.py               # Click CLI interface
    ├── config.py            # Configuration management
    ├── models.py            # Data models
    ├── api/
    │   ├── client.py       # PagerDuty API wrapper
    │   └── exceptions.py   # Custom exceptions
    ├── analysis/
    │   ├── analyzer.py     # Main analysis orchestrator
    │   └── calculator.py   # On-call days calculator
    └── output/
        └── formatter.py    # Output formatting

Running Tests

# Run all tests
uv run pytest

# Run with coverage
uv run pytest --cov=src/oncall_scheduler --cov-report=term-missing

# Run specific test file
uv run pytest tests/test_calculator.py

Code Quality

# Format code
uv run ruff format .

# Lint code (add --fix to apply safe fixes)
uv run ruff check .

# Type checking
uv run mypy src/oncall_scheduler

CI runs ruff format --check, ruff check and mypy on every pull request and fails if any of them reports anything.

Troubleshooting

Authentication Errors

Error: Authentication failed

Solution: Check that your PAGERDUTY_API_KEY is correct and has the necessary permissions.

Team Not Found

Error: Resource not found: Team PXXXXXX not found

Solution: Verify the team ID is correct. You can find team IDs in the PagerDuty web UI under Configuration → Teams.

No Schedules Found

Warning: No schedules found for teams: [PXXXXXX]

Solution: The team might not have any associated schedules. Check that the team has schedules configured in PagerDuty.

License

Apache-2.0

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Support

For issues and questions:

About

No description or website provided.

Topics

Resources

Code of conduct

Security policy

Stars

5 stars

Watchers

0 watching

Forks

Releases

Sponsor this project

Packages

Used by

Contributors

Languages