Monitors are the backbone of Phare Uptime, they’re your tireless digital assistants checking if your website or server is alive and kicking. Set them up once, and they’ll dutifully perform HTTP(S) and TCP checks according to your schedule.

​
HTTP(s)

When monitoring websites, Phare sends HTTP(s) requests and checks both the status code and (optionally) looks for specific text in the response. If HTTPS is used, Phare automatically keeps an eye on your SSL certificates too, making sure they’re valid and well configured.

You can find more details about the HTTP(s) monitoring settings below.

​
TCP

For services beyond websites, TCP monitoring establishes a direct connection to your server port to make sure everything’s responding correctly. Phare currently supports plain and TLS connections.

​
Interval

You decide how often Phare checks on your endpoints. The default is every minute, but you can adjust the frequency to suit your needs, anything between a quick 30 seconds or a relaxed 1 hour.

​
Confirmations & recovery

Nobody likes false alarms. That’s why Phare allows double-checks, or more, before creating an incident. A service is only marked as down (or recovered) after a configurable number of consecutive failures (or successes). The default is two confirmations, but you can set this anywhere from one (immediate alerts) to five checks.

​
Regions

With 11 monitoring regions worldwide, Phare lets you check your services from different parts of the globe. It is strongly recommended to use at least two regions, this helps avoid false positives and keeps your monitoring resilient even if one region experiences issues.

​
Alert Rules

Phare’s alert rules system allows you to decide which events trigger notifications. There are a few alert events that can be raised by monitors, you can find the full list in the alerting documentation.

​
Configuration options for HTTP(s)

​
Method

When creating HTTP monitors, you can choose which method to use for your requests. Phare currently supports:

  • HEAD: Ideal for quick availability checks. Retrieves only headers, saving bandwidth and processing time, perfect for frequent checks on large websites.
  • GET: The standard choice for most monitoring. Gets the full page content, allowing you to verify text content and properly test APIs content or websites that do not support HEAD requests.
  • OPTIONS: Useful for checking API configuration and CORS policies without triggering data changes. Great for testing API availability without actually using it.

​
User agent

Every HTTP request from Phare carries the following user-agent signature:

Mozilla/5.0 (compatible; Phare/1.0; +https://phare.io/products/uptime)

If you added a secret-value to the user agent header for firewall filtering, it will look like this:

Mozilla/5.0 (compatible; Phare/1.0; +https://phare.io/products/uptime; Secret=secret-value)

​
Headers

Need to get past authentication or test an API with specific requirements? Phare lets you add custom headers to your HTTP monitors. This feature opens up possibilities for:

  • Testing APIs with authorization tokens
  • Bypassing basic authentication on staging environments
  • Setting content negotiation preferences
  • Adding custom tracking identifiers

To prevent abuse, it is not possible to set a custom user agent headers, with the exception of adding a secret value for firewall filtering.

​
Suggestions

As you set up your monitors, Phare’s suggestion system analyzes your configuration and offers improvements to boost reliability, and reduce false positives:

  • Adding more monitoring regions for better global coverage
  • Increasing confirmation counts to prevent false positives
  • Alerting you to redirections that might impact performance reporting
  • Recommending more frequent checks for critical services

​
Statistics

Each monitor will show detailed statistics to help you understand your service performance at a glance:

  • Performance charts appear shortly after creating a monitor, showing response time trends
  • View the performance data by region to spot geographical performance variations
  • Zoom in on specific timeframes by clicking and dragging to select portions of the chart
  • Track key metrics including average response time, total downtime, and SLA percentages
  • Historical data from 24 hours up to 90 days lets you spot both immediate issues and long-term trends

Be aware that charts are rendered leveraging downsampling techniques to present meaningful visualizations without overwhelming your browser. This is why you won’t see every single data point on the chart.

​
Firewall protection

IP filtering is not reliable as the Phare Uptime monitoring agent is deployed on Bunny.net’s global network, meaning that IPs could be shared with other Bunny.net customers.

The easiest way to filter the Phare Uptime monitoring agent is to use the User-Agent header. You can also add a secret value to the user agent to make it unique to your organization.

Need something more robust? Contact support.

​
Private monitoring agent

Need specialized monitoring for complex requirements, large-scale operations, or access through private VPNs? A private monitoring agent can be setup in your preferred region.

Let’s talk about your specific needs.

​
Good practices

If you’re unsure about the best settings for your monitoring needs, you can find the Best practices to configure an uptime monitoring service on the Phare blog.

​
Limits

Phare’s monitoring infrastructure operates at an impressive scale, but comes with a few constraints:

  • No IP filtering: The monitoring agent uses Bunny.net’s shared IP range, making IP filtering unreliable.
  • Certain ports restricted: TCP monitoring doesn’t support ports 25, 465, 587, and 2525.

If these limitations are showstoppers for you, reach out to discuss alternatives.