canonical: https://jentic.com/apis/googleapis.com/datapipelines # Google Data pipelines API The Google Cloud Data pipelines API offers a thin scheduling layer over Dataflow templates, letting teams create, list, run, and stop recurring data analytics pipelines. It exposes 8 endpoints scoped to pipelines and their job history, with a built-in cron-style schedule on each pipeline. Pipelines are backed by Dataflow batch and streaming jobs, but this API removes the need to schedule them through Cloud Scheduler or Composer. ## For AI agents Create, list, run, and stop scheduled Dataflow pipelines through a managed scheduling layer. Lets agents launch recurring data analytics jobs without wiring up Cloud Scheduler. ## Scope Does not author Beam code, manipulate running Dataflow jobs directly, or schedule non-Dataflow targets - use for scheduling and lifecycle of Dataflow-template-backed pipelines only. ## Capabilities - Create scheduled batch pipelines backed by Dataflow templates - Trigger an on-demand run of a pipeline regardless of its schedule - Stop a streaming pipeline cleanly through the API - List historical jobs produced by a recurring pipeline - Update pipeline parameters, schedule, and worker configuration - Inspect each pipeline's last execution status and error reason ## Use cases ### Scheduled Dataflow Batch Pipelines Schedule recurring Dataflow batch jobs from a managed cron without standing up Cloud Scheduler or Composer. The API stores the template path, parameters, and schedule on a Pipeline resource, and creates a Dataflow job each tick. Agents can list past job runs and inspect failure causes from a single resource. Setup typically takes minutes once a template URL is available. Example prompt: Create a pipeline 'nightly-export' running gs://acme/templates/export.json with parameters region=us-central1, scheduled at '0 2 * * *' UTC ### On-Demand Pipeline Runs Trigger a pipeline run outside its normal schedule for backfills, ad-hoc reports, or operator-driven reruns after a failure. The API's run endpoint creates a fresh Dataflow job from the pipeline's template and parameters, and returns the resulting Dataflow job reference for downstream monitoring. This avoids re-deploying the template just to run it once. Example prompt: Run pipeline 'nightly-export' on demand and return the resulting Dataflow job ID for the operator to track ### Pipeline Lifecycle Stop and Resume Stop a running streaming pipeline, update its parameters or schedule, and recreate it cleanly without orphaning Dataflow jobs. The API's stop endpoint drains the underlying Dataflow job, freeing slots and stopping further charges. This supports cost-control automations that pause non-critical pipelines outside business hours. Example prompt: Stop pipeline 'real-time-clicks' to drain the underlying Dataflow streaming job, then update its workerCount and recreate ### AI Agent Pipeline Scheduler An AI agent can schedule, run, and stop recurring Dataflow pipelines through Jentic without writing OAuth or template-management code. Jentic search returns the matching createPipeline, runPipeline, or stopPipeline operation, the agent loads the schema, and Jentic executes against datapipelines.googleapis.com using vault-stored credentials. Example prompt: Use Jentic to search 'create a scheduled dataflow pipeline', load the createPipeline schema, and execute it with the template, parameters, and cron schedule ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | /v1/{+parent}/pipelines | Create a scheduled Dataflow pipeline | | GET | /v1/{+parent}/pipelines | List pipelines under a project location | | POST | /v1/{+name}:run | Trigger an on-demand pipeline run | | POST | /v1/{+name}:stop | Stop a running streaming pipeline | | GET | /v1/{+parent}/jobs | List historical jobs of a pipeline | ## Key resources - **Pipelines** — Recurring Dataflow pipelines with schedule, template, and parameters - **Jobs** — Historical Dataflow job runs produced by a pipeline ## Why Jentic - **Setup:** Wiring the Data pipelines API by hand means setting up Google OAuth, refreshing short-lived scoped tokens rather than holding client secrets in code, and addressing pipeline resources on datapipelines.googleapis.com. Through Jentic you install once, import the Data pipelines API from the API Directory, store the Google credential once, and your agent calls it. - **Permission scoping:** The API carries the parent and pipeline name in the URL path (/v1/{parent}/pipelines and /v1/{name}:run), so a rule can pin your agent to one pipeline. You choose the operations it may call, so it can list and run pipelines while stop is not included unless you add it. - **Credential handling:** Your Google OAuth credential is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'schedule a Dataflow pipeline', and Jentic returns the matching Data pipelines operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **Dataflow API** — Data pipelines schedules Dataflow jobs; Dataflow exposes the underlying job control plane. - **Cloud Scheduler API** — Cloud Scheduler is a generic cron service; Data pipelines is purpose-built for Dataflow templates. - **Cloud Composer API** — Composer is managed Airflow with full DAG semantics; Data pipelines schedules a single Dataflow job per pipeline. ## FAQ ### What authentication does the Data pipelines API use? The Data pipelines API uses OAuth 2.0 with the cloud-platform scope. Through Jentic the OAuth client and refresh tokens are stored in your Jentic One instance and the agent receives short-lived scoped tokens, so raw Google credentials never enter the agent context. ### Can I run on-demand jobs with the Data pipelines API? Yes. POST /v1/{+name}:run starts a fresh Dataflow job from the pipeline's template and parameters, independent of its cron schedule. The response includes the Dataflow job reference so the agent can follow the job through the Dataflow API. ### What are the rate limits for the Data pipelines API? Google enforces standard Cloud quotas on datapipelines.googleapis.com: per-project rate limits on read/write calls and Dataflow's underlying quota on concurrent jobs per region. Quotas are visible in the Cloud Console under IAM and admin, quotas, filtered to datapipelines.googleapis.com. ### How do I create a scheduled pipeline through Jentic? Search Jentic for 'create a scheduled dataflow pipeline', load the schema for POST /v1/{+parent}/pipelines, and execute with the template URL, parameters, schedule (cron string), and pipelineSources. Jentic returns the created Pipeline resource with its scheduleInfo set. ### Is the Data pipelines API free? API calls are free; pipeline runs are billed under standard Dataflow pricing (vCPU, memory, and PD-SSD seconds for batch and streaming jobs). There is no separate scheduling charge from Data pipelines itself. ### How do I stop a streaming pipeline cleanly? Call POST /v1/{+name}:stop with the pipeline resource name. The service issues a drain on the underlying Dataflow streaming job, allowing in-flight elements to complete before the workers shut down. The pipeline state moves to PAUSED. ### Can I limit what my agent is allowed to do with the Data pipelines API? Yes. Because you run Jentic One yourself, your own rules decide which operations and credentials the agent may use, and this API makes scoping straightforward since the parent and pipeline name travel in the URL path (/v1/{parent}/pipelines and /v1/{name}:run). You can pin the agent to a single pipeline and choose exactly which operations it may call, for example allowing it to list and run pipelines while withholding the stop operation. The Google OAuth credential is held by your instance and injected only at execution time, so the agent never sees the raw secret.