azure-cosmos-java

作者: microsoft

Azure Cosmos DB Java SDK。支持全局分布、多模型和响应式模式的NoSQL数据库操作。 触发词:“CosmosClient java”、“CosmosAsyncClient”、“cosmos database java”、“cosmosdb java”、“document database java”。

npx skills add https://github.com/microsoft/skills --skill azure-cosmos-java

Azure Cosmos DB SDK for Java

Client library for Azure Cosmos DB NoSQL API with global distribution and reactive patterns.

Installation

<dependency>
    <groupId>com.azure</groupId>
    <artifactId>azure-cosmos</artifactId>
    <version>LATEST</version>
</dependency>

Or use Azure SDK BOM:

<dependencyManagement>
    <dependencies>
        <dependency>
            <groupId>com.azure</groupId>
            <artifactId>azure-sdk-bom</artifactId>
            <version>{bom_version}</version>
            <type>pom</type>
            <scope>import</scope>
        </dependency>
    </dependencies>
</dependencyManagement>

<dependencies>
    <dependency>
        <groupId>com.azure</groupId>
        <artifactId>azure-cosmos</artifactId>
    </dependency>
</dependencies>

Environment Variables

COSMOS_ENDPOINT=https://<account>.documents.azure.com:443/
COSMOS_KEY=<your-primary-key>

Authentication

Key-based Authentication

import com.azure.cosmos.CosmosClient;
import com.azure.cosmos.CosmosClientBuilder;

CosmosClient client = new CosmosClientBuilder()
    .endpoint(System.getenv("COSMOS_ENDPOINT"))
    .key(System.getenv("COSMOS_KEY"))
    .buildClient();

Async Client

import com.azure.cosmos.CosmosAsyncClient;

CosmosAsyncClient asyncClient = new CosmosClientBuilder()
    .endpoint(serviceEndpoint)
    .key(key)
    .buildAsyncClient();

With Customizations

import com.azure.cosmos.ConsistencyLevel;
import java.util.Arrays;

CosmosClient client = new CosmosClientBuilder()
    .endpoint(serviceEndpoint)
    .key(key)
    .directMode(directConnectionConfig, gatewayConnectionConfig)
    .consistencyLevel(ConsistencyLevel.SESSION)
    .connectionSharingAcrossClientsEnabled(true)
    .contentResponseOnWriteEnabled(true)
    .userAgentSuffix("my-application")
    .preferredRegions(Arrays.asList("West US", "East US"))
    .buildClient();

Client Hierarchy

ClassPurpose
CosmosClient / CosmosAsyncClientAccount-level operations
CosmosDatabase / CosmosAsyncDatabaseDatabase operations
CosmosContainer / CosmosAsyncContainerContainer/item operations

Core Workflow

Create Database

// Sync
client.createDatabaseIfNotExists("myDatabase")
    .map(response -> client.getDatabase(response.getProperties().getId()));

// Async with chaining
asyncClient.createDatabaseIfNotExists("myDatabase")
    .map(response -> asyncClient.getDatabase(response.getProperties().getId()))
    .subscribe(database -> System.out.println("Created: " + database.getId()));

Create Container

asyncClient.createDatabaseIfNotExists("myDatabase")
    .flatMap(dbResponse -> {
        String databaseId = dbResponse.getProperties().getId();
        return asyncClient.getDatabase(databaseId)
            .createContainerIfNotExists("myContainer", "/partitionKey")
            .map(containerResponse -> asyncClient.getDatabase(databaseId)
                .getContainer(containerResponse.getProperties().getId()));
    })
    .subscribe(container -> System.out.println("Container: " + container.getId()));

CRUD Operations

import com.azure.cosmos.models.PartitionKey;

CosmosAsyncContainer container = asyncClient
    .getDatabase("myDatabase")
    .getContainer("myContainer");

// Create
container.createItem(new User("1", "John Doe", "john@example.com"))
    .flatMap(response -> {
        System.out.println("Created: " + response.getItem());
        // Read
        return container.readItem(
            response.getItem().getId(),
            new PartitionKey(response.getItem().getId()),
            User.class);
    })
    .flatMap(response -> {
        System.out.println("Read: " + response.getItem());
        // Update
        User user = response.getItem();
        user.setEmail("john.doe@example.com");
        return container.replaceItem(
            user,
            user.getId(),
            new PartitionKey(user.getId()));
    })
    .flatMap(response -> {
        // Delete
        return container.deleteItem(
            response.getItem().getId(),
            new PartitionKey(response.getItem().getId()));
    })
    .block();

Query Documents

import com.azure.cosmos.models.CosmosQueryRequestOptions;
import com.azure.cosmos.util.CosmosPagedIterable;

CosmosContainer container = client.getDatabase("myDatabase").getContainer("myContainer");

String query = "SELECT * FROM c WHERE c.status = @status";
CosmosQueryRequestOptions options = new CosmosQueryRequestOptions();

CosmosPagedIterable<User> results = container.queryItems(
    query,
    options,
    User.class
);

results.forEach(user -> System.out.println("User: " + user.getName()));

Key Concepts

Partition Keys

Choose a partition key with:

  • High cardinality (many distinct values)
  • Even distribution of data and requests
  • Frequently used in queries

Consistency Levels

LevelGuarantee
StrongLinearizability
Bounded StalenessConsistent prefix with bounded lag
SessionConsistent prefix within session
Consistent PrefixReads never see out-of-order writes
EventualNo ordering guarantee

Request Units (RUs)

All operations consume RUs. Check response headers:

CosmosItemResponse<User> response = container.createItem(user);
System.out.println("RU charge: " + response.getRequestCharge());

Best Practices

  1. Reuse CosmosClient — Create once, reuse throughout application
  2. Use async client for high-throughput scenarios
  3. Choose partition key carefully — Affects performance and scalability
  4. Enable content response on write for immediate access to created items
  5. Configure preferred regions for geo-distributed applications
  6. Handle 429 errors with retry policies (built-in by default)
  7. Use direct mode for lowest latency in production

Error Handling

import com.azure.cosmos.CosmosException;

try {
    container.createItem(item);
} catch (CosmosException e) {
    System.err.println("Status: " + e.getStatusCode());
    System.err.println("Message: " + e.getMessage());
    System.err.println("Request charge: " + e.getRequestCharge());
    
    if (e.getStatusCode() == 409) {
        System.err.println("Item already exists");
    } else if (e.getStatusCode() == 429) {
        System.err.println("Rate limited, retry after: " + e.getRetryAfterDuration());
    }
}

Reference Links

ResourceURL
Maven Packagehttps://central.sonatype.com/artifact/com.azure/azure-cosmos
API Documentationhttps://azuresdkdocs.z19.web.core.windows.net/java/azure-cosmos/latest/index.html
Product Docshttps://learn.microsoft.com/azure/cosmos-db/
Sampleshttps://github.com/Azure-Samples/azure-cosmos-java-sql-api-samples
Performance Guidehttps://learn.microsoft.com/azure/cosmos-db/performance-tips-java-sdk-v4-sql
Troubleshootinghttps://learn.microsoft.com/azure/cosmos-db/troubleshoot-java-sdk-v4-sql

来自 microsoft 的更多技能

oss-growth
microsoft
OSS增长黑客角色
agent-framework-azure-ai-py
microsoft
使用Microsoft Agent Framework Python SDK(agent-framework-azure-ai)构建Azure AI Foundry代理。在创建使用AzureAIAgentsProvider的持久化代理、使用托管工具(代码解释器、文件搜索、网络搜索)、集成MCP服务器、管理对话线程或实现流式响应时使用。涵盖函数工具、结构化输出和多工具代理。
development
airunway-aks-setup
microsoft
在AKS上设置AI Runway——从裸集群到运行模型。涵盖集群验证、控制器安装、GPU评估、提供商设置和首次部署。适用场景:“设置AI Runway”、“接入AKS集群”、“安装AI Runway”、“airunway设置”、“将模型部署到AKS”、“在AKS上进行GPU推理”、“在AKS上配置KAITO”、“在AKS上运行LLM”、“在AKS上使用vLLM”、“在AKS上设置模型服务”、“AI Runway控制器”。
devops
appinsights-instrumentation
microsoft
使用Azure Application Insights对Web应用进行插桩的指南。提供遥测模式、SDK设置和配置参考。适用场景:如何对应用进行插桩、App Insights SDK、遥测模式、什么是App Insights、Application Insights指南、插桩示例、APM最佳实践。
devops
applicationinsights-web-ts
microsoft
使用Application Insights JavaScript SDK(@microsoft/applicationinsights-web)为浏览器/Web应用添加检测。用于真实用户监控(RUM)——页面视图、点击、AJAX/fetch依赖项、异常、自定义事件,以及与后端OpenTelemetry追踪关联的浏览器端GenAI代理追踪。涵盖SDK加载器脚本和npm设置、框架扩展(React、React Native、Angular)、点击分析、遥测初始化器,以及从浏览器发出的代理/工具/模型跨度所遵循的OTel GenAI语义约定。
devops
azure-ai-anomalydetector-java
microsoft
使用适用于 Java 的 Azure AI 异常检测器 SDK 构建异常检测应用程序。在实现单变量/多变量异常检测、时间序列分析或 AI 驱动的监控时使用。
development
azure-ai-language-conversations-py
microsoft
使用azure-ai-language-conversations Python SDK实现对话语言理解(CLU)。当使用ConversationAnalysisClient分析对话意图和实体、构建NLP功能或将语言理解集成到应用程序中时使用。
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python。用于机器学习工作区、作业、模型、数据集、计算资源和管道。 触发词:“azure-ai-ml”、“MLClient”、“工作区”、“模型注册表”、“训练作业”、“数据集”。
development