Compute Resources and Alerting

Compute resources determine where SQL, ingestion, and workflow tasks initiated through the SDK run. Alerting notifies operators when those tasks develop problems. Both are scoped to a workspace, but they have different roles: compute resources execute work, while alert rules observe execution results and metrics.

SDK coverage

The official Python SDK and Go SDK currently expose Catalog, ingestion task, workflow, SQL, and health-check operations. Neither SDK currently exposes methods to:

• create, update, start, suspend, or delete a compute resource;

• create or update alert rules or notification recipients;

• list or update alert records.

An application can therefore submit and inspect work that consumes compute through the SDK, while compute resources and alerting are managed in the AI Studio console. Do not use generic HTTP helpers to construct undocumented internal endpoints. Internal paths and payloads are not part of the stable SDK contract.

Choose a path

Goal	Documentation	Available SDK operations
Check resource type and availability before SQL or processing	Compute resources	run_sql / RunSQL, task and workflow operations, health_check / HealthCheck
Investigate an ingestion or workflow task after an alert	Alerting	get_task / GetTask, get_workflow_job / GetWorkflowJob
Create, resize, suspend, or remove compute	AI Studio · Compute resources	Console
Configure rules and recipients or inspect alert records	AI Studio · Alerts	Console

The service URL, API key, and workspace used by the SDK must identify the same workspace you manage in the console. See SDK fundamentals for authentication, client initialization, and common error handling.