Skip to content

Monitor telemetry and health

Preview

The Workload API is a preview feature, on by default.

Feature flag: Enable Access to Workload API Experimental Features

Workloads emit OpenTelemetry logs, metrics, and traces; the platform also tracks request statistics, lifecycle events, and the readiness state that gates the running status. This section covers the telemetry model, readiness behavior, and the REST and console surfaces that expose them.

Resource Description
Monitoring concepts OpenTelemetry, statistics, events, retention, and Workload-level access patterns.
Application OpenTelemetry telemetry Where traces, logs, and metrics emitted by your container surface in the Console.
Instrument a Workload with OpenTelemetry (Python) Copy-ready SDK setup for traces, metrics, and logs from a Python container.
Health and readiness Readiness probes and the HTTP gates that control the transition to running.
REST: Monitor Workloads Stats, events, history, OTel paths, and aggregated Workload statistics.
UI: Monitor deployed Workloads Service health, resource and capacity monitoring, data exploration, and OTel metrics.
UI: View deployed Workload activity MLOps events, governance audit trail, and log viewers.