클라우드 이점 극대화 비법! 클라우드 네이티브 완벽 가이드

5 [Applause] 5 we [Music] We face various challenges when adopting the cloud.

  When moving to the cloud, physical servers disappear, so how will the existing operational staff's roles change in the cloud?   How will the on-premise environment and the cloud be connected to expand services?   How will communication between the increasingly emerging development and operation organizations be efficiently managed? Will  existing developed apps be usable in the cloud environment? Will there be difficulties in learning new tools? Are internal personnel prepared? If not, how will they be recruited or developed? There are many people who have concerns when considering the cloud.

 If we  organize these various considerations from the perspective of development and operation, they are as follows.

  First, when considering the infrastructure, you must choose whether to configure it with a single cloud or a multi-cloud structure that uses multiple cloud providers.

 The  biggest reason for such concerns is to  continue business even in the event of a failure.

   If you use a single cloud provider, you can safely apply services by distributing them across multiple regions rather than just one region.

   Yes, even so, the reason for choosing a multi-cloud is to safely operate it across multiple companies in case a problem occurs with the entire cloud provider service.

    Also, financial institutions or In cases where key services, such as public institutions, must be  used without connecting to the Internet, a private cloud must be built.

 In cases where Prime introduces the cloud,   not all services available in the public cloud are provided due to limited resources, so it must check what services are actually needed and configure the infrastructure.

 In the case of applications, it must be decided whether to use the application architecture structure used in the on-frame environment as it is or to  reconfigure it to fit the cloud environment.

 In the  case of development, it is important to create an automated deployment process so that the developed source code is  distributed from the developer's computer to the on-premises server at once.

 This is   because code updates are made dozens or even hundreds of times a day, and if all of this work is done manually, it consumes a lot of time and energy.

 In addition, a   solution that helps the development organization and the operation organization communicate efficiently must be selected.

 And above all, it is necessary to consider introducing 15 degrees to check the entire project at a glance, what problems there are, and whether the development progress is progressing according to schedule.

   [Music] my A new development operation method suitable for the cloud environment  We call this cloud native.

 In  other words, cloud native 1 It means a way to build and run applications so that the benefits of the cloud can be maximized.

   With the introduction of cloud services, compared to The introduction of optimized cloud native technology and working methods has also become essential.

 [Music]  In order to implement cloud native, it is  essential to understand the concept of app ops, which moves like a single entity without any division between the development organization and the operation organization, and the     methodology that automates the entire structure from source development to the actual service as one process.

  And when we actually operate the service, we need technology for a stable infrastructure that can be freely expanded to the multi-color of and on-place environments that are attracting attention these days without worrying about the infrastructure.

     That is exactly the technology of Funten and the microservice architecture.

 To understand cloud native, you need to learn about the infrastructure operating environment before the cloud.

  Before the cloud, we installed the operating system directly on the only server and  installed and used the necessary applications.

 As the service increased, we continued to expand the service in our dreams using the physical server.

  However, when installing and using only one operating system on the physical equipment, the  usage was low during normal times, so it was a waste of resources.

  And when using a physical server, it   often took at least several weeks and up to several months from disassembly to actual installation.

 So the technology that came out was server virtualization technology.

 With server virtualization technology, you can  use one physical server as 10 or 20 virtual servers.

  To virtualize a physical server, you install an operating system on the physical server and   use software called hypervisor to create and run the desired operating system on a virtual machine.

 With this server virtualization technology, you can use physical servers efficiently.

  And you do n't have to wait several weeks to add one server, you can use it within a few hours.

 The  cloud is all about cultivating virtual servers.

 In the early stages of cloud adoption, companies are not yet familiar with the cloud, so they want to transfer their on-premise environment to the cloud as is.

  However, virtual servers only run on the same hypervisor that created the virtual server, so they do not run on other companies' hypervisors.

 For example, if you move a service running on VMware's Piper Visor to  Microsoft's Hyper-V environment, it will not run.

 If you want to operate an existing virtual server on another hypervisor environment, you need to convert the virtual server to fit the hypervisor you want to transfer it to.

   Since all cloud companies use different hypervisors, you  need a different method to implement a flexible cloud-native environment.

  Not only will you not be dependent on the cloud company, Above all, the technology that enables a flexible cloud native environment that can be configured and operated quickly is the   container.

 Containers can be used in any cloud as long as the container engine that runs the container is the same.

  This is the technology that is absolutely necessary for multi-cloud cloud native.

 And after the infrastructure is prepared based on containers, methods  such as Daebeop Ssuk CI era microarchitecture are used for rapid application deployment.

 [Music] There  are four key elements of cloud native.

 Isn't it a container house? Microservices are the key elements.

 Here, we will look at only the core concepts of each element.

  [Music] The first container is the concept of operating system virtualization.

 As  mentioned earlier, we increased the efficiency of physical servers through server virtualization, but as we use virtual servers, we want to use virtual servers more efficiently.

  However, to use virtual servers efficiently, we  install hypervisors and install virtual servers in virtual servers, which results in a huge waste.

 So now we will no longer virtualize servers and make applications running on virtual servers operate like virtual servers.

 This is the container.

 If we used a  hypervisor to create a virtual server,  we used a joker to create a container.

 It uses a container engine.

  Since containers are not an operating system but a single executable file, they run much faster than virtual servers.

  And since containers can run anywhere where Docker is installed,   unlike virtual servers that depend on hypervisors, they can run in any cloud environment without any special changes.

 Of course, they can also run in on-premise environments.

 The  second MSA is a business  microservice blog that divides existing single applications into small pieces that can be executed, and communicates with each other through an API that transmits service requests in the middle,  so they can be distributed independently.

      You can think of an API as a service that transmits data between the person who requested the service and the final application, like a delivery service that connects restaurants and consumers when we order food delivery.

 [Music] The reason for using MSA is to reduce the inefficiency of stopping and restarting the entire service when trying to update a  single detailed function when implementing all services of a single application, and to     prevent the entire service from being stopped due to an individual service failure.

 And the third, Mr.

 Daebok, although there are various definitions, can be simply defined as follows.

 It is a     method of seamlessly connecting the processes between unified development and operation using the Great Law and maximizing efficiency through automation methods.

 You can  combine the work of the development team and the operation team to speed up development, and it     is a business innovation method that includes increasing efficiency through automation tools and reducing communication costs that are lacking in project schedule management.

 The last fourth is CD in Sia.

 CD in Sia can be seen as a series of processes to implement the philosophy of a large company.

 The   period of Sia is a term that combines CI continuous interview, that is, continuous integration, and CD continuous deployment.

 First of all, CI continuous integration.

 When a development project is carried out, it is rare for one developer to develop everything, and in most cases, several people develop their own functions and merge the source code.

 Then, the merged codes are packaged in a deployable state.

  Since this process occurs continuously, we call it continuous integration.

  Hit This means continuously deploying the deployable package created through the CI process  to the actual development or production environment.

  Hit in Sia This process includes not only deployment in application development but also continuous monitoring and testing.

  222 Proud Pun How to do it is based on the definition of lipstick, which best expresses the essence of the cloud.

 The   five features are: public disclosure of firepower, resources, self-operation, and pay-as-you-go system, level 1.

 Although they  show the special features well, security is difficult for schools.

 The  first feature, elasticity, means that the target of security can be changed.

  In the vent race, you only need to check the corrected assets, but you   need to track when a new source of the cloud line is created or deleted to waterproof the secret hole.

  Because of these characteristics of the cloud, represented by auto auto scaling, cloud security must also have elastic characteristics such as automated resource distance books, as well as the cloud phone itself, and agent Lee's Boa solution, which automatically distributes and blocks the cloud.

 The  feature of Roy network access is that you can manage the cloud anytime and anywhere, so you  can access it anywhere under the same conditions as hackers.

 If it were on-face, we could target a closed network and block external access at the source, but in the cloud, both me and the hiker can live with administrator rights under the same conditions, so you need to do some  revisions and see various things.

 The  most frequent cause of cloud security accidents is this security setting model.

 Because of the rayon p day, if it came later, it would be because the cloud virtualizes our long-term and  allows an unspecified number of women to use it at any time and bear 1.

  However, in the end, since the same water shares the equipment and network path,  encrypted data transmission and storage are essential in case a problem with the cell security occurs.

 The cold 0 oriental fee makes a difference because there is a technically on-frame behind it.

 The  possibility of self-operation means that the pre-review and approval process in the ombre-miss environment is no longer booming, and that is why a new world like Daewoo 3 is needed.

 The  ability to pay on a pay-as-you-go basis can also be a new shirin poem in cloud security, but  Claudio Gold is an important person, and because it is often a flat- rate system due to early investment in security, I  only introduced the principles of Squared Security, but in a word, if you look at it this way,  cloud security cloud-like and cloud-like Boa can only grow in a place that deals with security and cloud together.

 The   so-called SDS is a global language school that Samsung started in 2001.

 No, it is      constantly evolving as an adjunct to cloud security, from the commercialization of the first cloud security service in Korea in 2015 to the first AWS QT computer sector in Korea in 2018 to the first ID G Market Ski Bu registration in Korea in 2021.

   In the new fields of the various clouds introduced today, another new alternative is being created.

 I hope those who are watching this video do not worry alone and take my experience as a word.

  Thank you.

   I will explain the core elements of cloud native more easily through a simple example.

 For example, if you run a shopping mall,  there are many functions for processing service requests, but let's assume that there are many login and payment functions here.

  In the traditional method, a server only runs one application.

 Therefore, in the case of the above auxiliary,  if virtual server 1 in charge of login is not known, it will lead to an overall service failure.

  To eliminate this failure, in a physical server environment, multiple physical servers are prepared and a load balancer is connected.

 When the service increases, the wool server is connected to the load balancer, and when there is no service, the physical server is removed from the load balancer and used for another purpose.

 On  the other hand, if you design a service based on microarchitecture and utilize containers, you can improve the speed and stability of the service.

 You    can run login payment applications on a single virtual server as containers.

 In this way, even if the virtual server is full,  the service will be provided without any problems through virtual server admission.

  When the service increases, the virtual servers will   automatically increase as the service increases through a technology called auto scaling, and the number will automatically decrease as the service decreases.

  Of course, the containers running on the virtual servers will also increase and decrease as the virtual servers increase.

  [Music] The effort required for infrastructure management will be greatly reduced.

 And you can easily understand that individual applications are continuously developed and updated by the development team.

 The   login function is developed and updated independently by Development Team A, and the payment function is developed independently by Development Team B.

 Even if  Development Team A adds the login function or reflects modifications, it will  not affect the payment function being developed by Development Team B at all, and continuous development and updates are possible.

 [Music] There  are various advantages of adopting cloud native, but they can be summarized into four.

  The first is securing a competitive advantage.

 Cloud day tips can increase productivity due to rapid deployment and an automated infrastructure environment, and this can bring forward the time to market.

 In addition, the 5 o'clock customer part that came in through the customer center can be  reflected in actual product development and continuous service improvement.

 Through this, you  can improve service quality and maximize customer experience.

 The second is flexibility.

  When you introduce and use Cloud Native, you can build applications that can run on any cloud without additional development or code modification.

  [Music]  Since application deployment is possible in a multi-cloud hybrid environment, you can   quickly transfer services to other clouds in the event of a cloud service failure, and reduce customer complaints due to failures.

 The  third is the introduction of cloud native, which motivates developers to focus on development.

 In  other words, you can reduce the overhead of writing code to run or scale code on multiple cloud infrastructures, and   instead focus on writing code that can increase customer value without knowing it.

  The fourth is the ability to focus on business.

 You can    reduce failures caused by engineer errors through automation throughout the process, and employees can focus on application improvement and business innovation instead of simple operational tasks.

  So far, we have learned about the definition and four key elements of cloud native.

 We  also learned that you can secure business competitiveness through a flexible operating environment and development environment through the introduction of cloud native.

   In the next episode, we will look at the four key elements of cloud native one by one.