mirror of
https://github.com/LCTT/TranslateProject.git
synced 2024-12-26 21:30:55 +08:00
Merge pull request #8919 from MZqk/master
20180305 What-s next in IT automation- 6 trends to watch
This commit is contained in:
Ezio
GitHub
commit
f987c2cd6e
@ -1,132 +0,0 @@
|
||||
translating by MZqk
|
||||
|
||||
What’s next in IT automation: 6 trends to watch
|
||||
======
|
||||
|
||||
|
||||
|
||||
We’ve recently covered the [factors fueling IT automation][1], the [current trends][2] to watch as adoption grows, and [helpful tips][3] for those organizations just beginning to automate certain processes.
|
||||
|
||||
Oh, and we also shared expert advice on [how to make the case for automation][4] in your company, as well as [keys for long-term success][5].
|
||||
|
||||
Now, there’s just one question: What’s next? We asked a range of experts to share a peek into the not-so-distant future of [automation][6]. Here are six trends they advise IT leaders to monitor closely.
|
||||
|
||||
### 1. Machine learning matures
|
||||
|
||||
For all of the buzz around [machine learning][7] (and the overlapping phrase “self-learning systems”), it’s still very early days for most organizations in terms of actual implementations. Expect that to change, and for machine learning to play a significant role in the next waves of IT automation.
|
||||
|
||||
Mehul Amin, director of engineering for [Advanced Systems Concepts, Inc.][8], points to machine learning as one of the next key growth areas for IT automation.
|
||||
|
||||
“With the data that is developed, automation software can make decisions that otherwise might be the responsibility of the developer,” Amin says. “For example, the developer builds what needs to be executed, but identifying the best system to execute the processes might be [done] by software using analytics from within the system.”
|
||||
|
||||
That extends elsewhere in this same hypothetical system; Amin notes that machine learning can enable automated systems to provision additional resources when necessary to meet timelines or SLAs, as well as retire those resources when they’re no longer needed, and other possibilities.
|
||||
|
||||
Amin is certainly not alone.
|
||||
|
||||
“IT automation is moving towards self-learning,” says Kiran Chitturi, CTO architect at [Sungard Availability Services][9]. “Systems will be able to test and monitor themselves, enhancing business processes and software delivery.”
|
||||
|
||||
Chitturi points to automated testing as an example; test scripts are already in widespread adoption, but soon those automated testing processes may be more likely to learn as they go, developing, for example, wider recognition of how new code or code changes will impact production environments.
|
||||
|
||||
### 2. Artificial intelligence spawns automation opportunities
|
||||
|
||||
The same principles above hold true for the related (but separate) field of [artificial intelligence][10]. Depending on your definition of AI, it seems likely that machine learning will have the more significant IT impact in the near term (and we’re likely to see a lot of overlapping definitions and understandings of the two fields). Assume that emerging AI technologies will spawn new automation opportunities, too.
|
||||
|
||||
“The integration of artificial intelligence (AI) and machine learning capabilities is widely perceived as critical for business success in the coming years,” says Patrick Hubbard, head geek at [SolarWinds][11].
|
||||
|
||||
### 3. That doesn’t mean people are obsolete
|
||||
|
||||
Let’s try to calm those among us who are now hyperventilating into a paper bag: The first two trends don’t necessarily mean we’re all going to be out of a job.
|
||||
|
||||
It is likely to mean changes to various roles – and the creation of [new roles][12] altogether.
|
||||
|
||||
But in the foreseeable future, at least, you don’t need to practice bowing to your robot overlords.
|
||||
|
||||
“A machine can only consider the environment variables that it is given – it can’t choose to include new variables, only a human can do this today,” Hubbard explains. “However, for IT professionals this will necessitate the cultivation of AI- and automation-era skills such as programming, coding, a basic understanding of the algorithms that govern AI and machine learning functionality, and a strong security posture in the face of more sophisticated cyberattacks.”
|
||||
|
||||
Hubbard shares the example of new tools or capabilities such as AI-enabled security software or machine-learning applications that remotely spot maintenance needs in an oil pipeline. Both might improve efficiency and effectiveness; neither automatically replaces the people necessary for information security or pipeline maintenance.
|
||||
|
||||
“Many new functionalities still require human oversight,” Hubbard says. “In order for a machine to determine if something ‘predictive’ could become ‘prescriptive,’ for example, human management is needed.”
|
||||
|
||||
The same principle holds true even if you set machine learning and AI aside for a moment and look at IT automation more generally, especially in the software development lifecycle.
|
||||
|
||||
Matthew Oswalt, lead architect for automation at [Juniper Networks][13], points out that the fundamental reason IT automation is growing is that it is creating immediate value by reducing the amount of manual effort required to operate infrastructure.
|
||||
|
||||
Rather than responding to an infrastructure issue at 3 a.m. themselves, operations engineers can use event-driven automation to define their workflows ahead of time, as code.
|
||||
|
||||
“It also sets the stage for treating their operations workflows as code rather than easily outdated documentation or tribal knowledge,” Oswalt explains. “Operations staff are still required to play an active role in how [automation] tooling responds to events. The next phase of adopting automation is to put in place a system that is able to recognize interesting events that take place across the IT spectrum and respond in an autonomous fashion. Rather than responding to an infrastructure issue at 3 a.m. themselves, operations engineers can use event-driven automation to define their workflows ahead of time, as code. They can rely on this system to respond in the same way they would, at any time.”
|
||||
|
||||
### 4. Automation anxiety will decrease
|
||||
|
||||
Hubbard of SolarWinds notes that the term “automation” itself tends to spawn a lot of uncertainty and concern, not just in IT but across professional disciplines, and he says that concern is legitimate. But some of the attendant fears may be overblown, and even perpetuated by the tech industry itself. Reality might actually be the calming force on this front: When the actual implementation and practice of automation helps people realize #3 on this list, then we’ll see #4 occur.
|
||||
|
||||
“This year we’ll likely see a decrease in automation anxiety and more organizations begin to embrace AI and machine learning as a way to augment their existing human resources,” Hubbard says. “Automation has historically created room for more jobs by lowering the cost and time required to accomplish smaller tasks and refocusing the workforce on things that cannot be automated and require human labor. The same will be true of AI and machine learning.”
|
||||
|
||||
Automation will also decrease some anxiety around the topic most likely to increase an IT leader’s blood pressure: Security. As Matt Smith, chief architect, [Red Hat][14], recently [noted][15], automation will increasingly help IT groups reduce the security risks associated with maintenance tasks.
|
||||
|
||||
His advice: “Start by documenting and automating the interactions between IT assets during maintenance activities. By relying on automation, not only will you eliminate tasks that historically required much manual effort and surgical skill, you will also be reducing the risks of human error and demonstrating what’s possible when your IT organization embraces change and new methods of work. Ultimately, this will reduce resistance to promptly applying security patches. And it could also help keep your business out of the headlines during the next major security event.”
|
||||
|
||||
**[ Read the full article: [12 bad enterprise security habits to break][16]. ] **
|
||||
|
||||
### 5. Continued evolution of scripting and automation tools
|
||||
|
||||
Many organizations see the first steps toward increasing automation – usually in the form of scripting or automation tools (sometimes referred to as configuration management tools) – as "early days" work.
|
||||
|
||||
But views of those tools are evolving as the use of various automation technologies grows.
|
||||
|
||||
“There are many processes in the data center environment that are repetitive and subject to human error, and technologies such as [Ansible][17] help to ameliorate those issues,” says Mark Abolafia, chief operating officer at [DataVision][18]. “With Ansible, one can write a specific playbook for a set of actions and input different variables such as addresses, etc., to automate long chains of process that were previously subject to human touch and longer lead times.”
|
||||
|
||||
**[ Want to learn more about this aspect of Ansible? Read the related article:[Tips for success when getting started with Ansible][19]. ]**
|
||||
|
||||
Another factor: The tools themselves will continue to become more advanced.
|
||||
|
||||
“With advanced IT automation tools, developers will be able to build and automate workflows in less time, reducing error-prone coding,” says Amin of ASCI. “These tools include pre-built, pre-tested drag-and-drop integrations, API jobs, the rich use of variables, reference functionality, and object revision history.”
|
||||
|
||||
### 6. Automation opens new metrics opportunities
|
||||
|
||||
As we’ve said previously in this space, automation isn’t IT snake oil. It won’t fix busted processes or otherwise serve as some catch-all elixir for what ails your organization. That’s true on an ongoing basis, too: Automation doesn’t eliminate the need to measure performance.
|
||||
|
||||
**[ See our related article[DevOps metrics: Are you measuring what matters?][20] ]**
|
||||
|
||||
In fact, automation should open up new opportunities here.
|
||||
|
||||
“As more and more development activities – source control, DevOps pipelines, work item tracking – move to the API-driven platforms – the opportunity and temptation to stitch these pieces of raw data together to paint the picture of your organization's efficiency increases,” says Josh Collins, VP of architecture at [Janeiro Digital][21].
|
||||
|
||||
Collins thinks of this as a possible new “development organization metrics-in-a-box.” But don’t mistake that to mean machines and algorithms can suddenly measure everything IT does.
|
||||
|
||||
“Whether measuring individual resources or the team in aggregate, these metrics can be powerful – but should be balanced with a heavy dose of context,” Collins says. “Use this data for high-level trends and to affirm qualitative observations – not to clinically grade your team.”
|
||||
|
||||
**Want more wisdom like this, IT leaders?[Sign up for our weekly email newsletter][22].**
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: https://enterprisersproject.com/article/2018/3/what-s-next-it-automation-6-trends-watch
|
||||
|
||||
作者：[Kevin Casey][a]
|
||||
译者：[译者ID](https://github.com/译者ID)
|
||||
校对：[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]:https://enterprisersproject.com/user/kevin-casey
|
||||
[1]:https://enterprisersproject.com/article/2017/12/5-factors-fueling-automation-it-now
|
||||
[2]:https://enterprisersproject.com/article/2017/12/4-trends-watch-it-automation-expands
|
||||
[3]:https://enterprisersproject.com/article/2018/1/getting-started-automation-6-tips
|
||||
[4]:https://enterprisersproject.com/article/2018/1/how-make-case-it-automation
|
||||
[5]:https://enterprisersproject.com/article/2018/1/it-automation-best-practices-7-keys-long-term-success
|
||||
[6]:https://enterprisersproject.com/tags/automation
|
||||
[7]:https://enterprisersproject.com/article/2018/2/how-spot-machine-learning-opportunity
|
||||
[8]:https://www.advsyscon.com/en-us/
|
||||
[9]:https://www.sungardas.com/en/
|
||||
[10]:https://enterprisersproject.com/tags/artificial-intelligence
|
||||
[11]:https://www.solarwinds.com/
|
||||
[12]:https://enterprisersproject.com/article/2017/12/8-emerging-ai-jobs-it-pros
|
||||
[13]:https://www.juniper.net/
|
||||
[14]:https://www.redhat.com/en?intcmp=701f2000000tjyaAAA
|
||||
[15]:https://enterprisersproject.com/article/2018/2/12-bad-enterprise-security-habits-break
|
||||
[16]:https://enterprisersproject.com/article/2018/2/12-bad-enterprise-security-habits-break?sc_cid=70160000000h0aXAAQ
|
||||
[17]:https://opensource.com/tags/ansible
|
||||
[18]:https://datavision.com/
|
||||
[19]:https://opensource.com/article/18/2/tips-success-when-getting-started-ansible?intcmp=701f2000000tjyaAAA
|
||||
[20]:https://enterprisersproject.com/article/2017/7/devops-metrics-are-you-measuring-what-matters?sc_cid=70160000000h0aXAAQ
|
||||
[21]:https://www.janeirodigital.com/
|
||||
[22]:https://enterprisersproject.com/email-newsletter?intcmp=701f2000000tsjPAAQ
|
||||
@ -0,0 +1,130 @@
|
||||
IT自动化的下一步是什么: 6 大趋势
|
||||
======
|
||||
|
||||
|
||||
|
||||
我们最近介绍了 [促进自动化的因素][1] ,目前正在被人们采用的 [趋势][2], 以及那些刚开始使用自动化部分流程组织 [有用的技巧][3] 。
|
||||
|
||||
噢, 我们也分享了在你的公司[如何使用自动化的案例][4] , 以及 [长期成功的关键][5].
|
||||
|
||||
现在, 只有一个问题: 自动化的下一步是什么? 我们邀请一系列专家分享一下 [自动化][6]不远的将来。 以下是他们建议IT领域领导需密切关注的六大趋势。
|
||||
|
||||
### 1. 机器学习的成熟
|
||||
|
||||
对于关于 [机器学习][7]的讨论 (与“自我学习系统”相似的定义),对于绝大多数组织的项目来说，实际执行起来它仍然为时过早。但预计这将发生变化，机器学习将在下一次IT自动化浪潮中将扮演着至关重要的角色。
|
||||
|
||||
[Advanced Systems Concepts, Inc.][8]公司工程总监 Mehul Amin 指出机器学习是IT自动化下一个关键增长领域之一。
|
||||
|
||||
“随着数据化的发展, 自动化软件理应可以自我决策，否则这就是开发人员的责任了”， Amin 说。 “例如, 开发者需要执行构建内容, 但是识别系统最佳执行流程的，可能是由系统内软件分析完成。”
|
||||
|
||||
假设将这个系统延伸到其他地方中。Amin 指出，机器学习可以使自动化系统在必要的时候提供额外的资源，以需要满足时间线或SLA，同样在不需要资源的时候退出以及其他的可能性。
|
||||
|
||||
显然不只有 Amin 一个人这样认为。
|
||||
|
||||
[Sungard Availability Services][9] 公司首席架构师 Kiran Chitturi 表示,“IT自动化正在走向自我学习的方向” 。“系统将会能测试和监控自己，加强业务流程和软件交付能力。”
|
||||
|
||||
Chitturi 指出自动化测试就是个例子。脚本测试已经被广泛采用，但很快这些自动化测试流程将会更容易学习，更快发展，例如开发出新的代码或将更为广泛地影响生产环境。
|
||||
|
||||
### 2. 人工智能催生的自动化
|
||||
|
||||
上述原则同样适合 [人工智能][10]（但是为独立）的领域。假定新兴的人工智能技术将也会产生新的自动化机会。根据对人工智能的定义，机器学习在短时间内可能会对IT领域产生巨大的影响（并且我们可能会看到这两个领域的许多重叠的定义和理解）。
|
||||
|
||||
[SolarWinds][11]公司技术负责人 Patrick Hubbard说，“人工智能（AI）和机器学习的整合普遍被认为对未来几年的商业成功起至关重要的作用。”
|
||||
|
||||
### 3. 这并不意味着不再需要人力
|
||||
|
||||
让我们试着安慰一下那些不知所措的人：前两种趋势并不一定意味着我们将失去工作。
|
||||
|
||||
这很可能意味着各种角色的改变以及[全新角色][12]的创造。
|
||||
|
||||
但是在可预见的将来，至少，你不必需要机器人鞠躬。
|
||||
|
||||
“一台机器只能运行在给定的环境变量中它不能选择包含新的变量，在今天只有人类可以这样做，” Hubbard 解释说。“但是，对于IT专业人员来说，这将是需要培养AI和自动化技能的时代。如对程序设计、编程、管理人工智能和机器学习功能算法的基本理解，以及用强大的安全状态面对更复杂的网络攻击。”
|
||||
|
||||
Hubbard 分享一些新的工具或功能例子，例如支持人工智能的安全软件或机器学习的应用程序，这些应用程序可以远程发现石油管道中的维护需求。两者都可以提高效益和效果，自然不会代替需要信息安全或管道维护的人员。
|
||||
|
||||
“许多新功能仍需要人工监控，”Hubbard 说。“例如，为了让机器确定一些‘预测’是否可能成为‘规律’，人为的管理是必要的。”
|
||||
|
||||
即使你把机器学习和AI先放在一边，看待一般地IT自动化，同样原理也是成立的,尤其是在软件开发生命周期中。
|
||||
|
||||
[Juniper Networks][13]公司自动化首席架构师 Matthew Oswalt ，指出IT自动化增长的根本原因是它通过减少操作基础设施所需的人工工作量来创造直接价值。
|
||||
|
||||
在代码上，操作工程师可以使用事件驱动的自动化提前定义他们的工作流程，而不是在凌晨3点来应对基础设施的问题。
|
||||
|
||||
“它也将操作工作流程作为代码而不再是容易过时的文档或系统知识阶段，”Oswalt解释说。“操作人员仍然需要在[自动化]工具响应事件方面后发挥积极作用。采用自动化的下一个阶段是建立一个能够跨IT频谱识别发生的有趣事件的系统，并以自主方式进行响应。在代码上，操作工程师可以使用事件驱动的自动化提前定义他们的工作流程，而不是在凌晨3点来应对基础设施的问题。他们可以依靠这个系统在任何时候以同样的方式作出回应。”
|
||||
|
||||
### 4. 对自动化的焦虑将会减少
|
||||
|
||||
SolarWinds公司的 Hubbard 指出，“自动化”一词本身就产生大量的不确定性和担忧，不仅仅是在IT领域，而且是跨专业领域，他说这种担忧是合理的。但一些随之而来的担忧可能被夸大了，甚至是科技产业本身。现实可能实际上是这方面的镇静力：当自动化的实际实施和实践帮助人们认识到这个列表中的“3”时，我们将看到“4”的出现。
|
||||
|
||||
“今年我们可能会看到对自动化焦虑的减少，更多的组织开始接受人工智能和机器学习作为增加现有人力资源的一种方式，”Hubbard说。“自动化历史上的今天为更多的工作创造了空间,通过降低成本和时间来完成较小任务，并将劳动力重新集中到无法自动化并需要人力的事情上。人工智能和机器学习也是如此。”
|
||||
|
||||
自动化还将减少IT领导者神经紧张主题的一些焦虑：安全。正如[红帽][14]公司首席架构师 Matt Smith 最近[指出][15]的那样，自动化将越来越多地帮助IT部门降低与维护任务相关的安全风险。
|
||||
|
||||
他的建议是：“首先在维护活动期间记录和自动化IT资产之间的交互。通过依靠自动化，您不仅可以消除历史上需要大量手动操作和手术技巧的任务，还可以降低人为错误的风险，并展示当您的IT组织采纳变更和新工作方法时可能发生的情况。最终，这将迅速减少对应用安全补丁的抵制。而且它还可以帮助您的企业在下一次重大安全事件中摆脱头条新闻。”
|
||||
|
||||
**[ 阅读全文: [12个企业安全坏习惯要打破。][16] ] **
|
||||
|
||||
### 5. 脚本和自动化工具将持续发展
|
||||
|
||||
看到许多组织增加自动化的第一步 - 通常以脚本或自动化工具（有时称为配置管理工具）的形式 - 作为“早期”工作。
|
||||
|
||||
但是随着各种自动化技术的使用，对这些工具的观点也在不断发展。
|
||||
|
||||
[DataVision][18]首席运营官 Mark Abolafia 表示：“数据中心环境中存在很多重复性过程，容易出现人为错误，[Ansible][17]等技术有助于缓解这些问题。“通过 Ansible ，人们可以为一组操作编写特定的步骤，并输入不同的变量，例如地址等，使过去长时间的过程链实现自动化，而这些过程以前都需要人为触摸和更长的交货时间。”
|
||||
|
||||
**[想了解更多关于Ansible这个方面的知识吗？阅读相关文章:[使用Ansible时的成功秘诀][19]。 ]**
|
||||
|
||||
另一个因素是：工具本身将继续变得更先进。
|
||||
|
||||
“使用先进的IT自动化工具，开发人员将能够在更短的时间内构建和自动化工作流程，减少易出错的编码，” ASCI 公司的 Amin 说。“这些工具包括预先构建的，预先测试过的拖放式集成，API作业，丰富的变量使用，参考功能和对象修订历史记录。”
|
||||
|
||||
### 6. 自动化开创了新的指标机会
|
||||
|
||||
正如我们在此前所说的那样，IT自动化不是万能的。它不会修复被破坏的流程，或者以其他方式为您的组织提供全面的灵丹妙药。这也是持续不断的：自动化并不排除衡量性能的必要性。
|
||||
|
||||
**[ 参见我们的相关文章 [DevOps指标：你在衡量什么重要吗？][20] ]**
|
||||
|
||||
实际上,自动化应该打开新的机会。
|
||||
|
||||
[Janeiro Digital][21]公司架构师总裁 Josh Collins 说，“随着越来越多的开发活动 - 源代码管理，DevOps管道，工作项目跟踪 - 转向API驱动的平台 - 将这些原始数据拼接在一起以描绘组织效率提升的机会和图景”。
|
||||
|
||||
Collins 认为这是一种可能的新型“开发组织度量指标”。但不要误认为这意味着机器和算法可以突然预测IT所做的一切。
|
||||
|
||||
“无论是衡量个人资源还是整体团队，这些指标都可以很强大 - 但应该用大量的背景来衡量。”Collins说，“将这些数据用于高层次趋势并确认定性观察 - 而不是临床评级你的团队。”
|
||||
|
||||
**想要更多这样知识, IT领导者？[注册我们的每周电子邮件通讯][22]。**
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: https://enterprisersproject.com/article/2018/3/what-s-next-it-automation-6-trends-watch
|
||||
|
||||
作者：[Kevin Casey][a]
|
||||
译者：[MZqk](https://github.com/MZqk)
|
||||
校对：[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]:https://enterprisersproject.com/user/kevin-casey
|
||||
[1]:https://enterprisersproject.com/article/2017/12/5-factors-fueling-automation-it-now
|
||||
[2]:https://enterprisersproject.com/article/2017/12/4-trends-watch-it-automation-expands
|
||||
[3]:https://enterprisersproject.com/article/2018/1/getting-started-automation-6-tips
|
||||
[4]:https://enterprisersproject.com/article/2018/1/how-make-case-it-automation
|
||||
[5]:https://enterprisersproject.com/article/2018/1/it-automation-best-practices-7-keys-long-term-success
|
||||
[6]:https://enterprisersproject.com/tags/automation
|
||||
[7]:https://enterprisersproject.com/article/2018/2/how-spot-machine-learning-opportunity
|
||||
[8]:https://www.advsyscon.com/en-us/
|
||||
[9]:https://www.sungardas.com/en/
|
||||
[10]:https://enterprisersproject.com/tags/artificial-intelligence
|
||||
[11]:https://www.solarwinds.com/
|
||||
[12]:https://enterprisersproject.com/article/2017/12/8-emerging-ai-jobs-it-pros
|
||||
[13]:https://www.juniper.net/
|
||||
[14]:https://www.redhat.com/en?intcmp=701f2000000tjyaAAA
|
||||
[15]:https://enterprisersproject.com/article/2018/2/12-bad-enterprise-security-habits-break
|
||||
[16]:https://enterprisersproject.com/article/2018/2/12-bad-enterprise-security-habits-break?sc_cid=70160000000h0aXAAQ
|
||||
[17]:https://opensource.com/tags/ansible
|
||||
[18]:https://datavision.com/
|
||||
[19]:https://opensource.com/article/18/2/tips-success-when-getting-started-ansible?intcmp=701f2000000tjyaAAA
|
||||
[20]:https://enterprisersproject.com/article/2017/7/devops-metrics-are-you-measuring-what-matters?sc_cid=70160000000h0aXAAQ
|
||||
[21]:https://www.janeirodigital.com/
|
||||
[22]:https://enterprisersproject.com/email-newsletter?intcmp=701f2000000tsjPAAQ
|
||||
Loading…
Reference in New Issue
Block a user