// the find
OneUptime/oneuptime
Complete open-source monitoring and observability platform.
OneUptime is a self-hostable observability platform that tries to replace Datadog, PagerDuty, StatusPage, and Sentry with a single install. It covers uptime monitoring, incident management, on-call scheduling, log aggregation, APM with OpenTelemetry, and an AI agent that allegedly opens PRs with auto-fixes. Aimed at teams tired of paying five separate SaaS bills.
The OpenTelemetry integration is a legitimate architectural choice — agents ship to a standard protocol so you're not locked into proprietary instrumentation. The Helm chart path to production is real and actively maintained (GitHub Actions CI shows regular release cuts). Breadth of coverage is genuine: status pages, on-call rotation, and uptime monitoring are all present as first-class features, not bolted-on afterthoughts. The Proxmox and Ceph monitoring support is unusually specific and useful for teams running on-prem hypervisors.
The AI Copilot that 'opens PRs with code fixes automatically' is doing a lot of marketing work here — the actual implementation in AIAgent/ shows it handles exception-fixing tasks queued via the backend, not a general-purpose agent; don't deploy this expecting GitHub Copilot Workspace. The monorepo has accumulated serious complexity (dozens of top-level directories, each with its own package.json and Dockerfile template) that makes local dev setup genuinely painful and contributor onboarding slow. Self-hosted production deployments will feel the operational weight — this is not a 'deploy and forget' tool, it needs real Kubernetes capacity to run all its services. Test coverage appears sparse based on the GitHub Actions config; the test workflows exist but the suite looks thin for the surface area being claimed.