Azure Cosmos DB 的总拥有成本 (TCO)Total Cost of Ownership (TCO) with Azure Cosmos DB

Azure Cosmos DB 中设计了精细的多租户和资源调控。Azure Cosmos DB is designed with the fine grained multi-tenancy and resource governance. 这种设计可让 Azure Cosmos DB 以明显降低的成本运行,并帮助用户节省成本。This design allows Azure Cosmos DB to operate at significantly lower cost and help users save. 目前,Azure Cosmos DB 支持在一台计算机上运行 280 多个客户工作负荷(密度还是不断提升),以及在一个群集中运行数千个客户工作负荷。Currently Azure Cosmos DB supports more than 280 customer workloads on a single machine with the density continuously increasing, and thousands of customer workloads within a cluster. 它对群集中不同计算机上的客户工作负荷以及数据中心内多个群集中的客户工作负荷进行负载均衡。有关详细信息,请参阅 Azure Cosmos DB:扩大多区域分布式数据库的边界It load balances replicas of customers workloads across different machines in a cluster and across multiple clusters within a data center To learn more, see Azure Cosmos DB: Pushing the frontier of multiple-regionally distributed databases. 由于实施资源调控、多租户并与 Azure 基础结构的剩余部分原生集成,与 MongoDB、Cassandra 或者在 IaaS 上运行的其他 OSS NoSQL 相比,Azure Cosmos DB 的费用平均要便宜 4 到 6 倍;与本地运行的数据库引擎相比,则最多便宜 10 倍。Because of resource-governance, multi-tenancy, and native integration with the rest of Azure infrastructure, Azure Cosmos DB is on average 4 to 6 times cheaper than MongoDB, Cassandra, or other OSS NoSQL running on IaaS and up to 10 times cheaper than the database engines running on premises. 请参阅有关 NoSQL 数据库云服务的总(非)拥有成本的白皮书。See the paper on The total cost of (non) ownership of a NoSQL database cloud service.

Apache Cassandra、MongoDB、HBase、引擎等 OSS NoSQL 数据库解决方案是为本地设计的。The OSS NoSQL database solutions, such as Apache Cassandra, MongoDB, HBase, engines were designed for on-premises. 作为托管服务提供时,它们相当于一个资源管理器模板,其中包含用于管理预配群集和监视支持的租户数据库。When offered as a managed service they are equivalent to a Resource Manager template with a tenant database for managing the provisioned clusters and monitoring support. OSS NoSQL 体系结构需要很大的运营开销,了解其专业知识可能比较困难且费用不菲。OSS NoSQL architectures require significant operational overhead, and the expertise can be difficult and expensive to find. 另一方面,Azure Cosmos DB 是完全托管式的云服务,可让开发人员专注于业务创新,而不是管理和维护数据库基础结构。On the other hand, Azure Cosmos DB is a fully managed cloud service, which allows developers to focus on business innovation rather than on managing and maintaining database infrastructure.

与云原生的数据库服务 Azure Cosmos DB 不同,设计和构建的 OSS NoSQL 数据库引擎不会根据基本的体系结构原则包含资源调控或精细的多租户功能。Unlike a cloud-native database service Azure Cosmos DB, OSS NoSQL database engines were not designed and built with the resource governance or fine-grained multi-tenancy as the fundamental architectural principles. Cassandra 和 MongoDB 等 OSS NoSQL 数据库引擎做出这种基本假设:运行这些引擎的虚拟机的所有资源可供这些引擎使用。OSS NoSQL database engines like Cassandra and MongoDB make a fundamental assumption that all the resources of the virtual machine on which they are running are available for their use. 如果资源量降到特定的阈值以下,则其中的许多数据库引擎将无法正常运行。Many of these database engines cannot function if the amount of resources drops below a certain threshold. 例如,小型 VM 实例可用的资源根据供应商建议的配置而定,这意味着,大型 VM 的成本往往更高。For example, for small VM instances, and they are available with vendor-recommended configurations suggesting typically large-scale VMs with higher cost. 因此,无法托管 OSS NoSQL 或其他任何本地数据库引擎,并使用基于消耗量的计费模型(例如每秒请求数,或消耗的存储)来提供资源。So, it is not possible to host an OSS NoSQL or any other on-premises database engine and make it available by using a consumption-based charging model like requests per second or consumed storage.

Azure Cosmos DB 的总拥有成本Total cost of ownership of Azure Cosmos DB

Azure Cosmos DB 的无服务器预配模型消除了过度预配数据库基础结构的需求。The serverless provisioning model of Azure Cosmos DB eliminates the need to over-provision the database infrastructure. 无需专业化的配置或许可即可提供 Azure Cosmos DB 资源。Azure Cosmos DB resources are provided without any need for specialized configurations or licensing. 因此,与 OSS NoSQL 数据库相比,Azure Cosmos DB 支持的应用程序的总拥有成本最多可以节省 70%。As a result, the Azure Cosmos DB-backed applications can run with as much as a 70 percent Total cost of ownership savings when compared to OSS NoSQL databases. 有关部分实时示例,请参阅客户用例For some real-time examples, see customer use-cases. Azure Cosmos DB 定价模型的其他优势包括:Other benefits of the Azure Cosmos DB pricing model include:

  • 价格体现高价值: 市场分析师、客户和合作伙伴已经确认,与自行实施或通过其他供应商实施这些解决方案相比,Azure Cosmos DB 提供的所有功能的价值更高,且价格要低得多。Great value for the price: Market analysts, customers, and partners have confirmed a greater value of all the features that Azure Cosmos DB offers for a much lower price compared to what customers can get when implementing these solutions on their own or through other vendors. 使用 Azure Cosmos DB 可以大幅简化多区域分布、多主数据库、妥善定义的直观一致性模型和自动索引等数据库功能,且消除任何复杂性、开销或停机时间。The database features such multiple-region distribution, multi-master, well-defined and intuitive consistency models, automatic indexing are greatly simplified with Azure Cosmos DB without any complexity, overhead, or downtime.

  • 无需进行 NoSQL DevOps 管理: 使用 Azure Cosmos DB 时无需采用 DevOps 来管理部署以及执行维护、缩放或修补。No NoSQL DevOps administration is required: With Azure Cosmos DB one does not need to employ DevOps to manage deployments, perform maintenance, scale, or patch. 可以像对本地或云基础结构中托管的 OSS NoSQL 群集所做的那样来执行所有工作负荷。You can execute the all the workloads that you would do with OSS NoSQL cluster hosted on-premises or on cloud infrastructure.

    Azure Cosmos DB 拥有成本

  • 可以弹性缩放: 可以扩展和缩减 Azure Cosmos DB 吞吐量,从而在非高峰期降低拥有成本。Ability to elastically scale: Azure Cosmos DB throughput can be scaled up and down, allowing you to reduce the cost of ownership during non-peak times. 云基础结构中部署的 OSS NoSQL 群集提供有限的弹性,而本地部署在定义上不提供弹性。OSS NoSQL clusters deployed on cloud infrastructure offer limited elasticity, and on-premises deployments aren't elastic by definition. 在 Azure Cosmos DB 中,如果预配更高的吞吐量,则可以保证吞吐量呈线性扩展。In Azure Cosmos DB, if you provision more throughput, your throughput is guaranteed to scale linearly. 这一保证附带 SLA 的经济保障,在任何部署规模都能在 99% 的时间提供这种保证。This guarantee is backed up by financial SLAs and at the 99th percentile at any scale.

  • 规模效益: Azure Cosmos DB 等托管服务使用大量的节点来运行,原生与网络、存储和计算等资源集成。Economies of scale: A managed service like Azure Cosmos DB operates with a large number of nodes, integrated natively with networking, storage, and computes. 由于 Azure Cosmos DB 的规模较大,通过标准化可以节省成本。Because of Azure Cosmos DB's large scale, standardization you can save costs.

  • 针对云进行优化: Azure Cosmos DB 是从无到有设计出来的,提供精细的多租户控制和性能隔离。Optimized for the cloud: Azure Cosmos DB is designed from the ground-up with fine-grained multi-tenancy and performance isolation. 这样,便可以跨群集与数据中心以最佳方式放置、执行和均衡数千个租户及其工作负荷。This allows for optimally placing, executing, and balancing thousands of tenants and their workloads across clusters and data centers. 相比之下,最新一代的 OSS NoSQL 数据库使用假设运行单租户的工作负荷的虚拟机在本地工作。In contrast, the current generation of OSS NoSQL databases operate on-premises with the entire virtual machine assumed to run a single tenant's workload. 同样,这些数据库也不是设计为在最大程度上利用云提供商的基础结构和硬件。These databases are also not designed to leverage a cloud provider's infrastructure and hardware to the full extent. 例如,OSS NoSQL 数据库引擎不知道如何区分虚拟机关闭与例行映像升级,或者已三向复制高级磁盘这一事实。For example, an OSS NoSQL database engine isn't aware of the differences between a virtual machine being down Vs a routine image upgrade, or the fact that premium disk is already three-way replicated. 它不能利用这些优势,因此无法向客户传达这些优势和成本节省。It can't take advantage of these benefits and pass on the benefits and savings to customers.

  • 按小时付费: 需要在任意时间点缩放的大型工作负荷只需按小时付费。You pay by the hour: For large-scale workloads, that need to scale at any point in time, you are only charged by the hour. 应用程序中的工作负荷以及查询的数据在一年中的不同时间通常会有变化。The workloads on an application typically vary across times of the year, and by the data that is queried. 使用 Azure Cosmos DB 可按需进行扩展或缩减,只有所需的资源才产生费用。With Azure Cosmos DB, you can scale up or down as you need and pay only for what you need. 不能搭配本地或 IaaS 托管的系统使用此模型,因为没有任何方法可以每隔一小时解除硬件。With on-premises or IaaS-hosted systems, you can't match this model, because there isn't a way to decommission the hardware every hour. 在这种情况下,使用 Azure Cosmos DB 可能会将成本平均节省 10 到 14 倍。In such cases, you can potentially save between 10 to 14 times on an average with Azure Cosmos DB.

  • 获得大量免费功能: 在 Azure Cosmos DB 中,写入工作负荷的成本要比替代型数据库服务便宜得多。You get numerous features for free: In Azure Cosmos DB, write workloads are substantially cheaper compared to alternative database services. 此外,Azure Cosmos DB 免费提供自动索引生存时间 (TTL)更改源等功能,而其他数据库服务通常对此收费。In addition, Azure Cosmos DB offers features such as such as automatic indexing, Time to Live (TTL), Change Feed and others without any additional charges, something that other database services typically charge.

  • 对不同的工作负荷使用统一的货币: 与替代型产品/服务不同,举例而言,在 Azure Cosmos DB 中,无需将工作负荷划分为读取和写入。Uses unified currency for diverse workloads: Unlike alternative offerings, in Azure Cosmos DB, you do not need to segment workloads, for example, into reads and writes. 也不需要根据工作负荷类型预配吞吐量,即读取吞吐量与写入吞吐量。Or provision throughput on a per workload type that is read throughput vs. write throughput. 在 Azure Cosmos DB 中,预配的吞吐量将使用统一的规范化货币按每秒请求单位 (RU) 数予以保留。Azure Cosmos DB 不强制要求对工作负荷分配优先级、执行容量规划,或者单独为每种容量付费。In Azure Cosmos DB, provisioned throughput is reserved using a unified and normalized currency in terms of Request Units or RU/sec. Azure Cosmos DB doesn't force you to assign priority to your workloads, perform capacity planning or pay for each type of capacity separately. 借助这种方法可以在不同的操作和工作负荷类型之间轻松进行相同的 RU/秒换算。Such approach enables you to easily interchange the same RU/s between various operations and workload types.

  • 无需预配 VM 即可缩放: 大多数操作数据库要求使用大型虚拟机,以避免干扰性邻居并实现松散的资源调控(如果需要缩放)。Doesn't require provisioning VMs to scale: Most operational databases require you to go with large virtual machines to avoid noisy neighbors and for loose resource governance, if you want scale. 这给客户的前期成本承诺带来了负担。This puts the burden and the upfront commitment of cost on the customers. 使用 Azure Cosmos DB 可以从较小的规模着手,然后无缝扩展到大规模工作负荷,这不会造成任何停机,也不影响数据可用性。With Azure Cosmos DB, you can start small and grow into the large scale workload sizes seamlessly, and without any downtime or impact on data availability.

  • 可以尽最大限度利用预配的吞吐量: 凭借 Azure Cosmos DB 的子核心多路复用优势,可在更高的程度充分利用预配的吞吐量,而 IaaS 托管选项或第三方产品/服务却做不到这一点。You can utilize provisioned throughput to a maximum limit: By the virtue of sub-core multiplexing in Azure Cosmos DB, you can saturate the provisioned throughput to a greater extent than IaaS hosted options or third party offers. 与替代型解决方案相比,此方法可以节省大量的成本。This method saves a lot more than the alternative solutions.

  • Azure Cosmos DB 与其他 Azure 服务的深度集成。Deep integration of Azure Cosmos DB with other Azure services. Azure Cosmos DB 原生与网络、计算、Azure Functions(无服务器)、Azure IoT 和其他 Azure 服务相集成。Azure Cosmos DB has a native integration with Networking, Compute, Azure Functions (serverless), Azure IoT, and others Azure services. 这种集成可以提供最佳的性能、中国范围内的数据复制速度优势,并保证可靠。With this integration, you get the best performance, speed of data replication across China with robust guarantees. 第三方解决方案做不到这一点,或者通常以付费型高级版的形式提供此类功能。The third party solutions won't be able to match or would typically charge a premium to offer such features.

  • 默认情况下至少提供 10 到 20 个容错域,让你自动获得高可用性: Azure Cosmos DB 支持跨容错域的工作负荷分布,这项功能对于高可用性至关重要。You automatically get high availability, with at least 10-20 fault domains by default: Azure Cosmos DB supports the distribution of workloads across fault domains, a feature that is critical for high availability. 它在中国的任何位置,在 99% 的时间内提供 99.999 的读写高可用性。It offers 99.999 high availability for reads and writes at the 99th percentile across anywhere in China. 而你自行实现或通过第三方解决方案实现此类功能的成本则比较高。The cost of implementing something like this on your own or through a third-party solution, would be high.

  • 自动免费获得所有企业功能。You automatically get all enterprise capabilities, at no additional cost. Azure Cosmos DB 免费提供极其全面的合规性认证、安全性以及静态和动态加密(与我们的竞争产品相比)。Azure Cosmos DB offers the most comprehensive set of compliance certifications, security, and encryption at rest and in motion at no additional cost (compared to our competition). 在中国任何位置都可自动获得区域可用性。You automatically get regional availability anywhere in China. 在中国,可让数据库跨越任意数目的 Azure 区域,并随时可以添加或删除区域。You can span your database across any number of Azure regions and add or remove regions at any point in China.

后续步骤Next steps