博客

DBAA与Nutanix时代的TCO(或为什么从传统数据库操作中演变)

比尔·汤普森(Bill Thompson&) Steve Kaplan
| min

Over the years, we’ve seen IT positions evolve or even disappear as technology advances. It is often hard, for example, to find PC technicians in organizations running VDI. And former dedicated server technicians have evolved into virtualization administrators.

如今,HCI正在逐步淘汰存储管理员的需求。基于本机的基于Linux的管理程序正在逐步淘汰对专用虚拟化管理员的需求。没有人在AWS,阿里巴巴或Nutanix虚拟化中获得认证。

Nutanix时代为数据库管理带来了一种新的方法,该方法正在改变DBA角色的性质。与其将大部分时间花在数据库管理的低级任务上,例如数据库提供,修补和生命周期管理,而是ERA自动化并简化了这些负担,使DBA能够专注于应用程序。例如,复制数据库变得像copy and paste计算机上的操作。

传统数据库操作的挑战

Traditional database operations are disjointed, complex, and inefficient. Leveraging multiple solutions for database operations such as provisioning, cloning, and refreshing results in a complex environment that slows business agility and efficiency. When working with enterprise databases, for example, DBAs do not just clone a production instance exactly as-is, but also must ensure the ability to create these clones from any point in time.

令人沮丧的是,许多组织已转向云托管的数据库管理解决方案,尽管更敏捷,但仍引入了重大的操作变更要求,在大多数情况下,成本更高。Nutanix时代为公共云解决方案的敏捷性和简单性提供了一种本地替代品,但成本要低得多。

将DBA焦点从低级活动转移到高价值活动

We recently worked with an enterprise customer whose application development team was under constant pressure to add new functionality required by the line of business. The lone Oracle DBA struggled just to keep up with provisioning, tuning, and cloning of existing databases let alone handle new requests. The customer was still trying to find the specialized Oracle skill sets required to fill two open DBA head counts after 18 months.

While the IT admins, in theory, could have used scripting to automate many of the routine tasks, they lacked the time and skill set required. Instead, we showed them how Era could enable the single DBA to meet the organization’s database requests while saving $2.6M in TCO over five years. These savings resulted from four key areas: Labor efficiency, lower third-party technology costs, fast time to market, and reduced storage costs.

1. Labor Efficiency
Figure 1 shows where the customer’s lone DBA spent her day. Up to 75% of her time was consumed by routine, low-level requests, such as database provisioning, cloning/backups, and patching/upgrades. Less than one-quarter of her time was spent on more forward-looking and higher-value activities such as capacity planning, performance management, and documentation.

图1:像往常一样的业务努力

Reducing project pipeline and re-allocating effort towards higher-value activities would have previously required three DBAs at a $600K annual cost. Nutanix Era, however, removes the complexity of common database operations through one-click automation, shrinking the amount of time spent on low-level activities by approximately 87%. DBAs can shift their focus to higher-value, forward-looking activities such capacity planning, performance tuning, security compliance, and new projects as shown in Figure 2.

图2:Nutanix时代的DBA努力

Because the single DBA could now handle the new project requests, the customer realized an annual savings of $400K. The elimination of legacy inefficiencies impacted staff other than the DBA. Manually intensive DBA processes often cause additional negative downstream labor impacts (costs) for infrastructure administrators.

Developers particularly benefit from calling the Era API to create clones or refresh existing ones. They can include these API calls in whatever workflow tool they are using (Calm, chef, ansible, etc.) and make them completely independent from the DBA. This is especially interesting in CI/CD scenarios where a clone refresh may be required every time new code is checked in.

图3显示了每个估计的五十二个年度项目要求中的每一个都需要平均二十小时完成周期。使用全额负担65美元的劳动力成本,我们确定与项目延迟相关的年度成本为67,600美元。随着ERA自动化,客户可能期望将与DBA相关的过程周期减少50%,从而节省33,800美元的年度机会成本。

图3:系统管理机会成本节省

2.降低许可和支持成本
对Nutanix时代的投资可能使客户可以降低与数据库管理工具集有关的成本。例如,我们的客户正在为Oracle生命周期管理,云管理包和高级压缩支付11​​6,622美元的年度支持费用。如图4所示,Nutanix ERA授权客户减少组织对专业技能,自定义脚本和技术支持成本的依赖。

图4:降低第三方技术成本

3.上市时间
冗长的手动过程不可避免地会导致项目瓶颈和释放延迟。这些延迟的负面影响可以降低到客户满意度得分,最终用户效率低下以及潜在的收入损失。许多客户对这些下游业务影响表示担忧,但没有有效的方法来衡量它们。

图5显示了由于DBA相关的瓶颈影响更广泛的组织,我们的保守方法是计算损失收入的方法。假设只有52个年度项目中只有一项收入增长了100万美元,那么这项业务照常工作九十天的部署代表了246,575美元的收入机会成本。通过减少开发瓶颈,我们保守地认为,总部署周期可以减少30%(63天),转化为潜在的73,973美元的年收入。

Figure 5: Reduction in Time to Market

4.缩小克隆和备份的存储要求
DBA在开发,升级和测试周期期间处理许多数据库副本和备份的请求。这些完整的数据库副本通常会导致高存储成本以及较长的备份和恢复时间。对于我们的客户研究,如图6所示,我们计算了216TB的年度副本存储要求。每TB $ 800,这导致了每年172,800美元的成本。

Nutanix ERA复制数据管理服务为零字节数据库克隆提供到任何时间点,导致数据存储需求减少6倍,或每年$ 28,800。ERA使数据库的复制数据管理操作的大小与您的个人计算机上的复制和粘贴操作一样容易。

图6:数据库副本的降低存储成本

Public Cloud Goodness On-premises Driving a Compelling ROI

Nutanix ERA可以使数据库作为服务(DBAA)模型,用于管理和操作数据库。它可以在组织的数据库遗产中进行标准化,同时引入效率,敏捷性,可扩展性和成本节省。在图7中总结的我们的客户示例中,Nutanix ERA能够在5年内将成本降低62%,即260万美元。市场减少的时间增加了37万美元的收入。

图7:5年的简易成本比较

图8显示了DBA运营效率的260万美元TCO节省的细分,存储和管理技术成本的降低以及改善了部署时间。由于使用Nutanix Enterprise Cloud功能,包括压缩(Oracle Advanced Compression选项)和加密(Oracle Advanced Advanced Security Option)以及已经提到的Cloud Management Pack和数据库LifeCycle Management Pack,因此可能会导致对Oracle许可的需求减少,从而产生了其他重大节省。

Figure 8: 5-Year Savings with Era

Era Extends the Nutanix Enterprise Cloud Platform

nutanixEnterprise Cloud is the software-defined platform built upon the Nutanix hyperconverged infrastructure. The unique architecture of the Nutanix solution enables software automation of datacenter tasks such as file services, micro-segmentation, DevOps, and many others.

ERA利用Enterprise Cloud平台启用零字节架空限制的快照和克隆,这些快照和克隆不需要复制实际数据。这种方法使克隆和捕捉超级快速,并且在编写或覆盖新数据之前,它们不再使用任何实际的存储容量。此外,ERA都使用黄金图像进行配置和修补,从而使这些图像的版本管理既容易又有效。

ERA的即将到来的版本将进一步利用Nutanix平台在群集(生产到非生产)中实现敏捷数据管理(生命周期),并定期刷新非生产环境。与其将时间花在手动任务上,这些任务不会为组织增加价值(或真正地对自己的职业生涯),而是让DBA专注于为组织带来巨大收益的活动。

免责声明:此博客中表达的观点是作者,不一定是Nutanix,Inc。或其他任何员工或分支机构的观点。

也可以看看
DBAA时代。2019年。MaksimMalygin。nutanix.com。
使用Nutanix构建数据服务平台。2019年。SachinChheda和Chris Paap。nutanix.com。

©2020 Nutanix,Inc。保留所有权利。Nutanix,Nutanix徽标和所有Nutanix产品,此处提到的功能和服务名称是美国和其他国家的Nutanix,Inc。的注册商标或商标。本文提到的所有其他品牌名称仅用于识别目的,可能是其各自持有人的商标。这篇文章可能包含指向不属于Nutanix.com一部分的外部网站的链接。Nutanix不控制这些站点,并对任何外部站点的内容或准确性不承担所有责任。我们决定链接到外部站点的决定不应被视为对该站点上任何内容的认可。本文中包含的某些信息可能与从第三方来源以及我们自己的内部估计和研究获得的研究,出版物,调查和其他数据有关。尽管我们认为这些第三方研究,出版物,调查和其他数据是在本文之日起可靠的,但它们尚未独立验证,我们对任何信息的充分性,公平性,准确性或完整性都没有任何代表性从第三方来源获得。

Baidu