Observability and evidence

Immediate lifecycle, durable experiment state, and fleet telemetry

Use each evidence plane for its intended scope:

PlaneUse
TauGrid status and logsImmediate lifecycle of one submitted run
Kubernetes events and pod stateAdmission, scheduling, startup, and termination
Ray and GPU metricsRuntime and hardware behavior
Expstore and durable artifactsAuthoritative experiment, checkpoint, and recovery state
adx-mon and ADX/KustoOptional hosted scalar and fleet analysis
Dashboards and alertsConsumer views over telemetry

A workload reaching Scheduled or Running status does not prove the model is making useful progress. Treat each as a separate claim and verify independently. Preserve raw logs, profiles, and artifacts when a diagnosis depends on them.

Metrics-store methodology

TauGrid separates durable experiment ownership from fleet analytics:

LayerResponsibility
Immutable metric filesPreserve normalized scalar history in Parquet
Expstore SQLite indexTrack runs, metric files, artifacts, idempotency keys, and summaries
Metric summariesProvide finite counts, step ranges, min/max values, and the latest finite point per run and metric
ADX/Kusto projectionSupport cross-workspace dashboards and fleet queries

Expstore is authoritative. ADX ingestion uses checkpoints and query-time deduplication so retries preserve one logical scalar point. Configure ADX retention, hot-cache duration, identities, and database roles as platform policy. The workspace’s durable storage lifecycle governs local metric files and artifacts.

Use project, run_group_id, and run_id for experiment identity. Use metric_name, step, and wall_time to align and compare scalar histories. Tags add workspace and workload dimensions without changing those core keys.

For a runnable workload that publishes immutable loss and accuracy chunks, see Live experiment evidence. For the full local evidence and artifact contract, see Experiment evidence and artifacts.

For the optional ADX-backed data plane, see Prepare ADX/Kusto for TauGrid. For Portal deployment and its available boards, see Configure Portal.