### Chapter 33 Putting It All Together把它们放在一起 > The art of war teaches us to rely not on the likelihood of the enemy’s not coming,but on our own readiness to receive him; not on the chance of his not attacking,but rather on the fact that we have made our position unassailable.—Sun Tzu > 战争的艺术告诉我们,不要依赖敌人不来的可能性,而要依赖我们自己是否准备好迎接他;不是因为他不进攻的机会,而是因为我们已经使自己的地位变得无懈可击。——《孙子》 We started this book with a discussion of how scalability is a combination of art andscience. The art aspect of scaling is seen in the interactions between platforms, orga-nizations, and processes, which impact any structured approach in a company. Thescience of scalability is embodied within the method by which we measure our effortsand in the application of the scientific method. A particular company’s approach toscaling must be crafted around the ecosystem fashioned by the intersection of thetechnology platform, the uniqueness of the organization, and the maturity and capa-bilities of the existing processes. Because a one-size-fits-all implementation or answerdoes not exist, we have focused this book on providing skills and lessons regardingapproaches. 本书一开始就讨论了可扩展性如何成为艺术与科学的结合。扩展的艺术体现在平台、组织和流程之间的交互中,这会影响公司中的任何结构化方法。可扩展性的科学体现在我们衡量努力的方法和科学方法的应用中。特定公司的扩展方法必须围绕由技术平台、组织的独特性以及现有流程的成熟度和功能的交集形成的生态系统来制定。由于不存在一刀切的实施或答案,因此我们将本书的重点放在提供有关方法的技能和课程上。 It all begins with people. You can’t get better or even sustain without the rightteam, the right leadership, the right management, and the right organizational struc-ture. People are central to establishing and following processes as well as designingand implementing the technology. Having the right people in terms of skill set, moti-vation, and cultural fit are the essential building blocks. On top of this must beplaced the right roles. Even the best people must be placed in the right job that appro-priately utilizes their skills. Additionally, these roles must be organized in the rightorganizational structure and they must receive strong leadership and management inorder to perform at an optimal level. 一切始于人。如果没有合适的团队、合适的领导、合适的管理和合适的组织结构,你就无法变得更好,甚至无法维持下去。人是建立和遵循流程以及设计和实施技术的核心。拥有具备技能、积极性和文化契合度的合适人才是重要的组成部分。最重要的是必须放置正确的角色。即使是最优秀的人才也必须被安排在能够充分发挥其技能的正确岗位上。此外,这些角色必须以正确的组织结构进行组织,并且必须得到强有力的领导和管理,才能发挥最佳水平。 Although people develop and maintain the processes within an organization, theprocesses control how the individuals and teams behave. Processes are essentialbecause they allow your teams to react quickly to crisis, determine the root cause offailures, determine capacity of systems, analyze scalability needs, implement scalabilityprojects, and many more fundamental needs for a scalable system. There is no singleright answer when it comes to processes, and there are many wrong answers. Eachand every process must be evaluated first for general fit within the organization interms of its rigor or repeatability and then specifically for what steps are right foryour particular team in terms of complexity. Too much process can stifle innovationand strangle shareholder value, whereas if you are missing the processes that allowyou to learn from both your past mistakes and failures, you will very likely at leastunderachieve and potentially even fail as a company. 尽管人们在组织内开发和维护流程,但流程控制着个人和团队的行为方式。流程至关重要,因为它们使您的团队能够快速应对危机、确定故障的根本原因、确定系统容量、分析可扩展性需求、实施可扩展性项目以及可扩展系统的许多其他基本需求。对于流程来说,没有唯一正确的答案,而且有很多错误的答案。必须首先评估每个流程的严格性或可重复性,以了解其在组织中的总体适应性,然后根据复杂性具体评估哪些步骤适合您的特定团队。过多的流程会扼杀创新并扼杀股东价值,而如果你缺少能够从过去的错误和失败中吸取教训的流程,那么你的公司很可能至少会表现不佳,甚至可能会失败。 Last but not least is the technology that drives the business either as the productitself or the infrastructure allowing the product to be brought to market. There aremany methodologies to understand and consider when implementing your scalablesolution such as splitting by multiple axes, caching, using asynchronous calls, devel-oping a data management approach, and many others that we’ve covered in thisbook. The key is to develop expertise in these approaches so that you appropriatelyutilize them when necessary. The right people, who understand these approaches, andthe right processes, ones that insist on the evaluation of these approaches, are all puttogether to develop scalable technology. 最后但并非最不重要的一点是驱动业务的技术,无论是作为产品本身还是将产品推向市场的基础设施。在实现可扩展解决方案时,有许多方法需要理解和考虑,例如按多个轴拆分、缓存、使用异步调用、开发数据管理方法以及我们在本书中介绍的许多其他方法。关键是培养这些方法的专业知识,以便在必要时适当地利用它们。理解这些方法的合适的人,以及坚持评估这些方法的正确的流程,都聚集在一起开发可扩展的技术。 In the beginning of this book, we introduced the concept of virtuous and viciouscycles (refer to Figure I.1 in the Introduction). The lack of attention to the people andprocesses can cause poor technical decisions that are what we term a vicious cycle.After you start down this path, teams are likely to ignore the people and processmore because of the demands on energy and resources to fix the technology prob-lems. The exact opposite is what we called the virtuous cycle when the right peopleand the right process feed each other and produce excellent, scalable technology, thusfreeing up resources to continue to improve the overall ecosystem within the organi-zation. After you start down the vicious cycle, it is difficult to stop but with focusedintensity you can do it. In Figure 33.1, we have depicted this concept of stopping thedownward spiral and starting it back in the other direction. 在本书的开头,我们介绍了良性循环和恶性循环的概念(参见引言中的图I.1)。缺乏对人员和流程的关注可能会导致糟糕的技术决策,这就是我们所说的恶性循环。在你开始走这条路之后,团队可能会忽视更多的人员和流程,因为修复技术需要能源和资源问题。完全相反的是我们所说的良性循环,即合适的人员和合适的流程相互促进并产生优秀的、可扩展的技术,从而释放资源以继续改善组织内的整体生态系统。当你开始恶性循环后,就很难停止,但只要集中精力,你就可以做到。在图 33.1 中,我们描述了停止螺旋式下降并从另一个方向开始的概念。 What to Do Now? 现在做什么? The question that you might have now that you have made your way through all thisinformation about people, processes, and technology is “What to do now?” We havea simple four-step process that we recommend for putting all this information intopractice. Note that this process holds true whether you are making technical, process,organizational, or even personal changes (as in changes in you). These steps are 当您已经了解了有关人员、流程和技术的所有这些信息后,您可能会遇到的问题是“现在该怎么办?”我们建议采用一个简单的四步流程来将所有这些信息付诸实践。请注意,无论您是在进行技术、流程、组织,甚至个人的改变(如您自身的改变),这个过程都适用。这些步骤是 1.Assess your situation. 1.评估您的情况。 2.Determine where you need or want to be. 2.确定您需要或想要去的地方。 3.Annotate the difference between the two. 3.注明两者的区别。 4.Make a plan. 4.制定计划。 ![](https://blog.baidu-google.com/usr/uploads/2024/06/314754718.png) The first step is to assess your current situation. Look at all three aspects of peo-ple, process, and technology and make an honest assessment of where you are witheach. Answer questions such as what is the caliber of your people, is the structure ofyour organization optimal, what parts of your technology scale, and are your pro-cesses adequately supporting your goals? These assessments are difficult to do frominside the organization. Sometimes, you are just too close to the situation. Our firmAKF Partners is often asked to perform these assessments because we can provide anobjective third-party perspective. 第一步是评估您目前的情况。审视人员、流程和技术的所有三个方面,并对每个方面的情况做出诚实的评估。回答诸如您的人员素质如何、您的组织结构是否最佳、您的技术规模的哪些部分以及您的流程是否充分支持您的目标等问题?这些评估很难在组织内部进行。有时,你太接近实际情况了。我们公司 AKF Partners 经常被要求进行这些评估,因为我们可以提供客观的第三方观点。 After you have the assessment in hand, the second step is to determine where youneed to be along those three aspects of people, process, and technology. Do you needto focus on infusing more experience in the engineering team or do you need to dedi-cate time and energy to developing more robust processes? One approach to this is tostart with your projected growth rate and assess how this will place a demand on thetechnology, process, and people of your company. Additional factors such as acquisi-tions or additional funding should be considered when determining this ideal state. 完成评估后,第二步是确定在人员、流程和技术这三个方面您需要处于什么位置。您是否需要专注于为工程团队注入更多经验,或者是否需要投入时间和精力来开发更强大的流程?一种方法是从预计的增长率开始,评估这将如何对公司的技术、流程和人员提出要求。在确定这种理想状态时,应考虑其他因素,例如收购或额外资金。 The third step is to compare the real and ideal situations within your company.This will result in a list of items that are not at a sufficient level of skill, maturity, orscalability. For example, you may decide in Step 1 that you have a young engineeringteam with no experience in scaling large databases. In Step 2, you decide that basedon your projected growth rate, you will need to triple the amount of data stored inyour database in the next 18 months. In Step 3, you would identify this discrepancyand mark down that you need to add people on your engineering team who haveexperience scaling databases. The last part of Step 3 is to prioritize this list of itemsfrom most important to least. We often start by breaking the list into two sets: thosediscrepancies that we consider weaknesses and those that are opportunities toimprove. 第三步是比较公司内的真实情况和理想情况。这将产生一个技能、成熟度或可扩展性水平不够的项目列表。例如,您可能在步骤 1 中决定您有一个年轻的工程团队,没有扩展大型数据库的经验。在第 2 步中,您决定根据预计的增长率,在未来 18 个月内需要将数据库中存储的数据量增加两倍。在步骤 3 中,您将识别这种差异,并标记您需要在工程团队中添加具有扩展数据库经验的人员。步骤 3 的最后一部分是按从最重要到最不重要的顺序排列此项目列表的优先级。我们通常首先将清单分为两组:我们认为是弱点的差异和需要改进的机会。 The last step is to create an action plan for fixing the discrepancies identified in theprevious step. The timeframe for the action plan or plans will be determined by yourorganization’s needs, the available resources, and the severity of the discrepancies.We often suggest teams create two plans: one for addressing immediate concerns in a30- to 45-day plan and one for addressing longer term issues in a 180-day plan. Ifyou have categorized the items into weaknesses and opportunities, these will corre-spond to the short- and long-term action plans. 最后一步是制定行动计划来修复上一步中发现的差异。一个或多个行动计划的时间框架将取决于您组织的需求、可用资源以及差异的严重程度。我们通常建议团队制定两个计划:一个用于在 30 到 45 天的计划中解决眼前的问题,另一个用于解决当前问题。在 180 天计划中解决长期问题。如果您将这些项目分为弱点和机会,这些将对应于短期和长期行动计划。 #####Four Steps to Action 行动的四个步骤 A simple four-step process for putting this all together: 将所有这些整合在一起的简单四步过程 1.Perform an assessment. Rate your company on organization, processes, and architec-ture. Use an outside agent if necessary. 1.进行评估。对您公司的组织、流程和架构进行评分。如有必要,请使用外部代理。 2.Define where you need to be or your ideal state. How large is your company going to grow in 12 to 24 months? What does that mean for your organization, processes, and architecture? 2.定义您需要去的地方或您的理想状态。您的公司将在 12 到 24 个月内发展到多大?这对您的组织、流程和架构意味着什么? 3.List the differences between your actual and ideal situations. Rank order these differ-ences from most to least severe. 3.列出你的实际情况和理想情况之间的差异。将这些差异从最严重到最不严重进行排序。 4.Action plans. Put together an action plan to resolve the issues identified in the previous steps. This can be in the form of a single prioritized plan, a short-term and long-term plan, or any other planning increments that your company uses. 4.行动计划。制定行动计划来解决前面步骤中发现的问题。这可以是单一优先计划、短期和长期计划或公司使用的任何其他计划增量的形式。 ####Case Studies 实例探究 Throughout this book, we painted a picture of AllScale, a fictional company amal-gamating our experiences as advisors and executives, to provide real-world examples.As a last example of how our techniques can be employed to create scalable solu-tions, we are going to leave AllScale and discuss three companies with which we havepersonal experience. We chose these three companies to attempt to show how ourtechniques can be applied to companies in various stages of growth and maturity. 在本书中,我们描绘了 AllScale 的形象,这是一家虚构的公司,融合了我们作为顾问和高管的经验,以提供现实世界的例子。作为如何使用我们的技术来创建可扩展解决方案的最后一个例子,我们我们将离开 AllScale,并讨论与我们有个人经验的三家公司。我们选择这三个公司是为了展示我们的技术如何应用于处于不同成长和成熟阶段的公司。 #####eBay: Incredible Success and a Scalability Implosion eBay:令人难以置信的成功和可扩展性的爆炸 Search for “eBay outage 1999” and you’ll get a number of results centered betweenMay and July of 1999.Outages lasting more than five hours were fairly commonaccording to widely published news stories throughout the Internet. Had these out-ages continued, there is no doubt that eBay would have lost the advantage it enjoyedthrough at least 2005 as the largest online ecommerce site in gross merchandise sales. 搜索“eBay outage 1999”,您将得到许多集中在 1999 年 5 月至 7 月期间的结果。根据互联网上广泛发布的新闻报道,持续超过 5 小时的中断是相当常见的。如果这些中断持续下去,毫无疑问,eBay 将失去至少在 2005 年作为商品销售总额最大的在线电子商务网站所享有的优势。 In 1999, Internet commerce was still in its infancy. eBay, Amazon, and Yahoo wereblazing new trails in terms of commerce transaction volume on the Internet. Facilitat-ing online auctions also had some attributes unique from other online sites: Time wasalso a very important element. Regardless of the length of an auction, the most likelytime that the auction would be viewed and bid upon was the last several hours of thatauction’s life. If eBay was down for five hours, it is very likely that certain auctionswould close at a final price significantly lower than they would if the site remainedavailable. Unlike most Internet commerce sites, where demand would be scatteredover some subset of catalog items even during peak traffic, demand on eBay auctionswould be laser focused on an even smaller set of items at any given time. Imagine1,000 buyers in a 10,000 square foot store all attempting to view and purchase anitem in the back corner of the store and for which the store only has a single copy orpiece. Further imagine that the number of buyers entering the store doubled some-times daily and sometimes monthly. The items these buyers were interested in werealso doubling, or tripling. Although not unheard of today, this type of growth simplyhad not been experienced before. 1999年,互联网商务还处于起步阶段。 eBay、亚马逊和雅虎在互联网商业交易量方面开辟了新的道路。促进在线拍卖还具有一些不同于其他在线网站的独特属性:时间也是一个非常重要的因素。无论拍卖的时间长度如何,最有可能观看拍卖并出价的时间是该拍卖生命周期的最后几个小时。如果 eBay 宕机五个小时,某些拍卖的最终价格很可能会比该网站保持可用时的价格低得多。与大多数互联网商务网站不同的是,即使在高峰流量期间,需求也会分散在目录项目的某些子集上,而 eBay 拍卖的需求在任何给定时间都将集中在甚至更小的一组项目上。想象一下,在一家 10,000 平方英尺的商店里,有 1,000 位买家都试图在商店后角查看和购买一件商品,而该商店只有一个副本或一件。进一步想象一下,进入商店的买家数量有时每天翻一番,有时每月翻一番。这些买家感兴趣的商品也增加了一倍或三倍。尽管在今天并非闻所未闻,但这种增长方式以前从未经历过。 What happened? As we’ve attempted to point out in this book, it all starts withpeople. eBay had some of the brightest engineers around in 1999 before and duringthe outages. But these incredibly bright people didn’t have some of the experiencesnecessary to run platforms at such scale and at such rapid growth. In fact, no one hadthat type of experience. So, the executive team needed to find folks with different, yetapplicable experience to augment the existing team. Meg Whitman brought in May-nard Webb as the President of Technology and Maynard in turn brought in LynnReedy and Marty Abbott. The executives established a compelling yet achievablevision to be both highly available and highly scalable. Lynn and Marty hired a num-ber of people to further augment the team over a period of several years includingTom Keeven and Mike Fisher (both now partners in AKF), creating a perpetual cycleof “seed, feed, and weed.” Experiences were added to the team and talent wasboosted. Great individual contributors and managers were recognized and promotedand some people were asked to leave or left of their own volition if they did not fitinto the new culture. 发生了什么?正如我们在本书中试图指出的那样,一切都始于人。 1999 年停电之前和停电期间,eBay 拥有一些最聪明的工程师。但这些极其聪明的人并没有运营如此规模和如此快速增长的平台所需的经验。事实上,没有人有过这样的经历。因此,执行团队需要寻找具有不同但适用经验的人员来扩充现有团队。梅格?#24800;特曼 (Meg Whitman) 任命梅纳德?#38886;伯 (May-nard Webb) 担任技术总裁,而梅纳德又任命了林恩?#37324;迪 (LynnReedy) 和马蒂?#38463;博特 (Marty Abbott)。高管们制定了一个令人信服且可实现的愿景,即高度可用且高度可扩展。林恩和马蒂在几年的时间里聘请了一些人来进一步扩充团队,其中包括汤姆?#22522;文和迈克?#36153;舍尔(两人现在都是 AKF 的合伙人),创造了一个“种子、饲料和杂草”的永久循环。团队的经验得到了增加,人才也得到了提升。伟大的个人贡献者和管理者得到了认可和晋升,一些人被要求离开,或者如果他们不适应新的文化就自愿离开。 While continuing to focus on ensuring that the teams had the right skills and expe-riences, the executives simultaneously looked at processes. Most important were theprocesses that would allow the organizations to learn over time. Crisis management,incident management, postmortem, and change management and control processeswere all added within the first week. Morning operations meetings were added tofocus on open and recurring incidents and to drive incidents and problems to closure.Project management disciplines were added to keep business and scalability relatedprojects on track. 在继续专注于确保团队拥有正确的技能和经验的同时,高管们同时关注流程。最重要的是允许组织不断学习的流程。危机管理、事件管理、事后分析以及变更管理和控制流程均在第一周内添加。添加了晨间运营会议,重点关注未决和重复发生的事件,并推动事件和问题得到解决。添加了项目管理规则,以保持与业务和可扩展性相关的项目步入正轨。 And of course there was the focus on technology! It is important to understandthat although people, process, and technology were all simultaneously focused on,the most important aspects for long-term growth stem from people first and processsecond. As we’ve said time and time again, technology does not get better withouthaving the right team with the right experiences, and people do not learn without theright (and appropriately sized) processes to reinforce lessons learned, thereby keepingissues from happening repeatedly. 当然,还有对技术的关注!重要的是要理解,尽管人、流程和技术都同时受到关注,但长期增长最重要的方面源于人第一,流程第二。正如我们一再说过的,如果没有拥有正确经验的正确团队,技术就不会变得更好,如果没有正确(且规模适当)的流程来强化经验教训,人们就无法学习,从而防止问题重复发生。 Databases and applications were split on the x-, y-, and z-axes of scale. Whatstarted out as one monolithic database on the largest server available at the time wasnecessarily split to allow for the system to scale to user demand. Data elements withhigh read to write ratios were replicated using x-axis techniques. Customer informa-tion was split from product information, product information was split into severaldatabases, and certain functions like “feedback” were split into their own systemsover the period of a few years. 数据库和应用程序在 x、y 和 z 轴上进行划分。一开始是当时最大的服务器上的一个整体数据库,必须进行拆分,以便系统能够根据用户需求进行扩展。使用 x 轴技术复制具有高读写比率的数据元素。几年时间里,客户信息从产品信息中拆分出来,产品信息拆分到多个数据库中,某些功能(例如“反馈”)拆分到自己的系统中。 #####Quigo: A Young Product with a Scalability Problem Quigo:一款存在可扩展性问题的年轻产品 Quigo started out as a company offering a service based on technology. Relying on aproprietary relevance and learning engine, its first product promised to help increasethe returns in the nascent search engine marketing industry for direct response adver-tisers. Leveraging this existing technology, the company branched out into offering aprivate label contextual advertising platform for premium publishers. AdSonar wasborn. Early premium branded publishers loved the product and loved the capabilityto increase their revenue per page over the existing alternatives. Quigo 最初是一家提供基于技术的服务的公司。依靠专有的相关性和学习引擎,其第一个产品有望帮助直接响应广告商提高新兴搜索引擎营销行业的回报。利用现有技术,该公司涉足为优质出版商提供自有品牌上下文广告平台。 AdSonar 诞生了。早期的优质品牌出版商喜欢该产品,并且喜欢与现有替代品相比增加每页收入的能力。 However, within months, the new advertising platform had problems. It simplycouldn’t handle the demand of the new publishers. How did a new product fail soquickly? The product wasn’t anywhere near the scale of an eBay, Amazon, or Google;it wasn’t even near the scale of competing ad networks. What went wrong and howcould it be fixed? 然而,几个月后,新的广告平台就出现了问题。它根本无法满足新出版商的需求。一个新产品怎么这么快就失败了?该产品的规模远不及 eBay、亚马逊或谷歌;甚至还远不及竞争广告网络的规模。出了什么问题以及如何修复? Again, the answer starts with people. The existing team was smart and dedicated,just as with the eBay team. But it missed experience in large-scale operations anddesigning platforms for hyper growth. This is when two future AKF Partners werebrought onboard. The new executives didn’t have direct experience with advertisingtechnology, but their experience with commerce and payment platforms was directlyapplicable. More importantly, they knew how to focus teams on an objective andhow to create a culture that would support the needs of a highly scalable site. Consis-tent with the layout of this book, it all starts with people. The new team created met-rics and goals supporting availability, scalability, and cost. It created a compellingvision of the ideal future and gave the team hope that it could be achieved. Wherenecessary, it added great engineering and managerial talent. 再说一遍,答案首先要从人开始。现有团队聪明且敬业,就像 eBay 团队一样。但它错过了大规模运营和设计超增长平台的经验。就在这时,两位未来的 AKF 合作伙伴加入了。新高管没有广告技术方面的直接经验,但他们在商业和支付平台方面的经验可以直接应用。更重要的是,他们知道如何让团队专注于一个目标,以及如何创建一种支持高度可扩展站点需求的文化。与本书的布局一致,一切从人开始。新团队创建了支持可用性、可扩展性和成本的指标和目标。它创造了一个令人信服的理想未来愿景,并给了团队实现这一目标的希望。必要时,它增加了优秀的工程和管理人才。 The new executives also set about adding the right processes to support scalability.Scalability summits, operations meetings, incident management processes, andchange management processes were all added within a couple of weeks. Joint Appli-cation Design and Architecture Review Boards soon followed. Architectural princi-ples focusing the team on the critical elements of scale were introduced and usedduring Architectural Review Boards. 新高管还着手添加正确的流程来支持可扩展性。可扩展性峰会、运营会议、事件管理流程和变更管理流程均在几周内添加完毕。联合应用程序设计和架构审查委员会很快就成立了。在架构审查委员会期间引入并使用了使团队专注于规模关键要素的架构原则。 And of course the team focused on technology. Again, what ultimately became theAKF Scale Cube was employed to split services, resources, and (where necessary)data elements. Fault isolation was employed where possible to increase scalability. 当然,团队专注于技术。同样,最终成为 AKF Scale Cube 的东西被用来分割服务、资源和(如果需要)数据元素。尽可能采用故障隔离来提高可扩展性。 What were the results of all of this work? Within two years, the company hadgrown more than 100x in transactions and revenue and was successfully sold to AOL. 所有这些工作的结果是什么?两年内,该公司的交易量和收入增长了 100 倍以上,并成功出售给 AOL。 #####ShareThis: A Startup Story 分享这个:一个创业故事 ShareThis is a company that is all about sharing. Its products allow people to easilyshare the things they find online, by consolidating address books and friend lists, sothat anything can be shared immediately, without even leaving a Web page. Withinsix months of launching the ShareThis widget, there were already over 30,000 pub-lishers using it. Witnessing this hyper growth, the cofounder and CEO Tim Schigelmet with the AKF Partners to talk about guidance with scalability concerns. Tim is aseasoned veteran of startups having seen them for more than a decade as a venturecapitalist and was well aware of the need to address scalability early and from aholistic approach. Michael Fisher from AKF Partners worked with Tim to lay out ashort- and long-term plan for scalability. At the top of the list was filling some openpositions on his team with great people. One of these key hires was Nanda Kishore asthe chief technology officer. Prior to ShareThis, Nanda was a general manager atAmazon.com and knew firsthand about how to hire, lead, design, and develop scal-able organizations, processes, and products. ShareThis是一家注重分享的公司。其产品通过整合地址簿和朋友列表,让人们可以轻松地共享他们在网上找到的东西,这样任何东西都可以立即共享,甚至无需离开网页。在推出 ShareThis 小部件后的六个月内,已经有超过 30,000 家出版商使用它。见证了这种高速增长,联合创始人兼首席执行官 Tim Schigelmet 与 AKF Partners 讨论了可扩展性问题的指导。 Tim 是一位经验丰富的初创公司资深人士,十多年来一直以风险投资家的身份看待这些公司,并且非常清楚需要尽早解决可扩展性问题并采取整体方法。 AKF Partners 的 Michael Fisher 与 Tim 合作制定了短期和长期的可扩展性计划。最重要的是用优秀的人才来填补他团队中的一些空缺职位。其中一位关键雇员是担任首席技术官的南达?#22522;肖尔 (Nanda Kishore)。在加入 ShareThis 之前,Nanda 是 Amazon.com 的总经理,对如何招聘、领导、设计和开发可扩展的组织、流程和产品有着第一手的了解。 In addition to other key hires in operations, engineering, product management,and the data warehouse team, there was a dedicated focus on improving processes.Some of the processes that were put in place within the first few weeks were sourcecode control, on-call procedures, bug tracking, and product councils. After peopleand process were firmly established, they could properly address scalability withinthe technology. 除了运营、工程、产品管理和数据仓库团队中的其他关键员工外,还专门致力于改进流程。在最初几周内实施的一些流程包括源代码控制、随叫随到程序、错误跟踪和产品委员会。在人员和流程牢固建立后,他们可以正确解决技术内的可扩展性问题。 With a keen focus on managing cost and improving performance, the teamworked on reducing the widget payload. It implemented a content delivery network(CDN) solution for caching and moved all serving and data processing into Amazon’sEC2 cloud. Because of the ShareThis architecture and need for large amounts of com-pute processing for data, this combination of a CDN and public cloud worked excep-tionally well. Under Nanda’s leadership, the team reduced the serving cost by morethan 56% while experiencing growth rates in excess of 15% per month. All of thissharing activity resulted in terabytes of data that needs to be processed daily. Theteam has produced a data warehousing solution that can scale with the ever increas-ing amount of data while reducing the processing time by 1900% in the past sixmonths. 该团队专注于管理成本和提高性能,致力于减少小部件的有效负载。它实施了用于缓存的内容交付网络 (CDN) 解决方案,并将所有服务和数据处理转移到 Amazon 的 EC2 云中。由于 ShareThis 架构以及对大量数据计算处理的需求,CDN 和公共云的这种组合运行得非常好。在 Nanda 的领导下,该团队将服务成本降低了 56% 以上,同时每月增长率超过 15%。所有这些共享活动都会产生每天需要处理的数 TB 数据。该团队开发了一种数据仓库解决方案,可以随着数据量的不断增加而扩展,同时在过去六个月内将处理时间减少 1900%。 Less than two years after the launch, the ShareThis widget reached more than 200million unique users per month and more than 120,000 publisher sites. ShareThis is ascalability success story because of its focus on people, process, and technology. 推出后不到两年,ShareThis 小部件每月覆盖超过 2 亿唯一用户和超过 120,000 个发布商网站。分享这是可扩展性的成功故事,因为它关注人员、流程和技术。 Again, it’s worth repeating a recurring theme throughout this book: You can’tscale without focusing on all three elements of people, process, and technology. Toomany books and Web sites feature the flavor of the day technical implementation tofix all needs. Vision, mission, culture, team composition, and focus are the mostimportant elements to long-term success. Processes need to support the developmentof the team and need to reinforce lessons learned as well as rapid learning. Technol-ogy, it turns out, is the easiest piece of the puzzle, but unfortunately the one peopletend to focus on first. Just as with complex math equations, one simply needs to iter-atively simplify the equation until the component parts are easy to solve. 同样,值得在本书中重复一个反复出现的主题:如果不关注人员、流程和技术这三个要素,就无法扩展。许多书籍和网站都以当今的技术实现为特色来解决所有需求。愿景、使命、文化、团队组成和重点是长期成功的最重要要素。流程需要支持团队的发展,需要强化经验教训以及快速学习。事实证明,技术是这个难题中最简单的一块,但不幸的是,人们往往首先关注它。就像复杂的数学方程一样,我们只需要迭代地简化方程,直到各个组成部分易于求解为止。 People and organizations are more dynamic and demanding. Although there is nosingle right solution for them, there is an approach that is guaranteed to work everytime. Start with a compelling vision mixed with compassion and hope, and treat yourorganization as you would your garden. Add in goals and measurements and help theteam overcome obstacles. 人和组织更加充满活力和要求更高。尽管没有唯一正确的解决方案,但有一种方法保证每次都能奏效。从一个充满同情心和希望的令人信服的愿景开始,像对待花园一样对待你的组织。添加目标和衡量标准并帮助团队克服障碍。 Process development should focus on those things that help a company learn overtime and avoid repeating mistakes. Use process to help manage risks and repeat supe-rior results. Avoid process that becomes cumbersome or significantly slows downproduct development. 流程开发应该关注那些能够帮助公司不断学习并避免重复错误的事情。使用流程帮助管理风险并重复卓越的结果。避免流程变得繁琐或显着减慢产品开发速度。
没有评论