您现在访问的是微软AZURE全球版技术文档网站,若需要访问由世纪互联运营的MICROSOFT AZURE中国区技术文档网站,请访问 https://docs.azure.cn.

数据创新Data innovations

许多公司想要将其现有数据仓库迁移到云。Many companies want to migrate their existing data warehouse to the cloud. 它们有很多因素,包括:They are motivated by a number of factors, including:

  • 无硬件可用于购买或维护成本。No hardware to buy or maintenance costs.
  • 没有要管理的基础结构。No infrastructure to manage.
  • 能够切换到安全、可缩放且低成本的云解决方案。The ability to switch to a secure, scalable, and low-cost cloud solution.

例如,Azure 中的云本机即用即付服务(称为 Azure Synapse Analytics)为组织提供分析数据库管理系统。For example, the cloud-native, pay-as-you-go service from Azure called Azure Synapse Analytics provides an analytical database management system for organizations. Azure 技术在迁移后可帮助实现数据仓库的现代化,并扩展分析功能以促进新的商业价值。Azure technologies help modernize your data warehouse after it's migrated and extend your analytical capabilities to drive new business value.

数据仓库迁移项目涉及多个组件。A data warehouse migration project involves many components. 其中包括架构、数据、提取-转换-加载 (ETL) 管道、授权特权、用户、BI 工具语义访问层和分析应用程序。These include schema, data, extract-transform-load (ETL) pipelines, authorization privileges, users, BI tool semantic access layers, and analytic applications.

将数据仓库迁移到 Azure Synapse Analytics 后,可以利用 Microsoft 分析生态系统中的其他技术。After your data warehouse has been migrated to Azure Synapse Analytics, you can take advantage of other technologies in the Microsoft analytical ecosystem. 这样一来,不仅可以实现数据仓库的现代化,还可以在 Azure 上的其他分析数据存储中生成见解。Doing so allows you to not only modernize your data warehouse but also bring together insights produced in other analytical data stores on Azure.

可以扩大 ETL 处理,将任何类型的数据引入 Azure Data Lake Storage。You can broaden ETL processing to ingest data of any type into Azure Data Lake Storage. 可以使用 Azure 数据工厂,按比例准备并集成它。You can prepare and integrate it at scale by using Azure Data Factory. 这将生成可供数据仓库使用、且还可供数据科学家和其他应用程序使用的可信、常用的数据资产。This produces trusted, commonly understood data assets that can be consumed by your data warehouse, and also accessed by data scientists and other applications. 您可以生成实时的、面向批处理的分析管道。You can build real-time, batch-oriented analytical pipelines. 你还可以创建机器学习模型,这些模型可部署到实时在流式传输数据和按需运行的批处理中运行。You can also create machine learning models that can deploy to run in batch, in real time on streaming data, and on demand.

此外,还可以使用 PolyBase 超出数据仓库。In addition, you can use PolyBase to go beyond your data warehouse. 这简化了对在 Azure 上的多个基础分析平台中生成的见解的访问。This simplifies access to insights being produced in multiple underlying analytical platforms on Azure. 在逻辑数据仓库中创建全面的集成视图,以从 BI 工具和应用程序获取对流式处理、大数据和传统数据仓库见解的访问。You create holistic, integrated views in a logical data warehouse to gain access to streaming, big data, and traditional data warehouse insights from BI tools and applications.

许多公司的数据仓库多年来都在数据仓库中运行,以使用户能够生成商业智能。Many companies have had data warehouses running in their datacenters for years, to enable users to produce business intelligence. 数据仓库从已知事务系统中提取数据、暂存数据,然后清除、转换和集成数据仓库以填充数据仓库。Data warehouses extract data from known transaction systems, stage the data, and then clean, transform, and integrate it to populate data warehouses.

用例、业务案例和技术进步都支持 Azure Synapse 分析如何帮助你进行数据仓库迁移。Use cases, business cases, and technology advances all support how Azure Synapse Analytics can help you with data warehouse migration. 以下部分列出了其中的许多示例。The following sections list many of these examples.

用例Use cases

  • 连接的产品创新Connected product innovation
  • 未来工厂Factory of the future
  • 临床分析Clinical analytics
  • 相容性分析Compliance analytics
  • 基于开销的分析Cost-based analytics
  • 对全渠道优化Omnichannel optimization
  • 个性化设置Personalization
  • 智能供应链Intelligent supply chain
  • 动态定价Dynamic pricing
  • 采购分析Procurement analytics
  • 数字控制塔Digital control tower
  • 风险管理Risk management
  • 客户分析Customer analytics
  • 欺诈检测Fraud detection
  • 声明分析Claims analytics

业务案例Business cases

  • 使用单个分析服务构建端到端分析解决方案。Build end-to-end analytics solutions with a single analytics service.
  • 使用 Azure Synapse Analytics studio,它为数据准备、数据管理、数据仓库、大数据和 AI 任务提供统一的工作区。Use the Azure Synapse Analytics studio, which provides a unified workspace for data prep, data management, data warehousing, big data, and AI tasks.
  • 使用无代码的可视化环境构建和管理管道,自动执行查询优化,构建概念证明,并使用 Power BI,所有这些都来自同一分析服务。Build and manage pipeline with a no-code visual environment, automate query optimization, build proofs of concept, and use Power BI, all from the same analytics service.
  • 将数据见解传递到数据仓库和大数据分析系统。Deliver your data insights to data warehouses and big data analytics systems.
  • 对于任务关键型工作负荷,通过智能工作负荷管理、工作负荷隔离和无限并发来优化所有查询的性能。For mission-critical workloads, optimize the performance of all queries with intelligent workload management, workload isolation, and limitless concurrency.
  • 直接从 Azure Synapse Analytics 编辑并构建 Power BI 的仪表板。Edit and build Power BI dashboards directly from Azure Synapse Analytics.
  • 缩短 BI 和机器学习项目的项目开发时间。Reduce project development time for BI and machine learning projects.
  • 仅需几次单击,即可使用 Azure Synapse Analytics 中的 Azure 数据共享集成轻松共享数据。Easily share data with just a few clicks by using Azure Data Share integration within Azure Synapse Analytics.
  • 使用列级安全性和本机行级安全性实现精细的访问控制。Implement fine-grained access control with column-level security and native row-level security.
  • 通过动态数据掩码实时自动保护敏感数据。Automatically protect sensitive data in real time with dynamic data masking.
  • 具有内置安全功能(例如自动威胁检测和始终启用的数据加密)的行业领先的安全性。Industry-leading security with built-in security features like automated threat detection and always-on data encryption.

技术进步Technology advances

  • 无需购买硬件或维护费用,因此只需为所使用的内容付费。No hardware to buy or maintenance costs so you pay only for what you use.
  • 不需要管理基础结构,因此你可以专注于竞争性见解。No infrastructure to manage, so you can focus on competitive insights.
  • 在需要时,可以通过动态的可伸缩性进行大规模并行的 SQL 查询处理,并可以在不需要时关闭或暂停。Massively parallel SQL query processing with dynamic scalability when you need it, and the option to shut down or pause when you don't.
  • 能够从计算中独立缩放存储。Ability to independently scale storage from compute.
  • 你可以避免不必要的、昂贵的升级,因为数据仓库上的临时区域太大、占用了存储容量,并强制进行升级。You can avoid unnecessary, expensive upgrades caused by the staging areas on your data warehouse getting too big, taking up storage capacity, and forcing an upgrade. 例如,将暂存区域移动到 Azure Data Lake Storage。For example, move the staging area to Azure Data Lake Storage. 然后使用类似 Azure 数据工厂的 ETL 工具或在 Azure 上运行的现有 ETL 工具以较低的成本对其进行处理。Then process it with an ETL tool like Azure Data Factory or your existing ETL tool running on Azure at lower cost.
  • 通过使用 Azure Data Lake Storage 和 Azure 数据工厂在 Azure 中处理 ETL 工作负荷,以避免昂贵的硬件升级。Avoid expensive hardware upgrades by processing ETL workloads in Azure, by using Azure Data Lake Storage and Azure Data Factory. 这通常是比在现有数据仓库 DBMS 上运行的更好的解决方案。This is often a better solution than running on your existing data warehouse DBMS with SQL query processing doing the work. 随着暂存数据量的增加,更多的存储和计算能力将本地数据仓库用于 ETL。As staging data volumes increase, more storage and compute power underpinning your on-premises data warehouse is consumed by ETL. 这反过来会影响查询、报告和分析工作负荷的性能。This in turn affects the performance of query, reporting, and analysis workloads.
  • 避免生成在本地硬件上使用存储和数据库软件许可证的成本高昂的数据市场。Avoid building expensive data marts that use storage and databases software licenses on on-premises hardware. 可以改为在 Azure Synapse Analytics 中生成它们。You can build them in Azure Synapse Analytics instead. 如果数据仓库是数据保管库设计,这种设计会非常有用,这通常会导致更高的数据市场需求。This is especially helpful if your data warehouse is a data vault design, which often causes an increased demand for data marts.
  • 避免在本地硬件上分析和存储高速、高容量数据的成本。Avoid the cost of analyzing and storing high-velocity, high-volume data on on-premises hardware. 例如,如果需要分析实时、计算机生成的数据(如数据仓库中的单击流和流式处理 IoT 数据),可以使用 Azure Synapse 分析。For example, if you need to analyze real-time, machine generated data like click-stream and streaming IoT data in your data warehouse, you can use Azure Synapse Analytics.
  • 随着数据仓库的增长,你可以在数据中心的昂贵仓库硬件上存储数据,避免支付高级版。You can avoid paying a premium for storing data on expensive warehouse hardware in the datacenter as your data warehouse grows. Azure Synapse Analytics 可以更低的成本将数据存储在云存储中。Azure Synapse Analytics can store your data in cloud storage at a lower cost.

后续步骤Next steps

数据 democratizationData democratization