Frequently Asked Questions

• Inside-out monitoring: Your process reports its own health rather than being monitored externally
• Simple API: Just make an HTTP request when your task completes
• Smart Alerts: Machine learning detects pattern changes, not just failures
• Zero infrastructure: No agents to install or configure
• Custom messages: Include context about what your process actually did

Yes. SilentCanary is designed with security in mind:

• All API calls use HTTPS encryption
• Canary tokens act as secure API keys
• No sensitive data is required - only timing and optional messages
• Data is isolated per user account
• Regular security audits and updates

Best practices: Keep your canary tokens secure, don't include sensitive information in messages, and use environment variables for tokens in scripts.

Interval: How often your process normally runs or should check in.

Grace Period: Extra time to allow for normal variation.

Examples:

• Hourly cron job: 60 minutes interval, 10-15 minutes grace
• Daily backup: 1440 minutes (24 hours) interval, 30-60 minutes grace
• Continuous service: 5-10 minutes interval, 2-5 minutes grace
• Weekly report: 10080 minutes (7 days) interval, 120 minutes grace

Tip: Start with generous grace periods and tighten them based on actual performance.

Yes! You can create multiple canaries for different services, processes, or environments.

Organization strategies:

• By service: "Database Backup", "API Health Check", "Report Generation"
• By environment: "Production DB Backup", "Staging DB Backup"
• By frequency: "Hourly Process", "Daily Cleanup", "Weekly Reports"
• By criticality: "Critical - Payment Processing", "Non-Critical - Log Rotation"

Use tags to organize and filter your canaries in the dashboard.

It depends on your workflow:

Email Only:

• Personal projects or small teams
• Non-urgent monitoring
• When you want a permanent record

Slack Only:

• Team-based workflows
• Need immediate visibility
• Integration with existing ChatOps

Both:

• Critical systems (redundancy)
• Mixed team preferences
• 24/7 operations (email for off-hours, Slack during work)

If a check-in fails to reach SilentCanary, the canary will eventually be marked as failed when it exceeds the interval + grace period.

To handle this:

• Implement retries: Try the check-in multiple times with delays
• Set appropriate grace periods: Account for occasional network delays
• Use timeouts: Don't let check-in attempts hang indefinitely
• Log failures: Track when check-ins fail in your process logs

Example retry logic is available in our troubleshooting guide.

Yes, but be careful with timing expectations.

Common scenarios:

• Load balancer: Multiple instances of the same service
• Redundancy: Backup systems that can take over
• Multi-stage process: Different steps checking in

Considerations:

• Any successful check-in resets the timer
• Configure intervals based on the expected frequency of ANY check-in
• Use messages to identify which instance checked in
• Consider separate canaries for truly independent processes

Messages are limited to 500 characters to keep the system responsive and logs readable.

Best practices for messages:

• Include key metrics: "Processed 1,234 records in 5.2 seconds"
• Status indicators: "Backup completed successfully - 2.3GB"
• Error summaries: "3 warnings, 0 errors"
• Resource usage: "Memory: 45%, CPU: 12%"

Don't include:

• Sensitive data (passwords, API keys, personal information)
• Very long stack traces (summarize instead)
• Binary data or encoded content

Currently, there are no strict rate limits, but the system may throttle excessive requests to maintain performance for all users.

Best practices:

• Don't check in more frequently than your configured interval
• Use reasonable timeouts (5-10 seconds) in your client code
• Implement exponential backoff for retries
• Consider the actual needs of your monitoring

Typical usage patterns:

• Cron jobs: Once per interval (hourly, daily, etc.)
• Services: Every few minutes to every hour
• Batch processes: Once per run completion

Use Smart Alerts when:

• Your process doesn't run at exact intervals
• Execution time varies significantly
• You need early warning of performance degradation
• Your process has time-of-day or day-of-week patterns
• You want to detect subtle changes in behavior

Stick with regular monitoring when:

• Your process runs very consistently
• You only need to know about complete failures
• Your process is new (< 1 week of data)
• Simple timeout-based alerts are sufficient

Best approach: Use both! Regular alerts catch failures, Smart Alerts provide early warning.

Smart Alerts analyze your historical check-in data to understand normal patterns:

What it learns:

• Timing patterns: Average intervals and normal variation
• Hourly patterns: Different behavior by time of day
• Daily patterns: Weekday vs weekend differences
• Consistency: How much timing varies normally

How it detects anomalies:

• Statistical analysis of timing deviations
• Comparison against learned hourly/daily patterns
• Sensitivity-based thresholds
• Multiple factors combined for accurate detection

Requirements: Minimum 3 successful check-ins in the learning period, some consistency in patterns.

There are currently no strict limits on the number of canaries you can create. However, consider organizing your monitoring logically:

• One canary per distinct process or service
• Separate canaries for different environments (prod, staging)
• Group related checks using tags for better organization
• Consider the maintenance overhead of many canaries

If you need to monitor a very large number of processes, consider whether some could be grouped together or monitored at a higher level.

Yes, you can update your email address in your account settings.

Steps:

1. Go to your account Settings page
2. Contact support to change your email address (this requires admin assistance for security)
3. Verify the new email address

Note: Email verification is required for the new address before alerts will be sent there.

You can delete your account from the Settings page.

What gets deleted:

• Your user account
• All your canaries
• All check-in logs and history
• Smart Alert configurations

⚠️ Warning: This action is permanent and cannot be undone. Make sure to disable any scripts or processes that are still trying to check in to avoid errors.