How to set up monitors and tailor them to your use case

How to set up a monitor

Under Monitors → Create Monitor, you will find the configuration mask.

First, give the new monitor a name and link the service you want to monitor. Next, set the interval at which the monitor should check your service. Under Retries, specify how many times a check can fail before an alert is triggered. Now select your desired probe type and enter an address and timeout.

The timeout determines the maximum duration of your monitor probe. With a timeout of five seconds, the probe must not take longer than five seconds to complete.

In the HTTP Monitor, you can also customize the following settings.

Under Method, you can choose from the following options:

• GET: Retrieves the URI (Uniform Resource Identifier) of the target.

• POST: Sends data to a server to create a new resource or perform an action.

• PUT: Updates the target or creates content within it.

• DELETE: Removes the target.

• PATCH: Makes partial changes to the target.

• HEAD: Unlike GET, retrieves only the headers of the target; the body is not included.

• OPTIONS: Returns the available HTTP methods and other options for the target to determine its capabilities.

Optionally, you can specify which status codes are allowed for your monitor under Valid Status Codes.

Status codes are sent from your monitored target to your monitoring tool to indicate the target's current status. The most well-known status code is probably 404 – Not Found.

Under Headers, you can set keys and values (= information) that you want your monitor to query.

Once you've finalized your monitor under + Add, it becomes active. You can pause the monitor at any time afterward.

Differences between ICMP (Ping) and HTTP monitoring

ICMP/Ping Monitoring

ICMP (Internet Control Message Protocol) is a network protocol. Ping monitoring sends so-called echo request packets to a host and waits for echo reply responses. Three key metrics are measured in this process:

• Reachability – Does the host respond?

• Round-Trip Time (RTT) – How long does the round trip take?

• Packet Loss – What percentage of packets is lost?

ICMP can assess network latency, packet loss, and network connectivity, and determine whether the host is online or offline.

ICMP/Ping monitoring has two main issues that must be taken into account:

ICMP traffic is often blocked by default by firewalls or cloud providers. This can result in a server running perfectly fine, but ICMP being blocked, causing the check to be marked as failed.

Similarly, a host may be reachable even though the underlying services have failed.

HTTP monitoring

HTTP monitoring sends actual HTTP/HTTPS requests (using methods such as GET, POST, HEAD, etc.) to a URL and evaluates the complete response. HTTP monitoring can perform the following checks:

• Status codes

• Content check: Does the page contain the expected text, or is it missing (e.g., after a failed deployment)?

• Certificate monitoring (SSL/TLS)

• Response time

• Redirects: Correct redirects (HTTP → HTTPS)?

• Endpoint authentication

Which use case calls for which monitor type?

When ICMP/Ping is useful

• Network Infrastructure – Monitor routers, switches, and firewalls.

• Basic Server Monitoring – Is the physical/virtual host online?

• Latency trending – Measuring network quality over time.

• Internal devices – Printers, IoT devices, NAS systems without an HTTP interface.

• Basic alerting – First indicator before deeper diagnostics begin.

When HTTP should come into play

• Web Applications & APIs: A reliable way to verify whether a service is truly operational.

• E-commerce: Monitor checkout pages and payment endpoints.

• SSL Expiration Alerts: Notifications about certificate expirations.

• Content Monitoring: Was the homepage delivered correctly after a deployment?

• SLA Measurement: Real end-user experience.

• Multi-Step Flows: e.g., Login → API Call → Validate Response.

Best Practice: Combining ICMP and HTTP

In practice, these two methods complement each other: A robust monitoring setup uses ICMP for quick network-layer checks and HTTP for service-layer validation. Only this combination provides a complete view of availability.