Computer and server monitoring for MSPs

KairosLink monitors managed computers in near real time: CPU, memory, disks and availability, with warning and critical thresholds configurable per organization. Alerts arrive by email only when they open or escalate, without flooding the inbox, and resolve themselves when the problem goes away.

Metrics that matter, with nothing extra to install

The already installed agent reads CPU, memory, per-drive disk usage and uptime on every monitoring cycle. Readings use each system's native API: the CIM/WMI classes on Windows, /proc and systemd on Linux, and the system tools on macOS. In all three they are numeric values, so collection does not depend on the operating system's language: it makes no difference whether Windows is in Spanish or English. Data is stored as a time series and each computer shows its graphs for the last 24 hours, plus a Check now button to force an immediate reading when you are diagnosing live.

Alerts with judgment, not spam

For each computer and problem type there is a single open alert: it updates while the problem persists, escalates from warning to critical if it gets worse and resolves automatically when the value returns to normal. Email goes out only when the alert opens and when it escalates. A technician can acknowledge the alert to signal they are working on it, with a record of who and when.

Fine-grained control of notifications

Thresholds (for example, warning at 85% CPU and critical at 95%) and the check interval are configured per organization. Each computer also has its own monitoring switch and its own email alerts switch: you can keep measuring a noisy computer without it flooding your inbox. On the same managed computers you can add DNS web filtering to control what they browse and certificate monitoring so an expiration never catches you late.

Frequently asked questions