This topic describes the basic concepts of Intelligent Media Management, such as projects and data sources.
This document is no longer maintained. We recommend that you use the new version of Intelligent Media Management.
For a comparison between the new and earlier versions of Intelligent Media Management, see Usage guide for new and earlier versions.
For information about the basic concepts of the new version of Intelligent Media Management, see Basic concepts.
Projects
A project is the basic unit for using the Intelligent Media Management service. For example, when you develop a network disk app, you can create a project in the same region as the Object Storage Service (OSS) bucket where your data is stored. You must then grant the project permission to access the OSS bucket. This lets you use Intelligent Media Management features, such as document preview and image AI.
Intelligent Media Management includes the following project types:
Document Medium: supports format transformation and document preview features.
Image Medium: supports image AI features, such as content detection, QR code detection, and face detection.
Data sources
A data source provides data for Intelligent Media Management. Currently, only OSS is supported as a data source.
A data source can be an entire bucket or a bucket prefix. To control access to the data source, you can configure a role policy in the Resource Access Management (RAM) console. Then, you must specify the role when you create a project in Intelligent Media Management.
Request types
Sync request
After a client sends a request, the client is blocked until a result is returned. The result is usually returned within several hundred milliseconds. Examples include face detection and content detection in image instances.
Asynchronous request
Intelligent Media Management uses an asynchronous request mechanism to prevent clients from being blocked for long periods by long-running sync tasks. For example, a client connection may be disconnected if a task runs for more than 10 seconds.
After a client sends a request, the system quickly returns an ID. You can use this ID to query the status of the asynchronous request. For example, a format transformation request for a large file may take tens of seconds or even minutes to complete.
Tasks
A task is the execution entity of a single asynchronous data processing request. The ID that is returned when you call an asynchronous request API is the task ID. For example, a task is created to execute the asynchronous transformation of a single document.
After a task is created, you can use the Get operation to query its status, the Delete operation to delete the task, and the List operation to traverse task requests.