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.

混合和多云简介Introduction to hybrid and multicloud

Microsoft Azure 提供了帮助你在云中构建和操作技术解决方案所需的所有产品和功能。Microsoft Azure provides all of the products and features required to help you build and operate your technology solutions in the cloud. 我们也理解,有一些必要的业务原因可能导致需要使用多个私有和/或公有云。We also understand that there are sound business reasons that may drive the necessity of using multiple private and/or public clouds. 作为混合和多云之旅的第一步,本文概述并扩展了 Microsoft 对于重要云计算术语的独特观点。As a first step in your hybrid and multicloud journey, this article outlines and expands on Microsoft's unique perspective on important cloud computing terms.

定义混合和多云Defining hybrid and multicloud

混合云是一种云计算,它结合了私有云(本地基础结构)与公有云(第三方提供商通过公共 Internet 提供的计算服务)。A hybrid cloud is a type of cloud computing that combines a private cloud (on-premises infrastructure), with a public cloud (computing services offered by third-party providers over the public internet). 混合云允许数据和应用程序在两个云环境之间一致地移动。Hybrid clouds allow data and applications to consistently move between the two cloud environments. 许多组织因为业务需求而选择混合云策略,例如满足法规和数据主权要求、最大限度地利用本地技术投资或解决延迟问题。Many organizations choose a hybrid cloud strategy because of business requirements, such as meeting regulatory and data sovereignty requirements, maximizing on-premises technology investments, or addressing latency issues.

混合云正在发展,以包括边缘工作负载。The hybrid cloud is evolving to include edge workloads. 云管理的边缘计算设备将公有云的计算能力引入私有云,更接近 IoT 设备所在的位置,包括驻留在应用程序、连网设备和移动使用者服务中的数据。Cloud-managed edge computing devices bring the computing power of the public cloud to the private cloud, closer to where the IoT devices reside, including data residing in applications, connected devices, and mobile consumer services. 通过将工作负载移动到边缘来降低延迟,设备将花费更少的时间与云进行通信,并且可以在更长的离线时间内可靠运行。Reducing latency by moving workloads to the edge, devices spend less time communicating with the cloud, and can operate reliably in extended offline periods. 扩展的计算、存储和服务可用性提供了更贴近客户的体验驱动资源。Expanded compute, storage, and service availability provides experience-driven resources closer to your customers.

多云计算是指在异类环境中使用来自多个云提供商(包括私有云和公有云)的多个云计算服务。Multicloud computing refers to the use of multiple cloud computing services from more than one cloud provider (including private and public clouds), in a heterogeneous environment. 多云策略提供了更大的灵活性,并降低了风险。A multicloud strategy provides greater flexibility, and mitigates risk. 从不同的云提供商中选择最适合特定任务的服务,或者利用特定位置的特定云提供商提供的服务。Choose services from different cloud providers that are best suited for a specific task, or take advantage of services offered by a particular cloud provider in a specific location.

混合和多云叙述Hybrid and multicloud narrative

此方案遵循通用的混合和多云叙述,提供有关在组织采用云的过程中可以采取哪些做法来获得成功的指导。This scenario follows a common hybrid and multicloud narrative, and provides guidance on what you can do differently to be successful during your organization's cloud adoption effort. 这一一般叙述并不局限于单个云采用方法,而是着眼于整个云采用之旅。This general narrative is not restricted to a single cloud adoption methodology, but takes a view of the entire cloud adoption journey.

混合云平台为你的组织提供了许多优势:更大的灵活性、控制能力和可伸缩性,具有更多部署选项、全球规模、集成的跨平台安全性、统一的符合性,在整个组织内提升工作负载、运营和成本效率,并通过现有基础结构一致地实现更高价值。A hybrid cloud platform gives your organization many advantages: greater flexibility, control, and scalability, with more deployment options, global scale, integrated cross-platform security, unified compliance, and improved workload, operational, and cost efficiencies across the enterprise, consistently achieving more value from existing infrastructure. 当计算和处理需求波动时,你可以通过混合云计算将本地基础结构无缝扩展到公有云,以处理任何溢出,而无需让第三方数据中心访问你的全部数据。When computing and processing demand fluctuates, hybrid cloud computing enables you to seamlessly scale up your on-premises infrastructure to the public cloud to handle any overflow, without giving third-party datacenters access to the entirety of your data. 通过在云中运行某些工作负载,你的组织看可以获得公有云提供的灵活性和创新,同时在自己的数据中心内保留高度敏感的数据,以满足客户需求或保持符合法规要求。By running certain workloads in the cloud, your organization gains the flexibility and innovation the public cloud provides, while retaining highly sensitive data in your own datacenter to meet client needs, or remain in compliance with regulatory requirements.

这允许你扩展计算资源,同时现代化和保护任务关键型应用程序和数据This allows you to scale computing resources, while modernizing and protecting mission-critical applications and data. 无需为满足短期需求高峰而进行大规模资本支出,也无需被迫释放本地资源以获取更敏感的数据。Eliminate the need to make massive capital expenditures to meet short-term spikes in demand, or being forced to free up local resources for more sensitive data. 使用云计费模型,你的组织将只为临时使用的资源付费,而无需购买、计划和维护可能长期闲置的额外资源和设备。With cloud billing models, your organization will only pay for resources you temporarily use, instead of having to purchase, program, and maintain additional resources and equipment that could remain idle for long periods.

另一项可以消除的资本支出是异地灾难恢复和备份基础结构投资。Another capital expenditure that could be eliminated is in offsite disaster recovery and backup infrastructure investments. 对于本地工作负载和因驻留在公有云中而在某方面不受限制的相关数据,适用于 BCDR 的公有云策略是非常有吸引力的选择。Public cloud for BCDR strategies is a compelling option for those on-premises workloads and associated data not restricted in some way from residing in a public cloud. 通过使用适用于 BCDR 的公有云,客户可以利用在隐私和安全方面的主要投资、按需扩展以及简便快捷的恢复。By using public cloud for BCDR customers take advantage of the major investments in privacy and security, scale on demand and ease and speed of recovery.

各公司正在将资源分散到本地、多个云和边缘。Companies are spreading resources across on-premises, multiple clouds, and the edge. 我们的客户经常提出四种常规需求:Customers have four common needs that we often hear about:

  1. 在单个窗格中可以看到所有现有和未来基础结构和应用程序的运行状况。Visibility into the health of all existing and future infrastructure and applications in a single pane of glass.
  2. 难以将本地策略和更新与云基础结构集成。Difficulty integrating on-premises policies and updates with cloud infrastructure. 组织理解实现治理标准的需要,Organizations understand the need for implementing a governance standard,
  3. 需要跨本地和云的一系列技能,因为组织中常有不同的应用程序开发团队。A wide range of skills across on-premises and cloud, because there are often different application development teams in the organization. 客户希望两者之间具有一致的互操作性,以便能够统一开发实践。Customers are looking for consistent interoperability between the two so they can unify development practices.
  4. 希望在不大量修改当前操作的情况下管理安全态势。Desire to manage security posture, without heavily modifying current operations. 云和多云加剧了这一挑战,这可能会降低信任、增加忧虑。Cloud and multicloud compounds this challenge, which can decrease trust and increase apprehension.

考虑混合和多云环境中的云本地服务部署。Consider the deployment of cloud-native services in a hybrid and multicloud environment. 云服务常被限定为简单地“将数据和应用程序移动到公有云”。Cloud services are often strictly contextualized as simply "moving data and applications to the public cloud". 混合策略完全支持不能使用公有云处理某些工作负载的客户操作,如政府基础结构、医疗保健和金融服务等高度管制的行业。A hybrid strategy fully supports customer operations that preclude the use of the public cloud for some workloads, such as highly regulated industries like government infrastructure, healthcare, and financial services. 根据地理位置和数据主权法规,可能需要将内部和客户数据保留在本地数据中心的边界内。Depending on geography and data sovereignty regulations, internal and customer data may be required to remain within the boundaries of on-premises datacenters. 数据延迟敏感度要求在靠近本地数据中心内源数据的位置进行计算,预计会出现 Internet 连接中断,或者会产生严重影响。Data latency sensitivity requires compute to be close to source data in on-premises datacenters, and internet connectivity disruptions are expected, or have critical implications. 在这些方案中,可以在本地数据中心部署可提供云服务、减少管理开销(在本地维护这些服务)和即用即付云计费模型的混合解决方案。In these scenarios, hybrid solutions that bring cloud services, decreased management overhead (maintaining these services on-premises), and a pay-as-you-go cloud billing model can be deployed in on-premises datacenters.

混合和多云动机Hybrid and multicloud motivations

作为真正的企业级云提供商,Azure 支持你跨公有、混合和多云环境的业务目标。As a true enterprise-grade cloud provider, Azure supports your business objectives across public, hybrid, and multicloud environments. 本系列文章将讨论有助于促进各种云混合的各种最佳做法,从 100% 的 Azure 环境到很少或没有 Azure 基础结构的环境。This series will discuss different best practices that can help facilitate various cloud mixes ranging from 100% Azure environments to environments that have little, or no, Azure infrastructure in place.

我们认识到,客户有许多正当理由选择在混合和多云环境中分布其数字资产。We recognize that there are many valid reasons for customers to choose to distribute their digital estate across hybrid and multicloud environments. 下面是一些常见的业务驱动因素:Here are some common business drivers:

  • 最大程度地减少或避免使用单个云提供商后难以改换其他提供商的情况Minimize or avoid single cloud provider lock-in
  • 业务部门、子公司或被收购公司已经采用了不同的云平台Business units, subsidiaries, or acquired companies have already adopted different cloud platforms
  • 不同的云提供商在不同的国家/地区可能有法规和数据主权要求Different cloud providers may have regulatory and data sovereignty requirements in different countries
  • 通过在两个云提供商之间复制工作负载,来提高业务连续性和灾难恢复能力Improve business continuity and disaster recovery by duplicating workloads across two cloud providers
  • 通过在用户位置附近运行应用程序(可能需要采用混合或多云)来最大限度地提高性能Maximize performance by running applications close to user locations, which may require hybrid or multicloud adoption
  • 通过采用多云策略,为某些数据平台或特定于行业的应用程序实现轻松迁移Enable easy migration for some data platforms or industry-specific applications by adopting multicloud strategies

混合和多云考虑因素Hybrid and multicloud concerns

通过可靠的混合和多云采用策略,上面列出的一些动机可以推动业务转换。Some of the motivations listed above can become business transformations with a sound hybrid and multicloud adoption strategy.

另一些则需要在部署前和部署后完成大量工作才能实现这些创新优势。Others require significant effort predeployment and post-deployment efforts to realize those innovative benefits. 例如,可以避免难以改换云提供商的情况。Cloud provider lock-in, for instance, is possible. 但若要避免难以改换提供商的情况,就要求组织限制其在云采用方面的愿景。But to avoid lock-in, organizations are required to limit their vision for cloud adoption. 一家云提供商的许多有益的产品和功能不能移植到其他云提供商。Many of the most beneficial products and features in a cloud provider are not portable to other cloud providers. 若要实现可移植性和并尽量减少难以改换供应商的情况,组织通常需要将云采用限制于基础结构即服务 (IaaS) 功能,或者在使用容器或 Kubernetes 等云本机技术方面进行大量投资。To achieve portability and minimize lock-in, organizations are often required to limit cloud adoption to basic infrastructure as a service (IaaS) capabilities, or invest heavily in the use of cloud-native technologies like containers or Kubernetes.

在发布工作负载并将其投入生产后,与混合和多云采用相关的另一个常见问题出现了:组织试图在新环境中为工作负载提供操作管理支持时,通常必须迅速重新考虑现有做法。After workloads are released and are in production, another common concern associated with hybrid and multicloud adoption surfaces: when organizations attempt to provide operations management support to workloads in new environments, they often have to quickly rethink their practices. 现有的操作管理平台(包括现有的操作管理策略和流程)不是为这些类型的环境构建的。Existing operations management platforms (including existing operations management policies and processes), were not built for these types of environments. 为了解释云环境中的偏差,公司通常会使用不同的操作工具和操作实践,这使得运营成本随支持的云环境数量成倍数增长。To account for deviations in cloud environments, companies often end up with disparate operations tooling and operations practices, which multiplies the cost of operations by the number of cloud environments supported.

下一步:通过统一操作最大程度地减少混合和多云问题Next step: Minimize hybrid and multicloud concerns with unified operations

在开始混合和多云之旅之前,请先了解统一运营的概念:使用一个公共控制平面跨所有云环境的一致操作方法,有助于解决有关混合和多云策略的许多问题。Understand the concept of unified operations before starting your hybrid and multicloud journey; consistent operations practices across all of your cloud environments with a common control plane can help to address many concerns regarding hybrid and multicloud strategies.

在继续大规模采用混合和多云之前,请确定你是否需要为每个云提供商重复操作,或对云管理实施统一操作方法Determine whether you need to duplicate operations for each cloud provider or implement a unified operations approach to cloud management before proceeding with hybrid and multicloud adoption at scale.