The programmer was beaten.

A good workman does his work well , You must sharpen your tools first .

1. It's fighting …

“ Smiling brother , R & D and operation and maintenance are about to start , Go and have a look …”

One night in a certain year , I am sitting in a small corner of the company , Put on your headphones and knock the code happily , Suddenly the girl from the testing department came to me .

It is said that R & D and testing are about to start , Let me see .

At that time , The company is making every effort to develop the fourth collection platform , It adopts the latest micro service architecture , At that time, there were not many cases of large-scale use in China .

We have a lot of problems , Among them, there is the cooperation between R & D and operation and maintenance , Previously, O & M may only need to deploy more than a dozen applications .

After adopting the microservice Architecture , The platform expands directly to 100 Multiple application instances , And there are dependencies between applications and application deployment .

The online process is complex and prone to problems , Responsibilities are not clear after the online accident , This leads to frequent conflicts between R & D and operation and maintenance .

And the later architecture upgrade , Cause the contradiction to further escalate .

2. Architecture upgrade

One year after using the microservice Architecture , To make more efficient use of server resources , Avoid problems caused by inconsistent server environments .

We are in the architecture , Introduced container technology Docker.

And this period K8s Also slowly mature , In the face of massive container scheduling and management , And will be K8s Into the architecture .

and K8s The introduction of , It also helps us solve another problem from the bottom .

We use Spring Cloud Build the microservices architecture , Solved the problem of configuration center 、 Service discovery 、 gateway 、 Fuse 、 Architecture problems such as load balancing .

These new technologies are enough for a novice to learn for a long time , And it's true that these technologies have nothing to do with the business .

in addition , Communicate in the microservice application layer 、 Current limiting 、 Load efficiency is not high , When there are enough applications , The problem will worsen .

and K8s, It can help us realize these functions from the bottom , It can not only improve the communication efficiency between services ,（ Because less processing part of the work ） It can also improve the R & D efficiency of developers .

K8s Is good , But it triggered the development and operation and maintenance …

3. Conflicts escalate

At this point, the whole architecture is more perfect ,K8s Responsible for cluster management ,Dokcer Responsible for container deployment , R & D focuses on business development .

but K8s Good as it is , But it brings another problem .

It 's the whole thing K8s The ecology of is too big , And new projects are pouring in every year , To improve the development of the whole ecology .

I'll send you one ,2022 Panorama of annual cloud primordial , You should be able to feel one or two .

And we want to release cloud native , You have to continue to understand and learn K8s, The key is to learn both R & D and O & M .

If developers don't understand K8s There's no way to develop , The O & M personnel will not K8s Unable to complete application clustering deployment , And it is not clear who is responsible for the deployment .

Experienced N After the second contradiction , On an overtime night , After another online failure , The operation and maintenance personnel and developers even fought …

What do I do ？

The problem must be solved after all , So I did some research on the Internet , What tools can lower the threshold for Cloud Applications , And can manage the whole life cycle of the application .

I really found it .

4.Orbit

Investigated many domestic and foreign tools , Most of them cover only part of the application lifecycle .

We want to lower the threshold for cloud applications , It also covers application development 、 deliver 、 Operation and maintenance 、 Deploy the entire cycle , There is only one lightweight tool that meets the requirements and even exceeds the expectations .

That's it Orbit.

Orbit The overall design idea is as follows .

Under the condition of ensuring the security of resource permissions , Hand over the operation and maintenance of the application to R & D , Responsible for the whole life management of applications , For example, the creation of applications 、 Release and destroy .

This avoids research and development ⼈ The operator can only ask the O & M for ⽇ Records , Query relevant problem information through operation and maintenance cooperation in various environments , This leads to a long problem-solving chain and extremely low efficiency .

stay Orbit In the design of , Operation and maintenance ⼈ Members focus on the delivery of resources , Responsible for familiarizing k8s And it's ⽣ state , To write templates or Yunyuan ⽣⽣ State plug-ins .

Research and development ⼈ The clerk introduced ⽤ These templates and plug-ins , Fill in a small number of business-related parameters , You can complete Yunyuan ⽣ Should be ⽤ The reconstruction of , Adapt to Yunyuan ⽣ The latest energy of the community ⼒.

This effectively distinguishes the responsibility boundary between R & D and operation and maintenance , R & D personnel do not need to understand K8s All the details , Perfect shielding K8s Complexity .

5. It's not just that

Yes, of course ,Orbit It's not just that .

Orbit Application Center , Application creation can be completed through graphical interface operation 、 change , And various forms of release , Improve application delivery efficiency and reliability .

In response to ⽤ details ⻚ in , You can see that this should ⽤ By which services 、 To configure 、 The data table consists of .

Can see should ⽤ Distance ⼀ How many changes have been accumulated in this release , Can see this should ⽤ To what environment .

If something goes wrong , Different operation and maintenance tools can be accessed according to the usage scenarios , View the monitoring included in the environment page 、 journal 、 event 、 Call chain tracing and other information positioning problems .

After the problem is located , Use the application publishing capability of the application center , It can be realized quickly hotfix Automated release .

Orbit Publishing strategy for , Blue and green publishing can be easily realized 、 Release of canary , It even supports the atomic release across regions and applications 、 Atomization rollback .

Orbit The application center adopts GitOps To practice K8s Application release of , The benefits are obvious .

adopt GitOps The control loop of K8s The declarative form of API And final state control , Users only need to put K8s Of yaml Files are submitted to the code repository .

GitOps Will automatically sense yaml Changes in documents , Cooperate with the merge request function of the code warehouse , take yaml File changes are automatically pushed to the cluster .

The whole process does not require developers to learn K8s To issue orders for , There is no need to operate the cluster directly , This greatly reduces the threshold for using cloud native .

Through the above introduction , I believe everyone should be right Orbit Have a whole understanding of .

Orbit It can not only lower the threshold of cloud native use , It also provides the creation of a complete set of cloud applications 、 deliver 、 monitor 、 The ability to deploy .

adopt Orbit This is a platform , Completed the whole life cycle management of cloud applications .

I just attended yesterday Techo Day Online activities , The key point of the product release of this activity is “ Lightweight tools ”, As mentioned above Orbit This is the release 7 One of the products .

in addition to , Tencent's first open source simulation platform for higher education “ Tencent digital twin Tad Sim“、 The first one for industry /

IOT platform for energy asset insight ” IOT device insight IoT Insight“.

And the first lightweight machine learning platform for individual developers ” Tencent cloud TI Platform public cloud version “, It really makes me feel that Tencent cloud is trying to lower the development threshold 、 Full sincerity to improve R & D efficiency .

After watching the press conference, I can feel , Tencent is in “ Lightweight tools ” A lot of R & D resources have been invested , The whole press conference was full of technical men's style .

The trend of using cloud native technology on the Internet is irresistible , and Orbit From a higher dimension , Help us lower the threshold of using cloud native technology .

In particular, the feeling of demonstrating monitoring is quite deep , You can check with one click through the graphical interface Bug, With this function, troubleshooting is much easier .

in addition , The event also introduced the technical principles of relevant lightweight star products of Tencent's entire cloud native service system 、 Applicable scenarios and other courseware contents are sorted into 《 Tencent cloud lightweight tool guide 》.

（ Long press identification to download ）