Monitoring Agents

Overview

upti.my agents are lightweight, cross-platform processes that run directly on your infrastructure. Unlike external monitoring nodes that probe your services from the outside, agents operate from within your network. This gives them the ability to monitor internal services, private endpoints, and resources that are not exposed to the public internet.

Each agent registers with your upti.my account, receives its assigned checks, and continuously reports metrics and heartbeats back to the platform. If an agent stops reporting, upti.my automatically flags it as offline and triggers an alert.

Supported Platforms

Agents are designed to run anywhere your services live. upti.my provides installation packages and deployment manifests for all major platforms:

• Linux - Debian, Ubuntu, CentOS, RHEL, Alpine, and other distributions via binary or package manager
• macOS - Native binary for Intel and Apple Silicon architectures
• Windows - Runs as a Windows service with automatic startup
• Docker - Official container image for containerized environments
• Kubernetes DaemonSet - Deploy across all nodes in your cluster with a single manifest

curl -fsSL https://get.upti.my/agent | sh
uptimy-agent register --token YOUR_AGENT_TOKEN
uptimy-agent start

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: uptimy-agent
  namespace: monitoring
spec:
  selector:
    matchLabels:
      app: uptimy-agent
  template:
    metadata:
      labels:
        app: uptimy-agent
    spec:
      containers:
        - name: agent
          image: uptimy/agent:latest
          env:
            - name: UPTIMY_TOKEN
              valueFrom:
                secretKeyRef:
                  name: uptimy-agent
                  key: token

How Agents Work

Once registered, an agent establishes a persistent connection with the upti.my platform. It pulls its configuration, including which local checks to run and which recovery actions to perform when failures are detected. The agent then enters a continuous loop:

1. Execute checks at the configured interval for each assigned local check
2. Report results back to upti.my, including metrics like response time, CPU, memory, and disk usage
3. Send heartbeats at regular intervals to confirm the agent itself is alive and healthy
4. Trigger recovery actions when a check fails and a self-healing action is configured

Local System Checks

Agents can perform 8 types of local health checks directly on the host where they run. These checks monitor internal services, system resources, and local certificates without requiring any inbound network access.

Self-Healing Recovery Actions

When a local check fails, agents can automatically execute recovery actions to restore service health without human intervention. upti.my supports 10 recovery action types, from restarting a systemd service to running a custom shell script.

Plan Limits

The number of agents you can deploy depends on your upti.my plan:

Learn More

Dive deeper into agent capabilities with the detailed guides below.