Instiq
Chapter 4 · Extend the platform: connectors, APIs, automation·v1.0.0·Updated 6/30/2026·~14 min

What's changed: Created PL-400 Chapter 4 (domain: Extend the platform, part 2): custom connectors (Open API definition, auth OAuth/API key/Basic, policy templates, Azure Function, data transformation), platform APIs and Azure Functions (Dataverse Web API OData, Organization service .NET SDK, throttling/retry, OAuth, optimistic concurrency/ExecuteMultiple, Azure Functions long-running/triggers/managed identity), and Power Automate cloud flows (Dataverse connector, expressions, Azure Key Vault, error handling run after, trigger filter/retry, child flows, Entra ID service principal).

4.2Platform APIs and Azure Functions

Key points

Understand the Dataverse Web API, the Organization service, API limits and retries, OAuth, concurrency/transactions/bulk, and Azure Functions (long-running, triggers, managed identities).

There are multiple ways to operate Dataverse programmatically. Design which API to use and within which constraints.

4.2.1Web API and Organization service

The Dataverse Web API is an OData v4-based REST API callable over HTTP from any language/client. The Organization service uses the .NET SDK (IOrganizationService) from server-side/plug-ins. Authenticate with OAuth 2.0 (a Microsoft Entra ID app registration). APIs have request limits (throttling); handle 429 responses with retry (exponential backoff).

4.2.2Concurrency, bulk, Azure Functions

Detect update conflicts with optimistic concurrency (ETag/RowVersion) and keep integrity with transactions. Batch many operations with bulk (ExecuteMultipleRequest) for performance. Offload long-running work beyond plug-in timeouts to Azure Functions, started by scheduled/event-driven triggers. Azure Functions can authenticate to Power Platform with managed identities, removing secret management.

Exam point

Cues: "OData v4 REST from any language" = Dataverse Web API. ".NET SDK from server/plug-in" = Organization service. "retry on 429 (exponential backoff)" = throttling handling. "detect update conflicts" = optimistic concurrency (ETag). "batch many operations" = ExecuteMultiple. "heavy work beyond timeout" = Azure Functions. "secret-free auth" = managed identity.

Warning

Watch the mix-ups: (1) Web API (REST/any language) vs Organization service (.NET SDK/server). (2) Throttling (limit) vs retry (handling). (3) ExecuteMultiple (bulk) vs single requests. (4) Offload heavy work to Azure Functions, not a sync plug-in. (5) Managed identity (secret-free) vs API keys.

Diagram: the Dataverse Web API (OData v4 REST, any language) and the Organization service (.NET SDK, server/plug-ins); authenticate with OAuth (Entra ID app registration); retry on 429 throttling (exponential backoff); detect update conflicts via optimistic concurrency (ETag); bulk with ExecuteMultiple; offload long-running work beyond plug-in timeouts to Azure Functions, started by scheduled/event triggers, with managed-identity (secret-free) auth.
Operate within limits

4.2.3Section summary

  • Web API=OData REST/any language; Organization service=.NET SDK/server; OAuth auth
  • Retry on throttling (exponential backoff); optimistic concurrency; bulk via ExecuteMultiple
  • Long-running work via Azure Functions; triggers; managed-identity auth

Sign in to track progress — Log in.

Quick check

(just a quick review)

Q1. Which OData v4-based API operates Dataverse data over HTTP from any-language clients?

Q2. What is the proper response when API throttling returns 429?

Q3. You want to batch many create/update operations into one request for performance. Best?

Q4. You want to run long-running work beyond plug-in timeouts from Power Platform. Best?

Q5. From Azure Functions, you want secure auth to Power Platform without managing secrets. Best?

Check your understandingPractice questions for Chapter 4: Extend the platform: connectors, APIs, automation

Keep track of your progress

The full study guide is free to read. Sign up free to practice with the question bank, track what you have read, review your mistakes, and highlight passages.