Thank you for your answers, and give a unified answer: Why should nginx be restarted due to code changes?
Typical scenario, nginx-based microservices.
There must be some changes in service code upgrade. Any changes must restart the service. A and B must not affect each other.
(As to why nginx-based microservices are used, this is not the point. )
At present, the container service of cloud service provider can be configured, all 80 ports can be accessed, but I want to know how to realize it.
* * I don’t know why someone stepped on a problem.
Is there anything wrong with this problem, please point it out? **
What I don’t understand is why nginx should be restarted due to code changes.
If it is absolutely conflicting to map each container to 80, it is recommended that you use load balancing or reverse proxy to accomplish what you want.