Repo of the Day
henrygd/beszel
Published: Nov 16, 2026
Open repository ↗Lightweight server monitoring with historical data, docker stats, and alerts
Summary
Beszel is a self-hosted server monitoring platform with two components: a PocketBase-based web “hub” and an “agent” that runs on each system being monitored. It records historical host and container metrics, presents them through a web dashboard, and supports configurable alerts, making it a practical fit for homelabs and other self-managed infrastructure.
What it is useful for
Beszel gives engineers a central place to track server health and investigate changes over time. Host metrics include CPU and memory usage, disk usage and I/O, network activity, load average, temperature, GPU usage and power draw, and battery charge. For Docker and Podman workloads, it tracks container status along with CPU, memory, and network usage.
The historical view helps operators identify capacity trends, recurring resource spikes, and possible performance regressions. Alerts can be configured for most metrics and sent through supported notification services. Beszel also monitors network response times and interruptions, reports disk health through S.M.A.R.T. data, and can alert on drive failure. Its ZFS coverage includes pool health, capacity, I/O throughput, scrub status, and per-dataset usage.
It also provides practical administration features: users can manage their own systems while administrators share systems across accounts. OAuth/OIDC authentication is supported, and password authentication can be disabled. Automatic backups can be saved to disk or S3-compatible storage.
Metric availability depends on the host. Fan monitoring is explicitly Linux-based through /sys/class/hwmon; eMMC wear data and Linux mdraid array health are available through sysfs when supported. The README’s overview mentions API access, but its REST API feature entry is commented out, so engineers should verify the current API scope before relying on it for automation.
How engineers can use it
Start with Beszel’s quick start guide, which the README identifies as the intended setup path. The basic operating model is to run the hub as the central dashboard, deploy an agent on each machine to monitor, and connect those agents to the hub. From there, engineers can inspect individual system pages, review historical charts, configure alerts for the metrics that matter to their services, and set the backup destination.
The repository README does not include installation commands, environment variables, or service definitions, so the external guide is necessary for exact deployment steps. Before enabling alerts or relying on a particular metric, check that the target operating system and hardware expose the required interfaces. This makes Beszel most suitable as a focused monitoring layer that complements, rather than replaces, application-level telemetry.