OK, we start with this session, this is the last session for the long day.I briefly introduce myself, I'm a former pivotal among the cloud computing department, cloud platform architect.Today I will talk about the cloud native continuous people application on the platform.首先我們參加這個會叫QCon,CNCF, cloud native computing foundation.So let's first take a look at the cloud native, cloud native being the future of my team.It has four parts, the first one is DevOps, the integration of development and operation, they are not separate.We use some tools and actions to integrate development and ops.It's more about the process procedure.The second part in the cloud native concept is continuous delivery of a deliverable integration.It's more about the tools, we can use the tools and tool chains to realize the drivers.And also the microservice and the container technology, these are the other two parts of the cloud native.The microservice will also have a similar design of a microservice, especially for cloud migration.How can we turn it into cloud native microservice?And whether it's a Docker or the Kubernetes, they provide the container technology and the container orchestration technology.These are the four technologies in the cloud native and what is the relationship among them.We can see microservice and containers and technologies, what are the purposes?The purpose is to realize the automation of development and operation.So from the perspective of business value, the functions can be launched into the market as soon as possible.So that users can use main functions.And therefore the DevOps is more about the culture of the procedures through the CIC.The other tools we can realize that we can use the microservice and the containers and DevOpsand the continuous delivery to make sure the applications can first respond to the changes.And then we can launch the continuous delivery by building the microservice.In the different environments of production, we can pass continuous delivery.So today I will focus on the CIC.So we use our tours for the continuous delivery and continuous production.Now we take a look at the CIC.If you are engaged in a development operation, you know about the CIC,but to give you a deeper understanding, I can show you the definition.The CI will continue its operation,referring to the changes you will make on the cloud channel.And then the changes should be combined into a major launchby the automated tests and checks to make sure the application and procedure will not be disrupted by these changes.So the key part is the chain-underminated test.So it's more developed and oriented for any cold-chain delivery.It should make sure the code can be released to the environmentand set of tests.The continuous delivery,a part of the formula,the automated tests,will rebuild the test and release.So it should be done in a faster and more frequent way.So for the CI should cover the plan called build and test,then we can move on.如果我們講CIC,but for CD,it should cover the steps of the techniques.But also there is a continuous deployment.Continuous deployment,is closer than the release.So for the CI and CD,for the softwares,it will provide as a requirement for each feature of the software development.It is after the cold-chaindeployed into the production environmentand it shows that you or your softwarehas completed the feature.But to the topic,today,the topic ishow come the CI,CIC,DT,will help with the applications and the platformto continue the delivery.So we can have a two-part release.This slide will say the pure tall products我們用的是簡單介紹一下主要的項目,we have two PES,one is PT,the other one is PES,the other one is PES,the other one is PES,the other one is PES,the other one is PES,the other one is PES,the other one is PES,the other one is PES,we want to have a CIso we take the opt-out pathwaythrough GitHubthe resource doc to manage it. Later on,share you the concourseit's a very important for the 20thand I'll give you some down-down concourseso the applicationrelease the application for CI,C,DTthen we can use the opt-out path to the platform,就是我們現在什麼時候都做ICDOne way talk about CI-CD is more about applicationbut for the continuous delivery and continuous deploymentthere is a report to the part that is a platformwhere they show the as upgrading or ground funding upgradingfor the operators and containersit is really difficulthow can we have the applications for upgrading our company现在我们来看看Pyplon的功能在Pyplon的功能中它有三个功能一是Page二是James如果你做Pyplon的功能或者是Summable你会做Pyplon的功能您需要一些评论您需要一些评论您可以有些特别的评论但是这些评论的评论您需要新的评论您需要一个新的评论您需要一种新的评论您需要一个新的评论您需要新的评论您需要新的评论您需要新的评论您需要新的评论您需要新的评论您需要新的评论您需要新的评论您需要新的评论您需要新的评论那么我们来看这三个设计现在我们来看第一个概念的话这个RESOURCERESOURCE是一个概念例如所有的团体系统它可能是一个RESOURCE一般的RESOURCE是包括GIT也许是S3也许是Dunker所有的RESOURCE所有的RESOURCE所有的设计也可以做到一个发布这个材料可以 also be the issueRESOURCE最重要的发布是发布的预设发布的配上对发布的配上对发布的配上能够得到的另外一个概念这个设计其实是定义的在这个设计上分类到预设的设计参与了3D的分类可以把设计组织了或是解先了在设计中就有了设计 对立的配上来自这个冠状Key concept is the task itself, each task can help you with some real execution of the scripts, the new scripts can be exported to the catalog, they can be run in stand-along containers, so the concourse, the software itself is a cloud-native software based on container is different from the genus from the most sublaves, is different from that, the concourse is running in the container, each job can be running in the container, so, relatively speaking, the resource of stand-along is independent, the task running environment is all different, so we will show you a simple plan now例如,我們需要將我們搜索給一些單元設施,給一些united tasks,給一些搜索給一些leasing to the platform所以, you can see the discrimination for you, they define the resource, is from a git, and the resource, and there is the cf, this is the time, that is the cf of the final product release, and for the jobs, one job is united task, it has a one-plan, we need to get to the cloud and run tasks, this is the task它这个Task 第一件是在我们家里进行的Task进行的Yammer Fire we can define it然后这里有一个Output that is released所以这是一个很简单的图案那么我们再来看一下我们所有的图案都会运动Task will be running in their individual stand-alonecontentless whether it is a compilingor screenshot or any other tasksit will run in a relatively standard container所以你可以看到整个如果在做concoursefor the concourseand horizontal scaling or scale-upit's very easy based on the container technologyand this can help to improve the horizontal scaling capability那么我们说其实concourseconcourse is very suitable to build the cloudit's a very typical scenario to release a true serviceif we have many different microservicestypically we need to integrate with the external servicesbut for microservices we can form some ABI testsfor interface testsbut still in addition to mock and testwe also need integration testsduring the test we can do a lot of parallelyou can test and theninclude in one job to do the integration testif the microservice is very largewe can open many containers so that they can run togetherif you are running a complicated cloud-native microservicesystemthis is very suitable to usehere is the demoI hope you read the network biblenot wellbut let's have a tryso here we have a springboard instructionstructure that here the streamers of the toolsit is one of the applicationsfirst we define that there are different jobs after the build we are deployingand then we needso together there are three jobs它的速度的話the inputdocker imagethen you can see a docker image然后下了速度的话for the output那么中间的话还有in betweenthere is an architecturewe need to put it after the buildyou might be familiar with the maverick's nexus那么我们这边here we use the address flow architecture那它们这个具体的这个步骤的话there are these specific stepsso I also put it on a githubif you are interestedyou can search on githubit basically means of steps在concourse的话there is another advantagethere is time pointthe UI shows you the whole show stageslike you can see three jobs firstyou need deployed and then release前面的话我有两个速度一个是说我们原码previously we have two inputsone is artificial那么一旦有原码的变换之后we can trigger it那我现在的话那我稍微改一点点就好now I slightly modify the source code比如我现在是1.0so now it's 1.0version 1.0and then I submit it to githubso every one minutesthere will be a freshduring the deploymentI want to finish my follow up那我们可以看点如果我在做if I'm runningif I'm working on the application deliverythe workflow will be like thissomething wrong with the slideokthe workflow looks like thisso submit the code and the buildand there will be a batteryand then put it in our warehousewhich is the warehouseand then run the testyou can perform some testor some integrated testand then the package will beput in this stage for testingthere is healthwhether it can satisfy the end usersand then deploy to such a production environmentso this is how we deliver the applicationthe demo will present you withthe processbefore we release the new versionwe want to guarantee thatold version will not be affectedby switching routersthe new version can be smoothly updatedso this is about the delivery of new applicationsand this is about platformthe typical scenario is thatwe have several upgradesand CVE that needs to be updatedand also the platform level operationand thisis suitable for the operators to usethe typical scenario hereis that we have a new CVEafter we have itthe CVE will be uploadedinto our storagethere are two scenariosone is the publiclastto do a updatewe can update itand the other is the privateprivate lastwe can see thatsince weiterate ourprojects the pipelines will changeand become more complicatedwith the projectby using concourseand the configuration fileseven when treatingsuch complicated workflowwe can still seethe detail processand on the topit's switching different groupsafter having many jobsyou can categorize themand by using different buildsyou can see clearlylikewe can binding with different peopleplaying different rolesand now we canturn back to the demofinallyso I have to close the slide firstfinallywe release itin the cloud foundryin the cloud foundrywe can seethere are two applicationstwo applications have been launchedone is registry demoblue and greenare two versionsafter releasingblue and green versionsfor examplehere is the original versionwhich is the green versionin the green versionwe have the independence router我要去替换它的时候it's very substitutionsince it is binding with the MACin the DNSfor switching to blue versionI will first release the versionand thendo some testto see if the blue version works welland thenyou can see the meant routerwe switch it to the blue versionwith such a switchrouter switches very quicklywe can basicallyachieveswitch without interruptionfor those businessesnow we can click on itand visit itso it's the version 1.0you might still in the processin the releasingso after you finish the workflowit will be updatedso quickly sum upsomething wrongorit's not onlyit's like a home-cycle-basedpracticeit requiresthe effective operationamong all the involved peopleso all the dev and opsthey cancommunicate without any gapsat the same timein a very complicated environmentwe can release ordeliver our latest productit is with low riskfrequency,localand rapidand predictablefinally we can usethe cultural and organizationstructure changes to achieveandrealize practicesfor developments orITmissions inside companiesdifferent companiesthey have differentprocess or workflowsif you want todeliver micro-servicesto achieve continuousdeploymentthis willchange thecompany culturalto a certain extentbut it is not onlyinvolve pipelinehaving toolsis not equal tofinal successthere are other factorslike database migrationrisk management etcin the companyafter applying solutions of toywe need to control the riskotherwise the existing systemand we need tomanage the life cycleand the auditing workif wewant tousing cnc issuedeliver ourapplication anddeliver the platformthe main challengecould bethe company cultureand thenthe links to our resourcesthe linkfor a coursethe tutorial of a courseand alsosome samplesi want toshow you nowby clicking on ityou will see the blue version and green versionand also some upgradationfor example like pksor pkshow tonew version of pksandhow to interact withawshow to interact withthe platformwe have very rich sampleson thislink