Skip to main content
How-to guideApplies to: Unified engine deployments
Use the session operation API for features whose public definition is SQL. The HTTP transport carries the SQL statement; the SQL reference defines its syntax.

Preserve statement boundaries

The HTTP SQL operation request contains one SQL statement. For a script, submit statements as separate operations in the same session and wait for each required predecessor to succeed. Each new statement needs its own operation ID. A retry of one statement keeps that statement’s original ID. The product’s Run all control follows statement execution. Do not assume that posting an entire script creates a transaction across statements.

Inspect secret metadata

This read-only example lists secret definitions. It does not return plaintext values:

List secret definitions

Lists secret metadata through SQL operations. Plaintext secret values are not returned. Replace the placeholders in your client. Keep returned IDs for later requests. Paths below use the native API prefix; adjust it for your deployment.

1. Open a session

From the response, keep session_id from /session_id. Endpoint reference.

2. Submit SQL

Endpoint reference.

3. Follow execution

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

4. Read result metadata

Endpoint reference.

5. Fetch rows

Endpoint reference.

6. Close the session

Endpoint reference.

Inspect credential metadata

List credential definitions

Lists credential metadata through SQL operations without exposing underlying secret values. Replace the placeholders in your client. Keep returned IDs for later requests. Paths below use the native API prefix; adjust it for your deployment.

1. Open a session

From the response, keep session_id from /session_id. Endpoint reference.

2. Submit SQL

Endpoint reference.

3. Follow execution

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

4. Read result metadata

Endpoint reference.

5. Fetch rows

Endpoint reference.

6. Close the session

Endpoint reference.

Inspect streaming pipelines

Inspect streaming pipelines

Requires the native streaming path to be configured. Listing does not create or start pipelines. Replace the placeholders in your client. Keep returned IDs for later requests. Paths below use the native API prefix; adjust it for your deployment.

1. Open a session

From the response, keep session_id from /session_id. Endpoint reference.

2. Submit SQL

Endpoint reference.

3. Follow execution

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

4. Read result metadata

Endpoint reference.

5. Fetch rows

Endpoint reference.

6. Close the session

Endpoint reference. Creating a native streaming pipeline starts continuous execution by default. It needs its listed Kafka, Iceberg, and worker setup. A successful HTTP operation that creates a pipeline is separate from the pipeline reaching Running or publishing its first checkpoint. For stored-table examples, use a writable catalog. A context named default is not proof that an Iceberg catalog exists.

Create and remove a practice secret and credential

The instance needs a configured local secret store. Use unused names http_docs_token and http_docs_bearer, or replace them throughout. The token is a demonstration value and is not valid authentication. Run the explicit cleanup steps only for the objects created by this example.

Create a practice secret and credential

Send each statement separately in one session. Wait for success before its dependent step. Cleanup is explicit. Replace the placeholders in your client. Keep returned IDs for later requests. Paths below use the native API prefix; adjust it for your deployment.

1. Open a session

From the response, keep session_id from /session_id. Endpoint reference.

2. Create a demo secret

Endpoint reference.

3. Follow: Create a demo secret

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

4. Inspect the secret definition

Endpoint reference.

5. Follow: Inspect the secret definition

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

6. Read: Inspect the secret definition

Endpoint reference.

7. Create a bearer credential

Endpoint reference.

8. Follow: Create a bearer credential

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

9. Inspect the credential definition

Endpoint reference.

10. Follow: Inspect the credential definition

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

11. Read: Inspect the credential definition

Endpoint reference.

12. Remove the practice credential

Endpoint reference.

13. Follow: Remove the practice credential

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

14. Remove the practice secret

Endpoint reference.

15. Follow: Remove the practice secret

Repeat this read until /status reaches a terminal state: Succeeded, Failed, Cancelled, OutcomeUnknown, Skipped. Endpoint reference.

16. Close the session

Endpoint reference. For a complete pipeline definition over HTTP, follow Define a streaming pipeline over HTTP.
Last modified on November 4, 2026