Which languages are supported?
Go, Node.js, Python and Rust on the server, plus a browser SDK. All are built on OpenTelemetry.
Middle Monitor covers the core APM workflow — distributed tracing, root cause analysis, profiling and alerting — without the enterprise sales motion that typically surrounds these platforms.
New Relic and Dynatrace are deep, enterprise-grade platforms with extensive auto-instrumentation and AI-assisted analysis built over many years. At large scale, with dedicated observability teams, they offer more surface area than Middle Monitor does today.
Adding the SDK's HTTP middleware is a single line — net/http, Echo and Gin in Go, and the equivalent in Node.js, Python and Rust. From there incoming and outgoing requests are traced without further work, and W3C trace context propagates across services so a request's full path stays intact.
Heap and CPU flame graphs come from production through pprof integration in Go, included on the Pro plan rather than priced as a separate module. Profiling is Go-only today; the other SDKs cover traces, errors and logs.
When an error rate or latency threshold is crossed, the engine looks for infrastructure spikes and failed network checks in the same window and states a likely cause — a CPU spike on a named host, or an unreachable DNS — instead of only showing you a chart.
A Terraform provider manages hosts, services and install tokens, and a REST API covers hosts, services, errors, alert rules, incidents, channels and API keys.
Go, Node.js, Python and Rust on the server, plus a browser SDK. All are built on OpenTelemetry.
Continuous profiling is available on the Pro plan. In Go it integrates through net/http/pprof.