开发者学堂课程【深度解读服务网格的未来:深度解读服务网格的未来】学习笔记,与课程紧密联系,让用户快速学习知识。
课程地址:https://developer.aliyun.com/learning/course/966/detail/14887
深度解读服务网格的未来
In Conversation(对话)
Our Service Mesh is dedicated structural layer for making service communications safe, fast and reliable. Today we have Jeffery Harmand, vice president, principal analyst of Forster and the Xining Wang, senior technical expert who is in charge of ASM product; Dongsheng Li, who is chief architect from Dongfeng data Marketing Company. According to the sincere service of the last year, 27% of enterprises are using Service Mesh production and 23% are also evaluating Service Mesh. It is obvious that this number will continue to rise as enterprises continue to expand their micro service. So Jeffrey, according to you why Service Mesh is needed and what are the different values?
翻译:服务网格是一种专用结构层使得服务之间的通信安全快速和可靠。今天我们邀请到了Jeffery Harmand- Forrester公司的副总裁,兼首席分析师王夕宁,负责阿里云服务网格产品的高级技术专家以及李东生,东风目产数据服务有限公司的首席架构师。根据去年的诚信服务,27%的企业在使用服务网格产品,23%的企业还在对服务网格进行评估,很显然,由于企业不断扩展微服务,这一数字将继续上升。Jeffrey您认为为什么需要使用服务网格以及它有哪些不同的价值?
探讨软件平台的未来(Leap into the future of software platforms)
服务网格能否改变未来?(Can Service Mesh Transform the Future?)
探讨软件平台的未来
Leap into the future of software platforms
Jeffrey Hammond: So Service Mesh is a cloud, native technology and transition in my opinion. this is the year I think it really gets over the hump. For a few years, we’ve seen somewhat slow adoption. mainly because enterprise organizations weren’t sure what problem is solved and whether or not they had those problems. Service Mesh can be complex to roll out and maintain, especially if you’re trying to do it on your own instead of as a managed service. Teams can configure service meshes and married distinct styles, patterns and scopes of functionality depending on their needs, and the preferences of micro service development teams and their enterprises, so, to effectively plan and implement the service mesh, developers and solution architects must have a strong understanding of how it removes complexity in their environment. Now, what we’re seeing is that as organizations start to scale up their cloud native efforts, and in particular, as they begin to spend more time modernizing their existing applications, the benefits of service meshes are starting to become more apparent to enterprises. Some of the most compelling reasons that we see organizations adopting the service mesh include integration, especially if they are trying to integrate cloud native applications and micro services to traditional application constructs. So for example, I have a container and it needs to talk to virtual machine. I need to go from a cluster that’s in the public cloud to an application that’s running on Plumas. Securing communications between micro services. So as you start to break a monolithic application into a lot of independent services, it really increase it’s a test service. Every micro service is a potential entry point to protect from nefarious axis by bad actors. And so as service mesh is really good for securing those micro services, when health care technology from that we talked to, was building micro service and was very concerned about meeting their complaints requirements and they felt that implementing the service mesh was a good way to make sure that they would remaining complacent as they now announced. Traffic management is another reason that a service mesh can be very helpful and as more development organizations start to look at doing modern practices like multivariate testing, doing chaos testing and production, doing rolling updates of micro service. The benefits of a service mesh as enabling that and making it easier to do become more apparent. So the ability choose from multiple models for how they update those services and the complexity, and manage the complexity of those is really important. So, another use of service mesh is really good capabilities to identify what’s happening when a micro service misbehaves in the run time environment. When you’re got dozens or, even hundred of micro services running, there are many individual processes that need production monitoring and service level management to make sure that you’re meeting your availability in performance goals. So knowing exactly what’s going on is something that a service mesh can do by using the sidecar model in the Kubernetes world to get that telemetry and make it available. It can help developers understand what’s going on.
翻译:
我认为服务网格是一种云原生技术和中转站,它在今年摆脱了困境。几年来,我们可以看到它的推进有一些缓慢,这主要是因为企业组织不太确定它所解决的问题以及他们是否具有这方面的问题。服务网格的铺设和维护可能会很复杂,尤其是在您不采用托管服务,而是自行尝试进行铺设和维护的情况下,团队可以根据他们的需求,微服务开发团队及其企业的偏好,来配置服务网格,结合不同的方式、功能类型和功能范围。
因此,要有效地规划和实现服务网格,开发人员和解决方案架构师,必须真正地了解它如何在环境中消除复杂性。现在我们看到的是组织开始加大他们在云原生方面的投入,特别是他们开始花更多的时间,使现有的应用程序现代化,服务网格的优势开始对企业变得更加明显,我们看到组织采用服务网格的一些主要原因,包括“集成”,尤其是当他们试图将云原生应用程序和微服务集成到传统的应用程序结构中时,比如说,我有一个容器,它需要与虚拟机对话,我需要从公共云的一个集群,转移到在Plumas上运行的应用程序,以及实现微服务之间的安全通信。因此,当您开始将单独的应用程序分解成许多单独的服务时,这在实际上改善了测试服务,每项微服务都是能够防止危险分子恶意访问的潜在入口点,因此,由于服务网格非常有益于这些微服务的安全,当我们所讨论的卫生保健技术在构建微服务以及非常担心能否满足投诉需求时,他们认为采用服务网格有助于确保他们不会像现在宣布的那般沾沾自喜,流量管理是服务网格非常有用的另一个原因,由于更多的开发者开始寻求展开多变量测试,混沌测试和生产,滚动更新微服务等现代实践,服务网格能够促成这些实践,并使更易于实现这些实践的益处变得更加明显。因此从多个模型中选择如何更新这些服务,并管理这些服务的复杂性能力非常重要。故而,服务网格的另一种用途是识别微服务,在运行时环境中行为异常时发生的情况,当您有数十项甚至数百项微服务在运行时,需要对许多单独的流程,进行生产监测和服务水平管理,以确保您能够满足性能目标。因此,服务网格可以使用Kubernetes社区中的sidecar模型,来获得遥测数据并使其可用,来准确地了解正在发生的情况,它可以帮助开发人员了解正在发生的情况。
Ok, thank you. And Kubernetes is becoming a cloud operation system that supports distributed application while the industry generally believes that the service mesh is becoming a new batter ground in platform wars. So, where do you stand on the saying that we still need service mesh, when we already have Kubernetes, Jeffery?
翻译:好的 谢谢。Kubernetes正在成为能够支持分布式应用程序的云操作系统,行业普遍认为服务网格正在成为平台战的新战场。所以,当我们在已经可以使用Kubernetes的情况下,您对我们仍然需要使用服务网格怎么看呢Jeffery?
Jeffrey Hammond: I think you need both. To me, it’s kind of like saying I’ve got a really good hammer. So why do I need a screw driver? For most carpenters, the answer is: Yes, you need both because they do complementary things. So I don’t see a time in the next decade when most businesses won’t need to connect to thing besides containers. Bs are going to be around for a long time. I think there’s more development teams discover the need to test in production. They’re going to look for tools that give them fine grain traffic control. Doesn’t necessarily mean that lstio will be the service message choice. Though 2 years ago it looked like that’s the course that we were on, but Google got a bit squishy on the governance model for lstio and it caused enterprises, and I used to take a second look. And developers are staring to look at thing like how lstio comparison Linkerd other commercial offerings which of these can best meet my needs. And the services provided by service mesh can potentially be provided by other technologies too. My teammate David Murder recently wrote a blog post about Dapper and asked if it were a better alternative to the needs developers have compared to a service mesh. If you’re not familiar with Dapper, it’s a coding framework intended to solve the challenges of modern distributed applications. While a service mesh focuses on networking infrastructure concerns. Dapper focuses on what developers need to build micro service. So, I think this is an exciting area. I think service mesh is important. It’s a key tool. It’s a tool that we are going to use, lstio will very probably feature in that and be part of the solution, it may not be the exclusive solution. I think what we’ll have to do is to continue to watch what developers build here. I can tell you though, I think that any solution that makes it easier for developers to quickly build, deploy and operate micro service at scale while abstracting them from lower level, networking concerns is gonna prove popular, especially if it’s easy to install, run, managing use. So that’s why I think that hosted service mesh capabilities will be increasingly popular because it will make them easier to inject into enterprise organizations.
翻译:我认为我们这两种都需要,对我来说,这就像是我已经有了一把好锤子,那为什么我还需要螺丝刀呢?对于大多数木工来说答案是 是的。这两种都需要,因为它们是互补的。因此,我认为在未来十年内大多数企业都不需要连接到容器以外的东西,这种现象将会存在很长一段时间,我认为有更多地开发团队发现在生产过程中进行测试的必要,他们将寻找能够更好地控制流量的工具,但这并不一定意味着lstio将是服务消息的选择。虽然两年前,看起来这是我们正在努力的方向,但谷歌在lstio的管理模式上不太果断,造成企业选择观望。此外,开发人员开始关注lstio相比Linkerd和其他商业产品,如何最好地满足需求,服务网格提供的服务可能其他技术也可以提供,我的队友David Murder最近写了一篇关于Dapper的博客文章,问到相对服务网格Dapper是否符合开发人员需求的更好的替代方案,如果您还不太熟悉Dapper,它是旨在用于解决现代分布式应用程序的挑战的编码框架,服务网格侧重网络基础架构方面的问题,而Dapper则侧重开发人员在构建微服务方面所需的内容,因此,我认为这是一个令人兴奋的领域。我认为服务网格非常重要,它是一个关键工具,一个我们将使用的工具,lstio将非常有可能侧重这一方面,并将其视为解决方案的一部分,它可能不是唯一的解决方案。我认为我们要做的就是继续观察开发人员在这里构建的内容,不过,我可以说,我认为任何能使开发人员更容易地大规模地快捷构建部署和操作微服务,同时将这些微服努从较低级別的联网问题中抽象出来的解决方案都会很受欢迎,尤其是在它易于安装运行和管理使用的情况下,所以这就是为什么我认为托管式服务网格能力会越来越受欢迎,正是因为这将使这些能力更容易注入到企业组织。
Ok, Thank you. And how about you? Xinning.
翻译:好的,谢谢,您怎么看呢,夕宁。
Xinning Wang:
The major cloud vendors already provide the service mesh products and solution offers,for example last year Alibaba cloud had launched one managed service mesh product, named Alibaba Cloud Service Mesh, for short we call it ASM. ASM not only can help our customers lower the barrier to use, but also can provide a lot of enterprise-level capabilities, including enterprise security, scalability, observability, 24 and 7 online professional production, support and so on. Today I want to share one customer case with you to show how the customer use Alibaba Cloud Service Mesh product in their business applications. Let’s welcome Dongsheng li, who is Chief Architect from Dongfeng Nissan Data Marketing Co,Ltd. Hello, Dongsheng, would you like to have a brief about the background why you use service mesh in your business applications and what paint points have been resolved by service mesh technology?
翻译:主要的云供应商已经提供了服务网格产品和解决方案,比如说,在去年阿里云推出了一种托管式服务网格产品,名为“阿里云服务网格”。英文简称“ASM”。阿里云服务网格不仅能够帮助降低我们客户使用产品的障碍,而且能够提供许多企业级能力,包括企业安全性,可扩展性,可观测性,全天无休的在线专业生产和支持等。今天,我想和大家分享一个客户案例,来看看用户是如何在业务应用程序中使用阿里云的服务网格产品的,让我们欢迎李冬生,他是东风日产数据服务有限公司的,首席架构师,冬生您好,您能不能简单说一下您在业务应用程序中,使用服务网格的背景原因,以及您通过服务网格技术,解决了哪些痛点?
Dongsheng Li(李冬生):Okay. As a data service provider for the entire value chain of automobiles, we are providing digital service, sales service, customer service, operation service, omni-channel data value mining analysis and application for Dongfeng Nissan and other brands. With the development of business, the 12 complete test environments created early can no longer meets many concurrent needs. Therefore, the R&D department urgently needs to upgrade the 12 environments to the Infinite and Autoscaling environments. so that, in addition to automatically providing and development on demand, it also needs incremental deployment to save server resources. In the realization, one of the key points is the needs for refined management of layer 7 traffic, which can be forwarded according to the design situation, such as Header, cookies. Hitting the application corresponding to the demand or hitting the default application. This can not only solve environmental problems, but also reduce service resource cost.
翻译:好的。东风日产作为一家数据服务提供商,它在汽车的整个价值链中,为公司自身和其他汽车品牌提供针对全渠道数据价值,挖掘分析和应用程序方面的数字服务,销售服务,客户服务和运营服务。随着业务的发展,早期创建的十二生肖环境已不能满足许多并发需求因此研发部门迫切需要升级该环境以实现自动缩放的无限环境。因而除自动提供服务和按需发展外,还需要其他部署以节省服务器资源。在实现过程中,其中一个关键点是需要对第7层的流量进行精细化管理,这可以根据标头和cookie等设计环境进行转换,应用于符合需求的应用程序或者默认程序。这不但可以解决环境方面的问题,还可以减少服务器成本。
Xinning Wang :Yes, very nice. Would you like to share your service mesh journey with us.
翻译:是的,非常好。您能不能和大家说一下您使用服务网格的过程。
Dongsheng Li(李冬生):
Before we talking about anything about service mash in deeply. I want to explain how we need a service mesh. At the DNDC, we run the hundreds of micro service data do everything from the best customer experience. Running micro service at scale is not without its own challenges. When we compared to last year, our micro service size is repeatedly increasing. So, it become increasingly difficult to understand the interactions between all these service. When a problem occurs in a micro service world it can be really difficult to find where the problem is. Service mesh addresses these challenges.
翻译:
在深入讨论服务网格之前,我想说我们为什么需要使用服务网格,在东风日产,我们运行数百项微服务的数据,竭尽所能为客户提供最佳体验,大规模运行微服务并非没有挑战,与去年相比,我们的微服务规模在不断增加。因此,理解所有这些服务之间的交互变得越来越困难。在这些微服务出现问题时,我们很难找到问题所在,而服务网格能够解决这些挑战。
Xinning Wang : Yeah, when you decided that you are ready for a service mesh I think there are still several options to choose. For example, there are a lot to of service mesh products in the software world. For example, lstio, Linkerd, Linkerd2, Consul Connect and so on. What’s your decision about this?
翻译:好的,在您解决要使用服务网格的时候,我想您任然还有几个产品可供选择,比如说,软件行业有许多的服务网格产品。例如lstio, Linkerd, Linkerd2, Consul Connect等等。那么您是怎么作出选择的呢?
Dongsheng Li(李冬生):
Much like Kubernetes has become the de facto container orchestration standard, lstio is often seems as the de facto service mesh standard, and lstio is also the most popular service mesh from the recent surveys by CNCF. We made some of benchmark tests to show latency between our micro services communication, lstio and other service mesh products are nearly the same. In our use case, we decide to use lstio. Why are we don’t directly use open source community tools and we know that Alibaba cloud has launch one product named Alibaba Cloud Service Mesh, for short ASM.
翻译:就像Kubernetes已经成为有真正的容器编排标准一样,lstio也经常被视为实际的服务网格标准,且lstio还是云原生计算基金(CNCF)最近调差中,最受欢迎的服务网格,我们进行了一些基准测试,结果显示我们的微服务通信lstio和其他服务网格产品的延迟几乎是一样的,在我们的用例中,我们决定使用lstio。为什么我们不直接使用开源社区工具,我们知道阿里云推出了名为“阿里云服务网格”的产品,简称“ASM”。
Xinning Wang : So, why you finally choose Alibaba Cloud Service Mesh for your business applications?
翻译:那您为什么在业务应用程序中最终选择,我们阿里云服务网格呢?
Dongsheng Li(李冬生):
As the industry’s first fully managed lstio-compatible service mesh product, ASM has maintained consistency with the community and industry trends from the beginning. The components of the control plane are hosted on Alibaba Cloud side, and are independent of the user side's cluster on the data plane. By this managed mode, the life cycle management of the control plane components, and the data plane's K8s cluster is decoupled, making the architecture more flexible and improving the scalability of the system. only helps us lower the barrier to use, but also provides some enterprise level capabilities such as security, scalability, observability and 24/7 production support. ASM help us shorten the service mesh technology landing cycle, shortened by75%. The operate iteration efficiency is increased by 70%, the abnormal troubleshooting cost is reduced by 80%, and the control plane resource cost is reduced by 100%. Xining.
翻译:作为行业首个全面管理工具lstio的兼职服务网格产品,阿里云服务网格从一开始,就与社区和行业趋势保持了一致,控制平面组件由阿里云负责托管并在数据平面上,独立于用户规模集群。通过这种管理模式控制平面组件的生命周期管理,以及数据平面K8集群得以分离,使得架构更加灵活,继而提高了系统的可扩展性。阿里云服务网格不仅帮助我们降低了使用障碍还提供了一些企业级功能,比如说安全性可扩展性可观测性和全天无休的生产支持,它帮助我们缩短了服务网格技术的落地周期缩短了75%,运行迭代效率提高了70%,异常故障排除成本降低了80%,且控制平面资源成本降低了100%。夕宁。
Xinning Wang : Yes, as an important link between the upper-layer applications and the lower-layer computing infrastructure. ASM can provide the following capabilities from three perspectives. Firstly, from the perspective of a downward integration with infrastructure. ASM can be thought as Anywhere Service Mesh. Its goal is to enable service mesh capabilities for application services running on any computer infrastructure. Secondly, from the perspective of service mesh’s built-in feature enhancement, ASM can provide the unified traffic control, unified security protection and unified observability to our application services without most code change. Last one, from the perspective of supporting the application layer and the ability to be integrated upwards, for example, ASM can provide the easy-to-use ApIs for integration and also provide the declarative APIs, while are fully compatible with Istio community. ASM, fully leverage the advantage of Alibaba Cloud instruction to build the multi-regions and the multi-zones topology and build one delicated support team to guarantee the stability of managed service. Due to the limited time, I will close my topic here. Please visit the website of ASM products and go though the features and use case to understand what ASM can do for you. We appreciate our customer to share their service mesh adoption experience. Hi Dongsheng, thanks a lot for your time. See you later. Thank you.
翻译:是的,阿里云服务网格,作为上层应用程序和下层计算机基础之间放入重要连接。它从三个角度提供以下功能,首先,从向下基础整合来看,ASM可视为任何场所的服务网格,它的目标是为在任何计算机上运行的,应用服务,启用服务网格能力。其次,从服务网格内置特征增强来看, ASM可以为我们的应用服务提供统一的流量控制功能,统一的安全保护功能和统一的可观测性,而几乎无需修改代码。最后,从支持应用层和向上集成能力的角度,比如说,阿里云服务网格可以提供易于使用的API进行集成,也可以提供与Istio完全兼容的,声明式API,阿里云服务网格充分利用阿里云指令的优势,构建了多区域拓扑,并组建了一支专门的支持团队,保证托管服务的稳定运行。由于时间有限,我就说到这里了。请访问阿里云服务网格产品的网站,了解相关的功能和用例,如需了解阿里云服务网格能为您提供哪些服务,我们感谢我们的客户能够分享他们的服务网格使用经验,好的,冬生,非常的感谢。回见,谢谢。
Thank you everyone.
翻译:谢谢大家。