D toc |
---|
Review the changes to the publicly available REST APIs for
D s product | ||
---|---|---|
|
Changes for Release 8.2
None.
Changes for Release 8.1
Customize relational connectors via API
Beginning in Release 8.1, you can customize aspects of each relational connection type available in your product edition through a set of APIs. Some terms:
- connection: In the
and via API, you can create and manage connections between the platform a specific datastore. A connection is the user-defined object that enables the connection to the datastore.D s webapp connector: A connection interfaces with a connector, which is an underlying driver and its related configuration, that perform the actual connection. This configuration information includes runtime, publishing, and connection definitions.
Tip Tip: All connections of the same type use the same underlying connector, including its configuration. Overrides that you apply to a connector apply to all current and new connections of that type in the workspace.
Info | ||||||
---|---|---|---|---|---|---|
NOTE: In prior releases,
|
Get connector identifier
To use these API endpoints, you must acquire the connector identifier. This value is the vendor
value for a connection of the type. You can acquire this value in one of two ways:
Create a connection of the type in the
. Use theD s webapp /v4/connections/:id
endpoint with the GET method to acquire the connection information for your connection. Acquire thevendor
value.D s api refdoclink operation/listConnections
- You may find the
vendor
values listed in the documentation. See Connection Types.
Get connector metadata information - defaults
The following endpoint returns the default metadata information for a specified connector type. This information is stored in the Connector Configuration Service database.
For the :connectorId
value below, use the vendor
value that you acquired above. For example, to acquire connector type definitions for MySQL connection type, use the value mysql
.
Endpoint | /v4/connectorMetadata/:connectorId/defaults | ||
---|---|---|---|
Method | GET | ||
Description | Get the default metadata for a connector without applying custom overrides. This metadata is used to defined connectivity, ingestion, and publishing for the connector. | ||
Documentation |
|
Get connector metadata information - current values
The following endpoint acquires the current metadata for a specified connector type, which include the default values with any applicable overrides applied to them.
Endpoint | /v4/connectorMetadata/:connectorId/ | ||
---|---|---|---|
Method | GET | ||
Description | Get the consolidated metadata for a connector in a given workspace. This metadata is used to defined connectivity, ingestion, and publishing for the connector. | ||
Documentation |
|
Get connector metadata information - get overrides values
The following endpoint retrieves the overrides that have been applied to a specific connector.
Endpoint | /v4/connectorMetadata/:connectorId/overrides | ||
---|---|---|---|
Method | GET | ||
Description | Get the metadata overrides for a connector in a given workspace. These overrides are applied to the base configuration for connectivity operations. | ||
Documentation |
|
Create overrides for a connector
The following endpoint applies the specified value or values as overrides to the connector.
Endpoint | /v4/connectorMetadata/:connectorId/overrides | ||||
---|---|---|---|---|---|
Method | POST | ||||
Description | The specified overrides are merged into the current set of overrides for the connector. A new entry is created if no overrides currently exist. | ||||
Documentation |
|
Delete overrides for a connector
The following endpoint deletes all override values for a specified connector.
Endpoint | /v4/connectorMetadata/:connectorId/overrides | ||
---|---|---|---|
Method | DELETE | ||
Description | All overrides are deleted. The connector reverts to the base configuration. | ||
Documentation |
|
Changes for Release 8.0
Deprecation of prior asset transfer endpoint
The following endpoint has been available for transferring assets between users:
Code Block |
---|
/v4/people/{to-userid:}/assetTransfer/{from-userid:} |
This endpoint is not compatible with the forward direction of the platform.
Info |
---|
NOTE: The above endpoint will be deprecated in a future release. Please switch to using the new endpoint for transferring assets between users. |
For more information, see
D s api refdoclink |
---|
operation/transferAssets |
Changes for Release 7.7
credentialProvider no longer required for /v4/awsConfig
In prior releases, API requests to the /v4/awsConfig
endpoint using the PUT method could receive the following error:
Code Block |
---|
{ "exception": { "name": "ApiValidationFailed", "message": "Message does not adhere to API specification", "details": "'credentialProvider' field in request body must not be null" }} |
The credentialProvider
field is not required, and this requirement has been removed from the endpoint.
D s api refdoclink |
---|
operation/updateAwsConfig |
/v4/awsConfigs PUT method has been deprecated
In prior releases, the /v4/awsConfigs
endpoint supported the use of a PUT method to modify AWS configuration objects.
This endpoint-method combination has been deprecated. The new PATCH method is the following:
Info |
---|
NOTE: The PUT method for this endpoint is still accessible. In a future release, it will be removed from the platform. Please switch to using the PATCH method. |
Endpoint | /v4/awsConfigs | ||
---|---|---|---|
Method | PATCH | ||
Request Body |
|
D s api refdoclink |
---|
operation/updateAwsConfig |
For more information on configuring AWS access objects, see API Workflow - Manage AWS Configurations.
Changes for Release 7.5
Flow sharing API now accepts user email addresses
Beginning in Release 7.5, the API has been enhanced to allow insertion of a user's email address as part of the request body. Below is an example request:
Endpoint | /v4/flows/402/permissions/ | ||
---|---|---|---|
Method | POST | ||
Request Body |
|
If the above returns a 201 - Success
status message, then user@example.com
has been given the role of collaborator
, which uses the flow_editor
policy, for flowId 402.
Tip |
---|
Tip: You can still use the internal identifier for the user, too. |
D s api refdoclink |
---|
operation/shareFlow |
Changes for Release 7.1
Introducing in-app API reference documentation
Beginning in Release 7.1, API reference documentation is available directly from the
D s webapp |
---|
Key features:
- More endpoints and attributes are now available!
- Reference content comes directly from the codebase.
- Searchable and browsable
More content and features will be added to this new capability of the next few releases.
Info |
---|
NOTE: The API reference doc that was published with the product documentation is no longer available. Please use the in-app API reference documentation. |
Info |
---|
NOTE: The API reference documentation may require enablement in your environment. For more information, see API Reference. |
Workflow documentation is still available in the product documentation. For more information on these workflows, see API Reference.
API changes for authorization
Release 7.1 introduces authorization, which provides finer-grained access controls to user-defined objects in the platform. For more information, see Changes to User Management.
As a result of authorization, the following changes have been applied to the listed API endpoints.
Connection permissions - POST
Endpoint | /v4/connections/:id/permissions/ |
---|---|
Method | POST |
Change:
When setting permissions, the request body now must include the policyTag
parameter:
Info |
---|
NOTE: This is required. |
Code Block |
---|
"policyTag": "connection_viewer", |
Parameter | Description |
---|---|
policyTag | Defines the policy level to assign to the change in connection permissions. Accepted values:
|
For more information, see the API reference documentation:
D s api refdoclink |
---|
operation/createConnectionPermission |
Connection permissions - GET
Endpoint | /v4/connections/:id/permissions/ |
---|---|
Method | GET |
Change:
Response no longer includes connectionPermissions
parameter.
Info |
---|
NOTE: Above is removed. |
Response does include the following new parameters:
Code Block |
---|
"policyTag": "connection_viewer", "workspaceAdmin": false, "isCreatedBy": true, |
Parameter | Description |
---|---|
policyTag | See previous. |
workspaceAdmin | If true , then the permitted user is a workspace admin. |
isCreatedBy | If true , then the permitted user created the connection. |
For details on these parameters, please see the API Reference documentation:
D s api refdoclink |
---|
operation/getConnectionPermissions |
Connections - GET LIST
Endpoint | /v4/connections |
---|---|
Method | GET |
Change:
Response now includes an authorizationPermission
field in the associatedPeople
block for each connection:
Code Block |
---|
{ "data": [ { "id": 14, .... "associatedPeople": { "data": [ { "id": 1, ... "authorizationPermission": { "policyName": null, "policyTag": "workspace_admin", "resourceOperationList": [ { "operations": [ "execute", "delete", "update", "read", "create", "share" ], "policyTag": "plan_author", "resourceType": "plan" }, { "operations": [ "execute", "delete", "update", "read", "create", "share" ], "policyTag": "flow_author", "resourceType": "flow" }, |
For details on these parameters, please see the API Reference documentation:
D s api refdoclink |
---|
operation/listConnections |
Connections - GET
Endpoint | /v4/connections/:id |
---|---|
Method | GET |
Change:
Same change as previous.
For details on these parameters, please see the API Reference documentation:
D s api refdoclink |
---|
operation/getConnection |
People - GET LIST
Endpoint | /v4/people |
---|---|
Method | GET |
Change:
Response now includes the following attributes:
maximalPrivileges
authorizationRoles
Info |
---|
NOTE: Above attributes appear only if the authenticating user is an administrator or the specific user. |
Code Block |
---|
{ "data": [ { "id": 29, .... "maximalPrivileges": [ { "operations": [ "delete", "update", "read", "create", "share" ], "policyTag": "connection_author", "resourceType": "connection" }, { "operations": [ "execute", "delete", "update", "read", "create", "share" ], "policyTag": "plan_author", "resourceType": "plan" }, { "operations": [ "execute", "delete", "update", "read", "create", "share" ], "policyTag": "flow_author", "resourceType": "flow" } ], "authorizationRoles": [ { "policyId": 13, "name": "default", "tag": null, "workspaceId": 1, "createdAt": "2020-04-22T07:50:04.502+0000", "updatedAt": "2020-04-22T07:50:04.502+0000", "resourceOperations": [ { "operations": [ "execute", "delete", "update", "read", "create", "share" ], "policyTag": "plan_author", "resourceType": "plan" }, { "operations": [ "execute", "delete", "update", "read", "create", "share" ], "policyTag": "flow_author", "resourceType": "flow" }, { "operations": [ "delete", "update", "read", "create", "share" ], "policyTag": "connection_author", "resourceType": "connection" } ] } ], |
For details on these parameters, please see the API Reference documentation:
D s api refdoclink |
---|
operation/listPeople |
People - GET
Endpoint | /v4/people/:id |
---|---|
Method | GET |
Change:
Same change as previous.
For details on these parameters, please see the API Reference documentation:
D s api refdoclink |
---|
operation/getPerson |
Create Salesforce connection changes
Beginning in Release 7.1, the structure of a Salesforce connection object has been changed. In prior releases, a Salesforce connection could not be shared without also sharing credentials. These changes allow a Salesforce connection to be shared without requiring the sharing of credentials.
Prior to Release 7.1, when creating a Salesforce connection using the APIs, the /v4/connections
POST method required the insertion of a Salesforce security token as part of the params
section of the request:
Code Block |
---|
{ "vendor": "salesforce", "vendorName": "salesforce", "type": "jdbc", "credentialType": "basic", "ssl": true, "name": "Salesforce", "description": "example_salesforce_connection", "disableTypeInference": true, "isGlobal": true, "credentialsShared": true, "host": "salesforce.example.com", "params": { "servername": "myServer", "securityToken": "string" }, "credentials": [ { "username": "string", "password": "string" } ] } |
Beginning in Release 7.1, the following changes have been applied to this request structure:
Change
credentialType
tosecurityToken
Move
securityToken
key fromparams
tocredentials
area of the API request- A host and port number value are not required.
Code Block |
---|
{ "vendor": "salesforce", "vendorName": "salesforce", "type": "jdbc", "credentialType": "securityToken", "ssl": true, "name": "Salesforce", "description": "example_salesforce_connection", "disableTypeInference": true, "isGlobal": true, "credentialsShared": true, "host": "salesforce.example.com", "params": { "servername": "myServer" }, "credentials": [ { "username": "string", "password": "string", "securityToken": "string" } ] } |
For more information:
API:
D s api refdoclink operation/createConnection
Apply overrides to file-based data sources when you run a job
You can now apply overrides to your file-based data sources when you run a job using the APIs. See API Workflow - Run Job.
Changes for Release 7.0
v3 endpoints are no longer available
The v3 version of the API endpoints are no longer available in the platform.
You must use the v4 versions. See API Reference.
Path to file-based imported datasets can be specified by URI
In Release 6.8 and earlier, you could create an imported dataset from a file-based source using a request like the following:
Code Block |
---|
{ "path": "/tri-hdfs/uploads/1/4aee9852-cf92-47a8-8c6a-9ff2adeb3b4a/POS-r02.txt", "type": "hdfs", "bucket": null, "name": "POS-r02b.txt", "description": "POS-r02 - copy" } |
Beginning in this release, you can specify the source using a single uri
value:
Code Block |
---|
{ "uri": "hdfs:///tri-hdfs/uploads/1/4aee9852-cf92-47a8-8c6a-9ff2adeb3b4a/POS-r02.txt", "name": "POS-r02b.txt", "description": "POS-r02 - copy" } |
Info |
---|
NOTE: The prior format using |
For more information, see
D s api refdoclink |
---|
operation/createImportedDataset |
awsRoles GET method removed
Prior to Release 7.0, the platform supported the ability to retrieve awsRole objects based on awsRoleId value.
In Release 7.0 and later, this endpoint has been disabled in the product.
Tip | ||
---|---|---|
Workaround: You can still acquire awsRole identifiers and objects via API. For more information, see
|
Connections endpoints simplified
Prior to Release 7.0, the request to create a connection looked liked the following:
Code Block |
---|
{ "connectParams": { "vendor": "sqlserver", "vendorName": "sqlserver", "host": "sqlserver.example.com", "port": "1433" }, "host": "sqlserver.example.com", "port": 1433, "vendor": "sqlserver", "params": { "connectStrOpts": "" }, "ssl": false, "vendorName": "sqlserver", "name": "sqlserver_test2", "description": "", "type": "jdbc", "isGlobal": false, "credentialType": "basic", "credentialsShared": true, "disableTypeInference": false, "credentials": [ { "username": "<username>", "password": "<password>" } ] } |
In the above, the connectParams
information is duplicated elsewhere in the request.
Beginning in Release 7.0, the connectParams
elements of the Connections endpoint have been removed.
The above request now looks like the following:
Code Block |
---|
{ "host": "sqlserver.example.com", "port": 1433, "vendor": "sqlserver", "params": { "connectStrOpts": "" }, "ssl": false, "vendorName": "sqlserver", "name": "sqlserver_test2", "description": "", "type": "jdbc", "isGlobal": false, "credentialType": "basic", "credentialsShared": true, "disableTypeInference": false, "credentials": [ { "username": "<username>", "password": "<password>" } ] } |
Required changes:
- The
params
andname
attributes are now required when creating a new connection. - If you have downstream scripts that utilize the
connectParams
objects, these must be updated to reference the corresponding attributes after upgrade to Release 7.0 or later.
For more information, see
D s api refdoclink |
---|
operation/getConnection |
Publishing for a jobGroup has been simplified
After you have run a job, you can publish the results from the job using the following endpoint:
Code Block |
---|
/v4/jobGroups/<id>/publish |
Prior to Release 7.0, the request to submit to the endpoint looked like the following:
Code Block |
---|
{ "connection": { "id": 1 }, "path": ["default"], "table": "test_table", "action": "create", "inputFormat": "pqt", "flowNodeId": 27 } |
However, the flowNodeId
value is not necessary, as it can be inferred from the id
value that is part of the endpoint URI.
Beginning in Release 7.0, the request no longer requires the flowNodeId
value:
Code Block |
---|
{ "connection": { "id": 1 }, "path": ["default"], "table": "test_table", "action": "create", "inputFormat": "pqt" } |
Info |
---|
NOTE: No changes are required to existing scripts. Values submitted for the |
For more information, see
D s api refdoclink |
---|
operation/publishJobGroup |
WASB URI format has changed
In Release 6.8 and earlier, the supported format for WASB URIs was the following:
Code Block |
---|
wasbs://blobhost.blob.core.windows.net/container/path/to/file.csv |
This format was not an Azure standard. Beginning in Release 7.0, the following standards-based format is supported:
Code Block |
---|
wasbs://container@blobhost.blob.core.windows.net/path/to/file.csv |
Info |
---|
NOTE: Any references in your API tooling to the previous WASB UI format must be updated to the new format. |
Saving Databricks Personal Access Token has different response
In Release 6.8 and earlier, when saving a Databricks Personal Access Token, the response looked like the following:
Response Code | 204 - No Content |
---|---|
Response Body | Empty. |
Beginning in Release 7.0, the response looks like the following:
Response Code | 200 - Ok | ||
---|---|---|---|
Response Body |
|
For more information, see
D s api refdoclink |
---|
operation/saveDatabricksAccessToken |
Documentation for awsRoles object corrected
In Release 6.8 and earlier, the following issues appeared in the documentation for the awsRoles object. These issues have been corrected in this release:
Inadequate permissions:
The POST and PUT methods for the awsRoles endpoint require either of the following user permissions:
- Workspace admin
D s item item admin
Optional workspaceid/personid attributes:
For the POST and PUT methods for the awsRoles endpoint, you can optionally assign the role to a workspace or person.
See
D s api refdoclink |
---|
operation/createAwsRole |
See
D s api refdoclink |
---|
operation/updateAwsRole |
Changes for Release 6.8.1
Improvements to AssetTransfer API documentation
Before disabling or deleting a user's account, an admin may wish to transfer the user's assets to the ownership of a different user. For more information on how to transfer all assets via API, see
D s api refdoclink |
---|
operation/transferAssets |
Changes for Release 6.8
Overrides to relational sources and targets
Through the APIs, you can now apply overrides to relational sources and targets during job execution or deployment import.
jobGroups
When you are running a job, you can override the default publication settings for the job using overrides
in the request. For more information, see API Workflow - Run Job.
Deployments
When you import a flow package into a deployment, you may need to remap the source and output of the flow to use production versions of your data. This capability has been present in the product for file-based sources and targets. Now, it's available for relational sources and targets. For more information, see Define Import Mapping Rules.
New flow object definition
Release 6.8 introduces a new version of the flow object definition. This new version will support cross-product and cross-version import and export of flows in the future. For more information see Changes to the Object Model.
Info |
---|
NOTE: The endpoints to use to manage flow packages remain unchanged. Similarly, the methods used to define import mapping rules remains unchanged. The API responses that contain flow definitions has changed. See below. |
Export and import macros
Beginning in Release 6.8, you can export and import macro definitions via API.
Export:
D s api refdoclink operation/getMacroPackage
Import:
D s api refdoclink operation/importMacroPackage
Changes for Release 6.4
Request format for assigning connection permissions endpoint has changed
For this endpoint:
Code Block |
---|
/v4/connections/<cid>/permissions/ |
where:
<cid>
is in the internal connection identifier.
The request payload format has changed.
Before Release 6.4, the request format was as follows:
Code Block |
---|
[ { "personId": 3, "role": "readOnly" } ] |
Beginning in Release 6.4, the request format is as follows:
Code Block |
---|
{ "data": [ { "person": { "id": 3 }, "role": "readOnly" } ] } |
Info |
---|
NOTE: The old request format is no longer supported. |
For more information, see
D s api refdoclink |
---|
operation/createConnectionPermission |
v4 version of password reset request endpoint
To assist in migration from the command-line interface to using the APIs, a v4 version of an API endpoint has been made available to allow for administrators to generate password reset codes.
Changes to awsConfig object
Info |
---|
NOTE: No action is required. |
In Release 6.0, the awsConfig object was introduced to enable the assignment of AWS configurations to individual users (per-user auth) via API. This version of the awsConfig object supported a mapping of a single IAM role to an awsConfig object.
Beginning in Release 6.4, per-user authentication now supports mapping of multiple possible IAM roles to an individual user's configuration. To enable this one-to-many mapping, the awsRoles object was introduced.
- An awsRoles object creates a one-to-one mapping between an IAM role and an awsConfig object.
- An awsConfig object can have multiple awsRoles assigned to it.
Changes to awsConfig object:
- The
role
field in the object has been replaced byactiveRoleId
, which maps to the active role for the configuration object. - For each
role
reference in the awsConfig objects, a corresponding awsRole object has been created and mapped to it.
Beginning in Release 6.4, you can create, edit, and delete awsRoles objects, which can be used to map an AWS IAM role ARN to a specified AWSConfig object. You can map multiple awsRoles to a single awsConfig.
For more information, see API Workflow - Manage AWS Configurations.
Changes for Release 6.3
Assign AWSConfigs to a user at create time
Beginning in Release 6.3, you can assign an AWSConfig object to a user when you create the object. This shortcut reduces the number of REST calls that you need to make.
Info |
---|
NOTE: For security reasons, AWSConfig objects must be assigned to users at the time of creation. Admin users can assign to other users. Non-admin users are automatically assigned the AWSConfig objects that they create. |
Prior to Release 6.3, AWSConfig objects were assigned through the following endpoint. Example:
Code Block |
---|
/v4/people/2/awsConfigs/6 |
Info |
---|
NOTE: This endpoint has been removed from the platform. Please update any scripts that reference the above endpoint to manage AWS configuration assignments through the new method described in the following link. |
See API Workflow - Manage AWS Configurations.
Changes for Release 6.0
Error in Release 6.0.x API docs
In Release 6.0 - Release 6.0.2, the online and PDF versions of the documentation referenced the following endpoint: API JobGroups Get Status v4. According to the docs, this endpoint was triggered in this manner:
Method | GET | ||
---|---|---|---|
Endpoint |
|
This endpoint exists in v3 of the API endpoints. It does not exist in v4.
Instead, you should monitor the status
field for the base GET endpoint for jobGroups. For more information, see
D s api refdoclink |
---|
operation/getJobGroup |
Planned End of Life of v3 API endpoints
In Release 6.0, the v3 API endpoints are supported.
Warning | ||||
---|---|---|---|---|
In the next release of |
You must migrate to using the v4 API endpoints before upgrading to the next release after Release 6.0.
Changes for Release 5.9
Introducing Access Tokens
Each request to the API endpoints of the
D s platform |
---|
- A request could include clear-text username/password combinations. This method is not secure.
- A request could include a browser cookie. This method does not work for well for use cases outside of the browser (e.g. scripts).
Beginning in Release 5.9, API users can manage authentication using access tokens. These tokens obscure any personally identifiable information and represent a standards-based method of secure authentication.
Info |
---|
NOTE: All previous methods of API authentication are supported in this release. Access tokens is the preferred method of authentication. |
The basic process works in the following manner:
- API user requests generation of a new token.
- This initial request must contain a valid username and password.
- Request includes expiration.
- Token value is returned in the response.
- The token value inserted into the Authorization header of each request to the platform.
- User monitors current time and expiration time of the token. At any time, the user can request a new token to be generated using the same endpoint used in the initial request.
Access tokens can be generated via API or the
D s webapp |
---|
Info |
---|
NOTE: This feature must be enabled in your instance of the platform. See Enable API Access Tokens. |
API: For more information, see
D s api refdoclink operation/createApiAccessToken
: For more information, see Access Tokens Page.D s webapp
For more information on API authentication methods, see
D s api refdoclink |
---|
section/Authentication |
Changes for Release 5.1
None.
Changes for Release 5.0
Introducing v4 APIs
D beta |
---|
Release 5.0 signals the introduction of version 4 of the REST APIs.
Info |
---|
NOTE: At this time, a very limited number of v4 REST APIs are publicly available. Where possible, you should continue to use the v3 endpoints. For more information, see API Reference. |
v4 conventions
The following conventions apply to v4 and later versions of the APIs:
Parameter lists are consistently enveloped in the following manner:
Code Block { "data": [ { ... } ] }
- Field names are in
camelCase
and are consistent with the resource name in the URL or with theembed
URL parameter. From early API versions, foreign keys have been replaced with identifiers like the following:
v3 and earlier v4 and later Code Block "createdBy": 1,
Code Block "creator": { "id": 1 },
Code Block "updatedBy": 2,
Code Block "updater": { "id": 2 },
Publication endpoint references database differently. This change is to make the publishing endpoint for relational targets more flexible in the future.
v3 and earlier v4 and later Code Block "database": "dbName",
Code Block "path": ["dbName"],
Changes for Release 4.2
Create Hive and Redshift connections via API
You can create connections of these types via API:
- Only one global Hive connection is still supported.
- You can create multiple Redshift connections.
WrangledDataset endpoints are still valid
In Release 4.1.1 and earlier, the WrangledDataset endpoints enabled creation, modification, and deletion of a wrangled dataset object, which also created the associated recipe object.
In Release 4.2, wrangled datasets have been removed from the application interface. However, the WrangledDataset API endpoints remain, although they now apply to the recipe directly.
The following endpoints and methods are still available:
Info |
---|
NOTE: In a future release, these endpoints may be migrated to recipe-based endpoints. API users should review this page for each release. |
For more information, see Changes to the Object Model.