Cloud, Ops & Security · Frontier
Cloud Native
A whole way of building software that assumes from day one it will run in the cloud: split into small pieces, packed into containers, scheduled automatically, able to grow and shrink at any moment. It is not a specific product.
TL;DRIn 30 seconds
- A whole way of building software that assumes from the start it lives in the cloud — not a product you buy.
- Programs split into small pieces, packed into containers, scheduled automatically, able to grow and shrink at will.
- Lifting an old program onto a cloud server isn't cloud native — that's just a different machine.
OverviewIn detail
Cloud native isn't a product or a single technique; it's a whole way of building software. It requires a program, from the very first day of design, to assume it will run in the cloud: split into small pieces each with one responsibility, packed into containers, handed to a scheduler to manage automatically, able to spin up a batch quickly when needed and recover on its own when something fails.
The reason to redo things this way is that lifting an old-style program onto the cloud often reaps none of the cloud's benefits. Such a program is usually one block bound together: scaling means copying the whole thing, changing one part means redeploying all of it, and one stuck step drags everything down. Put it in a cloud environment and it can neither scale nor be fixed quickly.
Rebuilding the cloud way buys elasticity and resilience: traffic rises and instances are added, it falls and they are reclaimed, a machine dies and the service as a whole doesn't notice. These aren't properties you get by moving; they are designed in, which means splitting the program apart, moving state outside it, and making startup something that can be copied at any moment. That's why two systems both running in the cloud can differ so sharply when traffic peaks — the difference isn't how many machines, but whether the software was built the cloud way.
It's often misread as meaning that running a program on a cloud server makes it cloud native. Moving house only changes the machine; the way the program is organised hasn't changed. Reshaping it to suit the cloud is what the word really means. Between those two things lies the work of architectural change — and the difference between actually having elasticity and not.
MythsCommon misunderstandings
- ✕Thinking running it on the cloud makes it cloud native ✓Moving house only changes the machine — it counts only if the program is rebuilt to scale and recover on its own
- ✕Thinking it's a product you can buy ✓It's an approach and a set of habits, put into practice through containers and schedulers — there's nothing off the shelf
- ✕Thinking one round of rework settles it ✓Change the traffic pattern and you keep adjusting — it's a state you maintain, not a one-off delivery
In practiceWhere you will meet it
- Switching data centre or cloud provider and the program still runs, without a rewrite
- Traffic multiplying overnight and the system adding instances itself instead of someone buying machines
- Reading 'split into small services, autoscaling' in a deployment plan — that's this approach
SummaryThe key points
- A whole way of building software designed for the cloud, not a product
- Delivered through containers, automatic scheduling and elastic scaling
- Running in the cloud isn't the same as cloud native — what matters is whether it was reworked to scale and recover
- For teams at scale heading to production, not for a casual script
FAQ
What does Cloud Native mean?
A whole way of building software that assumes from the start it lives in the cloud — not a product you buy.Programs split into small pieces, packed into containers, scheduled automatically, able to grow and shrink at will.Lifting an old program onto a cloud server isn't cloud native — that's just a different machine.
Where will I run into Cloud Native?
Switching data centre or cloud provider and the program still runs, without a rewrite;Traffic multiplying overnight and the system adding instances itself instead of someone buying machines;Reading 'split into small services, autoscaling' in a deployment plan — that's this approach。
What do people most often get wrong about Cloud Native?
Thinking running it on the cloud makes it cloud native — but Moving house only changes the machine — it counts only if the program is rebuilt to scale and recover on its own;Thinking it's a product you can buy — but It's an approach and a set of habits, put into practice through containers and schedulers — there's nothing off the shelf。
What are the key points about Cloud Native?
A whole way of building software designed for the cloud, not a product;Delivered through containers, automatic scheduling and elastic scaling;Running in the cloud isn't the same as cloud native — what matters is whether it was reworked to scale and recover;For teams at scale heading to production, not for a casual script。
Common pitfalls To tell whether something is cloud native, look at whether it scales and recovers on its own — not at which machine it runs on.
Entry 171 of 186 · All entries
Keep readingRelated reading
DevOps(开发运维一体化)
让写代码的和管服务器的人共用一套流程和工具,把测试、打包、上线交给机器自动完成,改完代码很快就能安全发布。
镜像(Image,软件打包模板)
把软件和它运行需要的环境一起打包好的模板,取出来在任意一台机器上都能跑出一样的结果。
HTTPS(加密网页协议)
给浏览器和网站之间的通信上锁加密,网址前面那个小锁就是它;路上有人偷看,也只能看到一串乱码。
漏洞(安全漏洞)
软件里一处能被坏人利用的缺陷,能让对方做本不该被允许的事,比如偷数据、冒充管理员或把程序弄崩。
会话(Session)
服务器用来记住这一段连续互动还是同一个人的方式;从你登录到退出之间的操作,被串成一整条记录。
Linux(操作系统)
免费开源的电脑操作系统,全球大多数服务器都在跑它;通常没有图形界面,但稳定、省资源、能长期不关机。
Where to go nextExplore more
Back to the skills themselves
Most of the concepts covered here come from skills that install into your AI client with a single command.