I’ve been working as dev & ops for 25 years, essentially with small teams and companies. For years, we’ve self-hosted professional tools & public applications, and we used several apps to monitor and inspect stack and application health (api, web, db, tls, caches, etc...). So, like many people, we used grafana, Prometheus, uptimekuma, and more, so the monitoring stack ended up using more resources than what it monitored ! I wanted to have a small and single app to observe only what i need realy, with one single rule : less configuration is better.
So I created Maintenant, to regroup all that interested me : container status and update availability, HTTP and TCP checks directly declared by Docker labels, TLS certificate health, and heartbeats to monitor standalone scripts.
All in a single Go binary. No PromQL, no exporter required, SQLite by default and Postgres as option. What it is not: a metrics platform, if you like to query a very long history, stay with Prometheus.
Everything is on read-only : docker socket mounted with :ro or passed by socket-proxy (prefered on production), RBAC read-only on Kubernetes. Maintenant never start or stop any container.
Core app is community and on Apache 2.0 licence. Personal and Pro functionality (CVE and risk scoring, alerts routing and escalation) are paid but the source code is publicly available.
You can see what it looks like with online demo : https://demo.maintenant.dev (no real stack or Docker socket behind it) Next, working actualy to high availability, multi-tenancy with OIDC authentication and fine grained access.
Maintenant is laureat of Hyper Open X program, funded by France 2030 operated by Cap Digital with Bpifrance