Apache Unomi 2.x - Documentation

• 语言: 英语
• 格式: .pdf

摘要

文档详细介绍了Apache Unomi 2.x的功能和配置，包括会话处理、用户资料管理、数据模型、测试运行、SSH shell命令、规则管理、元数据、聚合以及Groovy动作的使用。文档还提供了如何通过命令行工具管理服务器生命周期、如何处理用户事件和配置文件等信息。Apache Unomi作为一个实时用户分析和个性化平台，支持数据收集、规则执行和扩展功能。

AI总结

### Apache Unomi 2.x Documentation Summary

#### 1. **Session and Profile Handling**
- **Session Management**:
- Unomi uses `sessionID` to load an existing session. If no session exists, a new one is created with the ID provided by the web server.
- If a session exists, the `profileID` is extracted. If not found, Unomi looks for a `context-profile-id` cookie.
- If no profile is found, a new profile ID (a UUID) is generated by Unomi.

- **Profile Creation**:
- If a valid `profileID` exists, the corresponding profile is loaded. If not, a new profile is created.

#### 2. **Context Request Flow**
- Unomi processes incoming requests to the `ContextServlet` by:
1. Checking for an existing session.
2. Extracting or creating a profile ID.
3. Processing events (if provided) against the profile.
4. Returning the updated profile in the response.

#### 3. **Data Model Overview**
- **Data Collection**:
- Unomi gathers user action data, which is processed and stored for personalization, behavior analysis, and segmentation.
- **Key Objects**:
- Includes user profiles, sessions, events, and rules.
- **Metadata**:
- Contains additional information about objects, such as IDs, names, descriptions, and tags.
- **Aggregations**:
- Used to collect and aggregate data based on specific conditions. Supported types include date, numeric range, and IP range aggregations.

#### 4. **Running Integration Tests**
- Integration tests require **JDK 11 or newer**.
- Tests can be run using the `integration-tests` profile.
- A default test page (`/cxs/context.js`) is provided to trigger context loading.

#### 5. **SSH Shell Commands**
- Connect to the Apache Karaf Shell using:

  ssh -p 8102 karaf@localhost

- Default credentials: `karaf/karaf` (change immediately).
- Commands are namespaced under `unomi:` and include:
- **Lifecycle Commands**: `migrate`, `start`, `stop`, `version`.
- **Runtime Commands**: `rule-list`, `rule-view`, `rule-remove`, `rule-reset-stats`, `rule-tail`.

#### 6. **Rule Management**
- Commands for managing rules:
- `rule-list`: Lists rules (max 100 entries, optional CSV output).
- `rule-view`: Dumps a single rule in JSON format.
- `rule-remove`: Deletes a rule (use with caution).
- `rule-reset-stats`: Resets rule statistics.
- `rule-tail`: Monitors executed rules in real-time.

#### 7. **Metadata**
- **Metadata Object**: Contains additional information about objects, including:
- `id`: Unique identifier (UUID).
- `name`: Display name.
- `description`: Optional description.
- `scope`: Administrative grouping identifier.
- `tags`: User-defined and system tags.

#### 8. **Connectors**
- **Salesforce CRM Connector**: Enables data synchronization between Unomi profiles and Salesforce Leads.
- **Contributors Needed**: Apache Unomi invites contributions for additional connectors.

#### 9. **Configuration File Example**
- Example configuration for Java and Karaf environments:

  JAVA_HOME="/usr/libexec/java_home -v 11.0.5"
  export JAVA_HOME
  java -version

#### 10. **Groovy Actions**
- Groovy actions allow developers to define dynamic actions via Groovy scripts.
- **Action Anatomy**:
- Uses `@Action` annotation to define action type and executor.
- Scripts are stored in Elasticsearch and executed when rules are triggered.
- Parameters are defined using annotations.

---

### Key Takeaways:
- **Session and Profile Management**: Unomi handles session creation and profile management using UUIDs for uniqueness and security.
- **Data Processing**: Collected user data is used for personalization, behavior analysis, and segmentation.
- **Integration**: Supports connectors like Salesforce and allows custom actions via Groovy scripts.
- **Testing and Monitoring**: Provides tools for running integration tests and managing rules in real-time.